17,635 11,094 14MB
Pages 5113 Page size 612 x 792 pts (letter) Year 2009
About this release
FRONT MATTER PART ONE - INTRODUCTION TO CLINICAL MEDICINE PART TWO - CARDINAL MANIFESTATIONS AND PRESENTATION OF DISEASE PART THREE - GENETICS AND DISEASE PART FOUR - CLINICAL PHARMACOLOGY PART FIVE - NUTRITION PART SIX - ONCOLOGY AND HEMATOLOGY PART SEVEN - INFECTIOUS DISEASES PART EIGHT - DISORDERS OF THE CARDIOVASCULAR SYSTEM PART NINE - DISORDERS OF THE RESPIRATORY SYSTEM PART TEN - DISORDERS OF THE KIDNEY AND URINARY TRACT PART ELEVEN - DISORDERS OF THE GASTROINTESTINAL SYSTEM PART TWELVE - DISORDERS OF THE IMMUNE SYSTEM, CONNECTIVE TISSUE, AND JOINTS PART THIRTEEN - ENDOCRINOLOGY AND METABOLISM PART FOURTEEN - NEUROLOGIC DISORDERS PART FIFTEEN - ENVIROMENTAL AND OCCUPATIONAL HAZARDS
Harrison's Principles Of Internal Medicine 15th Edition © 2001 by The McGraw-Hill Companies, Inc.
Version 1.0 Compiled to iSilo format from Harrison's Principles of Medicine CD-ROM by snickers brought to you by PalmWarez This is the complete text of Harrison's Principles of Medicine 15th Edition, formatted for use on a Palm handheld with iSilo 3. In order to minimize file size, tables and figures were not included, and some parts of the original book have been omitted, such as parts of the Front Matter, the Bibliographies, the Nobel Prize articles, Color Atlases, and Appendices. Go back to book
HARRISON'S PRINCIPLES OF INTERNAL MEDICINE - 15TH EDITION FRONT MATTER EDITORS OF PREVIOUS EDITIONS T. R. Harrison Editor-in-Chief, Editions 1, 2, 3, 4, 5 W. R. Resnick Editor, Editions 1, 2, 3, 4, 5 M. M. Wintrobe Editor, Editions 1, 2, 3, 4, 5 Editor-in-Chief, Editions 6, 7 G. W. Thorn Editor, Editions 1, 2, 3, 4, 5, 6, 7 Editor-in-Chief, Edition 8 R. D. Adams Editor, Editions 2, 3, 4, 5, 6, 7, 8, 9, 10 P. B. Beeson Editor, Editions 1, 2 I. L. Bennett, Jr. Editor, Editions 3, 4, 5, 6 E. Braunwald Editor, Editions 6, 7, 8, 9, 10, 12, 13, 14 Editor-in-Chief, Edition 11 K. J. Isselbacher Editor, Editions 6, 7, 8, 10, 11, 12, 14 Editor-in-Chief, Editions 9, 13 R. G. Petersdorf Editor, Editions 6, 7, 8, 9, 11, 12, 13 Editor-in-Chief, Edition 10 J. D. Wilson Editor, Editions 9, 10, 11, 13, 14 Editor-in-Chief, Edition 12 J. B. Martin Editor, Editions 10, 11, 12, 13, 14 A. S. Fauci
Editor, Editions 11, 12, 13 Editor-in-Chief, Edition 14 R. Root Editor, Edition 12 D. L. Kasper Editor, Edition 13, 14 S. L. Hauser Editor, Edition 14 D. L. Longo Editor, Edition 14 TITLE PAGE EDITORS EUGENE BRAUNWALD, MD, MA (HON), MD (HON), SCD (HON) Distinguished Hersey Professor of Medicine, Faculty Dean for Academic Programs at Brigham and Women's Hospital and Massachusetts General Hospital, Harvard Medical School; Vice-President for Academic Programs, Partners HealthCare Systems, Boston ANTHONY S. FAUCI, MD, SCD (HON) Chief, Laboratory of Immunoregulation; Director, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda DENNIS L. KASPER, MD, MA (HON) William Ellery Channing Professor of Medicine, Professor of Microbiology and Molecular Genetics; Executive Dean for Academic Programs, Harvard Medical School; Director, Channing Laboratory, Department of Medicine, Brigham and Women's Hospital, Boston STEPHEN L. HAUSER, MD Chairman and Betty Anker Fife Professor, Department of Neurology, University of California San Francisco, San Francisco DAN L. LONGO, MD Scientific Director, National Institute on Aging, National Institutes of Health, Gerontology Research Center, Bethesda and Baltimore J. LARRY JAMESON, MD, PHD Irving S. Cutter Professor and Chairman, Department of Medicine, Northwestern University Medical School; Physician-in-Chief, Northwestern Memorial Hospital, Chicago McGraw-Hill MEDICAL PUBLISHING DIVISION
New York San Francisco Washington, DC Auckland Bogota Caracas Lisbon London Madrid Mexico City Milan Montreal New Delhi San Juan Singapore Sydney Tokyo Toronto COPYRIGHT PAGE McGraw-Hill A Division of The McGraw-Hill Companies Note: Dr. Fauci and Dr. Longo's works as editors and authors were performed outside the scope of their employment as U.S. government employees. These works represent their personal and professional views and not necessarily those of the U.S. government. Harrison's Principles Of Internal Medicine 15th Edition Copyright© 2001, 1998, 1994, 1991, 1987, 1983, 1980, 1977, 1974, 1970, 1966, 1962, 1958 by The McGraw-Hill Companies, Inc. All rights reserved. Printed in the United States of America. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be reproduced or distributed in any form or by any means, or stored in a data base or retrieval system, without the prior written permission of the publisher. 1234567890 DOWDOW 0987654321 ISBN 0-07-007272-8 (Combo) 0-07-007273-6 (Vol. 1) 0-07-007274-4 (Vol. 2) 0-07-913686-9 (Set) Additional illustrations in Chapters 18, 19, 124, 126, 128, 132, 139-141, 146, 147, 159, 161, 163, 169, 170-172, 176, 177, 182, 183, 186-188, 193-195, 200, 201, 203-205, 208, 215, 219, 220, 222, 398 are courtesy of Color Atlas and Synopsis of Clinical Dermatology, 4th edition, T.B. Fitzpatrick et al., New York, McGraw-Hill, 2001. Additional illustrations in Chapters 19, 124, 126, 130, 131, 138-141, 147, 151, 160, 163, 165-167, 169, 174, 177-179, 184, 190, 193 are courtesy of Atlas of Infectious Diseases Volumes I-IX, G.L. Mandell (ed.), Current Medicine, Inc., Philadelphia, 1997. Additional illustrations in Chapters 229-233, 238, 239, 241, 243-245 are courtesy of Essential Atlas of Heart Diseases, E. Braunwald (ed.), Current Medicine, Inc., Philadelphia, 1997. FOREIGN LANGUAGE EDITIONS Arabic (Thirteenth Edition)¾McGraw-Hill Libri Italia srl (est. 1996) Chinese (Twelfth Edition)¾McGraw-Hill Book Company-Singapore © 1994 Croatian (Thirteenth Edition)¾Placebo, Split, Croatia
French (Fourteenth Edition)¾McGraw-Hill Publishing Co., Maidenhead, UK ©1999 German (Fourteenth Edition)¾McGraw-Hill Publishing Co., Maidenhead, UK ©1999 Greek (Fourteenth Edition)¾Parissianos, Athens, Greece ©2000 Italian (Fourteenth Edition)¾McGraw-Hill Libri Italia srl, Milan ©1999 Japanese (Eleventh Edition)¾Hirokawa©1991 Polish (Fourteenth Edition)¾Czelej Publishing Company, Lubin, Poland ©2000 Portuguese (Fourteenth Edition)¾McGraw-Hill Interamericana do Brasil Ltda ©1998 Turkish (Thirteenth Edition)¾McGraw-Hill Libri Italia srl (est. 1996) Romania (Fourteenth Edition)¾Teora Publishers, Bucharest, Romania (est. 2000) Spanish (Fourteenth Edition)¾McGraw-Hill Interamericana de Espana, Madrid ©1998 This book was set in Times Roman by Progressive Information Technologies. The editors were Martin Wonsiewicz and Mariapaz Ramos Englis. The production director was Robert Laffler. The index was prepared by Irving C. Tullar. The text and cover designer was Marsha Cohen/Parallelogram Graphics. R. R. Donnelley and Sons, Inc. was the printer and binder. Library of Congress Cataloging-in-Publication Data Harrison's principles of internal medicine¾15th ed./editors, Eugene Braunwald...[et al.] p. cm. Includes bibliographic references and index. ISBN 0-07-913686-9 (set)¾ISBN 0-07-007273-6 (v. 1)¾ISBN 0-07-0072744-4 (v. 2) 1. Internal medicine. I. Braunwald, Eugene, date RC46.H333 2001 616¾dc21 00-063809 INTERNATIONAL EDITION ISBN 0-07-118319-1 (Set); 0-07-118320-5 (Vol 1); 0-07-118321-3 (Vol 2) Copyright © 2001. Exclusive rights by The McGraw-Hill Companies, Inc., for manufacture and export. This book cannot be re-exported from the country to which it is consigned by McGraw-Hill. The International Edition is not available in North America. DEDICATION KURT J. ISSELBACHER
With this edition, the editors acknowledge the many contributions of our colleague Kurt J. Isselbacher, who served as an editor of Harrison's for nine editions, the sixth through the fourteenth, including Editor-in-Chief of the ninth and thirteenth editions. For more than three decades Dr. Isselbacher played a decisive role in ensuring that Harrison's epitomized the state of the art and science of internal medicine and the essence of accuracy and clarity. His indelible contributions to Harrison's are felt in the fifteenth edition and will endure into the future. Dr. Isselbacher is a graduate of Harvard College and of the Harvard Medical School. His further training included a residency in medicine at the Massachusetts General Hospital and a research fellowship at the National Institutes of Health. Chosen to lead the Gastrointestinal Unit of the MGH at the remarkable age of 31, over the ensuing 30 years as Chief of that Unit, he was a leader in advancing both the clinical specialty of gastroenterology and the basic understanding of gastrointestinal disease. Under his leadership, the MGH Gastrointestinal Unit became renowned for its training program in academic gastroenterology as well as for being one of the world's leading centers for clinical and research activities in gastroenterology. In 1987, Dr. Isselbacher undertook new challenges as the first Director of the Cancer Center at the MGH. Bringing his characteristic insight and leadership to this new task, in a relatively short time, the MGH Cancer Center has emerged as a premier cancer research institute. Dr. Isselbacher holds the Mallinckrodt Distinguished Professorship of Medicine at Harvard Medical School, and he has been a powerful force for excellence in scholarship at this institution since his graduation. For almost 30 years he served as Chairman of the Executive Committee of Harvard's Departments of Medicine and played a pivotal role in the departments' growth and quest for excellence. Dr. Isselbacher combines the attributes of an excellent scientist with those of a superb clinician and teacher. He has trained generations of physicians and investigators, including many who are now leaders in academic medicine. As the author of more than 400 scientific articles in leading journals, his research contributions include definition of enzymatic defects in absorptive disorders, delineation of biochemical mechanisms of absorption, malabsorption, protein synthesis, derangements of metabolism, and immunologic aspects of hepatic gastrointestinal disease. Kurt Isselbacher has been a recipient of many well-earned honors, including the Distinguished Achievement Award and the Friedenwald Medal of the American Gastro- enterological Association, the John Phillips Memorial Award for distinguished contributions to clinical medicine from the American College of Physicians, as well as the Kober Medal of the Association of American Physicians. He is a member of the National Academy of Sciences, of its Institute of Medicine, and has served as President of the American Gastroenterological Association, the American Association for the Study of Liver Disease, and the Association of American Physicians. Kurt Isselbacher exemplifies the highest values of medicine. A caring, empathic physician, he consistently combines compassion with incisive analysis in the care of patients. With contributions as a clinician, teacher, scientist, and editor, he has advanced the care of patients with gastrointestinal disorders and cancer while educating
generations of physicians. JEAN DONALD WILSON
Jean Wilson served as editor of the ninth through the fourteenth editions of Harrison's Principles of Internal Medicine from 1978 to 1998; he was Editor-in-Chief of the twelfth edition. A native of Texas, Jean Wilson attended the University of Texas at Austin and the University of Texas Southwestern Medical School in Dallas. He trained as a resident in internal medicine and as a fellow in endocrinology and metabolism at Parkland Hospital. With the exception of 2 years of research in biochemistry in the intramural program of the National Institutes of Health, Dr. Wilson spent his entire career at the University of Texas Southwestern Medical School, where he now holds the Charles Cameron Sprague Distinguished Chair in Biomedical Science. Dr. Wilson is one America's most distinguished biomedical scientists and is largely responsible for working out the mechanism of action and physiology of the male sex hormones from the embryo to the normal and diseased adult. Among his many important discoveries has been the 5-reductase reaction, whereby male target tissues convert testosterone to the more active androgen, dihydrotestosterone. He has been honored many times for his research, having received the Ernst Oppenheimer Memorial Award of the Endocrine Society, the Amory Prize of the American Academy of Arts and Sciences, the Lita Annenberg Hazen Award for Excellence in Clinical Research, the Henry Dale Medal of the Society for Endocrinology, the Gregory Pincus Award of the Worcester Foundation for Experimental Biology, the Fred Conrad Koch Award of the Endocrine Society, and the Kober Medal of the Association of American Physicians. Amongst his memberships are the National Academy of Sciences, the Institute of Medicine of the National Academy of Sciences, and the American Academy of Arts and Sciences. He is a Fellow of the Royal College of Physicians. He has served as President of the American Society for Clinical Investigation, the Association of American Physicians, and the Endocrine Society. For two decades, Dr. Wilson directed the enormously successful MD/PhD program and for 8 years, the highly esteemed Endocrine-Metabolism Division at Southwestern. Perhaps the finest thing that can be said of Jean Wilson is that he is a professor of internal medicine in the complete sense. He is, and always has been, a superb teacher and exemplary clinician while constantly maintaining a sterling career in research. Perhaps nothing describes him better than the enduring image of this renowned academic physician trimming callouses and ulcers in the Diabetic Foot Care Clinic at Parkland Memorial Hospital, his teaching hospital, where the patients are the medically indigent of Dallas. Dr. Wilson is a man of diverse interests, a true intellectual. One of his great gifts and loves is scientific and medical editing. He served as Editor-in-Chief of the Journal of Clinical Investigation and of William's Textbook of Endocrinology. As an editor of Harrison's for two decades, Dr. Wilson made ample use of his conspicuous strengths as clinician, teacher, and scientist. His meticulous scholarship and high standards have
had an enormous impact, not only on the Endocrinology, Metabolism, and Genetics sections, for which he had primary responsibility, but on the entire book. JOSEPH B. MARTIN
The editors wish to acknowledge the enormous contributions made by Joseph B. Martin, who edited the Neurology section of Harrison's from the tenth to the fourteenth editions. Dr. Martin followed Dr. Raymond D. Adams as editor of the Neurology section. In retrospect, the choice of Joseph Martin to replace Adams was prescient. It foresaw the transformation of neurology in the 1980s and 1990s from a largely descriptive discipline to one of the most dynamic and rapidly evolving areas of internal medicine. With his appointment as editor, the textbook had secured the foremost leader in the new field of molecular neurology to its ranks. Beginning with the tenth edition, Martin built upon the powerful didactic structure of the "syndromic approach" to neurology created by Adams and emphasized advances in molecular genetics and cell biology that reclassify neurologic diseases, clarify disease mechanisms, and offer new insights into clinical diagnosis and therapy. During his tenure, the neurology section of Harrison's became the best resource of its kind for the exposition of new discoveries in neurology and contributed substantially to the growing overall success of the textbook. Born in Bassano, Alberta, Canada, Dr. Martin received his premedical and medical education at the University of Alberta, Edmonton, trained in neurology at Case Western University, and received the PhD from the University of Rochester. His career in academic medicine began in 1971 at McGill University in Montreal, where he established an independent laboratory focused on hypothalamic regulation of pituitary hormone secretion, and where he quickly rose to become Chair of the Department of Neurology and Neurosurgery. In 1978, he joined the faculty at Harvard Medical School as Bullard (later Julianne Dorn) Professor of Neurology and Chief of the Neurology Service at Massachusetts General Hospital. While at Harvard, he established the Huntington's Disease Center Without Walls, which in 1984 reported the spectacular finding of a genetic marker linked to Huntington's disease, thereby inaugurating the modern era of molecular neurogenetics. In 1989 Dr. Martin joined the University of California, San Francisco, initially serving as Dean of the School of Medicine and subsequently as Chancellor. Among his many achievements at UCSF was the conception of a major new research campus in San Francisco, which is fast becoming a reality. In July, 1997, he returned to Harvard as the Caroline Shields Walker Professor of Neurobiology and Clinical Neuroscience and Dean of the Faculty of Medicine. A wonderful teacher and physician, Joe Martin has inspired a generation of housestaff, students, and colleagues at Harvard and UCSF. Dr. Martin has received many honors, including honorary degrees from five distinguished universities and the Abraham Flexner Award of the Association of American Medical Colleges. He serves or has served on the editorial boards of nineteen medical and neurology journals. He is a member of the Institute of Medicine of the National Academy of Sciences and has served as President of the American Neurological Association.
Dr. Martin's many contributions to this textbook were enhanced by his extraordinary organizational skills and by a clear and direct style of writing and editing that permitted him to distill complex concepts into easily readable prose accessible to a general medical readership. As an example, his chapter on neurogenetics has become an instant classic and a highlight of each new edition of the book. The editors greatly value their friendship with this remarkable man whose integrity and intellectual strengths have served Harrison's so well during the past two decades. Although he is no longer an editor, we are delighted that Dr. Martin will continue to contribute his expertise to Harrison's as an author. NOTICE Medicine is an ever-changing science. As new research and clinical experience broaden our knowledge, changes in treatment and drug therapy are required. The editors and the publisher of this work have checked with sources believed to be reliable in their efforts to provide information that is complete and generally in accord with the standards accepted at the time of publication. However, in view of the possibility of human error or changes in medical sciences, neither the editors nor the publisher nor any other party who has been involved in the preparation or publication of this work warrants that the information contained herein is in every respect accurate or complete, and they are not responsible for any errors or omissions or the results obtained from the use of such information. Readers are encouraged to confirm the information contained herein with other sources. For example and in particular, readers are advised to check the product information sheet included in the package of each drug they plan to administer to be certain that the information contained in this book is accurate and that changes have not been made in the recommended dose or in the contraindications for administration. This recommendation is particularly important in connection with new or infrequently used drugs. PREFACE The first edition of Harrison's Principles of Internal Medicine was published in the middle of the twentieth century, more than 50 years ago. In this fifteenth edition, the first of the new century, the text has undergone major revision to reflect further understanding of the biology and pathophysiology of disease and at the same time to retain those facts that, while not new, remain clinically useful and important. Virtually every chapter in this new edition has been completely or substantially rewritten, and a record 86 are new or have new authors. In this preface, we cannot describe all of these changes; however, we would like to call to the reader's attention those that are particularly noteworthy. Part One, "Introduction to Clinical Medicine," contains new chapters dealing with decision making and cost awareness in clinical medicine. A growing number of patients are turning to alternative therapies, and these are discussed in a new chapter. New authors describe contemporary approaches to medical problems associated with pregnancy and the peripartum period. The chapters on medical ethics and on segments of the population that often present special problems¾adolescents, women, and the elderly¾have been revised and updated. Part Two, "Cardinal Manifestations and Presentation of Disease," serves as a
comprehensive introduction to clinical medicine, examining current concepts of the pathophysiology and differential diagnosis to be considered in patients with these manifestations. Major symptoms are reviewed and correlated with specific disease states, and clinical approaches to patients presenting with these symptoms are summarized. New chapters have been prepared on chest discomfort, headache, hypothermia, shock, and disorders of smell, taste, and hearing. A new chapter succinctly outlines a rational approach to the febrile patient presenting to the emergency department. The sections on alterations in gastrointestinal and sexual function are almost entirely new. Given the explosive advances in human genetics, including the completion of a working draft of the sequence of the entire human genome and its growing relevance to clinical practice, Part Three, "Genetics and Disease," has been expanded and completely rewritten with new chapters on human genetics, chromosomal genetics, genetic defects, mitochondrial dysfunction, genetic screening and counseling, as well as gene therapy. Part Four, "Clinical Pharmacology," provides a sound theoretical basis for pharmacotherapy, so critical to every aspect of medical practice. Part Five, "Nutrition," has been extensively revised, with five new authors contributing chapters. This section covers nutritional considerations related to clinical medicine, including nutritional requirements, assessment of nutritional status, protein-energy malnutrition, and enteral and parenteral nutrition. It contains a new chapter on obesity, which incorporates the results of rapidly developing basic research in this important field. The core of Harrison's encompasses the disorders of the organ systems and is contained in Parts Six through Fifteen. These sections include succinct accounts of the pathophysiology of the diseases involving the major organ systems and emphasize clinical manifestations, diagnostic procedures, differential diagnosis, and treatment strategies. The treatment sections of virtually every chapter have been amplified and updated. They are supplemented by the liberal use of algorithms, and are clearly highlighted. Guidelines for disease management prepared by specialty societies are included for the first time. Part Six, "Oncology and Hematology," includes twelve chapters with new authors, including a new chapter by Judah Folkman on angiogenesis. In addition, a new chapter has been added on the medical problems that can arise in patients cured of cancer, including disease-related and treatment-related sequelae. The chapters on myeloid and lymphoid neoplasms include the new World Health Organization classification schemes. A conscientious effort has been made to provide specific, up-to-date treatment recommendations. Where appropriate, diagnostic and management algorithms have been incorporated. Changes in Part Seven, "Infectious Diseases," include the latest information on the pathology, genetics, and epidemiology of infectious diseases while focusing sharply on the needs of clinicians who must accurately diagnose and treat infections in their patients. Specific recommendations are offered for therapeutic regimens, including the drug of choice, dose, duration, and alternatives. Current figures and trends in
antimicrobial resistance are presented and considered in light of their impact on therapeutic choices. New authors cover the latest advances in the management of diseases such as infective endocarditis, meningococcal and gonococcal infections, and schistosomiasis. The overview of pathogenesis from earlier editions has been expanded to encompass viruses, fungi, and parasites as well as bacteria. The Atlas of Hematology includes a complete diagnostic set of spectacular color plates showing malaria-infected red blood cells. In Part Eight, "Disorders of the Cardiovascular System," a new chapter on the prevention of atherosclerosis focuses not only on the importance of the traditional risk factors but also on the novel risk factors that influence plaque stability. Global risk assessment and management are described. Both primary and secondary prevention of atherosclerosis are discussed. Myocardial imaging by means of ultrasound or radionuclide techniques, at rest and during stress, plays an ever more critical role in assessment of patients with ischemic heart disease, and a new chapter focuses on the clinical use of these important technologies. Despite major advances in its diagnosis and therapy, acute myocardial infarction remains the most common cause of death in industrialized nations. The chapter on acute myocardial infarction provides important new information on myocardial reperfusion therapy, thrombolysis, and primary coronary angioplasty and summarizes guidelines for acute coronary care and for risk stratification in the postinfarct patient. Unstable angina and congestive heart failure have emerged as two of the most common conditions leading to hospital admission in Western nations. Important advances in pathophysiology and therapy of these two very important conditions are included. Enormous strides have been made in the use of lung transplantation for selected patients with end-stage, irreversible, pulmonary parenchymal and vascular disease, and Part Nine, "Disorders of the Respiratory System," provides a chapter that focuses on patient selection for this therapy. New chapters on interstitial and granulomatous lung diseases as well as on sleep apnea provide contemporary views of these conditions at the interface between basic science and clinical pulmonology. In Part Ten, "Disorders of the Kidney and Urinary Tract," there has been considerable revision, with a new chapter on dialysis, incorporating the most recent advances. In Part Eleven, "Disorders of the Gastrointestinal System," several new authors have contributed to the section on liver and biliary tract disease, and all chapters have been extensively revised. The section is pivoted by a new chapter on "Approach to the Patient with Liver Disease." Recent advances in the therapy of hepatitis B and C have been highlighted. New authors have contributed chapters on endoscopy, peptic ulcer disease, disorders of absorption, inflammatory bowel disease, and irritable bowel syndrome. Our new contributors include the leaders in gastroenterology and hepatology. In Part Twelve, "Disorders of the Immune System, Connective Tissue, and Joints," the updating focuses on therapy. The chapter on "Introduction to the Immune System" has been completely rewritten and provides a comprehensive review of the human immune system, using the modern designations of innate versus adaptive immunity. The chapter on HIV disease and AIDS is comprehensive and up-to-date and includes coverage of
the natural history, epidemiology, and immunopathogenic mechanisms of HIV disease. In addition, the chapter contains both an organ system by organ system approach and a delineation of the major complications of HIV disease. The sections on therapy include a state-of-the-art discussion of the striking treatment advances of HIV infection with combinations of antiretroviral agents as well as the complications of such therapy. Profound changes can be found in Part Thirteen, "Endocrinology and Metabolism." Many new authors have been recruited, and all chapters have been extensively revised under the direction of our new editor, Dr. J. Larry Jameson. Nine of these chapters are completely new, including those on the pituitary, thyroid, diabetes mellitus, and osteoporosis. These clinically demanding topics retain a traditional pathophysiologic approach that characterizes the field of endocrinology. In addition, new insights from genetics permeate this section, and the results of evidence-based medicine provide a firm foundation for medical decision making and treatment. Part Fourteen, "Neurologic Disorders," has been thoroughly updated and expanded. The theme of genetics is emphasized throughout the section, and new chapters highlight the remarkable progress made during the "decade of the brain" in the 1990s that has elucidated the molecular basis of many neurologic and psychiatric diseases. One of the new chapters, written by 1997 Nobel Laureate Stanley B. Prusiner, summarizes the unique biology of prions and the clinical features of human prion disorders, including "mad cow disease." The very latest information can be found on treatment of epilepsy, Parkinson's disease, and Alzheimer's disease. Coverage of immune-mediated disorders of the nervous system has been greatly expanded to include the many new insights into pathogenesis and treatment that have appeared since the fourteenth edition. The chapter on cerebrovascular diseases offers state-of-the-art information on prevention and treatment of stroke, the third leading killer in the developed world; this chapter is a mini-textbook of stroke and stroke therapy. Another feature of the fifteenth edition is a discussion of the acute neurologic disorders encountered in the setting of critical illness; this chapter should be of value to all physicians who care for hospitalized patients. Throughout the book, there is an emphasis on the use of neuroimaging figures to illustrate the various disorders discussed. Harrison's exceptional collection of high-quality neuroimaging photographs sets a new standard for textbooks of medicine. Finally, Part Fifteen, "Environmental and Occupational Hazards," has been expanded and reorganized. In view of the requirements for continuing education for licensure and relicensure, as well as the emphasis on certification and recertification, a revision of the Pre-Test Self-Assessment and Review will again be published with this edition. It consists of several hundred questions based on Harrison's, along with answers and explanations for the answers. The Companion Handbook that was pioneered as a supplement to the eleventh edition of Harrison's has been updated and will appear shortly after the publication of this edition. A CD-ROM version of Harrison's has been available since the thirteenth edition. An expanded CD-ROM version of the fifteenth edition will be available and will be regularly updated. In 1998, Harrison's went online to provide a "living"
textbook of internal medicine. In addition to providing full search capabilities of the text, Harrison's Online offers daily updating, reports of clinical trials, practice guidelines, and concise reviews of timely topics, as well as new references with links to MEDLINE abstracts. The fifteenth edition of Harrison's welcomes a new editor, Dr. J. Larry Jameson, who has taken on principal responsibility for the sections on Nutrition, Genetics, Endocrinology, and Metabolism and whose impact on this edition is already clear. Dr. Kurt J. Isselbacher, Dr. Jean D. Wilson, and Dr. Joseph B. Martin have left the editorial group. Their enormous contributions to Harrison's are cited elsewhere. Special thanks go to Dr. Robert F. Schrier who has prepared biographies of Nobel Prize Laureates in Physiology or Medicine. These brief essays remind us how deeply our current knowledge and practice of medicine depends on seminal contributions to biomedical science and informs about the lives of some of the most outstanding contributors. We wish to express our appreciation to our many associates and colleagues, who, as experts in their fields, have helped us with constructive criticism and helpful suggestions. We acknowledge especially the contributions of: Donna Ambrosino, Peter Banks, Richard Blumberg, Douglas Brust, Myron Cohen, Jonathan Edlow, Christopher Fanta, Mary Gillam, Douglas Golenbach, Fred Gorelick, Charles Halsted, Lee Kaplan, Peter Kopp, Bruce Levy, Leo Liu, William Lowe, Lawrence Madoff, Josh Meeks, Mark Molitch, Chung Owyang, Eugene Pergament, Alice Pau, Gerald Pier, Peter Rice, Paul Sax, Tom Schnitzer, Julian Seifter, Anushua Sinha, Steven Weinberger, Michael Wessels, and Lee Wetzler. This book could not have been edited without the dedicated help of our co-workers in the editorial offices of the individual editors. We are especially indebted to Scott Cromer, Pat Duffey, Sarah Anne Matero, Julie McCoy, Elizabeth Robbins, Kathryn Saxon, Marie Scurti, and Julieta Tayco. Finally, we continue to be indebted to two outstanding members of the McGraw-Hill organization: Mariapaz Ramos Englis, Senior Managing Editor, and Martin J. Wonsiewicz, Publisher. They are an effective team who have given the editors constant encouragement and sage advice and have been of enormous help in bringing this edition to fruition in a timely manner. The Editors
Harrison's Principles Of Internal Medicine 15th Edition © 2001 by The McGraw-Hill Companies, Inc.
PART ONE - INTRODUCTION TO CLINICAL MEDICINE 1. THE PRACTICE OF MEDICINE 2. ETHICAL ISSUES IN CLINICAL MEDICINE 3. DECISION-MAKING IN CLINICAL MEDICINE 4. ECONOMIC ISSUES IN CLINICAL MEDICINE 5. INFLUENCE OF ENVIRONMENTAL AND OCCUPATIONAL HAZARDS ON DISEASE 6. WOMEN'S HEALTH 7. MEDICAL DISORDERS DURING PREGNANCY 8. ADOLESCENT HEALTH PROBLEMS 9. GERIATRIC MEDICINE 10. PRINCIPLES OF DISEASE PREVENTION 11. ALTERNATIVE MEDICINE
PART TWO - CARDINAL MANIFESTATIONS AND PRESENTATION OF DISEASE SECTION 1 - PAIN SECTION 2 - ALTERATIONS IN BODY TEMPERATURE SECTION 3 - NERVOUS SYSTEM DYSFUNCTION SECTION 4 - DISORDERS OF EYES, EARS, NOSE, AND THROAT SECTION 5 - ALTERATIONS IN CIRCULATORY AND RESPIRATORY FUNCTIONS SECTION 6 - ALTERATIONS IN GASTROINTESTINAL FUNCTION SECTION 7 - ALTERATIONS IN RENAL AND URINARY TRACT FUNCTION SECTION 8 - ALTERATIONS IN SEXUAL FUNCTION AND REPRODUCTION SECTION 9 - ALTERATIONS IN THE SKIN SECTION 10 - HEMATOLOGIC ALTERATIONS
SECTION 1 - PAIN 12. PAIN: PATHOPHYSIOLOGY AND MANAGEMENT 13. CHEST DISCOMFORT AND PALPITATIONS 14. ABDOMINAL PAIN 15. HEADACHE, INCLUDING MIGRAINE AND CLUSTER HEADACHE 16. BACK AND NECK PAIN SECTION 2 - ALTERATIONS IN BODY TEMPERATURE 17. FEVER AND HYPERTHERMIA 18. FEVER AND RASH 19. APPROACH TO THE ACUTELY ILL INFECTED FEBRILE PATIENT 20. HYPOTHERMIA AND FROSTBITE SECTION 3 - NERVOUS SYSTEM DYSFUNCTION 21. FAINTNESS, SYNCOPE, DIZZINESS, AND VERTIGO 22. WEAKNESS, MYALGIAS, DISORDERS OF MOVEMENT, AND IMBALANCE
23. NUMBNESS, TINGLING, AND SENSORY LOSS 24. ACUTE CONFUSIONAL STATES AND COMA 25. APHASIAS AND OTHER FOCAL CEREBRAL DISORDERS 26. MEMORY LOSS AND DEMENTIA 27. SLEEP DISORDERS SECTION 4 - DISORDERS OF EYES, EARS, NOSE, AND THROAT 28. DISORDERS OF THE EYE 29. DISORDERS OF SMELL, TASTE, AND HEARING 30. INFECTIONS OF THE UPPER RESPIRATORY TRACT 31. ORAL MANIFESTATIONS OF DISEASE SECTION 5 - ALTERATIONS IN CIRCULATORY AND RESPIRATORY FUNCTIONS 32. DYSPNEA AND PULMONARY EDEMA 33. COUGH AND HEMOPTYSIS 34. APPROACH TO THE PATIENT WITH A HEART MURMUR 35. APPROACH TO THE PATIENT WITH HYPERTENSION 36. HYPOXIA AND CYANOSIS 37. EDEMA 38. SHOCK 39. CARDIOVASCULAR COLLAPSE, CARDIAC ARREST, AND SUDDEN CARDIAC DEATH SECTION 6 - ALTERATIONS IN GASTROINTESTINAL FUNCTION 40. DYSPHAGIA 41. NAUSEA, VOMITING, AND INDIGESTION 42. DIARRHEA AND CONSTIPATION 43. WEIGHT LOSS 44. GASTROINTESTINAL BLEEDING 45. JAUNDICE 46. ABDOMINAL SWELLING AND ASCITES SECTION 7 - ALTERATIONS IN RENAL AND URINARY TRACT FUNCTION 47. AZOTEMIA AND URINARY ABNORMALITIES 48. INCONTINENCE AND LOWER URINARY TRACT SYMPTOMS 49. FLUID AND ELECTROLYTE DISTURBANCES 50. ACIDOSIS AND ALKALOSIS SECTION 8 - ALTERATIONS IN SEXUAL FUNCTION AND REPRODUCTION 51. ERECTILE DYSFUNCTION 52. DISTURBANCES OF MENSTRUATION AND OTHER COMMON GYNECOLOGIC COMPLAINTS IN WOMEN 53. HIRSUTISM AND VIRILIZATION 54. INFERTILITY AND FERTILITY CONTROL SECTION 9 - ALTERATIONS IN THE SKIN 55. APPROACH TO THE PATIENT WITH A SKIN DISORDER 56. ECZEMA, PSORIASIS, CUTANEOUS INFECTIONS, ACNE, AND OTHER COMMON SKIN DISORDERS 57. SKIN MANIFESTATIONS OF INTERNAL DISEASE 58. IMMUNOLOGICALLY MEDIATED SKIN DISEASES 59. CUTANEOUS DRUG REACTIONS 60. PHOTOSENSITIVITY AND OTHER REACTIONS TO LIGHT SECTION 10 - HEMATOLOGIC ALTERATIONS 61. ANEMIA AND POLYCYTHEMIA
62. BLEEDING AND THROMBOSIS 63. ENLARGEMENT OF LYMPH NODES AND SPLEEN 64. DISORDERS OF GRANULOCYTES AND MONOCYTES
PART THREE - GENETICS AND DISEASE 65. PRINCIPLES OF HUMAN GENETICS 66. CHROMOSOME DISORDERS 67. DISEASES CAUSED BY GENETIC DEFECTS OF MITOCHONDRIA 68. SCREENING, COUNSELING, AND PREVENTION OF GENETIC DISORDERS 69. GENE THERAPY
PART FOUR - CLINICAL PHARMACOLOGY 70. PRINCIPLES OF DRUG THERAPY 71. ADVERSE REACTIONS TO DRUGS 72. PHYSIOLOGY AND PHARMACOLOGY OF THE AUTONOMIC NERVOUS SYSTEM
PART FIVE - NUTRITION 73. NUTRITIONAL REQUIREMENTS AND DIETARY ASSESSMENT 74. MALNUTRITION AND NUTRITIONAL ASSESSMENT 75. VITAMIN AND TRACE MINERAL DEFICIENCY AND EXCESS 76. ENTERAL AND PARENTERAL NUTRITION THERAPY 77. OBESITY 78. EATING DISORDERS
PART SIX - ONCOLOGY AND HEMATOLOGY SECTION 1 - NEOPLASTIC DISORDERS SECTION 2 - DISORDERS OF HEMATOPOIESIS SECTION 3 - DISORDERS OF HEMOSTASIS
SECTION 1 - NEOPLASTIC DISORDERS 79. APPROACH TO THE PATIENT WITH CANCER 80. PREVENTION AND EARLY DETECTION OF CANCER
81. CANCER GENETICS 82. CELL BIOLOGY OF CANCER 83. ANGIOGENESIS 84. PRINCIPLES OF CANCER TREATMENT 85. INFECTIONS IN PATIENTS WITH CANCER 86. MELANOMA AND OTHER SKIN CANCERS 87. HEAD AND NECK CANCER 88. NEOPLASMS OF THE LUNG 89. BREAST CANCER 90. GASTROINTESTINAL TRACT CANCER 91. TUMORS OF THE LIVER AND BILIARY TRACT 92. PANCREATIC CANCER 93. ENDOCRINE TUMORS OF THE GASTROINTESTINAL TRACT AND PANCREAS 94. BLADDER AND RENAL CELL CARCINOMAS 95. HYPERPLASTIC AND MALIGNANT DISEASES OF THE PROSTATE 96. TESTICULAR CANCER 97. GYNECOLOGIC MALIGNANCIES 98. SOFT TISSUE AND BONE SARCOMAS AND BONE METASTASES 99. METASTATIC CANCER OF UNKNOWN PRIMARY SITE 100. PARANEOPLASTIC SYNDROMES 101. PARANEOPLASTIC NEUROLOGIC SYNDROMES 102. ONCOLOGIC EMERGENCIES 103. LATE CONSEQUENCES OF CANCER AND ITS TREATMENT SECTION 2 - DISORDERS OF HEMATOPOIESIS 104. HEMATOPOIESIS 105. IRON DEFICIENCY AND OTHER HYPOPROLIFERATIVE ANEMIAS 106. HEMOGLOBINOPATHIES 107. MEGALOBLASTIC ANEMIAS 108. HEMOLYTIC ANEMIAS AND ACUTE BLOOD LOSS 109. APLASTIC ANEMIA, MYELODYSPLASIA, AND RELATED BONE MARROW FAILURE SYNDROMES 110. POLYCYTHEMIA VERA AND OTHER MYELOPROLIFERATIVE DISEASES 111. ACUTE AND CHRONIC MYELOID LEUKEMIA 112. MALIGNANCIES OF LYMPHOID CELLS 113. PLASMA CELL DISORDERS 114. TRANSFUSION BIOLOGY AND THERAPY 115. BONE MARROW AND STEM CELL TRANSPLANTATION SECTION 3 - DISORDERS OF HEMOSTASIS 116. DISORDERS OF THE PLATELET AND VESSEL WALL 117. DISORDERS OF COAGULATION AND THROMBOSIS 118. ANTICOAGULANT, FIBRINOLYTIC, AND ANTIPLATELET THERAPY
PART SEVEN - INFECTIOUS DISEASES SECTION 1 - BASIC CONSIDERATIONS IN INFECTIOUS DISEASES SECTION 2 - CLINICAL SYNDROMES: COMMUNITY-ACQUIRED INFECTIONS SECTION 3 - CLINICAL SYNDROMES: NOSOCOMIAL INFECTIONS
SECTION 4 - APPROACH TO THERAPY FOR BACTERIAL DISEASES SECTION 5 - DISEASES CAUSED BY GRAM-POSITIVE BACTERIA SECTION 6 - DISEASES CAUSED BY GRAM-NEGATIVE BACTERIA SECTION 7 - MISCELLANEOUS BACTERIAL INFECTIONS SECTION 8 - MYCOBACTERIAL DISEASES SECTION 9 - SPIROCHETAL DISEASES SECTION 10 - RICKETTSIA, MYCOPLASMA, AND CHLAMYDIA SECTION 11 - VIRAL DISEASES SECTION 12 - DNA VIRUSES SECTION 13 - DNA AND RNA RESPIRATORY VIRUSES SECTION 14 - RNA VIRUSES SECTION 15 - FUNGAL AND ALGAL INFECTIONS SECTION 16 - PROTOZOAL AND HELMINTHIC INFECTIONS: GENERAL CONSIDERATIONS SECTION 17 - PROTOZOAL INFECTIONS SECTION 18 - HELMINTHIC INFECTIONS
SECTION 1 - BASIC CONSIDERATIONS IN INFECTIOUS DISEASES 119. INTRODUCTION TO INFECTIOUS DISEASES: HOST-PARASITE INTERACTIONS 120. MOLECULAR MECHANISMS OF MICROBIAL PATHOGENESIS 121. LABORATORY DIAGNOSIS OF INFECTIOUS DISEASES 122. IMMUNIZATION PRINCIPLES AND VACCINE USE 123. HEALTH ADVICE FOR INTERNATIONAL TRAVEL SECTION 2 - CLINICAL SYNDROMES: COMMUNITY-ACQUIRED INFECTIONS 124. SEPSIS AND SEPTIC SHOCK 125. FEVER OF UNKNOWN ORIGIN 126. INFECTIVE ENDOCARDITIS 127. INFECTIOUS COMPLICATIONS OF BITES AND BURNS 128. INFECTIONS OF THE SKIN, MUSCLE, AND SOFT TISSUES 129. OSTEOMYELITIS 130. INTRAABDOMINAL INFECTIONS AND ABSCESSES 131. ACUTE INFECTIOUS DIARRHEAL DISEASES AND BACTERIAL FOOD POISONING 132. SEXUALLY TRANSMITTED DISEASES: OVERVIEW AND CLINICAL APPROACH 133. PELVIC INFLAMMATORY DISEASE SECTION 3 - CLINICAL SYNDROMES: NOSOCOMIAL INFECTIONS 134. INFECTION CONTROL IN THE HOSPITAL 135. HOSPITAL-ACQUIRED AND INTRAVASCULAR DEVICE-RELATED INFECTIONS 136. INFECTIONS IN TRANSPLANT RECIPIENTS SECTION 4 - APPROACH TO THERAPY FOR BACTERIAL DISEASES 137. TREATMENT AND PROPHYLAXIS OF BACTERIAL INFECTIONS SECTION 5 - DISEASES CAUSED BY GRAM-POSITIVE BACTERIA 138. PNEUMOCOCCAL INFECTIONS 139. STAPHYLOCOCCAL INFECTIONS 140. STREPTOCOCCAL AND ENTEROCOCCAL INFECTIONS 141. DIPHTHERIA, OTHER CORYNEBACTERIAL INFECTIONS, AND ANTHRAX
142. INFECTIONS CAUSED BY LISTERIA MONOCYTOGENES 143. TETANUS 144. BOTULISM 145. GAS GANGRENE, ANTIBIOTIC-ASSOCIATED COLITIS, AND OTHER CLOSTRIDIAL INFECTIONS SECTION 6 - DISEASES CAUSED BY GRAM-NEGATIVE BACTERIA 146. MENINGOCOCCAL INFECTIONS 147. GONOCOCCAL INFECTIONS 148. MORAXELLA CATARRHALIS AND OTHER MORAXELLA SPECIES 149. HAEMOPHILUS INFECTIONS 150. INFECTIONS DUE TO THE HACEK GROUP AND MISCELLANEOUS GRAM-NEGATIVE BACTERIA 151. LEGIONELLA INFECTION 152. PERTUSSIS AND OTHER BORDETELLA INFECTIONS 153. DISEASES CAUSED BY GRAM-NEGATIVE ENTERIC BACILLI 154. HELICOBACTER PYLORI INFECTIONS 155. INFECTIONS DUE TO PSEUDOMONAS SPECIES AND RELATED ORGANISMS 156. SALMONELLOSIS 157. SHIGELLOSIS 158. INFECTIONS DUE TO CAMPYLOBACTER AND RELATED SPECIES 159. CHOLERA AND OTHER VIBRIOSES 160. BRUCELLOSIS 161. TULAREMIA 162. PLAGUE AND OTHER YERSINIA INFECTIONS 163. BARTONELLA INFECTIONS, INCLUDING CAT-SCRATCH DISEASE 164. DONOVANOSIS SECTION 7 - MISCELLANEOUS BACTERIAL INFECTIONS 165. NOCARDIOSIS 166. ACTINOMYCOSIS 167. INFECTIONS DUE TO MIXED ANAEROBIC ORGANISMS SECTION 8 - MYCOBACTERIAL DISEASES 168. ANTIMYCOBACTERIAL AGENTS 169. TUBERCULOSIS 170. LEPROSY (HANSEN'S DISEASE) 171. INFECTIONS DUE TO NONTUBERCULOUS MYCOBACTERIA SECTION 9 - SPIROCHETAL DISEASES 172. SYPHILIS 173. ENDEMIC TREPONEMATOSES 174. LEPTOSPIROSIS 175. RELAPSING FEVER 176. LYME BORRELIOSIS SECTION 10 - RICKETTSIA, MYCOPLASMA, AND CHLAMYDIA 177. RICKETTSIAL DISEASES 178. MYCOPLASMA INFECTIONS 179. CHLAMYDIAL INFECTIONS SECTION 11 - VIRAL DISEASES 180. MEDICAL VIROLOGY 181. ANTIVIRAL CHEMOTHERAPY, EXCLUDING ANTIRETROVIRAL DRUGS SECTION 12 - DNA VIRUSES
182. HERPES SIMPLEX VIRUSES 183. VARICELLA-ZOSTER VIRUS INFECTIONS 184. EPSTEIN-BARR VIRUS INFECTIONS, INCLUDING INFECTIOUS MONONUCLEOSIS 185. CYTOMEGALOVIRUS AND HUMAN HERPESVIRUS TYPES 6, 7, AND 8 186. SMALLPOX, VACCINIA, AND OTHER POXVIRUSES 187. PARVOVIRUS 188. HUMAN PAPILLOMAVIRUSES SECTION 13 - DNA AND RNA RESPIRATORY VIRUSES 189. COMMON VIRAL RESPIRATORY INFECTIONS 190. INFLUENZA SECTION 14 - RNA VIRUSES 191. THE HUMAN RETROVIRUSES 192. VIRAL GASTROENTERITIS 193. ENTEROVIRUSES AND REOVIRUSES 194. MEASLES (RUBEOLA) 195. RUBELLA (GERMAN MEASLES) 196. MUMPS - Anne Gershon 197. RABIES VIRUS AND OTHER RHABDOVIRUSES 198. INFECTIONS CAUSED BY ARTHROPOD- AND RODENT-BORNE VIRUSES 199. FILOVIRIDAE (MARBURG AND EBOLA VIRUSES) SECTION 15 - FUNGAL AND ALGAL INFECTIONS 200. DIAGNOSIS AND TREATMENT OF FUNGAL INFECTIONS 201. HISTOPLASMOSIS 202. COCCIDIOIDOMYCOSIS 203. BLASTOMYCOSIS 204. CRYPTOCOCCOSIS 205. CANDIDIASIS 206. ASPERGILLOSIS 207. MUCORMYCOSIS 208. MISCELLANEOUS MYCOSES AND ALGAL INFECTIONS 209. PNEUMOCYSTIS CARINII INFECTION SECTION 16 - PROTOZOAL AND HELMINTHIC INFECTIONS: GENERAL CONSIDERATIONS 210. APPROACH TO THE PATIENT WITH PARASITIC INFECTION 211. LABORATORY DIAGNOSIS OF PARASITIC INFECTIONS 212. THERAPY FOR PARASITIC INFECTIONS SECTION 17 - PROTOZOAL INFECTIONS 213. AMEBIASIS AND INFECTION WITH FREE-LIVING AMEBAS 214. MALARIA AND BABESIOSIS: DISEASES CAUSED BY RED BLOOD CELL PARASITES 215. LEISHMANIASIS 216. TRYPANOSOMIASIS 217. TOXOPLASMA INFECTION 218. PROTOZOAL INTESTINAL INFECTIONS AND TRICHOMONIASIS SECTION 18 - HELMINTHIC INFECTIONS 219. TRICHINELLA AND OTHER TISSUE NEMATODES 220. INTESTINAL NEMATODES 221. FILARIASIS AND RELATED INFECTIONS (LOIASIS, ONCHOCERCIASIS, AND
DRACUNCULIASIS) 222. SCHISTOSOMIASIS AND OTHER TREMATODE INFECTIONS 223. CESTODES
PART EIGHT - DISORDERS OF THE CARDIOVASCULAR SYSTEM SECTION 1 - DIAGNOSIS SECTION 2 - DISORDERS OF RHYTHM SECTION 3 - DISORDERS OF THE HEART SECTION 4 - VASCULAR DISEASE
SECTION 1 - DIAGNOSIS 224. APPROACH TO THE PATIENT WITH HEART DISEASE 225. PHYSICAL EXAMINATION OF THE CARDIOVASCULAR SYSTEM 226. ELECTROCARDIOGRAPHY 227. NONINVASIVE CARDIAC IMAGING: ECHOCARDIOGRAPHY AND NUCLEAR CARDIOLOGY 228. DIAGNOSTIC CARDIAC CATHETERIZATION AND ANGIOGRAPHY SECTION 2 - DISORDERS OF RHYTHM 229. THE BRADYARRHYTHMIAS: DISORDERS OF SINUS NODE FUNCTION AND AV CONDUCTION DISTURBANCES 230. THE TACHYARRHYTHMIAS SECTION 3 - DISORDERS OF THE HEART 231. NORMAL AND ABNORMAL MYOCARDIAL FUNCTION 232. HEART FAILURE 233. CARDIAC TRANSPLANTATION 234. CONGENITAL HEART DISEASE IN THE ADULT 235. RHEUMATIC FEVER 236. VALVULAR HEART DISEASE 237. COR PULMONALE 238. THE CARDIOMYOPATHIES AND MYOCARDITIDES 239. PERICARDIAL DISEASE 240. CARDIAC TUMORS, CARDIAC MANIFESTATIONS OF SYSTEMIC DISEASES, AND TRAUMATIC CARDIAC INJURY SECTION 4 - VASCULAR DISEASE 241. THE PATHOGENESIS OF ATHEROSCLEROSIS 242. PREVENTION AND TREATMENT OF ATHEROSCLEROSIS 243. ACUTE MYOCARDIAL INFARCTION 244. ISCHEMIC HEART DISEASE 245. PERCUTANEOUS CORONARY REVASCULARIZATION 246. HYPERTENSIVE VASCULAR DISEASE 247. DISEASES OF THE AORTA 248. VASCULAR DISEASES OF THE EXTREMITIES
PART NINE - DISORDERS OF THE RESPIRATORY SYSTEM SECTION 1 - DIAGNOSIS SECTION 2 - DISEASES OF THE RESPIRATORY SYSTEM
SECTION 1 - DIAGNOSIS 249. APPROACH TO THE PATIENT WITH DISEASE OF THE RESPIRATORY SYSTEM 250. DISTURBANCES OF RESPIRATORY FUNCTION 251. DIAGNOSTIC PROCEDURES IN RESPIRATORY DISEASE SECTION 2 - DISEASES OF THE RESPIRATORY SYSTEM 252. ASTHMA 253. HYPERSENSITIVITY PNEUMONITIS AND PULMONARY INFILTRATES WITH EOSINOPHILIA 254. ENVIRONMENTAL LUNG DISEASES 255. PNEUMONIA, INCLUDING NECROTIZING PULMONARY INFECTIONS (LUNG ABSCESS) 256. BRONCHIECTASIS 257. CYSTIC FIBROSIS 258. CHRONIC BRONCHITIS, EMPHYSEMA, AND AIRWAYS OBSTRUCTION 259. INTERSTITIAL LUNG DISEASES 260. PRIMARY PULMONARY HYPERTENSION 261. PULMONARY THROMBOEMBOLISM 262. DISORDERS OF THE PLEURA, MEDIASTINUM, AND DIAPHRAGM 263. DISORDERS OF VENTILATION 264. SLEEP APNEA 265. ACUTE RESPIRATORY DISTRESS SYNDROME 266. MECHANICAL VENTILATORY SUPPORT 267. LUNG TRANSPLANTATION
PART TEN - DISORDERS OF THE KIDNEY AND URINARY TRACT 268. DISTURBANCES OF RENAL FUNCTION 269. ACUTE RENAL FAILURE 270. CHRONIC RENAL FAILURE 271. DIALYSIS IN THE TREATMENT OF RENAL FAILURE 272. TRANSPLANTATION IN THE TREATMENT OF RENAL FAILURE 273. PATHOGENESIS OF GLOMERULAR INJURY 274. THE MAJOR GLOMERULOPATHIES 275. GLOMERULOPATHIES ASSOCIATED WITH MULTISYSTEM DISEASES 276. HEREDITARY TUBULAR DISORDERS 277. TUBULOINTERSTITIAL DISEASES OF THE KIDNEY 278. VASCULAR INJURY TO THE KIDNEY
279. NEPHROLITHIASIS 280. URINARY TRACT INFECTIONS AND PYELONEPHRITIS 281. URINARY TRACT OBSTRUCTION
PART ELEVEN - DISORDERS OF THE GASTROINTESTINAL SYSTEM SECTION 1 - DISORDERS OF THE ALIMENTARY TRACT SECTION 2 - LIVER AND BILIARY TRACT DISEASE SECTION 3 - DISORDERS OF THE PANCREAS
SECTION 1 - DISORDERS OF THE ALIMENTARY TRACT 282. APPROACH TO THE PATIENT WITH GASTROINTESTINAL DISEASE 283. GASTROINTESTINAL ENDOSCOPY 284. DISEASES OF THE ESOPHAGUS 285. PEPTIC ULCER DISEASE AND RELATED DISORDERS 286. DISORDERS OF ABSORPTION 287. INFLAMMATORY BOWEL DISEASE 288. IRRITABLE BOWEL SYNDROME 289. DIVERTICULAR, VASCULAR, AND OTHER DISORDERS OF THE INTESTINE AND PERITONEUM 290. ACUTE INTESTINAL OBSTRUCTION 291. ACUTE APPENDICITIS SECTION 2 - LIVER AND BILIARY TRACT DISEASE 292. APPROACH TO THE PATIENT WITH LIVER DISEASE 293. EVALUATION OF LIVER FUNCTION 294. BILIRUBIN METABOLISM AND THE HYPERBILIRUBINEMIAS 295. ACUTE VIRAL HEPATITIS 296. TOXIC AND DRUG-INDUCED HEPATITIS 297. CHRONIC HEPATITIS 298. ALCOHOLIC LIVER DISEASE 299. CIRRHOSIS AND ITS COMPLICATIONS 300. INFILTRATIVE, GENETIC, AND METABOLIC DISEASES AFFECTING THE LIVER 301. LIVER TRANSPLANTATION 302. DISEASES OF THE GALLBLADDER AND BILE DUCTS SECTION 3 - DISORDERS OF THE PANCREAS 303. APPROACH TO THE PATIENT WITH PANCREATIC DISEASE 304. ACUTE AND CHRONIC PANCREATITIS
PART TWELVE - DISORDERS OF THE IMMUNE SYSTEM, CONNECTIVE TISSUE, AND JOINTS SECTION 1 - DISORDERS OF THE IMMUNE SYSTEM
SECTION 2 - DISORDERS OF IMMUNE-MEDIATED INJURY SECTION 3 - DISORDERS OF THE JOINTS
SECTION 1 - DISORDERS OF THE IMMUNE SYSTEM 305. INTRODUCTION TO THE IMMUNE SYSTEM 306. THE MAJOR HISTOCOMPATIBILITY GENE COMPLEX 307. AUTOIMMUNITY AND AUTOIMMUNE DISEASES 308. PRIMARY IMMUNE DEFICIENCY DISEASES 309. HUMAN IMMUNODEFICIENCY VIRUS (HIV) DISEASE: AIDS AND RELATED DISORDERS SECTION 2 - DISORDERS OF IMMUNE-MEDIATED INJURY 310. ALLERGIES, ANAPHYLAXIS, AND SYSTEMIC MASTOCYTOSIS 311. SYSTEMIC LUPUS ERYTHEMATOSUS 312. RHEUMATOID ARTHRITIS 313. SYSTEMIC SCLEROSIS (SCLERODERMA) 314. SJOGREN'S SYNDROME 315. ANKYLOSING SPONDYLITIS, REACTIVE ARTHRITIS, AND UNDIFFERENTIATED SPONDYLOARTHROPATHY 316. BEHCET'S SYNDROME 317. THE VASCULITIS SYNDROMES 318. SARCOIDOSIS 319. AMYLOIDOSIS SECTION 3 - DISORDERS OF THE JOINTS 320. APPROACH TO ARTICULAR AND MUSCULOSKELETAL DISORDERS 321. OSTEOARTHRITIS 322. GOUT AND OTHER CRYSTAL ARTHROPATHIES 323. INFECTIOUS ARTHRITIS 324. PSORIATIC ARTHRITIS AND ARTHRITIS ASSOCIATED WITH GASTROINTESTINAL DISEASE 325. RELAPSING POLYCHONDRITIS AND OTHER ARTHRITIDES 326. PERIARTICULAR DISORDERS OF THE EXTREMITIES
PART THIRTEEN - ENDOCRINOLOGY AND METABOLISM SECTION 1 - ENDOCRINOLOGY SECTION 2 - DISORDERS OF BONE AND MINERAL METABOLISM SECTION 3 - DISORDERS OF INTERMEDIARY METABOLISM
SECTION 1 - ENDOCRINOLOGY 327. PRINCIPLES OF ENDOCRINOLOGY 328. DISORDERS OF THE ANTERIOR PITUITARY AND HYPOTHALAMUS 329. DISORDERS OF THE NEUROHYPOPHYSIS 330. DISORDERS OF THE THYROID GLAND 331. DISORDERS OF THE ADRENAL CORTEX
332. PHEOCHROMOCYTOMA 333. DIABETES MELLITUS 334. HYPOGLYCEMIA 335. DISORDERS OF THE TESTES 336. DISORDERS OF THE OVARY AND FEMALE REPRODUCTIVE TRACT 337. ENDOCRINE DISORDERS OF THE BREAST 338. DISORDERS OF SEXUAL DIFFERENTIATION 339. DISORDERS AFFECTING MULTIPLE ENDOCRINE SYSTEMS SECTION 2 - DISORDERS OF BONE AND MINERAL METABOLISM 340. INTRODUCTION TO BONE AND MINERAL METABOLISM 341. DISEASES OF THE PARATHYROID GLAND AND OTHER HYPER- AND HYPOCALCEMIC DISORDERS 342. OSTEOPOROSIS 343. PAGET'S DISEASE AND OTHER DYSPLASIAS OF BONE SECTION 3 - DISORDERS OF INTERMEDIARY METABOLISM 344. DISORDERS OF LIPOPROTEIN METABOLISM 345. HEMOCHROMATOSIS 346. THE PORPHYRIAS 347. DISORDERS OF PURINE AND PYRIMIDINE METABOLISM 348. WILSON'S DISEASE 349. LYSOSOMAL STORAGE DISEASES 350. GLYCOGEN STORAGE DISEASES AND OTHER INHERITED DISORDERS OF CARBOHYDRATE METABOLISM 351. INHERITED DISORDERS OF CONNECTIVE TISSUE 352. INHERITED DISORDERS OF AMINO ACID METABOLISM AND STORAGE 353. INHERITED DEFECTS OF MEMBRANE TRANSPORT 354. THE LIPODYSTROPHIES AND OTHER PRIMARY DISORDERS OF ADIPOSE TISSUE
PART FOURTEEN - NEUROLOGIC DISORDERS SECTION 1 - DIAGNOSIS OF NEUROLOGIC DISORDERS SECTION 2 - DISEASES OF THE CENTRAL NERVOUS SYSTEM SECTION 3 - DISORDERS OF NERVE AND MUSCLE SECTION 4 - CHRONIC FATIGUE SYNDROME SECTION 5 - PSYCHIATRIC DISORDERS SECTION 6 - ALCOHOLISM AND DRUG DEPENDENCY
SECTION 1 - DIAGNOSIS OF NEUROLOGIC DISORDERS 355. NEUROBIOLOGY OF DISEASE 356. APPROACH TO THE PATIENT WITH NEUROLOGIC DISEASE 357. ELECTROPHYSIOLOGIC STUDIES OF THE CENTRAL AND PERIPHERAL NERVOUS SYSTEMS 358. NEUROIMAGING IN NEUROLOGIC DISORDERS 359. MOLECULAR DIAGNOSIS OF NEUROLOGIC DISORDERS
SECTION 2 - DISEASES OF THE CENTRAL NERVOUS SYSTEM 360. SEIZURES AND EPILEPSY 361. CEREBROVASCULAR DISEASES 362. ALZHEIMER'S DISEASE AND OTHER PRIMARY DEMENTIAS 363. PARKINSON'S DISEASE AND OTHER EXTRAPYRAMIDAL DISORDERS 364. ATAXIC DISORDERS 365. AMYOTROPHIC LATERAL SCLEROSIS AND OTHER MOTOR NEURON DISEASES 366. DISORDERS OF THE AUTONOMIC NERVOUS SYSTEM, 367. COMMON DISORDERS OF THE CRANIAL NERVES 368. DISEASES OF THE SPINAL CORD 369. TRAUMATIC INJURIES OF THE HEAD AND SPINE 370. PRIMARY AND METASTATIC TUMORS OF THE NERVOUS SYSTEM 371. MULTIPLE SCLEROSIS AND OTHER DEMYELINATING DISEASES 372. BACTERIAL MENINGITIS AND OTHER SUPPURATIVE INFECTIONS 373. VIRAL MENINGITIS AND ENCEPHALITIS 374. CHRONIC AND RECURRENT MENINGITIS 375. PRION DISEASES 376. CRITICAL CARE NEUROLOGY SECTION 3 - DISORDERS OF NERVE AND MUSCLE 377. APPROACH TO THE PATIENT WITH PERIPHERAL NEUROPATHY 378. GUILLAIN-BARRE SYNDROME AND OTHER IMMUNE-MEDIATED NEUROPATHIES 379. CHARCOT-MARIE-TOOTH DISEASE AND OTHER INHERITED NEUROPATHIES 380. MYASTHENIA GRAVIS AND OTHER DISEASES OF THE NEUROMUSCULAR JUNCTION 381. APPROACH TO THE PATIENT WITH MUSCLE DISEASE 382. POLYMYOSITIS, DERMATOMYOSITIS, AND INCLUSION BODY MYOSITIS 383. MUSCULAR DYSTROPHIES AND OTHER MUSCLE DISEASES SECTION 4 - CHRONIC FATIGUE SYNDROME 384. CHRONIC FATIGUE SYNDROME SECTION 5 - PSYCHIATRIC DISORDERS 385. MENTAL DISORDERS SECTION 6 - ALCOHOLISM AND DRUG DEPENDENCY 386. BIOLOGY OF ADDICTION 387. ALCOHOL AND ALCOHOLISM 388. OPIOID DRUG ABUSE AND DEPENDENCE 389. COCAINE AND OTHER COMMONLY ABUSED DRUGS 390. NICOTINE ADDICTION
PART FIFTEEN - ENVIROMENTAL AND OCCUPATIONAL HAZARDS SECTION 1 - SPECIFIC ENVIRONMENTAL AND OCCUPATIONAL HAZARDS SECTION 2 - ILLNESSES DUE TO POISONS, DRUG OVERDOSAGE, AND ENVENOMATION
SECTION 1 - SPECIFIC ENVIRONMENTAL AND OCCUPATIONAL HAZARDS 391. SPECIFIC ENVIRONMENTAL AND OCCUPATIONAL HAZARDS 392. DROWNING AND NEAR-DROWNING 393. ELECTRICAL INJURIES 394. RADIATION INJURY 395. HEAVY METAL POISONING SECTION 2 - ILLNESSES DUE TO POISONS, DRUG OVERDOSAGE, AND ENVENOMATION 396. POISONING AND DRUG OVERDOSAGE 397. DISORDERS CAUSED BY REPTILE BITES AND MARINE ANIMAL EXPOSURES 398. ECTOPARASITE INFESTATIONS, ARTHROPOD BITES AND STINGS
PART ONE -INTRODUCTION TO CLINICAL MEDICINE 1. THE PRACTICE OF MEDICINE - The Editors WHAT IS EXPECTED OF THE PHYSICIAN The practice of medicine combines both science and art. The role of science in medicine is clear. Science-based technology and deductive reasoning form the foundation for the solution to many clinical problems; the spectacular advances in genetics, biochemistry, and imaging techniques allow access to the innermost parts of the cell and the most remote recesses of the body. Highly advanced therapeutic maneuvers are increasingly a major part of medical practice. Yet skill in the most sophisticated application of laboratory technology and in the use of the latest therapeutic modality alone does not make a good physician. One must be able to identify the crucial elements in a complex history and physical examination and extract the key laboratory results from the crowded computer printouts of laboratory data in order to determine in a difficult case whether to "treat" or to "watch." Deciding when a clinical clue is worth pursuing, or when it should be dismissed as a "red herring," and estimating in any given patient whether a proposed treatment entails a greater risk than the disease are essential to the decision-making process that the skilled clinician must exercise many times each day. This combination of medical knowledge, intuition, and judgment defines the art of medicine, which is as necessary to the practice of medicine as is a sound scientific base. The editors of the first edition of this book articulated what is expected of the physician in words that, although they reflect the gender bias of that era, still ring true as a universal principle: No greater opportunity, responsibility, or obligation can fall to the lot of a human being than to become a physician. In the care of the suffering he needs technical skill, scientific knowledge, and human understanding. He who uses these with courage, with humility, and with wisdom will provide a unique service for his fellow man, and will build an enduring edifice of character within himself. The physician should ask of his destiny no more than this; he should be content with no less. Tact, sympathy and understanding are expected of the physician, for the patient is no mere collection of symptoms, signs, disordered functions, damaged organs, and disturbed emotions. He is human, fearful, and hopeful, seeking relief, help and reassurance. THE PATIENT-PHYSICIAN RELATIONSHIP It may seem trite to emphasize that physicians need to approach patients not as "cases" or "diseases" but as individuals whose problems all too often transcend their physical complaints. Most patients are anxious and frightened. Physicians should instill confidence and reassurance, overtly and in their demeanor, but without an air of arrogance. A professional attitude, coupled with warmth and openness, can do much to alleviate the patients' anxiety and to encourage them to share parts of their history that may be embarrassing. Some patients "use" illness to gain attention or to serve as a
crutch to extricate themselves from a stressful situation; some even feign physical illness; others may be openly hostile. Whatever the patient's attitude, the physician needs to consider the setting in which an illness occurs -- in terms not only of the patients themselves but also of their families and social and cultural backgrounds. The ideal patient-physician relationship is based on thorough knowledge of the patient, on mutual trust, and on the ability to communicate with one another. The direct, one-to-one patient-physician relationship, which has traditionally characterized the practice of medicine, is increasingly in jeopardy because of the increasing complexity of medicine and change in health care delivery systems. Often the management of the individual patient is a team effort involving a number of several different physicians and professional personnel. The patient can benefit greatly from such collaboration, but it is the duty of the patient's principal physician to guide them through an illness. To carry out this difficult task, this physician must be familiar with the techniques, skills, and objectives of specialist physicians and of colleagues in the fields allied to medicine. In giving the patient an opportunity to benefit from scientific advances, the primary physician must, in the last analysis, retain responsibility for the major decisions concerning diagnosis and treatment. Patients are increasingly cared for by groups of physicians in clinics, hospitals, integrated health care delivery systems, and health maintenance organizations (HMOs). Whatever the potential advantages of such organized medical groups, there are also drawbacks, chiefly the loss of the clear identification of the physician who is primarily and continuously responsible for the patient. Even under these circumstances, it is essential for each patient to have a physician who has an overview of the problems and who is familiar with the patient's reaction to the illness, to the drugs given, and to the challenges that the patient faces. The practice of medicine in a "managed care" setting puts additional stress on the classic paradigm of the patient-physician relationship. Many physicians must deal with a patient within a restricted time frame, with limited access to specialists, and under organizational guidelines that may compromise their ability to exercise their individual clinical judgment. As difficult as these restrictions may be, it is the ultimate responsibility of the physician to determine what is best for the patient. This responsibility cannot be relinquished in the name of compliance with organizational guidelines. The physician must also bear in mind that the modern hospital constitutes an intimidating environment for most patients. Lying in a bed surrounded by air jets, buttons, and lights; invaded by tubes and wires; beset by the numerous members of the health care team -- nurses, nurses' aides, physicians' assistants, social workers, technologists, physical therapists, medical students, house officers, attending and consulting physicians, and many others; sharing rooms with other patients who have their own problems, visitors, and physicians; transported to special laboratories and imaging facilities replete with blinking lights, strange sounds, and unfamiliar personnel -it is little wonder that patients may lose their sense of reality. In fact, the physician is often the only tenuous link between the patient and the real world, and a strong personal relationship with the physician helps to sustain the patient in such a stressful situation.
Many trends in contemporary society tend to make medical care impersonal. Some of these have been mentioned already and include (1) vigorous efforts to reduce the escalating costs of health care; (2) the growing number of managed care programs, which are intended to reduce costs but in which the patient may have little choice in selecting a physician; (3) increasing reliance on technologic advances and computerization for many aspects of diagnosis and treatment; (4) increased geographic mobility of both patients and physicians; (5) the need for numerous physicians to be involved in the care of most patients who are seriously ill; and (6) an increasing tendency on the part of patients to express their frustrations with the health care system by legal means (i.e., by malpractice litigation). Given these changes in the medical care system, it is a major challenge for physicians to maintain the humane aspects of medical care. The American Board of Internal Medicine has defined humanistic qualities as encompassing integrity, respect, and compassion. Availability, the expression of sincere concern, the willingness to take the time to explain all aspects of the illness, and a nonjudgmental attitude when dealing with patients whose cultures, lifestyles, attitudes, and values differ from those of the physician are just a few of the characteristics of the humane physician. Every physician will, at times, be challenged by patients who evoke strongly negative (or strongly positive) emotional responses. Physicians should be alert to their own reactions to such patients and situations and should consciously monitor and control their behavior so that the patients' best interests remain the principal motivation for their actions at all times. An important aspect of patient care involves an appreciation of the "quality of life," a subjective assessment of what each patient values most. Such an assessment requires detailed, sometimes intimate knowledge of the patient, which can usually be obtained only through deliberate, unhurried, and often repeated conversations. It is in these situations that the time constraints of a managed care setting may prove problematic. The famous statement of Dr. Francis Peabody is even more relevant today than when delivered more than three-quarters of a century ago: The significance of the intimate personal relationship between physician and patient cannot be too strongly emphasized, for in an extraordinarily large number of cases both the diagnosis and treatment are directly dependent on it. One of the essential qualities of the clinician is interest in humanity, for the secret of the care of the patient is in caring for the patient. CLINICAL SKILLS History Taking The written history of an illness should embody all the facts of medical significance in the life of the patient. Recent events should be given the most attention. The patient should, at some point, have the opportunity to tell his or her own story of the illness without frequent interruption and, when appropriate, receive expressions of interest, encouragement, and empathy from the physician. The physician must be alert to the possibility that any event related by the patient, however trivial or apparently remote, may be the key to the solution of the medical problem. An informative history is more than an orderly listing of symptoms; something is always gained by listening to patients and noting the way in which they describe their
symptoms. Inflections of voice, facial expression, gestures, and attitude may reveal important clues to the meaning of the symptoms to the patient. Taking history often involves much data gathering. Patients vary in their medical sophistication and ability to recall facts. Medical history should therefore be corroborated whenever possible. The family and social history can also provide important insights into the types of diseases that should be considered. In listening to the history, the physician discovers not only something about the disease but also something about the patient. The process of history taking provides an opportunity to observe the patient's behavior and to watch for features to be pursued more thoroughly during the physical examination. The very act of eliciting the history provides the physician with the opportunity to establish or enhance the unique bond that is the basis for the ideal patient-physician relationship. It is helpful to develop an appreciation of the patient's perception of the illness, the patient's expectations of the physician and the medical care system, and the financial and social implications of the illness to the patient. The confidentiality of the patient-physician relationship should be emphasized, and the patient should be given the opportunity to identify any aspects of the history that should not be disclosed. Physical Examination Physical signs are objective indications of disease whose significance is enhanced when they confirm a functional or structural change already suggested by the patient's history. At times, however, the physical signs may be the only evidence of disease. The physical examination should be performed methodically and thoroughly, with consideration for the patient's comfort and modesty. Although attention is often directed by the history to the diseased organ or part of the body, the examination of a new patient must extend from head to toe in an objective search for abnormalities. Unless the physical examination is systematic, important segments may be omitted. The results of the examination, like the details of the history, should be recorded at the time they are elicited, not hours later when they are subject to the distortions of memory. Skill in physical diagnosis is acquired with experience, but it is not merely technique that determines success in eliciting signs. The detection of a few scattered petechiae, a faint diastolic murmur, or a small mass in the abdomen is not a question of keener eyes and ears or more sensitive fingers but of a mind alert to these findings. Since physical findings are subject to changes, the physical examination should be repeated as frequently as the clinical situation warrants. Laboratory Tests The availability of a wide array of laboratory tests has increased our reliance on these studies for the solution of clinical problems. The accumulation of laboratory data does not relieve the physician from the responsibility of careful observation, examination, and study of the patient. It is also essential to bear in mind the limitations of such tests. By virtue of their impersonal quality, complexity, and apparent precision, they often gain an aura of authority regardless of the fallibility of the tests themselves, the instruments used in the tests, and the individuals performing or interpreting them. Physicians must weigh the expense involved in the laboratory procedures they order relative to the value of the information they are likely to provide. Single laboratory tests are rarely ordered. Rather, they are generally obtained as "batteries" of multiple tests, which are often useful. For example, abnormalities of
hepatic function may provide the clue to such nonspecific symptoms as generalized weakness and increased fatigability, suggesting the diagnosis of chronic liver disease. Sometimes a single abnormality, such as an elevated serum calcium level, points to particular diseases, such as hyperparathyroidism or underlying malignancy. The thoughtful use of screening tests should not be confused with indiscriminate laboratory testing. The use of screening tests is based on the fact that a group of laboratory determinations can be carried out conveniently on a single specimen of blood at relatively low cost. Screening tests are most useful when they are directed towards common diseases or disorders in which the result directs other useful tests or interventions that would otherwise be costly to perform. Biochemical measurements, together with simple laboratory examinations such as blood count, urinalysis, and sedimentation rate, often provide the major clue to the presence of a pathologic process. At the same time, the physician must learn to evaluate occasional abnormalities among the screening tests that may not necessarily connote significant disease. An in-depth workup following a report of an isolated laboratory abnormality in a person who is otherwise well is almost invariably wasteful and unproductive. Among the more than 40 tests that are routinely performed on patients, one or two are often slightly abnormal. If there is no suspicion of an underlying illness, these tests are ordinarily repeated to ensure that the abnormality does not represent a laboratory error. If an abnormality is confirmed, it is important to consider its potential significance in the context of the patient's condition and other test results. Imaging Techniques The availability of ultrasonography, a variety of scans that employ isotopes to visualize organs heretofore inaccessible, computed tomography, and magnetic resonance imaging has opened new diagnostic vistas and has benefited patients because these new techniques have largely supplanted more invasive ones. While the enthusiasm for noninvasive technology is understandable, the expense entailed in performing these tests is often substantial and should be considered when assessing the potential benefits of the information provided. PRINCIPLES OF PATIENT CARE Medical Decision-Making Both during and in particular after the physician has taken the history, performed the physical examination, and reviewed the laboratory and imaging data, the challenging process of the differential diagnosis and medical decision-making begins. Formulating a differential diagnosis requires not only a broad knowledge base but also the ability to assess the relative probabilities of various diseases and to understand the significance of missing diagnoses that may be less likely. Arriving at a diagnosis requires the application of the scientific method. Hypotheses are formed, data are collected, and objective conclusions are reached concerning whether to accept or reject a particular diagnosis. Analysis of the differential diagnosis is an iterative process. As new information or test results are acquired, the group of disease processes being considered can be contracted or expanded appropriately. Medical decision-making occurs throughout the diagnostic and treatment process. It involves the ordering of additional tests, requests for consults, and decisions regarding prognosis and treatment. This process requires an in-depth understanding of the natural history and pathophysiology of disease, explaining why these features are strongly emphasized in this textbook. As described below, medical decision-making
should be evidence-based, thereby ensuring that patients derive the full benefit of the scientific knowledge available to physicians. Evidence-Based Medicine Sackett has defined evidence-based medicine as "the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients." Rigorously obtained evidence is contrasted with anecdotal experience, which is often biased. Even the most experienced physicians can be influenced by recent experiences with selected patients, unless they are attuned to the importance of using larger, more objective studies for making decisions. The prospectively designed, double-blind, randomized clinical trial represents the "gold standard" for providing evidence regarding therapeutic decisions, but it is not the only source. Valuable evidence about the natural history of disease and prognosis can come from prospective cohort studies and analytic surveys. Persuasive evidence on the accuracy of diagnostic tests can be derived from cross-sectional studies of patients in whom a specific disorder is suspected. Evidence is strengthened immensely when it has been confirmed by multiple investigations, which can be compared with one another and presented in a meta-analysis or systemic overview. In failing to apply the best and most current evidence, the physician places the patient at unnecessary risk. However, a knowledge of or rapid access to the best available evidence is not sufficient for optimal care. The physician must know whether the evidence is relevant to the patient in question and, when it is, the consequences of applying it in any particular situation. The skills and judgment required to apply sound evidence represent an increasing challenge. Indeed, one might redefine a "good doctor" as one who uses the ever-growing body of rigorously obtained evidence (the science of medicine) in a sensible, compassionate manner (the art of medicine). While an understanding of biologic and physiologic mechanisms forms the basis of contemporary medicine, when a therapeutic modality is selected, the highest priority must often be placed on improving clinical outcome rather than interrupting what is believed to be the underlying process. For example, for decades patients who had suffered myocardial infarction were treated intuitively with drugs that suppress frequent ventricular extrasystoles, since these were believed to be harbingers of ventricular fibrillation and sudden death. Clinical trials, however, have provided firm evidence that the antiarrhythmic agents actually increase the risk of death in such patients. This finding suggests that the extrasystoles are markers of high risk rather than the cause of fatal events. Practice Guidelines Physicians are faced with a large, increasing, and often bewildering body of evidence pointing to potentially useful diagnostic techniques and therapeutic choices. The intelligent and cost-effective practice of medicine consists of making selections most appropriate to a particular patient and clinical situation. Professional organizations and government agencies are developing formal clinical practice guidelines in an effort to aid physicians and other caregivers in this endeavor. When guidelines are current and properly applied, they can provide a useful framework for managing patients with particular diagnoses or symptoms. They can protect patients -- particularly those with inadequate health care benefits -- from receiving substandard care. Guidelines can also protect conscientious caregivers from inappropriate charges of malpractice and society from the excessive costs associated with the overuse of
medical resources. On the other hand, clinical guidelines tend to oversimplify the complexities of medicine. Different groups with differing perspectives may develop divergent recommendations regarding issues as basic as the need for periodic sigmoidoscopy in middle-aged persons. Furthermore, guidelines do not -- and cannot be expected to -- take into account the uniqueness of each individual and of his or her illness. The challenge for the physician is to integrate into clinical practice the useful recommendations offered by the experts who prepare clinical practice guidelines without accepting them blindly or being inappropriately constrained by them. Assessing the Outcome of Treatment Clinicians generally use objective and readily measurable parameters to judge the outcome of a therapeutic intervention. For example, findings on physical or laboratory examination -- such as the level of blood pressure, the patency of a coronary artery on an angiogram, or the size of a mass on a radiologic examination -- can provide information of critical importance. However, patients usually seek medical attention for subjective reasons; they wish to obtain relief from pain, to preserve or regain function, and to enjoy life. The components of a patient's health status or quality of life can include bodily comfort, capacity for physical activity, personal and professional function, sexual function, cognitive function, and overall perception of health. Each of these important areas can be assessed by means of structured interviews or specially designed questionnaires. Such assessments also provide useful parameters by which the physician can judge the patient's subjective view of his or her disability and the response to treatment, particularly in chronic illness. The practice of medicine requires consideration and integration of both objective and subjective outcomes. Care of the Elderly Over the next several decades, the practice of medicine will be greatly influenced by the health care needs of the growing elderly population. In the United States the population over age 65 will almost triple over the next 30 years. It is essential that we understand and appreciate the physiologic processes associated with aging; the different responses of the elderly to common diseases; and disorders that occur commonly with aging, such as depression, dementia, frailty, urinary incontinence, and fractures. The elderly have more adverse reactions to drugs, in large part due to altered pharmacokinetics and pharmacodynamics. Commonly used medications such as digoxin and aminoglycosides have prolonged half-lives in the elderly, and tissues such as the central nervous system are more sensitive to certain drugs, such as the benzodiazepines and narcotics. The large number of drugs used by the elderly increases the risk of unwanted interactions, especially when care is provided by several physicians in an uncoordinated manner. Diseases in Women versus Men In the past, many epidemiologic studies and clinical trials focused on men. It is now appreciated that there are significant gender differences in diseases that afflict both men and women. Mortality rates are substantially higher in women than in men under the age of 50 suffering acute myocardial infarction. Hypertension is more prevalent in African-American women than in their male counterparts (and in African-American than in white males); osteoporosis is more common in women, reflecting the menopausal loss of estrogen; diseases involving the immune system, such as lupus erythematosus, multiple sclerosis, and primary biliary cirrhosis, occur more frequently in women; and the average life expectancy of women is greater than that of men. Recently, considerable attention has been paid to women's
health issues, a subject that regrettably did not receive sufficient attention in the past. Ongoing study should enhance our understanding of the mechanisms of gender differences in the course and outcome of certain diseases. Iatrogenic Disorders In an iatrogenic disorder, the deleterious effects of a therapeutic or diagnostic maneuver cause pathology independent of the condition for which the intervention was performed. Adverse drug reactions occur in at least 5% of hospitalized patients, and the incidence increases with use of a large number of drugs. No matter what the clinical situation, it is the responsibility of the physician to use powerful therapeutic measures wisely, with due regard for their beneficial action, potential dangers, and cost. Every medical procedure, whether diagnostic or therapeutic, has the potential for harm, but it would be impossible to provide the benefits of modern scientific medicine if reasonable steps in diagnosis and therapy were withheld because of possible risks. Reasonable implies that the physician has weighed the pros and cons of a procedure and has concluded, on the basis of objective evidence whenever possible, that it is necessary for establishing a diagnosis, for the relief of discomfort, or for the cure of disease. However, the harm that a physician can do is not limited to the imprudent use of medication or procedures. Equally important are ill-considered or unjustified remarks. Many a patient has developed a cardiac neurosis because the physician ventured a grave prognosis on the basis of a misinterpreted finding of a heart murmur. Not only the diagnostic procedure or the treatment but the physician's words and behavior are capable of causing injury. Informed Consent Patients often require diagnostic and therapeutic procedures that are painful and that pose some risk. For many such procedures, patients are required to sign a consent form. The patient must understand clearly the risks entailed in these procedures; this is the definition of informed consent. It is incumbent on the physician to explain the procedures in a clear and understandable manner and to ascertain that the patient comprehends both the nature of the procedure and the attendant risks. The dread of the unknown that is inherent in hospitalization can be mitigated by such explanations. Incurability and Death No problem is more distressing than that presented by the patient with an incurable disease, particularly when premature death is inevitable. What should the patient and family be told, what measures should be taken to maintain life, what can be done to maintain the quality of life, and how is death to be defined? The concept of incurable illness and terminal care often evokes examples of cancer. However, patients with many other end-stage diseases including chronic obstructive pulmonary disease, congestive heart failure, renal or hepatic failure, and overwhelming infection face similar issues. The same principles of terminal care should be applied in each of these cases. Doing seemingly small things, focused on the needs of the patient, can do much to restore comfort or dignity during a person's final weeks or days. In the same way that pain should be attentively managed with analgesia, every effort should be made to alleviate shortness of breath and to provide good skin care. Although some would argue otherwise, there is no ironclad rule that the patient must immediately be told "everything," even if the patient is an adult with substantial family responsibilities. How much is told should depend on the individual's ability to deal with
the possibility of imminent death; often this capacity grows with time, and whenever possible, gradual rather than abrupt disclosure is the best strategy. A wise and insightful physician is often guided by an understanding of what a patient wants to know and when he or she wants to know it. The patient's religious beliefs may also be taken into consideration. The patient must be given an opportunity to talk with the physician and ask questions. Patients may find it easier to share their feelings about death with their physician, who is likely to be more objective and less emotional, than with family members. As William Osler wrote: One thing is certain; it is not for you to don the black cap and, assuming the judicial function, take hope away from any patient...hope that comes to us all. Even when the patient directly inquires, "Am I dying?" the physician must attempt to determine whether this is a request for information or a demand for reassurance. Only open communication between the patient and the physician can resolve this question and guide the physician in what to say and how to say it. The physician should provide or arrange for emotional, physical, and spiritual support and must be compassionate, unhurried, and open. There is much to be gained by the laying on of hands. Pain should be adequately controlled, human dignity maintained, and isolation from the family avoided. These aspects of care tend to be overlooked in hospitals, where the intrusion of life-sustaining apparatus can so easily detract from attention to the whole person and encourage concentration instead on the life-threatening disease, against which the battle will ultimately be lost in any case. In the face of terminal illness, the goal of medicine must shift from cure to care, in the broadest sense of the term. In offering care to the dying patient, the physician must be prepared to provide information to family members and to deal with their guilt and grief. It is important for the doctor to assure the family that everything possible has been done. "Do Not Resuscitate" Orders and Cessation of Therapy When carried out in a timely and expert manner, cardiopulmonary resuscitation is often useful in the prevention of sudden, unexpected death. However, unless there are reasons to the contrary, this procedure should not be used merely to prolong the life of a patient with terminal, incurable disease. The decision whether or not to resuscitate or even to treat an incurably and terminally ill patient must be reviewed frequently and must take into consideration any unexpected changes in the patient's condition. In this context, the administration of fluids or food is considered therapy that may be withdrawn or withheld. These decisions must also take into account both the underlying medical condition, especially its reversibility, and the wishes of the patient, especially if these have been expressed in a living will or advance directive. If the patient's wishes cannot be ascertained directly, a close relative or another surrogate who can be relied on to transmit the patient's wishes and to be guided by the patient's best interests should be consulted. The patient's autonomy -- whether the choice is to continue or discontinue treatment or to be resuscitated or not in the event of a cardiopulmonary arrest -- must be paramount. The courts have ruled that competent patients may refuse therapy and that an incompetent patient's previously stated wishes regarding life support should therefore be respected. The issues involving death and dying are among the most difficult in medicine. In approaching them rationally and consistently, the physician must
combine both the science and the art of medicine. THE EXPANDING ROLE OF THE PHYSICIAN Genetics and Medicine The genomic era is likely to lead to a revolution in the practice of medicine. Obtaining the DNA sequence of the entire human genome may help to elucidate the genetic components of common chronic diseases -- hypertension, diabetes, atherosclerosis, many cancers, dementias, and behavioral and autoimmune disorders. This information should make it possible to determine individual susceptibility to these conditions early in life and to implement individualized prevention programs. Subclassification of many diseases on a genetic basis may allow the selection of appropriate therapy for each patient. As the response to drugs becomes more predictable, pharmacotherapy should become more rational. In short, the completion of the Human Genome Project is likely to lead to a substantial increase in physicians' ability to influence their patient's health and well-being. Patients will be best served if physicians play an active role in applying this powerful, sensitive new information rather than being passive bystanders who are intimidated by the new technology. This is a rapidly evolving field, and physicians and other health care professionals must remain updated to apply this new knowledge. Genetic testing requires wise counsel based on an understanding of the value and limitations of the tests as well as the implications of their results for specific individuals. Medicine on the Internet The explosion in use of the Internet through personal computers is having an important impact on many practicing physicians. The Internet makes a wide range of information available to physicians almost instantaneously at any time of the day or night and from anywhere in the world. This medium holds enormous potential for delivering up-to-date information, practice guidelines, state-of-the-art conferences, journal contents, textbooks (including this text), and direct communications with other physicians and specialists, thereby expanding the depth and breadth of information available to the physician about the diagnosis and care of patients. Most medical journals are now accessible on-line, providing rapid and comprehensive sources of information. Patients, too, are turning to the Internet in increasing numbers to derive information about their illnesses and therapies and to join Internet-based support groups. Physicians are increasingly challenged by dealing with patients who are becoming more sophisticated in their understanding of illness. At this time, there is one critically important caveat. Virtually anything can be published on the Internet, thus circumventing the peer-review process that is an essential feature of quality publications. Physicians or patients who search the Internet for medical information must be aware of this danger. Notwithstanding this limitation, appropriate use of the Internet is revolutionizing information access for physicians and is a positive force in the practice of medicine. Delivering Cost-Effective Medical Care As the cost of medical care has risen, it has become necessary to establish priorities in the expenditure of resources. In some instances, preventive measures offer the greatest return for the expenditure; outstanding examples include vaccination, improved sanitation, reduction in accidents and occupational hazards, and biochemical- and DNA-based screening of newborns. For example, the detection of phenylketonuria by newborn screening may result in a net
saving of many thousands of dollars. As resources become increasingly constrained, the physician must weigh the possible benefits of performing costly procedures that provide only a limited life expectancy against the pressing need for more primary care for those persons who do not have adequate access to medical services. For the individual patient, it is important to reduce costly hospital admissions as much as possible if total health care is to be provided at a cost that most can afford. This policy, of course, implies and depends on close cooperation among patients, their physicians, employers, payers, and government. It is equally important for physicians to know the cost of the diagnostic procedures they order and the drugs and other therapies they prescribe and to monitor both costs and effectiveness. The medical profession should provide leadership and guidance to the public in matters of cost control, and physicians must take this responsibility seriously without being or seeming to be self-serving. However, the economic aspects of health care delivery must not interfere with the welfare of patients. The patient must be able to rely on the individual physician as his or her principal advocate in matters of health care. Accountability Medicine is a satisfying but demanding profession. Physicians must understand the characteristics of the populations they serve, and they must appreciate their patients' social and cultural attitudes to health, disease, and death. As the public has become more educated and more sophisticated regarding health matters, their expectations of the health system in general and of their physicians in particular have risen. Physicians are expected to maintain mastery of their rapidly advancing fields (the science of medicine) while considering their patient's unique needs (the art of medicine). Thus, physicians are held accountable not only for the technical aspects of the care that they provide but also for their patient's satisfaction with the delivery and costs of care. In the United States, there are increasing demands for physicians to account for the way in which they practice medicine by meeting certain standards prescribed by federal and state governments. The hospitalization of patients whose health care costs are reimbursed by the government and other third parties is subjected to utilization review. Thus the physician must defend the cause for and duration of a patient's hospitalization if it falls outside certain "average" standards. Authorization for reimbursement is increasingly based on documentation of the nature and complexity of an illness, as reflected by recorded elements of the history and physical examination. The purpose of these regulations is both to improve standards of health care and to contain spiraling health care costs. This type of review is being extended to all phases of medical practice and is profoundly altering the practice of medicine. Physicians are also expected to give evidence of their continuing competence through mandatory continuing education, patient-record audits, recertification by examination, or relicensing. Continued Learning The conscientious physician must be a perpetual student because the body of medical knowledge is constantly expanding and being refined. The profession of medicine should be inherently linked to a career-long thirst for new knowledge that can be used for the good of the patient. It is the responsibility of a physician to pursue continually the acquisition of new knowledge by reading, attending conferences and courses, and consulting colleagues and the Internet. This is often a difficult task for a busy practitioner; however, such a commitment to continued learning is an integral part of being a physician and must be given the highest priority.
Research and Teaching The title doctor is derived from the Latin docere, "to teach," and physicians should share information and medical knowledge with colleagues, with students of medicine and related professions, and with their patients. The practice of medicine is dependent on the sum total of medical knowledge, which in turn is based on an unending chain of scientific discovery, clinical observation, analysis, and interpretation. Advances in medicine depend on the acquisition of new information, i.e., on research, which often involves patients; improved medical care requires the transmission of this information. As part of broader societal responsibilities, the physician should encourage patients to participate in ethical and properly approved clinical investigations if they do not impose undue hazard, discomfort, or inconvenience. To quote Osler once more: To wrest from nature the secrets which have perplexed philosophers in all ages, to track to their sources the causes of disease, to correlate the vast stores of knowledge, that they may be quickly available for the prevention and cure of disease -- these are our ambitions. (Bibliography omitted in Palm version) Back to Table of Contents
2. ETHICAL ISSUES IN CLINICAL MEDICINE - Bernard Lo Physicians frequently confront ethical issues in clinical practice that are perplexing, time-consuming, and emotionally draining. Experience, common sense, and simply being a good person do not guarantee that physicians can identify or resolve ethical dilemmas. Knowledge about common ethical dilemmas is also essential. FUNDAMENTAL ETHICAL GUIDELINES Physicians should follow two fundamental but frequently conflicting ethical guidelines: respecting patient autonomy and acting in the patient's best interests. RESPECTING PATIENT AUTONOMY Competent, informed patients may refuse recommended interventions and choose among reasonable alternatives. Informed Consent Informed consent requires physicians to discuss with patients the nature of the proposed care, the alternatives, the risks and benefits of each, the likely consequences, and to obtain the patient's agreement to care. Informed consent involves more than obtaining signatures on consent forms. Physicians need to educate patients, answer questions, make recommendations, and help them deliberate. Patients can be overwhelmed with medical jargon, needlessly complicated explanations, or too much information at once. Nondisclosure of Information Physicians may consider withholding a serious diagnosis, misrepresenting it, or limiting discussions of prognosis or risks because they fear that a patient will develop severe anxiety or depression or refuse needed care. Patients should not be forced to receive information against their will. Most people, however, want to know their diagnosis and prognosis, even if they are terminally ill. Generally, physicians should provide relevant information, offer empathy and hope, and help patients cope with bad news. Emergency Care Informed consent is not required when patients cannot give consent and when delay of treatment would place their life or health in peril. People are presumed to want such emergency care, unless they have previously indicated otherwise. Futile Interventions Autonomy does not entitle patients to insist on whatever care they want. Physicians are not obligated to provide futile interventions that have no physiologic rationale or have already failed. For example, cardiopulmonary resuscitation would be futile in a patient with progressive hypotension despite maximal therapy. But physicians should be wary of using the term "futile" in looser senses to justify unilateral decisions to forego interventions when they believe that the probability of success is too low, no worthwhile goals can be achieved, the patient's quality of life is unacceptable, or the costs are too high. Such looser usages of the term are problematic because they may be inconsistent and mask value judgments. ACTING IN THE BEST INTERESTS OF PATIENTS
The guideline of beneficence requires physicians to act for the patient's benefit. Laypeople do not possess medical expertise and may be vulnerable because of their illness. They justifiably rely on physicians to provide sound advice and to promote their well-being. Physicians encourage such trust. Hence, physicians have a fiduciary duty to act in the best interests of their patients. The interests of the patient should prevail over physicians' self-interest or the interests of third parties, such as hospitals or insurers. These fiduciary obligations of physicians contrast sharply with business relationships, which are characterized by "let the buyer beware," not by trust and reliance. The guideline of "do no harm" forbids physicians from providing ineffective interventions or acting without due care. This precept, while often cited, provides only limited guidance, because many beneficial interventions also have serious risks. CONFLICTS BETWEEN BENEFICENCE AND AUTONOMY Patients' refusals of care may thwart their own goals or cause them serious harm. For example, a young man with asthma may refuse mechanical ventilation for reversible respiratory failure. Simply to accept such refusals, in the name of respecting autonomy, seems morally constricted. Physicians can elicit patients' expectations and concerns, correct misunderstandings, and try to persuade them to accept beneficial therapies. If disagreements persist after discussions, the patient's informed choices and view of his or her best interests should prevail. While refusing recommended care does not render a patient incompetent, it may lead the physician to probe further to ensure that the patient is able to make informed decisions. PATIENTS WHO LACK DECISION-MAKING CAPACITY Patients may not be able to make informed decisions because of unconsciousness, dementia, delirium, or other conditions. Physicians should ask two questions regarding such patients: Who is the appropriate surrogate? What would the patient want done? ASSESSING CAPACITY TO MAKE MEDICAL DECISIONS All adults are considered legally competent unless declared incompetent by a court. In practice, physicians usually determine that patients lack the capacity to make health care decisions and arrange for surrogates to make them, without involving the courts. By definition, competent patients can express a choice and appreciate the medical situation, the nature of the proposed care, the alternatives, and the risks, benefits, and consequences of each. Their choices should be consistent with their values and should not result from delusions or hallucinations. Psychiatrists may help in difficult cases because they are skilled at interviewing mentally impaired patients and can identify treatable depression or psychosis. When impairments are fluctuating or reversible, decisions should be postponed if possible until the patient recovers decision-making capacity. CHOICE OF SURROGATE If a patient lacks decision-making capacity, physicians routinely ask family members to serve as surrogates. Most patients want their family members to be surrogates, and
family members generally know the patient's preferences and have the patient's best interests at heart. Patients may designate a particular individual to serve as proxy; such choices should be respected. Some states have established a prioritized list of which relative may serve as surrogate if the patient has not designated a proxy. STANDARDS FOR SURROGATE DECISION MAKING Advance Directives These are statements by competent patients to direct care if they lose decision-making capacity. They may indicate (1) what interventions they would refuse or accept or (2) who should serve as surrogate. Following the patient's advance directives, surrogate respects patients' autonomy. Oral conversations are the most frequent form of advance directives. While such conversations are customarily followed in clinical practice, casual or vague comments may not be trustworthy. Living wills direct physicians to forego or provide life-sustaining interventions if the patient develops a terminal condition or persistent vegetative state. Generally patients may refuse only interventions that "merely prolong the process of dying." A health care proxy is someone appointed by the patient to make health care decisions if he or she loses decision-making capacity. It is more flexible and comprehensive than the living will, applying whenever the patient is unable to make decisions. Physicians can encourage patients to provide advance directives, to indicate both what they would want and who should be surrogate, and to discuss their preferences with surrogates. In discussions with patients, physicians can ensure that advance directives are informed, up-to-date, and address likely clinical scenarios. Such discussions are best carried out in the ambulatory setting. The federal Patient Self-Determination Act requires hospitals and health maintenance organizations to inform patients of their right to make health care decisions and to provide advance directives. Substituted Judgment In the absence of clear advance directives, surrogates and physicians should try to decide as the patient would under the circumstances, using all information that they know about the patient. While such substituted judgments try to respect the patient's values, they may be speculative or inaccurate. A surrogate may be mistaken about the patient's preferences, particularly when they have not been discussed explicitly. Best Interests When the patient's preferences are unclear or unknown, decisions should be based on the patient's best interests. Patients generally take into account the quality of life as well as the duration of life when making decisions for themselves. It is understandable that surrogates would also consider quality of life of patients who lack decision-making capacity. Judgments about quality of life are appropriate if they reflect the patient's own values. Bias or discrimination may occur, however, if others project their values onto the patient or weigh the perceived social worth of the patient. Most patients with chronic illness rate their quality of life higher than their family members and physicians do.
Legal Issues Physicians need to know pertinent state laws regarding patients who lack decision-making capacity. A few state courts allow doctors to forego life-sustaining interventions only if patients have provided written advance directives or very specific oral ones. Disagreements Disagreements may occur among potential surrogates or between the physician and surrogate. Physicians can remind everyone to base decisions on what the patient would want, not what they would want for themselves. Consultation with the hospital ethics committee or with another physician often helps resolve disputes. Such consultation is also helpful when patients have no surrogate and no advance directives. The courts should be used only as a last resort when disagreements cannot be resolved in the clinical setting. DECISIONS ABOUT LIFE-SUSTAINING INTERVENTIONS Although medical technology can save lives, it can also prolong the process of dying. Competent, informed patients may refuse life-sustaining interventions. Such interventions may also be withheld from patients who lack decision-making capacity on the basis of advance directives or decisions by appropriate surrogates. Courts have ruled that foregoing life-sustaining interventions is neither suicide nor murder. MISLEADING DISTINCTIONS People commonly draw distinctions that are intuitively plausible but prove untenable on closer analysis. Extraordinary and Ordinary Care Some physicians are willing to forego "extraordinary" or "heroic" interventions, such as surgery, mechanical ventilation, or renal dialysis, but insist on providing "ordinary" ones, such as antibiotics, intravenous fluids, or feeding tubes. However, this distinction is not logical because all medical interventions have both risks and benefits. Any intervention may be withheld, if the burdens for the individual patient outweigh the benefits. Withdrawing and Withholding Interventions Many health care providers find it more difficult to discontinue interventions than to withhold them in the first place. Although such emotions need to be acknowledged, there is no logical distinction between the two acts. Justifications for withholding interventions, such as refusal by patients or surrogates, are also justifications for withdrawing them. In addition, an intervention may prove unsuccessful or new information about the patient's preferences or condition may become available after the intervention is started. If interventions could not be discontinued, patients and surrogates might not even attempt treatments that might prove beneficial. DO NOT RESUSCITATE (DNR) ORDERS When a patient suffers a cardiopulmonary arrest, cardiopulmonary resuscitation (CPR) is initiated unless a DNR order has been made. Although CPR can restore people to vigorous health, it can also disrupt a peaceful death. After CPR is attempted on a general hospital service, only 14% of patients survive to discharge, and even fewer in
certain subgroups. DNR orders are appropriate if the patient or surrogate requests them or if CPR would be futile. To prevent misunderstandings, physicians should write DNR orders and the reasons for them in the medical record. "Slow" or "show" codes that merely appear to provide CPR are deceptive and therefore unacceptable. Although a DNR order signifies only that CPR will be withheld, the reasons that justify DNR orders may lead to a reconsideration of other plans for care. ASSISTED SUICIDE AND ACTIVE EUTHANASIA Proponents of these controversial acts believe that competent, terminally ill patients should have control over the end of life and that physicians should relieve refractory suffering. Opponents assert that such actions violate the sanctity of life, that suffering can generally be relieved, that abuses are inevitable, and that such actions are outside the physician's proper role. These actions are illegal throughout the United States, except that physician-assisted suicide is legal in Oregon under certain circumstances. Whatever their personal views, physicians should respond to patients' inquiries with compassion and concern. Physicians should elicit and address any underlying problems, such as physical symptoms, loss of control, or depression. Often, additional efforts to relieve distress are successful, and after this is done patients generally withdraw their requests for these acts. CARE OF DYING PATIENTS Patients often suffer unrelieved pain and other symptoms during their final days of life. Physicians may hesitate to order high doses of narcotics and sedatives, fearing they will hasten death. Relieving pain in terminal illness and alleviating dyspnea when patients forego mechanical ventilation enhances patient comfort and dignity. If lower doses of narcotics and sedatives have failed to relieve suffering, increasing the dose to levels that may suppress respiratory drive is ethically appropriate because the physician's intention is to relieve suffering, not hasten death. Physicians can also relieve suffering by spending time with dying patients, listening to them, and attending to their psychological distress. CONFLICTS OF INTEREST Acting in the patient's best interests may conflict with the physician's self-interest or the interests of third parties such as insurers or hospitals. The ethical ideal is to keep the patient's interests paramount. Even the appearance of a conflict of interest may undermine trust in the profession. FINANCIAL INCENTIVES In managed care systems, physicians may serve as gatekeepers or bear financial risk for expenditures. Although such incentives are intended to reduce inefficiency and waste, there is concern that physicians may withhold beneficial care in order to control costs. In contrast, physicians have incentives to provide more care than indicated when they receive fee-for-service reimbursement or when they refer patients to medical facilities in which they have invested. Regardless of financial incentives, physicians should recommend available care that is in the patient's best interests -- no more and
no less. DENIALS OF COVERAGE Utilization review programs designed to reduce unnecessary services may also deny coverage for care that the physician believes will benefit the patient. Physicians should inform patients when a plan is not covering standard care and act as patient advocates by appealing such denials of coverage. Patients may ask physicians to misrepresent their condition to help them obtain insurance coverage or disability. While physicians understandably want to help patients, such misrepresentation undermines physicians' credibility and violates their integrity. GIFTS FROM PHARMACEUTICAL COMPANIES Physicians may be offered gifts ranging from pens and notepads to lavish entertainment. Critics worry that any gift from drug companies may impair objectivity, increase the cost of health care, and give the appearance of conflict of interest. A helpful rule of thumb is to consider whether patients would approve if they knew physicians had accepted such gifts. OCCUPATIONAL RISKS Some health care workers, fearing fatal occupational infections, refuse to care for persons with HIV infection or multidrug-resistant tuberculosis. Such fears about personal safety need to be acknowledged, and institutions should reduce occupational risk by providing proper training, equipment, and supervision. Physicians should provide appropriate care within their clinical expertise, despite personal risk. MISTAKES Mistakes are inevitable in clinical medicine. They may cause serious harm to patients or result in substantial changes in management. Physicians and students may fear that disclosing such mistakes could damage their careers. Without disclosure, however, patients cannot understand their clinical situation or make informed choices about subsequent care. Similarly, unless attending physicians are informed of trainees' mistakes, they cannot provide optimal care and help trainees learn from mistakes. LEARNING CLINICAL SKILLS Learning clinical medicine, particularly learning to perform invasive procedures, may present inconvenience or risk to patients. To ensure patient cooperation, students may be introduced as physicians, or patients may not be told that trainees will be performing procedures. Such misrepresentation undermines trust, may lead to more elaborate deception, and makes it difficult for patients to make informed choices about their care. Patients should be told who is providing care, what benefits and burdens can be attributed to trainees, and how trainees are supervised. Most patients, when informed, allow trainees to play an active role in their care. IMPAIRED PHYSICIANS
Physicians may hesitate to intervene when colleagues impaired by alcohol abuse, drug abuse, or psychiatric or medical illness place patients at risk. However, society relies on physicians to regulate themselves. If colleagues of an impaired physician do not take steps to protect patients, no one else may be in a position to do so. CONFLICTS FOR TRAINEES Medical students and residents may fear that they will receive poor grades or evaluations if they act on the patient's behalf by disclosing mistakes, avoiding misrepresentation of their role, and reporting impaired colleagues. Discussing such dilemmas with more senior physicians can help trainees check their interpretation of the situation and obtain advice and assistance. ADDITIONAL ETHICAL ISSUES MAINTAINING CONFIDENTIALITY Maintaining the confidentiality of medical information respects patients' autonomy and privacy, encourages them to seek treatment and to discuss their problems candidly, and prevents discrimination. Physicians need to guard against inadvertent breaches of confidentiality, as when talking about patients in elevators. Maintaining confidentiality is not an absolute rule. The law may require physicians to override confidentiality in order to protect third parties, for example, reporting to government officials persons with specified infectious conditions, such as tuberculosis and syphilis; persons with gunshot wounds; and victims of elder abuse and domestic violence. Computerized medical records raise additional concerns because breaches of confidentiality may affect many patients. ALLOCATING RESOURCES JUSTLY Allocation of limited health care resources is problematic. Ideally, allocation decisions should be made as public policy, with physician input. At the bedside, physicians generally should act as patient advocates within constraints set by society, reasonable insurance coverage, and sound practice. Ad hoc rationing by the individual physician at the bedside may be inconsistent, discriminatory, and ineffective. In some cases, however, two patients may compete for the same limited resources, such as physician time or a bed in intensive care. When this occurs, physicians should ration their time and resources according to patients' medical needs and the probability of benefit. ASSISTANCE WITH ETHICAL ISSUES Discussing perplexing ethical issues with other members of the health care team, colleagues, or the hospital ethics committee often clarifies issues and suggests ways to improve communication and to deal with strong emotions. When struggling with difficult ethical issues, physicians may need to reevaluate their basic convictions, tolerate uncertainty, and maintain their integrity while respecting the opinions of others. (Bibliography omitted in Palm version)
Back to Table of Contents
3. DECISION-MAKING IN CLINICAL MEDICINE - Daniel B. Mark To the medical student who requires 2 h to collect a patient's history and perform a physical examination, and several additional hours to organize them into a coherent presentation, the experienced clinician's ability to reach a diagnosis and decide on a management plan in a fraction of the time seems extraordinary. While medical knowledge and experience play a significant role in the senior clinician's ability to arrive at a differential diagnosis and plan quickly, much of the process involves skill in clinical decision-making. The first goal of this chapter is to provide an introduction to the study of clinical reasoning. Equally bewildering to the student are the proper use of diagnostic tests and the integration of the results into the clinical assessment. The novice medical practitioner typically uses a "shotgun" approach to testing, hoping to a hit a target without knowing exactly what that target is. The expert, on the other hand, usually has a specific target in mind and efficiently adjusts the testing strategy to it. The second goal of this chapter is to review briefly some of the crucial basic statistical concepts that govern the proper interpretation and use of diagnostic tests; quantitative tools available to assist in clinical decision-making will also be discussed. CLINICAL DECISION-MAKING CLINICAL REASONING The most important clinical actions are not procedures or prescriptions but the judgments from which all other aspects of clinical medicine flow. In the modern era of large randomized trials, it is easy to overlook the importance of this elusive mental activity and focus instead on the algorithmic practice guidelines constructed to improve care. One reason for this apparent neglect is that much more research has been done on how doctors should make decisions (e.g., using a Bayesian model discussed below) than on how they actually do. Thus, much of what we know about clinical reasoning comes from empirical studies of nonmedical problem-solving behavior. Despite the great technological advances of the twentieth century, uncertainty still plays a pivotal role in all aspects of medical decision-making. We may know that a patient does not have long to live, but we cannot be certain how long. We may prescribe a potent new receptor blocker to reverse the course of a patient's illness, but we cannot be certain that the therapy will do so without side effects. Uncertainty in medical outcomes creates the need for probabilities and other mathematical/statistical tools to help guide decision-making. (These tools are reviewed later in the chapter.) Uncertainty is compounded by the information overload that characterizes modern medicine. Today's experienced clinician needs close to 2 million pieces of information to practice medicine. Doctors subscribe to an average of 7 journals, representing over 2500 new articles each year. Computers offer the obvious solution both for management of information and for better quantitation and management of the daily uncertainties of medical care. While the technology to computerize medical practice is available, many practical problems remain to be solved before patient information can be standardized and integrated with medical evidence on a single electronic platform.
The following three examples introduce the subject of clinical reasoning: · A 46-year-old man presents to his internist with a chief complaint of hemoptysis. The physician knows that the differential diagnosis of hemoptysis includes over 100 different conditions, including cancer and tuberculosis (Chap. 33). The examination begins with some general background questions, and the patient is asked to describe his symptoms and their chronology. By the time the examination is completed, and even before any tests are run, the physician has formulated a working diagnostic hypothesis and planned a series of steps to test it. In an otherwise healthy and nonsmoking patient recovering from a viral bronchitis, the doctor's hypothesis would be that the acute bronchitis is responsible for the small amount of blood-streaked sputum the patient observed. In this case, a chest x-ray and purified protein derivative (PPD) skin test may be sufficient. · A second 46-year-old patient with the same chief complaint who has a 100-pack-year smoking history, a productive morning cough, and episodes of blood-streaked sputum may generate the principal diagnostic hypothesis of carcinoma of the lung. Consequently, along with the chest x-ray andPPD skin test, the physician refers this patient for bronchoscopy. · A third 46-year-old patient with hemoptysis who is from a developing country is evaluated with an echocardiogram as well, because the physician thinks she hears a soft diastolic rumble at the apex on cardiac auscultation, suggesting rheumatic mitral stenosis. These three vignettes illustrate two aspects of expert clinical reasoning: (1) the use of cognitive shortcuts, or heuristics, as a way to organize the complex unstructured material that is collected in the clinical evaluation; and (2) the use of diagnostic hypotheses to consolidate the information and indicate appropriate management steps. THE USE OF COGNITIVE SHORTCUTS Heuristics reduce the complexity of a problem to a manageable level. Psychologists have found that people rely on three basic types of heuristics. For example, when assessing a patient, clinicians often weigh the probability that this patient's clinical features match those of the class of patients with the leading diagnostic hypotheses being considered. In other words, the clinician is searching for the diagnosis for which the patient appears to be a representative example; this cognitive shortcut is called the representativeness heuristic. It may take only a few characteristics from the history for an expert clinician using the representativeness heuristic to arrive at a sound diagnostic hypothesis. For example, an elderly patient with new-onset fever, cough productive of copious sputum, unilateral pleuritic chest pain, and dyspnea is readily identified as fitting the pattern for acute pneumonia, probably of bacterial origin. Evidence of focal pulmonary consolidation on the physical examination will increase the clinician's confidence in the diagnosis because it fits the expected pattern of acute bacterial pneumonia. Knowing this allows the experienced clinician to conduct an efficient, directed, and therapeutically productive patient evaluation although there may be little else in the history or physical examination of direct relevance. The inexperienced medical student or resident, who has not yet learned the patterns most prevalent in
clinical medicine, must work much harder to achieve the same result and is often at risk of missing the important clinical problem in a sea of compulsively collected but unhelpful data. However, physicians using the representativeness heuristic can reach erroneous conclusions if they fail to consider the underlying prevalence of two competing diagnoses. Consider a patient with pleuritic chest pain, dyspnea, and a low-grade fever. A clinician might consider acute pneumonia and acute pulmonary embolism to be the two leading diagnostic alternatives. Clinicians using the representativeness heuristic might judge both diagnostic candidates to be equally likely, although to do so would be wrong if pneumonia was much more prevalent in the underlying population. Mistakes may also result from a failure to consider that a pattern based on a small number of prior observations will likely be less reliable than one based on larger samples. A second commonly used cognitive shortcut, the availability heuristic, involves judgments made on the basis of how easily prior similar cases or outcomes can be brought to mind. For example, the experienced clinician may recall 20 elderly patients seen over the past few years who presented with painless dyspnea of acute onset and were found to have acute myocardial infarction. The novice clinician may spend valuable time seeking a pulmonary cause for the symptoms before considering and discovering the cardiac diagnosis. In this situation, the patient's clinical pattern does not fit the expected pattern of acute myocardial infarction, but experience with this atypical presentation, and the ability to recall it, can help direct the physician to the diagnosis. Errors with the availability heuristic can come from several sources of recall bias. For example, rare catastrophes are likely to be remembered with a clarity and force out of proportion to their value, and recent experience is, of course, easier to recall and therefore more influential on clinical judgments. The third commonly used cognitive shortcut, the anchoring heuristic, involves estimating a probability by starting from a familiar point (the anchor) and adjusting to the new case from there. For example, a clinician may judge the probability of colorectal cancer to be extremely high after an elevated screening carcinoembryonic antigen (CEA) result because the prediction of colorectal cancer is anchored to the test result. Yet, as discussed below, this prediction would be inaccurate if the clinical picture of the patient being tested indicates a low probability of disease (for example, a 30-year-old woman with no risk factors). Anchoring can be a powerful tool for diagnosis but is often used incorrectly (see "Measures of Disease Probability and Bayes' Theorem," below). DIAGNOSTIC HYPOTHESIS GENERATION Cognitive scientists studying the thought processes of expert clinicians have observed that clinicians group data into packets or "chunks," which are stored in their memories and manipulated to generate diagnostic hypotheses. Because short-term memory can typically hold only 7 to 10 items at a time, the number of packets that can be actively integrated into hypothesis-generating activities is similarly limited. The cognitive shortcuts discussed above play a key role in the generation of diagnostic hypotheses, many of which are discarded as rapidly as they are formed.
A diagnostic hypothesis sets a context for diagnostic steps to follow and provides testable predictions. For example, if the enlarged and quite tender liver felt on physical examination is due to acute hepatitis (the hypothesis), certain specific liver function tests should be markedly elevated (the prediction). If the tests come back normal, the hypothesis may need to be discarded or substantially modified. One of the factors that makes teaching diagnostic reasoning so difficult is that expert clinicians do not follow a fixed pattern in patient examinations. From the outset, they are generating, refining, and discarding diagnostic hypotheses. The questions they ask in the history are driven by the hypotheses they are working with at the moment. Even the physical examination is driven by specific questions rather than a preordained checklist. While the student is palpating the abdomen of the alcoholic patient, waiting for a finding to strike him, the expert clinician is on a focused search mission. Is the spleen enlarged? How big is the liver? Is it tender? Are there any palpable masses or nodules? Each question focuses the attention of the examiner to the exclusion of all other inputs until answered, allowing the examiner to move on to the next specific question. Negative findings are often as important as positive ones in establishing and refining diagnostic hypotheses. Chest discomfort that is not provoked or worsened by exertion in an active patient reduces the likelihood that chronic ischemic heart disease is the underlying cause. The absence of a resting tachycardia and thyroid gland enlargement reduces the likelihood of hyperthyroidism in a patient with paroxysmal atrial fibrillation. While the representativeness and availability heuristics may play the major roles in shaping early diagnostic hypotheses, the acuity of a patient's illness can also be very influential. For example, clinicians are taught to consider aortic dissection routinely as a possible cause of acute severe chest discomfort along with myocardial infarction, even though the typical history of dissection is different from myocardial infarction and dissection is far less prevalent (Chap. 247). This recommendation is based on the recognition that a relatively rare but catastrophic diagnosis like aortic dissection is very difficult to make unless it is explicitly considered. If the clinician fails to elicit any of the characteristic features of dissection by history and finds equivalent blood pressures in both arms and no pulse deficits, he or she may feel comfortable in discarding the aortic dissection hypothesis. If, however, the chest x-ray shows a widened mediastinum, the hypothesis may be reinstated and a diagnostic test ordered [e.g., thoracic computed tomography (CT) scan, transesophageal echocardiogram] to evaluate it more fully. In noncritical situations, the prevalence of potential alternative diagnoses should play a much more prominent role in diagnostic hypothesis generation. The value of conducting a rapid systematic clinical survey of symptoms and organ systems to avoid missing important but inapparent clues cannot be overstated. Because the generation and evaluation of appropriate diagnostic hypotheses is a skill that not all clinicians possess to an equal degree, errors in this process can occur, and in the patient with serious acute illness these may lead to tragic consequences. Consider the following hypothetical example. A 45-year-old male patient with a 3-week history of a "flulike" upper respiratory infection (URI) presented to his physician with symptoms of dyspnea and a productive cough. Based on the presenting complaint, the clinician pulled out a "URI Assessment Form" to improve quality and efficiency of care. The physician quickly completed the examination components outlined on this
structured form, noting in particular the absence of fever and a clear chest examination. He then prescribed an antibiotic for presumed bronchitis, showed the patient how to breathe into a paper bag to relieve his "hyperventilation," and sent him home with the reassurance that his illness was not serious. After a sleepless night with significant dyspnea unrelieved by rebreathing into a bag, the patient developed nausea and vomiting and collapsed. He was brought into the Emergency Department in cardiac arrest and could not be resuscitated. Autopsy showed a posterior wall myocardial infarction and a fresh thrombus in an atherosclerotic right coronary artery. What went wrong? The clinician decided, even before starting the history, that the patient's complaints were not serious. He therefore felt confident that he could perform an abbreviated and focused examination using the URI assessment protocol rather than considering the full range of possibilities and performing appropriate tests to confirm or refute his initial hypotheses. In particular, by concentrating on the "URI," the clinician failed to elicit the full dyspnea history, which would have suggested a far more serious disorder, and did not even search for other symptoms that could have directed him to the correct diagnosis. This example illustrates how patients can diverge from textbook symptoms and the potential consequences of being unable to adapt the diagnostic process to real-world challenges. The expert, while recognizing that common things occur commonly, approaches each evaluation on high alert for clues that the initial diagnosis may be wrong. Patients often provide information that "does not fit" with any of the leading diagnostic hypotheses being considered. Distinguishing real clues from false trails can only be achieved by practice and experience. A less experienced clinician who tries to be too efficient (as in the above example) can make serious judgment errors. MAJOR INFLUENCES ON CLINICAL DECISION-MAKING More than a decade of research on variations in clinician practice patterns has shed much light on forces that shape clinical decisions. The use of heuristic "shortcuts," as detailed above, provides a partial explanation, but several other key factors play an important role in shaping diagnostic hypotheses and management decisions. These factors can be grouped conceptually into three overlapping categories: (1) factors related to physician personal characteristics and practice style, (2) factors related to the practice setting, and (3) economic incentive factors. Practice Style Factors One of the key roles of the physician in medical care is to serve as the patient's agent to ensure that necessary care is provided at a high level of quality. Factors that influence this role include the physician's knowledge, training, and experience. It is obvious that physicians cannot practice evidence-based medicine if they are unfamiliar with the evidence. As would be expected, specialists generally know the evidence in their field better than do generalists. Surgeons may be more enthusiastic about recommending surgery than medical doctors because their belief in the beneficial effects of surgery is stronger. For the same reason, invasive cardiologists are much more likely to refer chest pain patients for diagnostic catheterization than are noninvasive cardiologists or generalists. The physician beliefs that drive these different practice styles are based on personal experience, recollection, and interpretation of the available medical evidence. For example, heart failure specialists are much more likely than generalists to achieve target angiotensin-converting enzyme (ACE) inhibitor
therapy in their heart failure patients because they are more familiar with what the targets are (as defined by large clinical trials), have more familiarity with the specific drugs (including dosages and side effects), and are less likely to overreact to foreseeable problems in therapy such as a rise in creatinine levels or symptomatic hypotension. Other intriguing research has shown a wide distribution of acceptance times of antibiotic therapy for peptic ulcer disease following widespread dissemination of the "evidence" in the medical literature. Some gastroenterologists accepted this new therapy before the evidence was clear (reflecting, perhaps, an aggressive practice style), and some gastroenterologists lagged behind (a conservative practice style, associated in this case with older physicians). As a group, internists lagged several years behind gastroenterologists. The opinion of influential leaders can also have an important effect on practice patterns. Such influence can occur at both the national level (e.g., expert physicians teaching at national meetings) and the local level (e.g., local educational programs, "curbside consultants"). Opinion leaders do not have to be physicians. When conducting rounds with clinical pharmacists, physicians are less likely to make medication errors and more likely to use target levels of evidence-based therapies. The patient's welfare is not the only concern that drives clinical decisions. The physician's perception about the risk of a malpractice suit resulting from either an erroneous decision or a bad outcome creates a style of practice referred to as defensive medicine. This practice involves using tests and therapies with very small marginal returns to preclude future criticism in the event of an adverse outcome. For example, a 40-year-old woman who presents with a long-standing history of intermittent headache and a new severe headache along with a normal neurologic examination has a very low likelihood of structural intracranial pathology. Performance of a headCT or magnetic resonance imaging (MRI) scan in this situation would constitute defensive medicine. On the other hand, the results of the test could provide reassurance to an anxious patient. Practice Setting Factors Factors in this category relate to the physical resources available to the physician's practice and the practice environment. Physician-induced demand is a term that refers to the repeated observation that physicians have a remarkable ability to accommodate to and employ the medical facilities available to them. A classic early study in this area showed that physicians in Boston had an almost 50% higher hospital admission rate than did physicians in New Haven, despite there being no obvious differences in the health of the cities' inhabitants. The physicians in New Haven were not aware of using fewer hospital beds for their patients, nor were the Boston physicians aware of using less stringent criteria to admit patients. Other environmental factors that can influence decision-making include the local availability of specialists for consultations and procedures, "high tech" facilities such as angiography suites, a heart surgery program, andMRImachines. Economic Incentives Economic incentives are closely related to the other two categories of practice-modifying factors. Financial issues can exert both stimulatory and inhibitory influences on clinical practice. In general, physicians are paid on a fee-for-service, capitation, or salary basis (Chap. 4). In fee-for-service, the more the physician does, the more the physician gets paid. The incentive in this case is to do
more. When fees are reduced (discounted fee-for-service), doctors tend to increase the number of services billed for. Capitation, in contrast, provides a fixed payment per patient per year, encouraging physicians to take on more patients but to provide each patient with fewer services. Expensive services are more likely to be affected by this type of incentive than inexpensive preventive services. Salary compensation plans pay physicians the same regardless of the amount of clinical work performed. The incentive here is to see fewer patients. Recognizing these powerful shapers of physician behavior, managed care plans have begun to explore combinations of the three reimbursement types with the goal of improving individual physician productivity while restraining their use of expensive tests and therapies. In summary, expert clinical decision-making can be appreciated as a complex interplay between cognitive devices used to simplify large amounts of complex information interacting with physician biases reflecting education, training, and experience, all of which are shaped by powerful, sometimes perverse, external forces. In the next section, we will review a set of statistical tools and concepts that can assist in making clinical decisions under uncertainty. QUANTITATIVE METHODS TO AID CLINICAL DECISION-MAKING The process of medical decision-making can be divided into two parts: (1) defining the available courses of action and estimating the likely outcomes with each, and (2) assessing the desirability of the outcomes. The former task involves integrating key information about the patient along with relevant evidence from the medical literature to create the structure of a decision problem. The remainder of this chapter will present some quantitative tools to assist the clinician in these activities. These tools can be divided into those that assist the clinician in making better outcome predictions, which are then used to make decisions, and those that support the decision process directly. While these tools are not yet used routinely in daily clinical practice, the computerization of medicine is creating the required substrate for their future widespread dissemination. QUANTITATIVE MEDICAL PREDICTIONS Diagnostic Testing The purpose of performing a test on a patient is to reduce uncertainty about the patient's diagnosis or prognosis and to aid the clinician in making management decisions. Although diagnostic tests are commonly thought of as laboratory tests (e.g., measurement of serum amylase level) or procedures (e.g., colonoscopy or bronchoscopy), any technology that changes our understanding of the patient's problem qualifies as a diagnostic test. Thus, even the history and physical examination can be considered a form of diagnostic test. In clinical medicine, it is common to reduce the results of a test to a dichotomous outcome, such as positive or negative, normal or abnormal. In many cases, this simplification results in the waste of useful information. However, such simplification makes it easier to demonstrate some of the quantitative ways in which test data can be used. To characterize the accuracy of diagnostic tests, four terms are routinely used (Table 3-1). The true-positive rate, i.e., the sensitivity, provides a measure of how well the test correctly identifies patients with disease. The false-negative rate is calculated as (1sensitivity). The true-negative rate, i.e., the specificity, reflects how well the test
correctly identifies patients without disease. The false-positive rate is (1- specificity). A perfect test would have a sensitivity of 100% and a specificity of 100% and would completely separate patients with disease from those without it. Calculating sensitivity and specificity require selection of a cutpoint value for the test to separate "normal" from "diseased" subjects. As the cutpoint is moved to improve sensitivity, specificity typically falls and vice versa. This dynamic tradeoff between more accurate identification of subjects with versus those without disease is often displayed graphically as a receiver operating characteristic (ROC) curve. An ROC curve plots sensitivity (y-axis) versus 1 -specificity (x-axis). Each point on the curve represents a potential cutpoint with an associated sensitivity and specificity value. The area under the ROC curve is often used as a quantitative measure of the information content of a test. Values range from 0.5 (no diagnostic information at all, test is equivalent to flipping a coin) to 1.0 (perfect test). In the diagnostic testing literature, ROC areas are often used to compare alternative tests. The test with the highest area (i.e., closest to 1.0) is presumed to be the most accurate. However, ROC curves are not a panacea for evaluation of diagnostic test utility. Like Bayes' theorem, they are typically focused on only one possible test parameter (e.g., ST segment response in a treadmill exercise test) to the exclusion of other potentially relevant data. In addition, ROC area comparisons do not simulate the way test information is actually used in clinical practice. Finally, biases in the underlying population used to generate the ROC curves (e.g., related to an unrepresentative test sample) can bias the ROC area and the validity of a comparison among tests. Measures of Disease Probability and Bayes' Theorem Unfortunately, there are no perfect tests; after every test is completed the true disease state of the patient remains uncertain. Quantitating this residual uncertainty can be done with Bayes' theorem. This theorem provides a simple mathematical way to calculate the posttest probability of disease from three parameters: the pretest probability of disease, the test sensitivity, and the test specificity (Table 3-2). The pretest probability is a quantitative expression of the confidence in a diagnosis before the test is performed. In the absence of more relevant information it is usually estimated from the prevalence of the disease in the underlying population. For some common conditions, such as coronary artery disease (CAD), nomograms and statistical models have been created to generate better estimates of pretest probability from elements of the history and physical examination. The posttest probability, then, is a revised statement of the confidence in the diagnosis, taking into account both what was known before and after the test. To understand how Bayes' theorem creates this revised confidence statement, it is useful to examine a nomogram version of Bayes' theorem that uses the same three parameters to predict the posttest probability of disease (Fig. 3-1). In this nomogram, the accuracy of the diagnostic test in question is summarized by the likelihood ratio for a positive test, which is the ratio of the true-positive rate to the false-positive rate [or sensitivity/(1 - specificity)]. For example, a test with a sensitivity of 0.90 and a specificity of 0.90 has a likelihood ratio of 0.90/(1 - 0.90), or 9. Thus, for this hypothetical test, a "positive" result is 9 times more likely in a patient with the disease than in a patient without it. The more accurate the test, the higher the likelihood ratio. However, if sensitivity is excellent but specificity is less so, the likelihood ratio will be substantially
reduced (e.g., with a 90% sensitivity but a 60% specificity, the likelihood ratio is 2.25). Most tests in medicine have likelihood ratios for a positive result between 1.5 and 20. Consider two tests commonly used in the diagnosis ofCAD, an exercise treadmill and an exercise thallium-201 single photon emission CT (SPECT) test (Chap. 244). Meta-analysis has shown the treadmill to have an average sensitivity of 66% and an average specificity of 84%, yielding a likelihood ratio of 4.1 [0.66/(1 - 0.84)]. If we use this test on a patient with a pretest probability of CAD of 10%, the posttest probability of disease following a positive result rises only to about 30%. If a patient with a pretest probability of CAD of 80% has a positive test result, the posttest probability of disease is about 95%. The exercise thalliumSPECTtest is a more accurate test for the diagnosis ofCAD. For our purposes, assume that it has both a sensitivity and specificity of 90%, yielding a likelihood ratio of 9.0 [0.90/(1 - 0.90)]. If we again test our low pretest probability patient and he has a positive test, using Fig. 3-1 we can demonstrate that the posttest probability of CAD rises from 10 to 50%. However, from a decision-making point of view, the more accurate test has not been able to improve diagnostic confidence enough to change management. In fact, the test has moved us from being fairly certain that the patient did not have CAD to being completely undecided (a 50:50 chance of disease). In a patient with a pretest probability of 80%, using the more accurate thallium SPECT test raises the posttest probability to 97% (compared with 95% for the exercise treadmill). Again, the more accurate test does not provide enough improvement in posttest confidence to alter management, and neither test has improved much upon what was known from clinical data alone. If the pretest probability is low (e.g., £20%), even a positive result on a very accurate test will not move the posttest probability to a range high enough to rule in disease (e.g., ³80%). Conversely, with a high pretest probability, a negative test will not adequately rule out disease. Thus, the largest gain in diagnostic confidence from a test occurs when the clinician is most uncertain before performing it (e.g., pretest probability between 30 and 70%). For example, if a patient has a pretest probability forCAD of 50%, a positive exercise treadmill test will move the posttest probability to 80% and a positive exercise thalliumSPECTtest will move it to 90% (Fig. 3-1). Bayes' theorem, as presented above, employs a number of important simplifications that should be considered. First, few tests have only two useful outcomes, positive or negative, and many tests provide numerous pieces of data about the patient. Even if these can be integrated into a summary result, multiple levels of useful information may be present (e.g., strongly positive, positive, indeterminate, negative, strongly negative). While Bayes' theorem can be adapted to this more detailed test result format, it is computationally complex to do so. Second, Bayes' theorem assumes that the information from the test is completely unique and nonoverlapping with information used to estimate the pretest probability. This independence assumption, however, is often wrong. In many cases, test results are correlated with patient characteristics. For example, the findings of cardiomegaly and pulmonary edema on chest x-ray are correlated with the historic features of heart failure and with the physical findings of a displaced left ventricular apical impulse, an S3gallop, and rales. The unique predictive information contributed by the test in this case (the chest x-ray) is only a fraction of its
total information because much had already been learned about the probability of heart failure before the test was done. Finally, it has long been thought that sensitivity and specificity are prevalence-independent parameters of test accuracy, and many texts still make this assertion. This statistically useful assumption, however, is clinically wrong. For example, a treadmill exercise test has a sensitivity in a population of patients with one-vesselCAD of around 30%, whereas the sensitivity in severe three-vessel CAD approaches 80%. Thus, the best estimate of sensitivity to use in a particular decision will often vary depending on the distribution of disease stages present in the tested population. A hospitalized population typically has a higher prevalence of disease and in particular a higher prevalence of more advanced disease stages than an outpatient population. As a consequence, test sensitivity will tend to be higher in hospitalized patients, whereas test specificity will be higher in outpatients. Statistical Prediction Models Bayes' theorem, as presented above, deals with a clinical prediction problem that is unrealistically simple relative to most problems a clinician faces. Prediction models, based on multivariable statistical models, can handle much more complex problems and substantially enhance predictive accuracy for specific situations. Their particular advantage is the ability to take into account many overlapping pieces of information and assign a relative weight to each based on its unique contribution to the prediction in question. For example, a logistic regression model to predict the probability ofCADtakes into account all of the relevant independent factors from the clinical examination and diagnostic testing instead of the small handful of data that clinicians can manage in their heads or with Bayes' theorem. However, despite this strength, the models are too complex computationally to use without a calculator or computer (although this limit may be overcome when medicine is practiced from a fully computerized platform.) To date, only a handful of prediction models have been developed and properly validated. The importance of independent validation in a population separate from the one used to develop the model cannot be overstated. Unfortunately, most published models have not been properly validated, making their utility in clinical practice uncertain at best. When statistical models have been compared directly with expert clinicians, they have been found to be more consistent, as would be expected, but not significantly more accurate. Their biggest promise, then, would seem to be to make less-experienced clinicians more accurate predictors of outcome. DECISION SUPPORT TOOLS DECISION SUPPORT SYSTEMS Over the past 30 years, many attempts have been made to develop computer systems to help clinicians make decisions and manage patients. Conceptually, computers offer a very attractive way to handle the vast information load that today's physicians face. The computer can help by making accurate predictions of outcome, simulating the whole decision process, or providing algorithmic guidance. Computer-based predictions using Bayesian or statistical regression models inform a clinical decision but do not actually reach a "conclusion" or "recommendation." Artificial intelligence systems attempt to
simulate or replace human reasoning with a computer-based analogue. To date, such approaches have achieved only limited success. Reminder or protocol-directed systems do not make predictions but use existing algorithms, such as practice guidelines, to guide clinical practice. In general, however, decision support systems have shown little impact on practice. Reminder systems, although not yet in widespread use, have shown the most promise, particularly in correcting drug dosing and in promoting guideline adherence. The full potential of these approaches will only be achieved when computers are fully integrated into medical practice. DECISION ANALYSIS Compared with the methods discussed above, decision analysis represents a completely different approach to decision support. Its principal application is in decision problems that are complex and involve a substantial risk, a high degree of uncertainty in some key area, or an idiosyncratic feature that does not "fit" the available evidence. Three general steps are involved. First, the decision problem must be clearly defined. Second, the elements of the decision must be made explicit. This involves specifying the alternatives being considered, their relevant outcomes, the probabilities attached to each outcome, and the relative desirability (called "utility") of each outcome. Cost can also be assigned to each branch of the decision tree, allowing calculation of cost-effectiveness (Chap. 4). An example of a decision tree used to evaluate strategies for management of the risk of infective endocarditis after catheter-associated Staphylococcus aureus bacteremia is shown in Fig. 3-2. Approximately 35,000 cases of S. aureus bacteremia occur each year in the United States. The development of complicating endocarditis, which occurs in about 6% of cases, is associated with high morbidity (31% mortality, 21% stroke rate) and medical costs. The three choices for management of the bacteremia are (1) transesophageal echocardiography (TEE), (2) a 4-week course of intravenous antibiotics (long-course), or (3) a 2-week course of intravenous antibiotics (short-course). In the TEE strategy, a 4-week course of antibiotics is given if endocarditis is evident and a 2-week course is given if it is not. With each strategy, there is a risk that the patient will develop endocarditis with or without major complications. In this analysis, the longest quality-adjusted survival (5.47 quality-adjusted life-years) was associated with the 4-week antibiotic course strategy, which also had the highest costs ($14,136 per patient), whereas the lowest costs ($9830 per patient) and worst outcomes (5.42 quality-adjusted life-years) were associated with the 2-week antibiotic course strategy. From a clinical point of view (ignoring costs), the 4-week antibiotic course was best. From a cost-effectiveness point of view, the TEE strategy (5.46 quality-adjusted life-years and $10,051 per patient costs) provided the best balance of added benefits and costs. Thus, decision analysis can be extremely helpful in clarifying tradeoffs in outcomes and costs in difficult management areas such as the above where it is highly unlikely that an adequate randomized trial will ever be done. The data needed to fill in a decision tree (Fig. 3-2) are typically cobbled together from a variety of sources, including the literature (randomized trials, meta-analyses, observational studies) and expert opinion. Once the decision tree is finished, the decision is "analyzed" by calculating the average value of each limb of the tree. The decision arm with the highest net value (or expected utility) is the preferred choice. The
value of this exercise, however, is not so much in developing a prescription for action as it is in exploring the key elements and pressure points of a complex or difficult decision. The process of building the decision tree forces the analyst to be explicit about the choices being considered and all their relevant outcomes. Areas of high uncertainty are readily identified. Sensitivity analyses are an integral part of decision analysis and involve systematically varying the value of each key parameter in the model alone (one-way sensitivity analysis,) in pairs (two-way), or in higher combinations (multivariable) to assess the impact on choice of preferred management strategy. In the above example, varying the incidence of endocarditis resulting from S. aureus bacteremia from 3% to over 50% had no impact on the choice ofTEE as the preferred strategy. User friendly personal computer-based software packages now make the creation and analysis of decision trees much more straightforward than in the past. However, the process is still too cumbersome and time-consuming to be used on a routine basis. When medicine is practiced from a fully computerized platform, a library of prestructured decision trees with user modifiable values can be made available to support practitioners working with individual patients. CONCLUSIONS In this era of evidence-based medicine, it is tempting to think that all the difficult decisions practitioners face have been or soon will be solved and digested into practice guidelines and computerized reminders. For the foreseeable future, however, such is not the case. Meta-analyses cannot generate evidence where there are no adequate randomized trials, and most of what clinicians face will never be thoroughly tested in a randomized trial. Excellent clinical reasoning skills and experience supplemented by well-designed quantitative tools and a keen appreciation for individual patient preferences will continue to be of paramount importance in the professional life of medical practitioners for years to come. (Bibliography omitted in Palm version) Back to Table of Contents
4. ECONOMIC ISSUES IN CLINICAL MEDICINE - Daniel B. Mark The United States has the distinction of having some of the best medical care of any technologically advanced country. We have many of the best hospitals and doctors in the world. The research pipeline is full of significant new therapeutic advances, with revolutionary genetic-based therapies perhaps only a decade away. Our citizens largely subscribe to the principle that excellent medical care should be available to all, regardless of ability to pay. Yet we also have over 43 million people (most of them employed and earning minimal wages) without any health insurance and many more who are inadequately insured. Since the collapse of the Clinton health care reform efforts in 1994, U.S. health policy has been directed by marketplace forces that have created powerful and sometimes perverse incentives in medicine: Health insurance companies that use every available means to avoid insuring sick people; "managed care" programs that really only manage costs; doctors who are provided incentives to provide less medical care; and pharmaceutical companies that develop powerful and expensive new drugs priced beyond the reach of many of the elderly and chronically ill who need them most. Facing such powerful and chaotic forces, physicians tend to focus narrowly on what they are most comfortable with, taking care of individual patients and conducting academic investigations. Many doctors consider economics too arcane for them to grasp and therefore do not even try. Consequently, when presented with economic arguments and evidence they are often unable to discriminate the legitimate from the fallacious. More importantly, they are ill equipped to defend their patients' interests in the crucible of cost containment that characterizes the modern managed care era. This chapter has two goals: first, to provide a brief introduction to some of the larger economic forces that shape modern medical practices, and second, to introduce the economic tools that are used for assessing the value of medical practices, including cost effectiveness analysis. HEALTH CARE SPENDING AND FINANCING HOW MUCH IS SPENT ON HEALTH CARE? In 1997, the United States spent $1.1 trillion on its health care system, representing 13.5% of the gross domestic product (GDP) (a crude measure of national income). Most of this ($969 billion) was spent on personal health care: 34% went to hospitals, 20% to physicians, 7% to nursing homes, and 8% to outpatient pharmaceuticals. In comparison, Canada and Western European countries spend a substantially smaller portion (6 to 10%) of their national income on health care but their citizens appear to be equally healthy, at least by crude metrics such as life expectancy and infant mortality rates. Economists and politicians have for years used such data to argue that the United States spends too much on health care. The issue of how much to spend is an inherently political one, however, and the discipline of economics has little to say about it. WHO PAYS FOR HEALTH CARE?
Two major factors are continually driving up the costs of medical care: introduction into medical practice of new medical technologies (drugs, devices, procedures) that have a high price tag, and the aging of the U.S. population (since older people require more medical care than younger ones). These costs are distributed unevenly across society. In 1997, the government paid about 47% of the total national health care bill (75% federal, 25% states), private insurance paid about 32%, and individuals paid 17%. The government, of course, gets its money from taxpayers and uses the health care segment of its budget to pay for the Medicare and Medicaid programs (discussed below). To respond to rising medical costs, the government can increase taxes or redistribute funds from other programs such as defense and education. Neither of these options are politically attractive. Alternatively, because of its size in the medical marketplace, the government can impose lower prices on providers to make the available funds go farther (see "Cost-Containment Strategies," below). Much of the private insurance bill is subsidized by employers through their employee benefits packages. As medical costs go up, health insurance costs also rise and businesses must either pass on higher premium and copayment costs to their employees, raise their prices (potentially impairing their competitive position in the marketplace), or reduce their profit margin (a very unpopular move with stockholders). Like the government, businesses may also negotiate lower prices with health care providers and/or health insurance plans. PUBLIC FINANCING OF HEALTH CARE The public sector (i.e., government as an agent for society) finances the Medicare and Medicaid programs as well as the Veteran's Administration Hospital system, the Department of Defense military care system, the Public Health Service, and the Indian Health Service. Of these, Medicare is by far the largest and most influential, with 39 million people receiving health insurance at a total cost in 1997 of $214.6 billion (20% of total national health expenditures). The Medicare program was enacted in 1965 by Congress as an amendment to the Social Security Act of 1935 and was envisioned by President Lyndon B. Johnson as a first step toward universal health insurance in the United States, a key part of his "great society" plan. Its impact on the evolution of the U.S. health care system has been profound. The original congressional act provided health care insurance for the elderly (defined as those 65 and older) who were eligible for social security (i.e., retired workers who had paid into the system during their working years and their dependents). Amendments in 1972 extended coverage to the disabled of all ages (currently numbering around 5 million) and to patients with chronic renal failure (who currently number about 284,000). Medicare consists of two related insurance programs. The Medicare Hospital Insurance Trust Fund (also known as Part A) covers hospital care and skilled nursing home care and is funded by compulsory federal payroll taxes on employers and employees. Medicare Part B, the Medical Supplementary Insurance Program, covers physician fees as well as laboratory and other diagnostic tests and is funded by general federal tax revenues and patient premiums. Both programs have substantial gaps in coverage, necessitating supplemental insurance (so-called Medigap policies) for those who can afford them. Because of its compulsory income redistribution feature, taking tax money from current workers to pay for health care for elderly citizens (many of whom are on fixed income close to the poverty level), Medicare is both a health insurance program
and a social welfare program designed to combat poverty in the disabled and elderly. In exchange for their tax money, the 150 million workers funding the program are promised the same type of social security when they become elderly (paid for by future generations of workers). Medicaid is a social insurance program for the poor that is jointly run by the federal and state governments. The federal government gives each state a grant of money for the program based on that state's per capita income (in 1997, this amount totaled $95 billion), and the states pay for the rest ($65 billion in 1997). The program, like Medicare, was enacted by Congress in 1965 as a part of President Johnson's "great society" program. It is larger than Medicare in terms of eligible beneficiaries (41 million people) but smaller in terms of budget ($160 billion, or 12% of the total national health expenditures). Because the requirements to qualify for Medicaid are stringent, many low-income individuals under age 65 (especially the working poor) do not qualify. Eligibility criteria are set by each state within general federal guidelines, and the income and asset tests individuals must meet to qualify vary widely among states. Many of the dollars in the Medicaid program actually pay for care for elderly and disabled Medicare beneficiaries who also qualify for Medicaid on the basis of poverty. PRIVATE FINANCING OF HEALTH CARE Approximately 70% of the non-elderly U.S. population is covered by some form of private medical insurance. The feasibility of group insurance for medical care was initially demonstrated in the 1930s by Blue Cross, a franchise of nonprofit groups providing hospitalization insurance in order to help prop up the financially strapped U.S. hospital industry. Blue Shield, a separate organization modeled after Blue Cross, started providing insurance for in-hospital physician services in 1939. During World War II, employee wages were frozen by the government and to entice workers, who were in short supply, some employers started offering health insurance as a fringe benefit. With the feasibility of employer-sponsored group health insurance demonstrated by the experience of the "Blues," commercial insurers began to enter the market. To win the support of doctors and hospitals, insurers agreed to pay "reasonable and customary charges" and to defer all medical management decisions to doctors. This "fee-for-service" reimbursement system, created in the post-World War II era, sowed the seeds of the tremendous inflation observed in the U.S. medical system during the 1970s and 1980s. The original focus of indemnity insurance plans was to cover individuals against catastrophic financial losses from high medical care bills. Insurance is a contract for protection against specific hazards that are unpredictable for individuals but can be defined with confidence for large groups. "Major medical" health insurance was designed to provide coverage for catastrophic illness, a relatively rare event in most populations. Group coverage is less expensive than individual coverage because it allows the insurance company to diffuse the risk of a large payout among a big pool of individuals who will pay premiums but make no claims. When coverage is shifted from a focus on rare catastrophes to routine maintenance medical care (comprehensive insurance policies), health insurance becomes a means for payment of expected rather than unexpected care. The consequence is higher health insurance premiums. The early appeal of health maintenance organizations (HMOs) was that they appeared to
offer an economically efficient way to provide routine preventive care and to manage the occasional catastrophic illness. MANAGED CARE Managed care is a generic term that embraces a wide spectrum of systems for integrating the financing and delivery of health care. Managed care organizations (MCOs) contract with doctors and hospitals to provide comprehensive care to enrolled members for a fixed, prospectively set, premium.HMOs are a form of managed care originally organized between the 1940s and 1960s as an alternative to the prevailing fee-for-service-based private insurance. With the advent of serious medical inflation in the 1970s, the HMO model was promoted by the federal government as a way to control the growth in medical spending. Early enthusiasm for this initiative was limited; in 1984, only 5% of individuals with employer-based health insurance were in an HMO. However, by 1998 that figure had risen to 85%. The exponential growth of managed care started in the 1990s in part as an employer-driven response to the uncontrolled medical inflation of the previous two decades. The massive increase in demand for managed care by employers and by the Medicare program produced a rapid, and sometimes bewildering, evolution in the managed care industry. One important trend has been the growth of for-profit (i.e., investor owned) managed care companies. Over half ofHMOmembers now belong to a for-profit plan. Investment dollars from Wall Street have made it easier for these plans to respond quickly to increased employer demand for managed care options. However, compared with their not-for-profit counterparts, for-profit HMOs spend a smaller proportion of each premium dollar paying for health care for members (the paradoxically named "medical loss ratio"), since stockholders also have to be paid. As a result, for-profit HMOs are less successful than not-for-profit plans in providing preventive care (a presumed strength of managed care). Another prevalent trend of the 1990s was the move from traditionalHMOmodels to virtual HMOs, built from contractual relationships with community physicians and hospitals. The three HMO models are the staff model, the group model, and the Independent Practice Association (IPA). The staff model HMO is a vertically integrated organization. That is, it owns its own hospitals, employs all its physicians full time for a set salary, and is focused in a particular geographic area. The group model HMO, exemplified by Group Health Cooperative of Puget Sound, contracts with one or more large multispecialty group practices to care for its patients for a preset capitated reimbursement. These physicians do not care for non-HMO patients. In the IPA model, the HMO contracts with an association of self-employed physicians who maintain their own offices and see both HMO and non-HMO patients. The network model refers to a hybrid of the other three forms of HMO. IPA and network model HMOs now have the majority of HMO membership in the United States. The other portion of the managed care industry is represented by point of service (POS) plans and preferred provider organizations (PPOs). POS plans incorporate key features of bothHMOs and traditional fee-for-service plans. A patient may choose care from a provider network or go outside the network. Care within network requires only a minimal copayment, while care outside the network requires a deductible and a large (e.g., 30%)
copayment. The goal of the plan is to offer patients a choice but to provide major financial incentives to stay within the HMO portion of the plan. PPOs use a defined provider network (physicians, hospitals) that has agreed to accept discounted fee-for-service to care for enrolled members. PPOs may incorporate various managed care features, such as physician gatekeepers and utilization review. THE UNINSURED AND UNDERINSURED Data from the U.S. Census Bureau indicate that 43.4 million people had no health insurance for all of 1997 and 71.5 million people were without insurance for at least part of the year. The great majority of uninsured individuals either work for small employers who do not offer a health insurance benefit or, more commonly, cannot afford the premiums of the plan(s) that are offered. Underinsurance also has a significant impact on the working poor by requiring them to pay an excessive proportion of their family's income for health insurance premiums and out-of-pocket medical costs (deductibles, copayments, and uninsured care). Outpatient prescription medications are a major source of underinsurance. Prescription drug costs are now the fastest growing segment of the national medical budget and the least likely segment to be covered by insurance. The elderly are particularly affected, since Medicare does not currently cover outpatient prescriptions and even Medigap policies have limited coverage. Some states have experimented with expanded coverage through their Medicaid programs to help the uninsured poor (such as the Oregon Medicaid program). For the forseeable future, however, it does not appear that the federal government will address this problem comprehensively. COST-CONTAINMENT STRATEGIES Current projections from the federal government's Health Care Finance Administration (HCFA) are that health care expenditures will double (to $2.2 trillion, or 16.2% of theGDP) by 2008. Over the past 30 years, the U.S. health care system has experimented with a vast array of cost-containment approaches. Conceptually, there are four major ways to control medical spending: (1) control prices, (2) control volume of care provided, (3) control the total budget available to pay for care, and (4) shift costs to another payer. Two of the most important price control initiatives in medicine have been the Medicare Hospital Prospective Payment System and the Medicare Fee Schedule for physicians. In 1983, Medicare replaced its retrospective cost-based hospital reimbursement system with a prospective payment system. In this system, all hospitalizations are classified into one of approximately 500 Diagnosis Related Groups (DRGs) based on the principal discharge diagnosis for the hospitalization and a few selected additional factors such as age, the performance of surgery, and the presence of complications. Each DRG is assigned an average reimbursement (adjusted annually). If the hospital can provide care for less than this amount, they make a profit. If they spend more than this amount, they lose money. The DRG system was designed to promote efficiency and cost containment in hospital-based care. While it has helped to control Medicare costs, it has not reduced overall U.S. health care costs, probably because of substantial cost-shifting by hospitals to the private insurance sector.
Between 1975 and 1987, Medicare payments to physicians increased at an annual rate of 18%, well above the rate of inflation. While total spending for physician services accounts for less than 25% of the Medicare budget, physicians have control over aspects of care (use of procedures, length of stay, hospital admission) that extend their direct influence to over 75% of the Medicare budget. Recognizing the importance of physicians in cost containment, Congress directed the development of a new physician payment system based on the use of a resource-based relative value scale (RBRVS). The Medicare Fee Schedule, which was first used in 1992, has three components: (1) a measure of the total work (time and complexity) involved in each physician service and standardized across all specialties, (2) a practice expense to cover the cost of running an office, and (3) an amount to cover malpractice insurance costs. The Medicare Fee Schedule classifies all physician services using the American Medical Association's Current Procedural Terminology (CPT) codes. Each CPT code has an associated relative value units (RVUs) weight. The RVU weights are multiplied by a national conversion factor to generate the actual physician fee associated with the service in question. Price controls are attractive for cost containment because they are less expensive administratively than volume controls and don't involve micromanagement of clinical care. Price controls alone, however, don't generally achieve control of costs because of compensatory responses of providers. For example, under Medicare prospective payment, hospitals have shifted much care to the outpatient setting, whereDRGs are not used. Physicians have responded to lower fees by an increased volume and intensity of service. Volume controls include various programs to limit the diffusion of expensive technologies (such as heart surgery) or extra hospital beds. Limits can be operationalized using either a regulatory approach [such as certificate of need (CON) programs] or a budgetary approach. Utilization review approaches attempt to discern which expensive care items are medically necessary and which are not. Budgetary controls are simpler than either price or volume control approaches. In Canada, for example, hospitals have global annual budgets. How the money is spent is decided by each hospital. If the budget is exceeded, there are no guarantees that the shortfall will be covered. Finally, payers can control their costs by cost-shifting to other willing payers. For example, as health insurance premiums rise, employers can choose to pass these costs on to employees. Hospitals and doctors who lose money caring for Medicare patients can try to make up their losses by charging more to private insurance patients. Insurance companies can choose to offer limited or no coverage for outpatient pharmaceuticals, shifting the full cost of expensive new medicines directly to patients. MEDICAL ECONOMIC CONCEPTS AND TOOLS MEDICAL COST CONCEPTS Medical cost analysis is a field that borrows heavily from both economics and
accounting. Economics provides the theoretical structure that defines the key questions to be addressed, and accounting provides many of the measurement tools. Traditional economics has as one of its major axioms that societal resources are finite. For this reason, society must choose from among the many ways that resources can be used and not all of society's goals can be fulfilled. Economics has devised a theoretical framework and a set of tools (including cost-effectiveness analysis) to help define the major competing goals for societal resources and to assist in selecting from among the ones that most efficiently fulfill societal needs. "Cost" in economics refers not so much to money but rather to the lost opportunities that occur when the limited societal resources are expended in a particular way. For example, if our medical armamentarium is enhanced over the next decade by discovery of powerful but expensive therapies and these are incorporated into standard clinical practice, the ability of the country to invest in education, defense, or transportation may be compromised. This notion of cost as a lost opportunity to use resources in alternative ways is referred to as opportunity cost. While representing the purest economic notion of cost, there is no practical way to measure it. Accountants, who are much more concerned with issues of measurement, have proposed a "gold standard" of cost measurement, true accounting cost, that involves enumerating all the individual resources consumed in the production of a particular medical good or service and assigning market prices for each of them. The total cost is then the sum of the dollar costs for all the component resources. Even this calculation, however, may be prohibitively difficult in "real world" applications, for several reasons. First, all medical care requires not only the easily identifiable components of personnel time and disposable supplies but also the infrastructure components such as the rent on the office building where the care is provided, the cost of utilities, and the expense of an office staff. Second, even if all the components can be identified, enumeration of exactly what is used may be prohibitively expensive. Finally, medicine does not have publicly available "market prices" that can be readily obtained for a medical cost analysis, the way one can obtain prices for automobiles or refrigerators. The reasons for this relate to the lack of a true competitive free market in medicine along with the severe price distortion created in medical charges by cost-shifting practices. KEY COST TERMS Several key sets of cost terms are used in medicine. As the volume of health care produced is increased or decreased, costs may exhibit either variable or fixed "behavior." Variable costs change with each unit shift in production volume (up or down). For example, each vaccination administered to a group of children increases costs (related to the dose of vaccine and the disposable syringe) in a predictable linear fashion. Fixed costs do not shift with short-term changes in the volume of care provided. For example, the rent on the clinical building and the cost of heating, lighting, and so forth do not change according to the number of individuals vaccinated per day. Some types of costs display hybrid features of both variable and fixed components. For example, clinic personnel costs (e.g., nurses, secretaries) may be fixed if these personnel are paid a salary regardless of clinic volume. If the clinic volume goes up so much that evening hours must be added, either new personnel must be hired or existing personnel must work overtime. Either of these changes would graft a variable component onto the fixed personnel costs.
Marginal cost is a concept often used by economists to refer to the cost of producing one more unit of a given health care good or service. For example, the costs of doing one more or one less diagnostic cardiac catheterization would be its marginal cost. For all practical purposes, this is the same as its variable costs (since fixed costs do not change with small changes in volume). While the concept of unit changes in volume is theoretically interesting, a more pragmatic issue is the cost effect of changing a group of patients from one strategy to another. Many experts use the term incremental costs to refer to this type of shift (although some use marginal and incremental synonymously). Incremental analysis is a key component of cost-effectiveness analysis (see below). Another set of cost terms relates to the traceability of costs to the production of health care goods and services. Direct costs, such as nursing and physician personnel and disposable supplies, can be clearly linked to the health care provided and are under the control of the health care providers. Indirect costs, sometimes known as overhead, cannot. For example, the utility, laundry, maintenance, and administration costs of a hospital cannot be linked with the care of an individual patient and are generally not under the control of the physicians and nurses providing the medical care. The distinction of direct versus indirect is useful in cost-containment efforts, where the first step is to identify all major cost components and decide how they are to be controlled. One common error in the evaluation of medical costs is to focus on the cost of a test or therapy in isolation. Virtually every major medical management decision creates downstream consequences. For example, if physicians order a screening diagnostic test and the result is abnormal, they will need to do a confirmatory or more definitive test. If they order a potent new antibiotic and a fraction of patients develop liver failure as an unexpected toxicity, the total cost of that course of antibiotic includes not only the cost of the drug itself but also the costs of treating the liver failure in the fraction of patients who develop it. Extra costs added as a consequence of some diagnostic or therapeutic decision are referred to as induced costs. Similarly, if a management decision produces downstream savings, these would be referred to as induced savings. For example, administration of HMG CoA reductase inhibitors to patients with hypercholesterolemia can prevent future myocardial infarctions and revascularization procedures, both of which entail expensive hospitalizations. One final important cost concept relates to the societal costs of lost productivity (primarily lost time from work) due to illness. While economists often refer to these as indirect costs, confusion with the accounting concept of indirect costs (overhead) has led many to prefer the alternative term, productivity costs. COST MEASUREMENT Using varying degrees of simplification, medical costs can be measured using either bottom-up or top-down approaches. Bottom-up approaches build from component resources to calculate total cost for an episode or type of care. Microcosting is the gold standard approach. It involves careful enumeration of all resources consumed and detailed cost-accounting estimation of the costs for each component resource. A number of medical centers have now installed computer-based cost-accounting systems that perform a modified type of microcosting analysis. For difficult-to-obtain resource
use data (such as time required for a particular type of care by a given type of personnel), these systems use expert opinion in place of empirical data. The other extreme of the bottom-up category of approaches involves enumeration and costing for only the "big ticket" or expensive items, such as hospitalization episodes and costly tests and procedures. The top-down methods of medical cost estimation calculate a cost estimate from aggregated data. One such approach uses hospital billing charge data and charge-to-cost conversion ratios (which each hospital produces annually in its Medicare Cost Report) to estimate hospital costs. Despite the approximations involved, this approach, which can be used for most nonfederal U.S. hospitals, has provided good agreement with bottom-up estimates in the few instances where formal comparisons have been made. The other top-down approach is the use ofDRGassignments and reimbursement rates to provide standard cost weights for hospitalization episodes. COST-EFFECTIVENESS ANALYSIS Given a finite budget (for health care overall or for a particular health system), how can we use the available money to provide the most health benefits for our patients? For the clinician, who is less concerned with such policy issues, a prevalent question is whether a new treatment is economically attractive. The analysis method used to address this question is dependent on how the effectiveness and costs of the new therapy compare with those of "standard care" (Fig. 4-1). Cost-effectiveness analysis is used when effectiveness of the new treatment is greater and its costs are higher. This analysis calculates the ratio of added (or incremental) health benefits to added costs produced by a new therapy or strategy relative to some reference standard. The general formula is:
where C = costs and E= effectiveness. The cost-effectiveness ratio provides a quantitative statement of the amount of money required to produce a single extra unit of benefit with the new therapy relative to usual care or some other relevant reference standard. The benefit can be calculated in any meaningful clinical unit, such as added survivors or extra patients with a correct diagnosis. However, the vast majority of cost-effectiveness analyses use the epidemiologic concept of life-years to express incremental benefit. Virtually all benchmarks for cost effectiveness relate to this endpoint. Because some therapies affect quality of life but not quantity, a more generally relevant effectiveness measure combines qualify of life and life expectancy into a single composite metric, the quality-adjusted life year (QALY). Calculation of incremental dollars required to add an extra QALY is called cost-utility analysis. The QALY is a useful concept, but many details regarding measurement and interpretation remain controversial. The third form of economic efficiency analysis, cost-benefit analysis, requires conversion of health benefits into monetary equivalents. Because such conversions are controversial, this form of analysis is rarely used in medicine. In theory, the time horizon of a cost-effectiveness analysis should be long enough to capture all important cost and health consequences of the therapy or strategy being evaluated. Most often, analysts
use a lifetime time frame. Because very few empirical studies are long enough to observe lifetime outcomes (especially when chronic diseases are being studied), models are required to extrapolate from available data. A cost-effectiveness analysis can be done from a variety of perspectives, but the most widely applicable perspective is societal. Other perspectives are often much narrower and may include unattractive qualities. For example, a managed care organization may be interested only in short-term costs and outcomes, knowing that patients tend to change their health insurance every few years. The benchmarks for cost-effectiveness ratios are determined by comparison with other well-accepted therapies in widespread medical use. A useful benchmark is hemodialysis for chronic renal failure, since the federal government has paid for all renal failure patients to get dialysis since 1973 through the End Stage Renal Disease Program. Recent estimates are that it costs this Medicare program about $50,000 to add 1 life-year to a chronic renal failure patient. Partly for this reason, many analysts use a cost-effectiveness ratio of $100,000 per added life-year are deemed economically unattractive and therapies between $50,000 and $100,000 per added life-year are in the economic "gray zone." Several caveats about cost-effectiveness analysis should be noted. First, cost-effectiveness analysis is descriptive, not prescriptive. It measures value that could be produced with available health care dollars but does not mandate how these dollars are to be used. If an expensive new therapy is introduced and is found to be very economically attractive by the above benchmarks, it will still not get used if there is no money in the budget to pay for it. Second, a cost-effectiveness ratio is only as good as the data that were used to calculate it. High-quality results can be obtained if economic analysis is prospectively incorporated into the design of large-scale multicenter randomized trials. Third, although cost-effectiveness ratios are often presented as deterministic (i.e., no variability), they often incorporate large amounts of uncertainty. This should be examined either with sensitivity analyses (varying each key parameter through a plausible range to see if the results are materially changed) or calculation of confidence limits. MEDICAL ECONOMICS AND CLINICAL PRACTICE In evaluating new therapies, three issues must be addressed: (1) is the new therapy significantly better than what is currently available? (2) how much does it cost and is it economically attractive? and (3) how many patients will need this therapy and is it affordable? The clinician should be primarily concerned with the answer to the first question. Although cost issues are now a reality of daily clinical life and cost-containment pressures are often substantial, decisions by clinicians that are based primarily on economic rather than clinical considerations put the physician in the role of the double agent (i.e., acting on behalf of both the patient and the payer) and compromise our fiduciary obligation to patients. The second question addresses cost effectiveness and, if favorable, can be used to support an argument by clinicians for adoption of the therapy. In the ideal world, at least, therapies that have a large database
of evidence demonstrating effectiveness and economic attractiveness should be given preference over therapies that do not have such supporting data. The final question is of primary concern to payers and health policy analysts. An effective therapy that is too expensive to use is of little more value than a therapy that has yet to be discovered. (Bibliography omitted in Palm version) Back to Table of Contents
5. INFLUENCE OF ENVIRONMENTAL AND OCCUPATIONAL HAZARDS ON DISEASE - Howard Hu, Frank E. Speizer Exposures to hazardous materials and processes in the home, the workplace, and the community can cause or exacerbate a multitude of diseases. Physicians commonly treat the sequelae of such diseases in the practice of medicine; however, unless the underlying connection with hazardous exposures is identified and mitigated, treatment of manifestations rather than the cause at best only ameliorates the condition. At worst, the neglect of hazardous exposures may lead to both failure of treatment and failure to recognize a public health problem with wide significance. No existing surveillance or reporting system can estimate the total contribution of hazardous exposures to morbidity and mortality. However, careful histories have identified occupational factors as etiologic in more than 10% of all admissions to general internal medicine wards in hospitals, with even higher percentages when the primary illness is either respiratory or musculoskeletal. Estimates of the number of new cases of disease due to work in the United States range from 125,000 to 350,000 per year; these cases do not include 5.3 million work-related injuries. Environmental exposures are increasingly associated with decrements in measures of health whose outcomes range from subclinical to clinically catastrophic. For example, exposure to lead at levels that are common in the general population has been associated with increased blood pressure and decreased creatinine clearance. Ambient air pollution with respect to levels of ozone and fine-particulate matter has been related to increased rates of hospital admission for respiratory and cardiovascular diseases and to increased mortality rates, respectively. Indoor exposure to radon and passive indoor exposure to environmental tobacco smoke have been linked with an increased risk of lung cancer. There is pressure on clinicians to be aware of and act on this type of information, which is suggestive but not necessarily conclusive with respect to causation. Patients are becoming increasingly concerned about hazardous exposures. More than 15% of patients seen in one study conducted in a primary care clinic expressed the opinion that their health problems were work-related, and 75% of this subgroup of patients reported exposure to one or more recognized toxic agents. Patients often want answers to very specific questions, such as: Is the water in our town safe to drink? Could my breathing problem be related to the new roofing sealant used in my building at work? Physicians are consulted because they are the most trusted sources of information on health risks, including chemical risks. Unfortunately, few physicians have more than rudimentary training in environmental and occupational medicine. Therefore, it becomes important for primary care physicians to be able to recognize symptoms precipitated by exposure to environmental or occupational hazards and either to manage these cases or to make appropriate referrals. Many manifestations of exposure-related illnesses are nonspecific (e.g., dizziness, headache) or are commonly encountered in general internal medicine (e.g., myocardial infarction, cancer). The establishment of a connection with an environmental or occupational hazard requires a high index of suspicion and the application of fundamental concepts of environmental/occupational medicine. Furthermore, early
recognition by physicians of unusual patterns of illness or of evidence of asymptomatic exposure to toxins with low-level effects (e.g., an elevated blood lead level) can alert health officials to the need for control measures. Case reports either sent to local authorities or published in the literature often prompt follow-up studies that can lead to the identification of new hazards. In many states and countries, the reporting by physicians of occupational/environmental diseases is mandatory. For instance, beginning in 1992, physicians in Massachusetts were required to report cases of pneumoconiosis, occupational asthma, carpal tunnel syndrome, and carbon monoxide poisoning, among other conditions. Identification of an environmental/occupational etiology of an illness may have important economic ramifications for the patient (e.g., the awarding of worker's compensation, which covers medical bills as well as lost wages). Finally, physicians are frequently asked to provide expert medical testimony during litigation on the causal relationship between toxic exposures and diseases. In this setting, the more knowledgeable the physician is about potential hazardous exposures, the better prepared he or she is to serve the patient. THE ENVIRONMENTAL/OCCUPATIONAL HISTORY For a physician, the most critical steps toward recognizing these disorders are remembering to consider them in the differential diagnosis and taking an appropriate environmental/occupational history as part of the medical workup. The level of detail that is called for depends on the clinical situation. Information should always be obtained on current and major past occupations, and patients should be asked whether they think their health problem is related to their work or to any particular environment or exposure. In the review of systems, patients should be asked if they have been exposed to dusts, fumes, chemicals, radiation, or loud noise. When patient and physician are confronted with an illness of uncertain etiology, these factors should be explored in more detail, with the environmental/occupational history as the point of departure. (A brief outline of a sample history is shown in Table 5-1.) The identification of specific chemical exposures can be difficult. Household products must list chemical ingredients on their labels, and this information may prove useful. For workplace exposures, the U.S. Occupational Safety and Health Administration (OSHA) requires chemical suppliers to provide material safety data sheets with their products and requires employers to retain these sheets and make them available to employees. The data sheets can be obtained by the physician or employee by a telephoned or written request; failure of an employer to provide them within 30 days of such a request is a violation of OSHA regulations and is punishable by fines. In addition to providing information on chemical ingredients and percent composition, the material safety data sheets provide basic information on toxicity. This information is seldom adequate from a clinical perspective but may indicate the general type of toxicity to be anticipated. EVALUATION OF POSSIBLE CHEMICAL OR ENVIRONMENTAL HAZARDS Given the wide variety of toxic exposures that may be uncovered during a workup, a clinician should routinely consult additional reference material to evaluate whether particular hazards may be associated with the illness at hand. Many sources of information exist.OSHA and some regional poison-control centers have extensive information on hazards and brief summary documents that can be transmitted over the
Internet or by telephone or facsimile. Depending on the area, other resources may include county and state health departments; regional offices of the National Institute for Occupational Safety and Health and the Environmental Protection Agency; the Consumer Products Safety Commission in Washington, DC; academic institutions; websites of these institutions; and individual toxicologists, occupational/environmental medicine specialists, or industrial hygienists. Sophisticated computerized databases are also available, including detailed listings on CD-ROM information systems. MEDLARS, the electronic database maintained by the National Library of Medicine, is accessible by modem or the Internet and is familiar to many physicians. Files other than MEDLINE, such as the Hazardous Substances Databank, provide specific toxicity information on chemicals and include toxicologic references not covered by MEDLINE. Many of these databases can also be accessed through the Internet. As with any other illness, laboratory investigation may be crucial. For example, tests of carboxyhemoglobin level to document carbon monoxide exposure or of serum anticholinesterase level to document organophosphate pesticide absorption should be performed within hours of exposure. As in cases of acute drug overdose, it is useful to freeze samples of urine and serum from any patient suspected of having had an acute chemical exposure; such specimens can be analyzed at a later date by sensitive methods of detection. Use of other tests must rely on knowledge of the specific hazard or illness in question. SUSPICIOUS SCENARIOS Some medical problems or clinical scenarios demand a particularly high degree of suspicion of occupational or environmental factors as causative or contributing agents. Respiratory Disease The contribution of occupational/environmental factors to respiratory disease is generally underrecognized, particularly among patients who smoke and among the elderly (Chap. 254). For instance, asthma related to chemical exposure may be treated without regard to cause or may be erroneously diagnosed as acute tracheobronchitis. A study of new-onset asthma among HMO members in Massachusetts found that 21% of these individuals met criteria for clinically significant asthma attributable to occupational exposures. The types of exposures and jobs in these cases varied widely; examples include exposure to smoke in a firefighter, to welding fumes in a technical school student, to cleaning compounds in a bartender, and to epoxy in an archery repairman. No single type of job or exposure predominated. Other examples of etiologic errors include shortness of breath from asbestosis that is attributed to chronic obstructive pulmonary disease and chemical pneumonitis that is misdiagnosed as a bacterial infection. Cancer Many cancers are thought to be causally related to occupational and environmental factors in addition to tobacco. Some are particularly likely to have a chemical etiology or another environmental cause, including cancers of the skin (solar radiation, arsenic, coal tar, soot); lung (asbestos, arsenic, nickel, radon); pleura (almost exclusively asbestos); nasal cavity and sinuses (chromium, nickel, wood and leather dusts); liver (arsenic, vinyl chloride); bone marrow (benzene, ionizing radiation); and bladder (aromatic amines).
Coronary Disease and Hypertension Carbon monoxide exposure is common, particularly in homes with malfunctioning furnaces or in workplaces close to motor vehicle exhaust. By reducing oxygen transport by hemoglobin and inhibiting mitochondrial metabolism, carbon monoxide can aggravate coronary disease. Methylene chloride, a solvent used in paint stripping, is converted to carbon monoxide and thus poses the same risk. Exposure to carbon disulfide, a chemical used in the production of rayon, accelerates the rate of atherosclerotic plaque formation. Chronic lead exposure, even at modest levels, is a risk factor for the development of hypertension as well as abnormalities of cardiac conduction. Hepatitis/Chronic Liver Disease In the absence of evidence that a viral infection, alcohol ingestion, or drug use is the main cause of hepatitis (Chaps. 295,296, and297), the involvement of a toxin must be considered. Toxin-induced hepatic injury may be cytotoxic, cholestatic, or both. The list of hepatotoxic agents is long, including organic synthetic compounds such as carbon tetrachloride (used in solvents and cleaning fluids) and methylene diamine (a resin hardener); pesticides such as chlordecone (Kepone); metals, particularly arsenic (used in pesticides and paints and found in well water); and natural toxins such as the pyrrolidizine alkaloids. Kidney Disease Many chemical and environmental factors can cause renal injury (Chap. 269). The etiology of much chronic kidney disease, however, remains unknown. An increasing body of evidence now links chronic renal failure with hypertension to lead exposure. One study demonstrated that chelation therapy with EDTA slowed the progression of renal insufficiency in patients with a mildly elevated body lead burden. Some studies suggest that chronic exposure to hydrocarbons (e.g., gasoline, paints, solvents) may lead to various types of glomerulonephritis, including Goodpasture's syndrome. Environmental cadmium exposure has been found to promote calcium loss via urinary excretion, which results in skeletal demineralization and thus in an increased risk of fractures. Peripheral Neuropathy Organic solvents such as n-hexane, heavy metals such as lead and arsenic, and some organophosphate compounds can damage the axons of peripheral nerves. Dimethylaminopropionitrile, an industrial catalyst, causes bladder neuropathy. Nerve entrapment syndromes of the upper extremity, such as carpal tunnel syndrome, may be caused by jobs that involve repetitive motion, especially those requiring the maintenance of awkward positions. Central Nervous System Disorders Fatigue, memory loss, difficulty in concentration, and emotional lability have been linked to chronic exposure to solvents such as toluene and perchloroethylene. Painters, metal degreasers, plastics workers, and cleaners are commonly exposed to solvents and develop these symptoms at a high rate. Among the features that distinguish these patients are characteristic patterns on formal neurobehavioral testing and stabilization of symptoms with gradual improvement after discontinuation of the exposure. Other substances associated with neurobehavioral dysfunction include metals, particularly lead, mercury, arsenic, and manganese; pesticides, such as organophosphates and organochlorines; polychlorinated biphenyls (PCBs); and gases such as carbon monoxide. Environmental factors are also suspected of contributing to other neurologic diseases,
such as degenerative disorders, motor neuron diseases, and extrapyramidal disorders. For example, a study in monozygotic and dizygotic twin pairs found a similarity in concordance indicating that environmental (as opposed to genetic) factors play a major etiologic role in cases of typical Parkinson's disease beginning after the age of 50 years. Teratogenesis and Reproductive Problems Toxins can impair successful reproduction at a variety of levels. Examples include insecticides and herbicides,PCBs and polybrominated biphenyls (PBBs), ethylene oxide (a sterilizing gas used in hospitals), metals (lead, arsenic, cadmium, mercury), and solvents. Dibromochloropropane, a nematocide, suppresses spermatogenesis. Some toxins, such as PCBs, PBBs, and chlorinated pesticides, are concentrated in milk. Concern has arisen over the ability of specific organic pollutants, particularly pesticides, to persist in the environment and accumulate in human tissues. Some of these chemicals may disrupt endocrine function, and these effects may be related to phenomena such as the observed increases in the incidences of testicular cancer, breast cancer, and hypospadias. Immunosuppression, Autoimmunity, and Hypersensitivity Evidence is increasing that exposures to some chemical agents can compromise the immune system, thereby leading to a generalized increase in the incidence of tumors (e.g., exposure toPBBs) or infections (e.g., respiratory infections after exposure to common air pollutants). Mercury, dieldrin, and methylcholanthrene are known to elicit autoimmune responses. Some chemicals are potent allergic sensitizers that cause dermal and respiratory problems (Chaps. 60 and254). BIOLOGICAL MARKERS An increasing number of methods are available for measuring and interpreting toxic exposure, including (1) the internal dose of specific toxins and (2) markers of the biologic effects of toxins. Internal-dose markers are relevant for toxins that are sequestered in the human body, such as lead (in blood), arsenic (in hair), and other metals (Chap. 395), and for halogenated compounds (such asPCBs). Examples of markers of the biologic effects of toxins include depressed levels of acetylcholinesterase in serum after exposure to organophosphate pesticides, sister chromatid exchanges in peripheral lymphocytes after exposure to the carcinogen ethylene oxide, and DNA adducts after exposure to tobacco smoke carcinogens. MANAGING A HAZARD-RELATED ILLNESS Once a chemical or another environmental hazard has been identified as an important contributor to an illness, the next step is to prevent further exposure. Although for chronic diseases such as cancer this step may be irrelevant for the patient in question, prevention of further exposure may still be critical for other persons who have been similarly exposed. When prevention of further exposure is important, the physician must be willing to become an active advocate for the patient. This advocacy may involve writing a letter stating that the patient should no longer be exposed to a hazard or should remain out of work. Alternatively, it may involve contacting appropriate officials in government, industry, or labor or other advocates who can deal with a hazardous exposure. Treatment is dependent on the specific hazard.
In few areas of medicine does a physician deal with more scientific uncertainty. Comprehensive information on toxicants is available for only a small percentage of chemicals. In general, the physician should take a conservative approach (i.e., advise the patient to avoid a hazard likely to have contributed to illness) and should use common sense and up-to-date information to evaluate causal relationships. LOW-LEVEL EXPOSURES AND THEIR EFFECTS The subclinical effects of toxins that are widespread in our environment and our workplaces are of increasing concern. Given the absence of any demonstrable effect threshold, low-level exposure to carcinogens should be avoided; not only carcinogenic but also noncarcinogenic effects of chronic low-level exposure to these substances are important. Perhaps lead provides the most important example of low-level noncarcinogenic effects that constitute a major public health problem. Multiple pathways of exposure, including the combustion of leaded gasoline, the use of lead-based paints and solder, and the presence of lead in cans containing food, have contributed to exposure of the entire population. Such low-level exposures can impair neurobehavioral development in infants and children and can raise blood pressure in adults. Furthermore, absorbed lead is stored in the skeleton and may reenter the circulation at times of heightened bone turnover (e.g., pregnancy, lactation, osteoporosis, hyperthyroidism). Subclinical toxic effects can be prevented if chronic low-level exposure is detected early and curtailed. In the case of lead, such exposure is detected by tests of blood lead level, which should be performed regularly in young children living in old housing and as a precautionary measure in adults with a history of lead exposure. (Bibliography omitted in Palm version) Back to Table of Contents
6. WOMEN'S HEALTH - Anthony L. Komaroff, Celeste Robb-Nicholson, Andrea E. Dunaif In recent years, the medical problems and health care of women have received increasing attention. There are poorly understood differences between men and women, both in morbidity and mortality and in the expression of diseases. Many research studies of disease prevention and pathophysiology have included only male subjects; most illnesses that can affect both sexes have not been as well studied in women. It also appears that women receive different care than men for certain common health problems. Finally, an increasing number of women are seeking health care in multidisciplinary women's health units that combine the expertise of gynecology, psychiatry, and internal medicine or family medicine. MORBIDITY AND MORTALITY IN WOMEN Morbidity Past studies have found that women experience more days of restricted activity than men at all ages, over and above the restricted activity caused by obstetric and gynecologic conditions. However, a study in 1998 concluded differently. Women make more visits to physicians, particularly for acute self-limited illnesses. Mortality In the developed nations, women live longer than men. In the United States, as of 1996, the projected average life expectancy from birth is 79.1 years for females, and 73.1 years for males. Although there are more male fetuses conceived than female fetuses, females have a survival advantage when compared to males, in all age groups. The longer life expectancy of women versus men in developed countries is due in large part to the difference in mortality caused by ischemic heart disease (IHD). As shown inTable 6-1, the leading causes of death among young women in the United States are accidents, homicide, and suicide. During the middle years, breast cancer is a slightly more common cause of death thanIHD and lung cancer. In women between ages 65 and 74, IHD, lung cancer, and cerebrovascular disease supercede breast cancer as the leading causes of death. Among women of all ages, IHD is the leading cause of death by a substantial margin, with a mortality rate five to sixfold higher than the rate for either lung or breast cancer. Nevertheless, polls find that U.S. women believe breast cancer poses the greatest threat to their lives. Social Factors Influencing Morbidity and Mortality Gender differences in morbidity and mortality may be explained in part by psychosocial factors such as socially-defined gender roles, poverty, participation in the work force, health insurance, and lifestyle. In the past 30 years in the United States, there has been a "feminization of poverty." One-third of families headed by women currently live in poverty, and the fraction is greater than one-half for African-American and Latino women. Almost a fifth of women over age 65 live below the poverty level. People of lower socioeconomic status experience poorer health and a higher mortality rate than those in higher income groups. The poor are more likely to smoke and less likely to have recommended preventive measures, including cancer screening. Lack of adequate health insurance is a major problem for many women; in general, they are more likely than men to have low-paying, part-time, non-union jobs that do not provide health insurance. Women who
are divorced or widowed may also lose health insurance that they had through their husbands. PREVENTION (See alsoChap. 10) Primary prevention and screening are crucial elements in improving the health of women. Based upon available literature and the consensus of experts, various authoritative organizations have published guidelines on preventive practices in women. Most physicians believe that a baseline history and physical examination is useful to set the stage for preventive measures appropriate to each patient. In general, authorities recommend that blood pressure be measured every other year throughout life. Counseling on diet, smoking cessation, exercise, and use of seatbelts are of demonstrated value in the primary prevention of diseases and accidents. Counseling about safe sexual practices, alcohol abuse, and violence are also recommended. Screening for glaucoma is recommended for African-American women over age 40 and for Caucasian women over age 50. Yearly examinations to test visual acuity are recommended for women over age 70. Regular screening for breast, cervical, and colorectal cancer is recommended, but how often tests should be performed and which tools to use are still being debated. Most authorities recommend annual clinical breast examination in all women beginning at age 35 to 40. There is strong evidence to support the efficacy of annual mammography in women age 50 to 59. For women age 60 or older, the evidence for screening is less strong. The benefits of screening for women between the ages of 40 and 49 are still being debated. Most authorities recommend Pap smear screening beginning at age 18 or when a woman becomes sexually active. After two or three consecutive normal Pap smears, most groups recommend Pap smear testing every three years. If Pap smears have been normal for 10 years, they can be discontinued in women after age 65. Recommendations for colorectal cancer screening vary. For patients over 50, the American Cancer Society recommends yearly fecal occult blood testing and rectal examination combined with flexible sigmoidoscopy every 5 years, colonoscopy every 10 years, or double-contrast barium enema every 5 to 10 years. Bone mineral testing has gained rapid acceptance as a screening tool for detecting osteoporosis, as well as for predicting the likelihood of the condition in the future. With the advent of multiple preventive and therapeutic strategies for osteoporosis, many authorities now recommend bone mineral testing to screen for the condition. A bone mineral density test is recomended for all women over age 65 as well as for all postmenopausal women who are at increased risk for developing osteoporosis (Chap. 342). Cigarette smoking, a major risk factor for cardiovascular diseases and cancers in women, has been well studied (Chap. 390). Over the past 60 years there has been a sharp decline in smoking among men, but not among women; teenage women smoke at
higher rates than their male counterparts. "Low-yield" cigarettes are marketed heavily to women. The Nurses' Health Study showed that one-third of the excess risk of ischemic heart disease was eliminated two years after smoking cessation, and that all of the excess risk was eliminated by 10 to 14 years after smoking cessation. The National Cholesterol Education Program recommends that total cholesterol and high-density lipoprotein (HDL) levels be measured once. If both are normal, a repeat test after 5 years is recommended. A meta-analysis of several small studies of women showed an increased risk ofIHD in women with serum cholesterol greater than 265, a ratio of total cholesterol to HDL cholesterol greater than 4, or an elevated fasting triglyceride. In various case-control and observational studies, postmenopausal estrogen therapy is associated with a 40 to 50% reduction in deaths due toIHD, but its value in a prospective, randomized trial has not yet been documented. Calcium and estrogen, as well as alendronate and the selective estrogen receptor modulators, tamoxifen and raloxifene, slow the development of osteoporosis and reduce the frequency of hip and vertebral fracture in postmenopausal women. In randomized clinical trials, both tamoxifen and raloxifene have been shown to reduce the risk of breast cancer in postmenopausal women. Considerable research indicates that a relatively high dietary intake of various antioxidants (including vitamins E and C) is associated with lower rates of vascular disease and malignancies. Randomized trials of supplemental antioxidants are under way. Preliminary research indicates that regular aspirin use is associated with reduced rates ofIHD and colorectal carcinoma. GENDER DIFFERENCES IN DISEASE Obviously, some diseases and conditions occur exclusively (or nearly exclusively) in women -- e.g., menopause and various breast and gynecological disorders. These are discussed elsewhere in this book (Chaps. 52,89,336,337). In this chapter, we seek to highlight some gender differences in diseases that occur in both women and men. Ischemic Heart Disease (See also Chap. 244) Many persons think ofIHD as a primary problem for men rather than women, perhaps because men have more than twice the total incidence of cardiovascular morbidity and mortality between the ages of 35 and 84. However, as stated earlier, in the United States IHD is among the leading causes of death among women as well as men (Table 6-1). The curve for the IHD mortality rate in women lags behind that for men by about a decade. Nevertheless, nearly 250,000 women die annually from IHD; after age 40, one in three women will die from heart disease. Although IHD mortality has been falling in men in the United States over the past 30 years, it has been increasing in women. Why areIHDrates lower in women? They have a more favorable risk profile in some respects: higherHDLcholesterol levels, lower triglyceride levels, and less upper-body obesity than men. But women also have a less favorable risk profile in other respects: more obesity, higher blood pressure, higher plasma cholesterol levels, higher fibrinogen
levels, and more diabetes. The simplest explanation for the sex differential in IHD is the "cardioprotective" effect of estrogen, which can be due to improvement of the lipid profile, a direct vasodilatory effect, and perhaps other factors. HDL cholesterol levels appear to be a particularly important risk factor for IHD in women. HDL levels are higher in all age groups in women compared to men, and are higher in premenopausal and estrogen-treated postmenopausal women. Smoking is the most important risk factor for IHD in women. IHDpresents differently in men and women. In the Framingham study, angina was the most frequent initial symptom of IHD in females, occurring in 47% of women, whereas myocardial infarction was the most frequent initial symptom in males, occurring in 46% of men. The exercise electrocardiogram has a substantial false positive as well as false negative rate for women, compared to men. Women, particularly African-American women, have a higher risk of morbidity and mortality than men following a myocardial infarction. Compared to men, women who obtain coronary artery bypass graft surgery have more advanced disease, a higher perioperative mortality rate, less relief of angina, and less graft patency; however, 5and 10-year survival rates are similar. Women undergoing percutaneous transluminal coronary angioplasty have lower rates of clinical and angiographic success than men, but also a lower rate of restenosis and a better long-term outcome. Women may benefit less and have more frequent serious bleeding complications from thrombolytic therapy than do men. Factors such as older age, more comorbid conditions, and more severeIHD in women at the time of events or procedures appear to account for at least part of the gender differences observed. Women with IHD benefit at least as much as men, and perhaps more, from reductions in cholesterol level. The incidence ofIHDincreases markedly at menopause, consistent with the hypothesis that estrogens are cardioprotective. A number of observational studies have supported this hypothesis by demonstrating significant decreases in IHD in women on hormone replacement therapy (HRT), both estrogen alone and estrogen-progestin combination therapy. However, the HERS, a recent clinical trial of HRT for the secondary prevention of IHD, showed no significant difference in cardiovascular events between therapy with combined continuous conjugated equine estrogen (0.625 mg qd) and that with medroxyprogesterone acetate (2.5 mg qd), compared to placebo over four years. Indeed, in the HRT group, there was about a 50% increase in cardiovascular events in the first year of the trial. The Women's Health Initiative is investigating directly the impact of various HRT modalities as a primary prevention of IHD risk. Until further data are available, caution should be exercised in prescribing HRT to women with a history of IHD, or for cardioprotection alone. Hypertension (See alsoChap. 246) Hypertension is more common in U.S. women than men, largely owing to the high prevalence of hypertension in older age groups and the longer survival rate for women. Both the effectiveness and the adverse effects of various antihypertensive drugs appear to be comparable in women and men. Benefits of treatment for severe hypertension have been dramatic in both women and men. However, in clinical trials of the treatment of mild to moderate hypertension, women have had a smaller decrease in morbidity and mortality than men, perhaps because women have a lower risk of myocardial infarction and stroke than men to begin with.
Older women benefit at least as much as men from treatment, as demonstrated by the Systolic Hypertension in Elderly study. The incidence of hypertension (above 140/90) appears to be low (less than 5%) with the current low-dose oral contraceptives. Postmenopausal estrogen therapy is not associated with increases in blood pressure. Immunologically Mediated Diseases Several immunologically mediated diseases -e.g., rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis, Graves' disease, and thyroiditis -- occur much more frequently in women than in men. In animal models of rheumatoid arthritis -- lupus and multiple sclerosis, for example -- it is the females of the species that are predominantly affected. On the other hand, animal studies indicate that females are less susceptible to infection. In short, female animals appear to have more vigorous immune responses, with both beneficial and adverse consequences. Increasing evidence indicates that estrogens upregulate both cellular and humoral immunity. Also, some immunocytes contain estrogen, progestin and androgen receptors, and the uterus produces a variety of cytokines, suggesting a complex interaction between the reproductive and immune systems. Osteoporosis (See alsoChap. 342) This condition is much more prevalent in postmenopausal women than in men of similar age. Osteoporotic hip fractures are a major cause of morbidity in elderly women. Men accumulate more bone mass and lose bone more slowly than women. Gender differences in bone mass are found as early as infancy. Calcium intake, vitamin D and estrogen all play important roles in osteoporosis; calcium intake is an important determinant of peak bone mass, particularly during adolescence. Vitamin D deficiency is surprisingly common in elderly women. Receptors for estrogens and androgens have been identified in bone. The aromatase enzyme system, which converts androgens to estrogens, is also present in bone. Therapy withHRT, or with calcium and vitamin D, has been shown to reduce the risk of osteoporotic fractures. Newer modalities, such as bisphosphonates (alendronate), calcitonin, and raloxifene, a selective estrogen receptor modulator, prevent bone loss and reduce the risk of osteoporotic fractures. Alzheimer's Disease (See alsoChap. 362) Alzheimer's disease (AD) affects approximately twice as many women as men, in part because women live longer. Several observational studies suggest thatHRT may decrease the risk of AD and improve cognitive function in older women. These benefits are seen in both current as well as past HRT users. In a few experimental studies, estrogen replacement has been shown to be associated with improved memory compared to placebo treatment. Estrogens enhance neuronal growth and activity, providing a biologic basis for these putative cognitive effects of HRT. Prospective clinical trials, including the Women's Health Initiative, are underway to pursue these intriguing observations. Diabetes Mellitus (See alsoChap. 333) Estrogens enhance insulin sensitivity in women but not in men. Despite this, the prevalence of type 2 diabetes mellitus (DM) is higher in women, which is related in part to the higher prevalence of female obesity. Premenopausal women with DM lose the cardioprotective effect of female gender and have identical rates ofIHD to those in males. This is partially explained by the presence
of several IHD risk factors in women with DM: obesity, hypertension and dysplipidemia. Recent evidence suggests that vascular responses differ in women with DM, as compared to normal women. Polycystic ovary syndrome and gestational diabetes mellitus -- common conditions in premenopausal women -- are associated with a significantly increased risk for type 2 DM. Psychological Disorders (See also Chap. 385) Depression, anxiety panic disorder and eating disorders (bulimia and anorexia nervosa) occur more often in women than in men. Epidemiologic studies from both developed and developing nations consistently find major depression to be twice as common in women as in men, with the gender disparity becoming evident in early adolescence. Depression occurs in 10% of women during pregnancy and in 10 to 15% of women during the first several months of the postpartum period. The incidence of major depression diminishes after age 45, and does not increase with the onset of menopause. Depression in women also appears to have a worse prognosis than in men; episodes of depression last longer and there is a lower rate of spontaneous remission. Social factors may account for the greater prevalence of some disorders in women; the traditionally subordinate role of women in society may generate feelings of helplessness and frustration which contribute to psychiatric illness. In addition, it is likely that biological factors, including hormonally influenced neurochemical changes, also play a role. The limbic system and hypothalamus -- areas of the brain thought to subserve appetite, satiety and emotion -- contain estradiol and testosterone receptors. Alcohol and Drug Abuse (See alsoChap. 387) One-third of Americans who suffer from alcoholism are women. Women alcoholics are less likely to be diagnosed than men; a greater proportion of men than women seek help for alcohol and drug abuse. Men are more likely to go to an alcohol or drug treatment facility, while women tend to approach a primary care physician or mental health professional for help under the guise of a psychosocial problem. Late-life alcoholism is more common in women than men. In 1997, an epidemiologic survey reported that, among women over age 59, an estimated 1.8 million were addicted to or abused alcohol, and over 2.8 million were addicted to or abused psychoactive or mood-altering prescription drugs. On average, alcoholic women drink less than alcoholic men, but exhibit the same degree of impairment. Blood alcohol levels are higher in women than in men after drinking equivalent amounts of alcohol, adjusted for body weight. This greater bioavailability of alcohol in women is probably due to the higher proportion of body fat and lower total body water. Women also have a lower gastric "first-pass metabolism" of alcohol, associated with lower activity of gastric alcohol dehydrogenase. In addition, alcoholic women are more likely than alcoholic men to abuse tranquilizers, sedatives, and amphetamines. Women alcoholics have a higher mortality rate than do nonalcoholic women and alcoholic men. Compared to men, women also appear to develop alcoholic liver disease and other alcohol-related diseases with shorter drinking histories and lower levels of alcohol consumption. Alcohol abuse also poses special risks to women who are or wish to become pregnant, adversely affecting fertility and the health of the baby (fetal alcohol syndrome). Finally, there is growing evidence that for several illicit drugs, women proceed more
rapidly to drug dependence than do men. Human Immunodeficiency Virus Infection (See also Chap. 309) As of September 1998, the Centers for Disease Control and Prevention estimate that between 120,000 and 160,000 adolescent and adult women in the United States were living with HIV infection, including those with AIDS (Table 6-1). Between 1985 and 1998, the proportion of all U.S. AIDS cases reported among women more than tripled, from 7 to 23%. HIV infection was the fourth leading cause of death among U.S. women age 25 to 44 in 1997, and the second leading cause of death among African-American women in this age group. The CDC estimates that 30% of the approximately 40,000 new HIV infections in the United States each year are among women. Between 1996 and 1997 the incidence of new AIDS cases in the United States decreased by 18% and that of AIDS-related deaths by 42%, largely because of advances in HIV therapies. The decline continued between 1997 and 1998, albeit at a slower rate. AIDS incidence and AIDS-related mortality fell by 11 and 20%, respectively. However, AIDS incidence and deaths are not decreasing as rapidly among women as among men. HIV and AIDS continue to affect women in racial/ethnic minorities and lower socioeconomic classes disproportionately. CDC estimates that 64% of new HIV infections in 1998 occured among African-American women, 18% among Hispanic women, and 18% among white women. Of the new HIV infections among women in the United States in 1998, CDC estimates that 75% of women were infected through heterosexual sex and 25% of women through injection drug use. Violence Against Women Violence against women in the United States is an enormous problem. Incidents of both rape and domestic violence are vastly underreported. Sexual assault is one of the most common crimes against women. One in five adult women in the United States reports having experienced sexual assault during her lifetime. Adult women are much more likely to be raped by a spouse, ex-spouse, or acquaintance than by a stranger. Domestic violence is defined in the American Medical Association guidelines as "an ongoing, debilitating experience of physical, psychologic, and/or sexual abuse in the home, associated with increasing isolation from the outside world and limited personal freedom and accessibility to resources." It affects women of all ages, ethnic orientations, and socioeconomic groups. Based upon national crime statistics, every year an estimated 2 million women in the United States are severely injured and more than 1000 are killed by their current or former male partner. Domestic violence is the most common cause of physical injury in women, exceeding the combined incidence of all other types of injury (such as from rape, mugging, and auto accidents). Women who are young, single, pregnant, recently separated or divorced, or who have a history of substance abuse or mental illness, or a partner with substance abuse or mental illness, are at increased risk of domestic violence. Domestic violence and sexual assault are associated with increased rates of physical and psychologic symptoms, medical office visits, and hospitalizations. Given this indirect presentation of the consequences of violence, and the high prevalence of unreported violence, clinicians should have a low threshold for pursuing the possibility of violence in female patients, particularly those with vague symptoms and psychological disorders.
The immediate treatment of rape and domestic violence focuses on assessing and treating physical injuries; providing emotional support; assessing and dealing with the risks of sexually transmitted infection and pregnancy; evaluating the safety of the patient and other family members; and documenting the patient's history and physical examination findings. In addition to dealing with the medical and psychological issues, appropriate care includes providing information about legal services, shelters and safe houses, hotlines, support groups, and counseling services. RESEARCH IN WOMEN'S HEALTH The growing recognition of the importance of women's health has spawned a number of research efforts, including large observational studies and clinical trials. The U.S. National Institutes of Health has introduced guidelines to mandate the inclusion of women in clinical studies, and the reporting of gender-specific data. Studies of Prevention Large observational studies of men and women, such as the Rancho Bernardo Study and the Framingham Study, designed to analyze data specific to women have been on the increase. The Nurses' Health Study has been following more than 200,000 women, many for more than 20 years, prospectively collecting data to study the impact of smoking, diet, physical activity, medications, prevention and screening behaviors, and some psychosocial factors on the risk of various medical disorders, including breast cancer,IHD, stroke, diabetes, and fracture, as well as causes of mortality. These studies have set the stage for clinical trials such as the Postmenopausal Estrogens/Progestins Intervention (PEPI) Trial, the first multicenter, randomized, double-blind, placebo-control trial of the effects of three estrogen/progestin regimens on risk factors for cardiovascular disease, bone mineral density, and endometrial hyperplasia. The study found that estrogen, alone or in combination with progestin, increased serum levels ofHDL and decreased low-density lipoprotein (LDL) and fibrinogen levels. While unopposed estrogen (without progestins) resulted in the most beneficial effects on lipids, it was also associated with an increased risk of endometrial hyperplasia. In 1992, the NIH funded the Women's Health Initiative (WHI), a study of the health of postmenopausal women. The WHI, the largest research study ever funded by the NIH, involves over 160,000 postmenopausal women participating at 45 clinical centers across the United States through the year 2002. The WHI study includes both a prospective observational study and an interventional randomized trial involving over 63,000 women, which is designed to test the effects of a low-fat diet, hormone replacement therapy, and calcium and vitamin D supplementation on the risks for cardiovascular disease, breast cancer, and osteoporotic fractures. Many other studies currently in progress promise new insights into the health of women within the next decade. Pharmacologic Studies Historically, women have been underrepresented in drug trials, even though the majority of pharmaceuticals sold in the United States each year
are used by women. However, this has been rapidly changing. The FDA requires information on the safety and effectiveness of experimental drugs in women, on the effects of the menstrual cycle and menopause on a drug's pharmacokinetics, and on a drug's influence on the effectiveness of oral contraceptives. The increased emphasis on entering women into drug trials is likely to yield important information. Studies that have included women indicate that there are clinically significant differences in the way women respond to a number of frequently prescribed pharmaceuticals, including sedative-hypnotics, antidepressants, antipsychotics, anticonvulsants, and b-adrenergic blocking agents. The 1992 FDA Adverse Experience Report found that women have a higher frequency of adverse drug reactions than men. Other studies suggest that the efficacy of many drugs may be different in women compared to men. For example, women require lower doses of neuroleptics to control schizophrenia than men do. Women awaken from anesthesia faster than do men who are given the same doses of anesthetics, and they have a more powerful response to certain classes of analgesics than men. The reasons for these differences are not clear. However, these observations have spurred researchers to consider separating out the effects of gender in future clinical research in an effort to define "gender-based" biologic processes. CONCLUSION At the same time that the health of women is undergoing more rigorous study and women's clinics are becoming increasingly common and popular, a growing fraction of health professionals are women. The number of women physicians has increased by 300% between 1970 and 1990, and more than 40% of all U.S. medical students now are women. This infusion of women into the physician work force is likely to lead to a still greater recognition of the unique aspects of health and disease in women. (Bibliography omitted in Palm version) Back to Table of Contents
7. MEDICAL DISORDERS DURING PREGNANCY - Robert L. Barbieri, John T. Repke Approximately 4 million births occur in the United States each year. A significant proportion of these are complicated by one or more medical disorders. Two decades ago, many medical disorders were contraindications to pregnancy. Advances in obstetrics, neonatology, obstetric anesthesiology, and medicine have increased the expectation that pregnancy will result in an excellent outcome for both mother and fetus despite most of these conditions. Successful pregnancy requires important physiologic adaptations, such as a marked increase in cardiac output. Medical problems that interfere with the physiologic adaptations of pregnancy increase the risk for poor pregnancy outcome; conversely, in some instances pregnancy may adversely impact an underlying medical disorder. HYPERTENSION (See also Chap. 246) In pregnancy, cardiac output increases by 40%, most of which is due to an increase in stroke volume. Heart rate increases by approximately 10 beats per minute during the third trimester. In the second trimester of pregnancy, systemic vascular resistance decreases and this is associated with a fall in blood pressure. During pregnancy, a blood pressure of 140/90 mmHg is considered to be abnormally elevated and is associated with a marked increase in perinatal morbidity and mortality. In all pregnant women, the measurement of blood pressure should be performed in the sitting position, because for many the lateral recumbent position is associated with a blood pressure lower than that recorded in the sitting position. The diagnosis of hypertension requires the measurement of two elevated blood pressures, at least 6 h apart. Hypertension during pregnancy is usually caused by preeclampsia, chronic hypertension, gestational hypertension, or renal disease. PREECLAMPSIA Approximately 5 to 7% of all pregnant women develop preeclampsia, the new onset of hypertension (blood pressure > 140/90 mmHg), proteinuria (>300 mg per 24 h), and pathologic edema. Although the precise placental factors that cause preeclampsia are unknown, the end result is vasospasm and endothelial injury in multiple organs. Preeclampsia is associated with abnormalities of cerebral circulatory autoregulation, which increase the risk of stroke at near-normal blood pressures. Risk factors for the development of preeclampsia include nulliparity, diabetes mellitus, a history of renal disease or chronic hypertension, a prior history of preeclampsia, extremes of maternal age (>35 years or 160/110 mmHg), severe proteinuria (>5 g per 24 h), oliguria or renal failure, pulmonary edema, hepatocellular injury (ALT >2´ the upper limits of normal), thrombocytopenia (platelet count < 100,000/uL), or disseminated intravascular coagulation. Women with mild preeclampsia are those with the diagnosis of new-onset hypertension, proteinuria, and edema without evidence of severe preeclampsia. The HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome is a special subgroup of severe preeclampsia and is a major cause of morbidity and mortality in this disease. The presence of platelet dysfunction and coagulation disorders further increases the risk of stroke. TREATMENT Preeclampsia resolves within a few weeks after delivery. For pregnant women with preeclampsia prior to 37 weeks' gestation, delivery reduces the mother's morbidity but exposes the fetus to the risk of premature delivery. The management of preeclampsia is challenging because it requires the clinician to balance the health of both mother and fetus simultaneously and to make management decisions that afford both the best opportunities for infant survival. In general, prior to term, women with mild preeclampsia can be managed conservatively with bed rest, close monitoring of blood pressure and renal function, and careful fetal surveillance. For women with severe preeclampsia, delivery is recommended after 32 weeks' gestation. This reduces maternal morbidity and slightly increases the risks associated with prematurity for the newborn. Prior to 32 weeks' gestation, the risks of prematurity for the fetus are great, and some authorities recommend conservative management to allow for continued fetal maturation. Expectant management of severe preeclampsia remote from term affords some benefits for the fetus with significant risks for the mother. Such management should be restricted to tertiary care centers where maternal-fetal medicine, neonatal medicine, and critical care medicine expertise are available. The definitive treatment of preeclampsia is delivery of the fetus and placenta. For women with severe preeclampsia, aggressive management of blood pressures> 160/110 mmHg reduces the risk of cerebrovascular accidents. Intravenous labetalol or hydralazine are the drugs most commonly used to manage preeclampsia. Alternative agents such as calcium channel blockers may be used. Elevated arterial pressure should be reduced slowly to avoid hypotension and a decrease in blood flow to the fetus. Angiotensin-converting enzyme (ACE) inhibitors as well as angiotensin-receptor blockers should be avoided in the second and third trimesters of pregnancy because of their adverse effects on fetal development. Pregnant women treated with ACE inhibitors often develop oligohydramnios, which may be caused by decreased fetal renal function. Magnesium sulfate is the treatment of choice for the prevention and treatment of
eclamptic seizures. Two large randomized clinical trials have demonstrated the superiority of magnesium sulfate over phenytoin and diazepam. Magnesium may prevent seizures by interacting with N-methyl-D-asparate (NMDA) receptors in the central nervous system. Given the difficulty of predicting eclamptic seizures on the basis of disease severity, it is recommended that once the decision to proceed with delivery is made, all patients carrying a diagnosis of preeclampsia be treated with magnesium sulfate (seeGuideline). CHRONIC ESSENTIAL HYPERTENSION Pregnancy complicated by chronic essential hypertension is associated with intrauterine growth restriction and increased perinatal mortality. Pregnant women with chronic hypertension are at increased risk for superimposed preeclampsia and abruptio placenta. Women with chronic hypertension should have a thorough prepregnancy evaluation, both to identify remediable causes of hypertension and to ensure that the prescribed antihypertensive agents are not associated with adverse pregnancy outcome (e.g.,ACEinhibitors, angiotensin-receptor blockers).a-Methyldopa and labetalol are the most commonly used medications for the treatment of chronic hypertension in pregnancy. Baseline evaluation of renal function is necessary to help differentiate the effects of chronic hypertension versus superimposed preeclampsia should the hypertension worsen during pregnancy. There are no convincing data that demonstrate that treatment of mild chronic hypertension improves perinatal outcome. GESTATIONAL HYPERTENSION This is the development of elevated blood pressure during pregnancy or in the first 24 h post partum in the absence of preexisting chronic hypertension and other signs of preeclampsia. Uncomplicated gestational hypertension that does not progress to preeclampsia has not been associated with adverse pregnancy outcome or adverse long-term prognosis. RENAL DISEASE (See also Chap. 268) Normal pregnancy is characterized by an increase in glomerular filtration rate and creatinine clearance. This occurs secondary to a rise in renal plasma flow and increase glomerular filtration pressures. Patients with underlying renal disease and hypertension may expect a worsening of hypertension during pregnancy. If superimposed preeclampsia develops, the additional endothelial injury results in a capillary leak syndrome that may make the management of these patients challenging. In general, patients with underlying renal disease and hypertension benefit from more aggressive management of blood pressure than do those with gestational hypertension. Preconception counseling is also essential for these patients so that accurate risk assessment can occur prior to the establishment of pregnancy and important medication changes and adjustments be made. In general, a prepregnancy serum creatinine level 75% relief of back and leg pain. Up to 25% develop recurrent stenosis at the same spinal level or an adjacent level 5 years after the initial surgery; recurrent symptoms usually respond to a second surgical decompression. Facet joint hypertrophy can produce unilateral radicular symptoms, due to bony compression, that are indistinguishable from disk-related radiculopathy. Patients may exhibit stretch signs, focal motor weakness, hyporeflexia, or sensory loss. Hypertrophic superior or inferior facets can often be visualized radiologically. Foraminotomy results in long-term relief of leg and back pain in 80 to 90% of patients. Lumbar adhesive arachnoiditis with radiculopathy is the result of a fibrotic process following an inflammatory response to local tissue injury within the subarachnoid space. The fibrosis results in nerve root adhesions, producing back and leg pain associated with motor, sensory, and reflex changes. Myelography-induced arachnoiditis has become rare with the abandonment of oil-based contrast. Other causes of arachnoiditis include multiple lumbar operations, chronic spinal infections, spinal cord injury, intrathecal hemorrhage, intrathecal injection of steroids and anesthetics, and foreign bodies. The spineMRIappearance of arachnoiditis includes nerve roots clumping together centrally and adherent to the dura peripherally, or loculations of cerebrospinal fluid (CSF) within the thecal sac that obscure nerve root visualization. Treatment is often unsatisfactory. Microsurgical lysis of adhesions, dorsal rhizotomy, and dorsal root ganglionectomy have resulted in poor outcomes. Dorsal column stimulation for pain relief has produced varying results. Epidural steroid injections have been of limited value. ARTHRITIS Arthritis is a major cause of spine pain.
Spondylosis Osteoarthritic spine disease typically occurs in later life and primarily involves the cervical and lumbosacral spine. Patients often complain of back pain that is increased by motion and associated with stiffness or limitation of motion. The relationship between clinical symptoms and radiologic findings is usually not straightforward. Pain may be prominent when x-ray findings are minimal; alternatively, large osteophytes can be seen in asymptomatic patients in middle and later life. Hypertrophied facets and osteophytes may compress nerve roots in the lateral recess or intervertebral foramen. Osteophytes arising from the vertebral body may cause or contribute to central spinal canal stenosis. Loss of intervertebral disk height reduces the vertical dimensions of the intervertebral foramen; the descending pedicle may compress the nerve root exiting at that level. Osteoarthritic changes in the lumbar spine may rarely compress the cauda equina. Ankylosing Spondylitis (See alsoChap. 315) This distinctive arthritic spine disease typically presents with the insidious onset of low back and buttock pain. Patients are often males below age 40. Associated features include morning back stiffness, nocturnal pain, pain unrelieved by rest, an elevated sedimentation rate, and the histocompatibility antigen HLA-B27. The differential diagnosis includes tumor and infection. Onset at a young age and back pain characteristically improving with exercise suggest ankylosing spondylitis. Loss of the normal lumbar lordosis and exaggeration of thoracic kyphosis are seen as the disease progresses. Inflammation and erosion of the outer fibers of the annulus fibrosus at the point of contact with the vertebral body are followed by ossification and bone growth. Bony growth (syndesmophyte) bridges adjacent vertebral bodies and results in reduced spine mobility in all planes. The radiologic hallmarks of the disease are periarticular destructive changes, sclerosis of the sacroiliac joints, and bridging of vertebral bodies by bone to produce the fused "bamboo spine." Similar restricted movement may accompany Reiter's syndrome, psoriatic arthritis, and chronic inflammatory bowel disease. Stress fractures through the spontaneously ankylosed posterior bony elements of the rigid, osteoporotic spine may result in focal spine pain, spinal cord compression or cauda equina syndrome. Occasional atlantoaxial subluxation with spinal cord compression occurs. Bilateral ankylosis of the ribs to the spine and a decrease in the height of axial thoracic structures may cause marked impairment of respiratory function. OTHER DESTRUCTIVE DISEASES Neoplasm (See alsoChap. 370) Back pain is the most common neurologic symptom among patients with systemic cancer. One-third of patients with undiagnosed back or neck pain and known systemic cancer have epidural extension or metastasis of tumor, and one-third have pain associated with vertebral metastases alone. About 11% have back pain unrelated to metastatic disease. Metastatic carcinoma (breast, lung, prostate, thyroid, kidney, gastrointestinal tract), multiple myeloma, and non-Hodgkin's and Hodgkin's lymphomas frequently involve the spine. Back pain may be the presenting symptom because the primary tumor site may be overlooked or asymptomatic. The pain tends to be constant, dull, unrelieved by rest, and worse at night. In contrast, mechanical low back pain is usually improved with rest. Plain x-rays usually, though not always, show destructive lesions in one or several vertebral bodies without disk space involvement.MRI orCT-myelography are the studies of choice in the setting of suspected spinal metastasis, but the trend of evidence favors the use of MRI. The procedure of
choice is the study most rapidly available because the patient may worsen during a diagnostic delay. Infection Vertebral osteomyelitis is usually caused by staphylococci, but other bacteria or the tubercle bacillus (Pott's disease) may be the responsible organism. A primary source of infection, most often from the urinary tract, skin, or lungs, can be identified in 40% of patients. Intravenous drug use is a well-recognized risk factor. Back pain exacerbated by motion and unrelieved by rest, spine tenderness over the involved spine segment, and an elevated erythrocyte sedimentation rate are the most common findings. Fever or elevated white blood cell count are found in a minority of patients. Plain radiographs may show a narrowed disk space with erosion of adjacent vertebrae; these diagnostic changes may take weeks or months to appear.MRI andCT are sensitive and specific for osteomyelitis; MRI definition of soft tissue detail is exquisite. CT scan may be more readily available and better tolerated by some patients with severe back pain. Spinal epidural abscess (Chap. 368) presents with back pain (aggravated by palpation or movement) and fever. The patient may exhibit nerve root injury or spinal cord compression accompanied by a sensory level, incontinence, or paraplegia. The abscess may track over multiple spinal levels and is best delineated by spineMRI. Osteoporosis and Osteosclerosis Considerable loss of bone may occur with or without symptoms in association with medical disorders, including hyperparathyroidism, chronic glucocorticoid use, or immobilization. Compression fractures occur in up to half of patients with severe osteoporosis. The risk of osteoporotic vertebral fracture is 4.5 times greater over 3 years among patients with a baseline fracture compared with osteoporotic controls. The sole manifestation of a compression fracture may be focal lumbar or thoracic aching (often after a trivial injury) that is exacerbated by movement. Other patients experience thoracic or upper lumbar radicular pain. Focal spine tenderness is common. When compression fractures are found, treatable risk factors should be sought. Compression fractures above the midthoracic region suggest malignancy. Osteosclerosis is readily identifiable on routine x-ray studies (e.g., Paget's disease) and may or may not produce back pain. Spinal cord or nerve root compression may result from bony encroachment on the spinal canal or intervertebral foramina. Single dual-beam photon absorptiometry or quantitativeCT can be used to detect small changes in bone mineral density.*For further discussion of these bone disorders, see Chaps. 341 to 343. REFERRED PAIN FROM VISCERAL DISEASE Diseases of the pelvis, abdomen, or thorax may produce referred pain to the posterior portion of the spinal segment that innervates the diseased organ. Occasionally, back pain may be the first and only sign. In general, pelvic diseases refer pain to the sacral region, lower abdominal diseases to the lumbar region (around the second to fourth lumbar vertebrae), and upper abdominal diseases to the lower thoracic or upper lumbar region (eighth thoracic to the first and second lumbar vertebrae). Local signs (pain with spine palpation, paraspinal muscle spasm) are absent, and minimal or no pain
accompanies normal spine movements. Low Thoracic and Upper Lumbar Pain in Abdominal Disease Peptic ulcer or tumor of the posterior stomach or duodenum typically produces epigastric pain (Chaps. 285 and90), but midline back or paraspinal pain may occur if retroperitoneal extension is present. Back pain due to peptic ulcer may be precipitated by ingestion of an orange, alcohol, or coffee and relieved by food or antacids. Fatty foods are more likely to induce back pain associated with biliary disease. Diseases of the pancreas may produce back pain to the right of the spine (head of the pancreas involved) or to the left (body or tail involved). Pathology in retroperitoneal structures (hemorrhage, tumors, pyelonephritis) may produce paraspinal pain with radiation to the lower abdomen, groin, or anterior thighs. A mass in the iliopsoas region often produces unilateral lumbar pain with radiation toward the groin, labia, or testicle. The sudden appearance of lumbar pain in a patient receiving anticoagulants suggests retroperitoneal hemorrhage. Isolated low back pain occurs in 15 to 20% of patients with a contained rupture of an abdominal aortic aneurysm (AAA). The classic clinical triad of abdominal pain, shock, and back pain in an elderly man occurs in fewer than 20% of patients. Two of these three features are present in two-thirds of patients, and hypotension is present in half. Ruptured AAA has a high mortality rate; the typical patient is an elderly male smoker with back pain. The diagnosis is initially missed in at least one-third of patients because the symptoms and signs can be nonspecific. Common misdiagnoses include nonspecific back pain, diverticulitis, renal colic, sepsis, and myocardial infarction. A careful abdominal examination revealing a pulsatile mass (present in 50 to 75% of patients) is an important physical finding. Lumbar Pain with Lower Abdominal Diseases Inflammatory bowel disorders (colitis, diverticulitis) or colonic neoplasms may produce lower abdominal pain, midlumbar back pain, or both. The pain may have a beltlike distribution around the body. A lesion in the transverse or initial descending colon may refer pain to the middle or left back at the L2-L3 level. Sigmoid colon disease may refer pain to the upper sacral or midline suprapubic regions or left lower quadrant of the abdomen. Sacral Pain in Gynecologic and Urologic Disease Pelvic organs rarely cause low back pain, except for gynecologic disorders involving the uterosacral ligaments. The pain is referred to the sacral region. Endometriosis or uterine carcinoma may invade the uterosacral ligaments; malposition of the uterus may cause uterosacral ligament traction. The pain associated with endometriosis begins during the premenstrual phase and often continues until it merges with menstrual pain. Malposition of the uterus (retroversion, descensus, and prolapse) may lead to sacral pain after standing for several hours. Menstrual pain may be felt in the sacral region. The poorly localized, cramping pain can radiate down the legs. Other pelvic sources of low back pain include neoplastic invasion of pelvic nerves, radiation necrosis, and pregnancy. Pain due to neoplastic infiltration of nerves is typically continuous, progressive in severity, and unrelieved by rest at night. Radiation therapy of pelvic tumors may produce sacral pain from late radiation necrosis of tissue or nerves. Low back pain with radiation into one or both thighs is common in the last weeks of pregnancy.
Urologic sources of lumbosacral back pain include chronic prostatitis, prostate carcinoma with spinal metastasis, and diseases of the kidney and ureter. Lesions of the bladder and testes do not usually produce back pain. The diagnosis of metastatic prostate carcinoma is established by rectal examination, spine imaging studies (MRI orCT), and measurement of prostate-specific antigen (PSA) (Chap. 95). Infectious, inflammatory, or neoplastic renal diseases may result in ipsilateral lumbosacral pain, as can renal artery or vein thrombosis. Ureteral obstruction due to renal stones may produce paraspinal lumbar pain. Postural Back Pain There is a group of patients with chronic, nonspecific low back pain in whom no anatomic or pathologic lesion can be found despite exhaustive investigation. These individuals complain of vague, diffuse back pain with prolonged sitting or standing that is relieved by rest. The physical examination is unrevealing except for "poor posture." Imaging studies and laboratory evaluations are normal. Exercises to strengthen the paraspinal and abdominal muscles are sometimes therapeutic. Psychiatric Disease Chronic low back pain (CLBP) may be encountered in patients with compensation hysteria, malingering, substance abuse, chronic anxiety states, or depression. Many patients with CLBP have a history of psychiatric illness (depression, anxiety, substance abuse) or childhood trauma (physical or sexual abuse) that antedates the onset of back pain. Preoperative psychological assessment has been used to exclude patients with marked psychological impairment who are at high risk for a poor surgical outcome. It is important to be certain that the back pain in these patients does not represent serious spine or visceral pathology in addition to the impaired psychological state. Unidentified The cause of low back pain occasionally remains unclear. Some patients have had multiple operations for disk disease but have persistent pain and disability. The original indications for surgery may have been questionable with back pain only, no definite neurologic signs, or a minor disk bulge noted onCT orMRI. Scoring systems based upon neurologic signs, psychological factors, physiologic studies, and imaging studies have been devised to minimize the likelihood of unsuccessful surgical explorations and to avoid selection of patients with psychological profiles that predict poor functional outcomes. TREATMENT Acute Low Back Pain A practical approach to the management of low back pain is to consider acute and chronic presentations separately.ALBP is defined as pain of less than 3 months' duration. Full recovery can be expected in 85% of adults with ALBP unaccompanied by leg pain. Most of these patients exhibit "mechanical" symptoms -pain that is aggravated by motion and relieved by rest. Observational, population-based studies have been used to justify a minimalist approach to individual patient care. These studies share a number of limitations: (1) a true placebo control group is often lacking; (2) patients who consult different provider groups (generalists, orthopedists, neurologists) are assumed to have similar etiologies
for their back pain; (3) no information is provided about the details of treatment within each provider group or between provider groups; and (4) no attempt to tabulate serious causes ofALBP is made. The appropriateness of specific diagnostic procedures or therapeutic interventions for low back pain cannot be assessed from these studies. The proposed algorithms (Fig. 16-6) for management ofALBP in adults draw considerably from published guidelines. However, it must be emphasized that currentCPGs for the treatment of low back pain are based on incomplete evidence -- for example, there is a paucity of well-designed studies documenting the natural history of disk lesions associated with a focal neurologic deficit. Guidelines should not substitute for sound clinical judgment. The initial assessment excludes serious causes of spine pathology that require urgent intervention, including infection, cancer, and trauma. Risks factors for a possible serious underlying cause of back pain include: age > 50 years, prior diagnosis of cancer or other serious medical illness, bed rest without relief, duration of pain >1 month, urinary incontinence or recent nocturia, focal leg weakness or numbness, pain radiating into the leg(s) from the back, intravenous drug use, chronic infection (pulmonary or urinary), pain increasing with standing and relieved by sitting, history of spine trauma, and glucocorticoid use. Clinical signs associated with a possible serious etiology include unexplained fever, well-documented and unexplained weight loss, positiveSLR sign or reverse SLR sign, crossed SLR sign, percussion tenderness over the spine or costovertebral angle, an abdominal mass (pulsatile or nonpulsatile), a rectal mass, focal sensory loss (saddle anesthesia or focal limb sensory loss), true leg weakness, spasticity, and asymmetric leg reflexes. Laboratory studies are unnecessary unless a serious underlying cause (Fig. 16-6, Algorithms A and B) is suspected. Plain spine films are rarely indicated in the first month of symptoms unless a spine fracture is suspected. The roles of bed rest, early exercise, and traction in the treatment of acute uncomplicated low back pain have been the subject of recent prospective studies. Clinical trials fail to demonstrate any benefit of prolonged (>2 days) bed rest forALBP. There is evidence that bed rest is also ineffective for patients with sciatica or for acute back pain with findings of nerve root injury. Theoretical advantages of early ambulation for ALBP include maintenance of cardiovascular conditioning, improved disk and cartilage nutrition, improved bone and muscle strength, and increased endorphin levels. A recent trial did not show benefit from an early vigorous exercise program, but the benefits of less vigorous exercise or other exercise programs remain unknown. The early resumption of normal physical activity (without heavy manual labor) is likely to be beneficial. Well-designed clinical studies of traction that include a sham traction group have failed to show a benefit of traction for ALBP. Despite this knowledge, one survey of physicians' perceptions of effective treatment identified strict bed rest for >3 days, trigger point injections (see below), and physical therapy (PT) as beneficial for more than 50% of patients with ALBP. In many instances, the behavior of treating physicians does not reflect the current medical literature. Proof is lacking to support the treatment of acute back and neck pain with acupuncture, transcutaneous electrical nerve stimulation, massage, ultrasound, diathermy, or electrical stimulation. Cervical collars can be modestly helpful by limiting spontaneous and reflex neck movements that exacerbate pain. Evidence regarding the efficacy of ice
or heat is lacking, but these interventions are optional given the lack of negative evidence, low cost, and low risk. Biofeedback has not been studied rigorously. Facet joint, trigger point, and ligament injections are not recommended in the treatment ofALBP. A beneficial role for specific exercises or modification of posture has not been validated by rigorous clinical studies. As a practical matter, temporary suspension of activity known to increase mechanical stress on the spine (heavy lifting, prolonged sitting, bending or twisting, straining at stool) may be helpful. Patient education is an important part of treatment. Studies reveal that patient satisfaction and the likelihood of follow-up increase when patients are educated about prognosis, treatment methods, activity modifications, and strategies to prevent future exacerbations. In one study, patients who felt they did not receive an adequate explanation for their symptoms wanted more diagnostic tests. Evidence for the efficacy of structured education programs ("back school") is inconclusive; in one controlled study, patients attending back school had a shorter duration of sick leave during the initial episode but not during subsequent episodes. Recent large, controlled, randomized studies of back school for primary prevention of low back injury and pain have failed to demonstrate a benefit. Medications used in the treatment ofALBPincludeNSAIDs, acetaminophen, muscle relaxants, and opioids. NSAIDs are superior to placebo for back pain relief. Acetaminophen is superior to placebo in the treatment of other types of pain but has not been compared against placebo for low back pain. Muscle relaxants provide short-term (4 to 7 days) benefit compared with placebo, but drowsiness often limits their daytime use. The efficacy of muscle relaxants compared to NSAIDs or in combination with NSAIDs is unclear. Opioid analgesics have not been shown to be more effective than NSAIDs or acetaminophen for relief of ALBP or likelihood of return to work. Short-term use of opioids in selected patients unresponsive to or intolerant of acetaminophen or NSAIDs may be helpful. There is no evidence to support the use of oral glucocorticoids or tricyclic antidepressants in treatment of ALBP. The role of diagnostic and therapeutic nerve root blocks for patients with acute back or neck pain remains controversial. Equivocal data suggests that epidural steroids may occasionally produce short-term pain relief in patients withALBP and radiculopathy, but proof is lacking for pain relief beyond 1 month. Epidural anesthetics, steroids, or opioids are not indicated as initial treatment for ALBP without radiculopathy. Diagnostic selective nerve root blocks have been advocated to determine if pain originates from a nerve root. However, these studies may be falsely positive due to a placebo effect, in patients with a painful lesion located distally along the peripheral nerve, or from anesthesia of the sinuvertebral nerve. Therapeutic selective nerve root blocks are an option after brief conservative measures fail, particularly when temporary relief of pain may be important for patient function. Needle position is confirmed under fluoroscopic guidance with nonionic contrast before injection of glucocorticoid and local anesthetic. A short course of spinal manipulation orPT for symptomatic relief of uncomplicatedALBP is an option. A prospective, randomized study comparing PT, chiropractic manipulation, and education interventions for patients with ALBP found
modest trends toward benefit with both PT and chiropractic manipulation at 1 year. Costs per year were equivalent in the PT/chiropractic group and ~$280 less for the group treated with the education booklet alone. The extent to which this modest improvement in symptoms and outcome is worth the cost must be determined for each patient. Extended duration of treatment or treatment of patients with radiculopathy is of unknown value and carries potential risk. The appropriate frequency or duration of spinal manipulation has not been addressed adequately. Chronic Low Back Pain CLBP is defined as pain lasting longer than 12 weeks. Patients with CLBP account for 50% of back pain costs. Overweight individuals appear to be at particular risk. Other risk factors include: female gender, older age, prior history of back pain, restricted spinal mobility, pain radiating into a leg, high levels of psychological distress, poor self-rated health, minimal physical activity, smoking, job dissatisfaction, and widespread pain. Combinations of these premorbid factors have been used to predict which individuals withALBP are likely to develop CLBP. The initial approach to these patients is similar to that for ALBP, and the differential diagnosis of CLBP includes most of the conditions described in this chapter. Treatment of this heterogeneous group of patients is directed toward the underlying cause when possible; the ultimate goal is to restore function to the greatest extent possible. Many conditions that produceCLBP can be identified by the combination of neuroimaging and electrophysiologic studies. SpineMRI orCT-myelography are the techniques of choice but are generally not indicated within the first month after initial evaluation in the absence of risk factors for a serious underlying cause. Imaging studies should be performed only in circumstances where the results are likely to influence surgical or medical treatment. Diskography is of questionable value in the evaluation of back pain. No additional anatomic information is provided beyond what is available byMRI. Reproduction of the patient's typical pain with the injection is often used as evidence that a specific disk is the pain generator, but it is not known whether this information has any value in selecting candidates for surgery. There is no proven role for thermography in the assessment of radiculopathy. The diagnosis of nerve root injury is most secure when the history, examination, results of imaging studies, and theEMG are concordant. The correlation betweenCT and EMG for localization of nerve root injury is between 65 and 73%. Up to one-third of asymptomatic adults have a disk protrusion detected by CT orMRIscans. Thus, surgical intervention based solely upon radiologic findings and pain increases the likelihood of an unsuccessful outcome. CLBPcan be treated with a variety of conservative measures. Acute and subacute exacerbations are managed withNSAIDsand comfort measures. There is no good evidence to suggest that one NSAID is more effective than another. Bed rest should not exceed 2 days. Activity tolerance is the primary goal, while pain relief is secondary. Exercise programs can reverse type II muscle fiber atrophy in paraspinal muscles and strengthen trunk extension. Supervised, intensive physical exercise or "work hardening" regimens (under the guidance of a physical therapist) have been effective in returning some patients to work, improving walking distances, and diminishing pain. The benefit
can be sustained with home exercise regimens; compliance with the exercise regimen strongly influences outcome. The role of manipulation, back school, or epidural steroid injections in the treatment of CLBP is unclear. Up to 30% of "blind" epidural steroid injections miss the epidural space even when performed by an experienced anesthesiologist. There is no strong evidence to support the use of acupuncture or traction in this setting. A reduction in sick leave days, long-term health care utilization, and pension expenditures may offset the initial expense of multidisciplinary treatment programs. In one study comparing 3 weeks of hydrotherapy versus routine ambulatory care, hydrotherapy resulted in diminished duration and intensity of back pain, reduced analgesic drug consumption, improved spine mobility, and improved functional score. Functional score returned to baseline at the 9-month follow-up, but all other beneficial effects were sustained. Percutaneous electrical nerve stimulation (PENS) has been shown to provide significant short-term relief of CLBP, but additional studies regarding long-term efficacy and cost are necessary. PAIN IN THE NECK AND SHOULDER Approach to the Patient In one recent epidemiologic survey, the 6-month prevalence of disabling neck pain was 4.6% among adults. Neck pain commonly arises from diseases of the cervical spine and soft tissues of the neck. Neck pain arising from the cervical spine is typically precipitated by neck movements and may be accompanied by focal spine tenderness and limitation of motion. Pain arising from the brachial plexus, shoulder, or peripheral nerves can be confused with cervical spine disease, but the history and examination usually identify a more distal origin for the pain. Cervical spine trauma, disk disease, or spondylosis may be asymptomatic or painful and can produce a myelopathy, radiculopathy, or both. The nerve roots most commonly affected are C7 and C6. TRAUMA TO THE CERVICAL SPINE Unlike injury to the low back, trauma to the cervical spine (fractures, subluxation) places the spinal cord at risk for compression. Motor vehicle accidents, violent crimes, or falls account for 87% of spinal cord injuries, which can have devastating consequences (Chap. 369). Emergency immobilization of the neck prior to complete assessment is mandatory to minimize further spinal cord injury from movement of unstable cervical spine segments. Whiplash injury is due to trauma (usually automobile accidents) causing cervical musculoligamental sprain or strain due to hyperflexion or hyperextension. This diagnosis should not be applied to patients with fractures, disk herniation, head injury, or altered consciousness. One prospective study found that 18% of patients with whiplash injury had persistent injury-related symptoms 2 years after the car accident. Such patients were older, had a higher incidence of inclined or rotated head position at impact, greater intensity of initial neck and head pain, greater number of initial symptoms, and more osteoarthritic changes on cervical spine x-rays at baseline compared to patients who ultimately recovered. Objective data on the pathology of neck soft tissue injuries is lacking. Patients with severe initial injury are at increased risk for poor long-term outcome.
CERVICAL DISK DISEASE Herniation of a lower cervical disk is a common cause of neck, shoulder, arm, or hand pain. Neck pain (worse with movement), stiffness, and limited range of neck motion are common. With nerve root compression, pain may radiate into a shoulder or arm. Extension and lateral rotation of the neck narrows the intervertebral foramen and may reproduce radicular symptoms (Spurling's sign). In young individuals, acute cervical nerve root compression from a ruptured disk is often due to trauma. Subacute radiculopathy is less likely to be related to a specific traumatic incident and may involve both disk disease and spondylosis. Cervical disk herniations are usually posterolateral near the lateral recess and intervertebral foramen. The usual patterns of reflex, sensory, and motor changes that accompany specific cervical nerve root lesions are listed inTable 16-2. When evaluating patients with suspected cervical radiculopathy it is important to consider the following: (1) overlap in function between adjacent nerve roots is common, (2) the anatomic pattern of pain is the most variable of the clinical features, and (3) the distribution of symptoms and signs may be evident in only part of the injured nerve root territory. Surgical management of cervical herniated disks usually consists of an anterior approach with diskectomy followed by anterior interbody fusion. A simple posterior partial laminectomy with diskectomy is an alternative approach. The risk of subsequent radiculopathy or myelopathy at cervical segments adjacent to the fusion is 3% per year and 26% at 10 years. Although the risk is sometimes portrayed as a late complication of cervical surgery, it may also reflect the natural history of degenerative cervical spine disease in this subpopulation of patients. CERVICAL SPONDYLOSIS Osteoarthritis of the cervical spine may produce neck pain that radiates into the back of the head, shoulders, or arms. Arthritic or other pathologic conditions of the upper cervical spine may be the source of headaches in the posterior occipital region (supplied by the C2-C4 nerve roots). Cervical spondylosis with osteophyte formation in the lateral recess or hypertrophic facet joints may produce a monoradiculopathy (Fig. 16-7). Narrowing of the spinal canal by osteophytes, ossification of the posterior longitudinal ligament, or a large central disk may compress the cervical spinal cord. In some patients, a combination of radiculopathy and myelopathy occur. An electrical sensation elicited by neck flexion and radiating down the spine from the neck (Lhermitte's symptom) usually indicates cervical or upper thoracic (T1-T2) spinal cord involvement. When little or no neck pain accompanies the cord compression, the diagnosis may be confused with amyotrophic lateral sclerosis (Chap. 365), multiple sclerosis (Chap. 371), spinal cord tumors (Chap. 368), or syringomyelia (Chap. 368). The possibility of this treatable cervical spinal cord disease must be considered even when the patient presents with leg complaints only. Furthermore, lumbar radiculopathy or polyneuropathy may mask an associated cervical myelopathy.MRI orCT-myelography can define the anatomic abnormalities, andEMG and nerve conduction studies can quantify the severity and localize the levels of motor nerve root injury. OTHER CAUSES OF NECK PAIN
Rheumatoid arthritis (RA) (Chap. 312) of the cervical apophyseal joints results in neck pain, stiffness, and limitation of motion. In typical cases with symmetric inflammatory polyarthritis, the diagnosis of RA is straightforward. In advanced RA, synovitis of the atlantoaxial joint (C1-C2;Fig. 16-2) may damage the transverse ligament of the atlas, producing forward displacement of the atlas on the axis (atlantoaxial subluxation). Radiologic evidence of atlantoaxial subluxation occurs in 30% of patients with RA. Not surprisingly, the degree of subluxation correlates with the severity of erosive disease. When subluxation is present, careful neurologic assessment is important to identify early signs of myelopathy. Occasional patients develop high spinal cord compression leading to quadriparesis, respiratory insufficiency, and death. Although low back pain is common among RA patients, the frequency of facet disease, fracture, and spondylolisthesis is no greater than among age- and sex-matched controls with mechanical low back pain. Ankylosing spondylitis can cause neck pain and on occasion atlantoaxial subluxation; when spinal cord compression is present or threatened, surgical intervention is indicated. Herpes zoster produces neck and posterior occipital pain in a C2-C3 distribution prior to the outbreak of vesicles. Neoplasms metastatic to the cervical spine, infections (osteomyelitis and epidural abscess), and metabolic bone diseases may also be the cause of neck pain. Neck pain may also be referred from the heart in the setting of coronary artery ischemia (cervical angina syndrome). THORACIC OUTLET The thoracic outlet is an anatomic region containing the first rib, the subclavian artery and vein, the brachial plexus, the clavicle, and the lung apex. Injury to these structures may result in posture or task-related pain around the shoulder and supraclavicular region. There are at least three subtypes of thoracic outlet syndrome (TOS). True neurogenic TOS results from compression of the lower trunk of the brachial plexus by an anomalous band of tissue connecting an elongate transverse process at C7 with the first rib. Neurologic deficits include weakness of intrinsic muscles of the hand and diminished sensation on the palmar aspect of the fourth and fifth digits.EMG and nerve conduction studies confirm the diagnosis. Definitive treatment consists of surgical division of the anomalous band compressing either the lower trunk of the brachial plexus or ventral rami of the C8 or T1 nerve roots. The weakness and wasting of intrinsic hand muscles typically does not improve, but surgery halts the insidious progression of weakness. The arterial TOS results from compression of the subclavian artery by a cervical rib; the compression results in poststenotic dilatation of the artery and thrombus formation. Blood pressure is reduced in the affected limb, and signs of emboli may be present in the hand; neurologic signs are absent. Noninvasive ultrasound techniques confirm the diagnosis. Treatment is with thrombolysis or anticoagulation (with or without embolectomy) and surgical excision of the cervical rib compressing the subclavian artery or vein. The disputed TOS includes a large number of patients with chronic arm and shoulder pain of unclear cause. The lack of sensitive and specific findings on physical examination or laboratory markers for this condition frequently results in diagnostic uncertainty. The role of surgery in disputed TOS is controversial; conservative approaches often include multidisciplinary pain management. Treatment is often unsuccessful.
BRACHIAL PLEXUS AND NERVES Pain from injury to the brachial plexus or arm peripheral nerves can occasionally be confused with pain of cervical spine origin. Neoplastic infiltration of the lower trunk of the brachial plexus may produce shoulder pain radiating down the arm, numbness of the fourth and fifth fingers, and weakness of intrinsic hand muscles innervated by the ulnar and median nerves. Postradiation fibrosis (breast carcinoma is the most common setting) or a Pancoast tumor of the lung (Chap. 88) may produce similar findings. A Horner's syndrome is present in two-thirds of patients with a Pancoast tumor. Suprascapular neuropathy may produce severe shoulder pain, weakness, and wasting of the supraspinatous and infraspinatous muscles. Acute brachial neuritis is often confused with radiculopathy. It consists of the acute onset of severe shoulder or scapular pain followed over days to weeks by weakness of the proximal arm and shoulder girdle muscles innervated by the upper or middle trunks or cords of the brachial plexus. The onset is often preceeded by an infection or immunization. Separation of this syndrome from cervical radiculopathy is important because slow, complete recovery of brachial neuritis occurs in 75% of patients after 2 years and in 89% after 3 years. Occasional cases of carpal tunnel syndrome produce pain and paresthesia extending into the forearm, arm, and shoulder resembling a C5 or C6 root lesion. Lesions of the radial or ulnar nerve can mimic a radiculopathy at C7 or C8, respectively.EMG and nerve conduction studies can accurately localize lesions to the nerve roots, brachial plexus, or nerves.*For further discussion of peripheral nerve disorders, see Chap. 377. SHOULDER Pain in the shoulder region can be difficult to separate clearly from neck pain. If the symptoms and signs of radiculopathy are absent, then the differential diagnosis includes mechanical shoulder pain (tendonitis, bursitis, rotator cuff tear, dislocation, adhesive capsulitis, and cuff impingement under the acromion) and referred pain (subdiaphragmatic irritation, angina, Pancoast tumor). Mechanical pain is often worse at night, associated with local shoulder tenderness, and aggravated by abduction, internal rotation, or extension of the arm. The pain of shoulder disease may at times radiate into the arm or hand, but the sensory, motor, and reflex changes that indicate disease of the nerve roots, plexus, or peripheral nerves are absent. TREATMENT A paucity of well-designed clinical trials exists for the treatment of neck pain. Symptomatic treatment of neck pain can include the use of analgesic medications and/or a soft cervical collar. Current indications for cervical disk surgery are similar to those for lumbar disk surgery; because of the risk of spinal cord injury with cervical spine disease, an aggressive approach is generally indicated whenever spinal cord injury is threatened. Surgical management of cervical herniated disks usually consists of an anterior approach with diskectomy followed by anterior interbody fusion. A simple posterior partial laminectomy with diskectomy is an acceptable alternative approach. The cumulative risk of subsequent radiculopathy or myelopathy at cervical segments adjacent to the fusion is approximately 3% per year and 26% per decade. Although this
risk is sometimes portrayed as a late complication of surgery, it may also reflect the natural history of degenerative cervical spine disease. Nonprogressive cervical radiculopathy (associated with a focal neurologic deficit) due to a herniated cervical disk may be treated conservatively with a high rate of success. Cervical spondylosis with bony, compressive cervical radiculopathy is generally treated with surgical decompression to interrupt the progression of neurologic signs. Cervical spondylotic myelopathy is typically managed with either anterior decompression and fusion or laminectomy. Outcomes in both surgical groups vary, but late functional deterioration occurs in 20 to 30% of patients; a prospective, controlled study comparing different surgical interventions is sorely needed. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 2 -ALTERATIONS IN BODY TEMPERATURE 17. FEVER AND HYPERTHERMIA - Charles A. Dinarello, Jeffrey A. Gelfand Body temperature is controlled by the hypothalamus. Neurons in both the preoptic anterior hypothalamus and the posterior hypothalamus receive two kinds of signals: one from peripheral nerves that reflect warmth/cold receptors and the other from the temperature of the blood bathing the region. These two types of signals are integrated by the thermoregulatory center of the hypothalamus to maintain normal temperature. In a neutral environment, the metabolic rate of humans consistently produces more heat than is necessary to maintain the core body temperature at 37°C. Therefore, the hypothalamus controls temperature by mechanisms of heat loss. A normal body temperature is ordinarily maintained, despite environmental variations, because the hypothalamic thermoregulatory center balances the excess heat production derived from metabolic activity in muscle and the liver with heat dissipation from the skin and lungs. According to recent studies of healthy individuals 18 to 40 years of age, the mean oral temperature is 36.8°± 0.4°C (98.2° ±0.7°F), with low levels at 6 A.M. and higher levels at 4 to 6 P.M. The maximum normal oral temperature is 37.2°C (98.9°F) at 6 A.M. and 37.7°C (99.9°F) at 4 P.M.; these values define the 99th percentile for healthy individuals. In light of these studies, an A.M. temperature of>37.2°C (98.9°F) or a P.M. temperature of>37.7°C (99.9°F) would define a fever. The normal daily temperature variation is typically 0.5°C (0.9°F). However, in some individuals recovering from a febrile illness, this daily variation can be as great as 1.0°C. During a febrile illness, diurnal variations are usually maintained but at higher levels. Daily temperature swings do not occur in patients with hyperthermia (see below). Rectal temperatures are generally 0.4°C (0.7°F) higher than oral readings. The lower oral readings are probably attributable to mouth breathing, which is a particularly important factor in patients with respiratory infections and rapid breathing. Lower esophageal temperatures closely reflect core temperature. Tympanic membrane (TM) thermometers measure radiant heat energy from the tympanic membrane and nearby ear canal and display that absolute value (unadjusted mode) or a value automatically calculated from the absolute reading on the basis of nomograms relating the radiant temperature measured to actual core temperatures obtained in clinical studies (adjusted mode). These measurements, although convenient, may be more variable than directly determined oral or rectal values. Studies in adults show that readings are lower with unadjusted-mode than with adjusted-mode TM thermometers and that unadjusted-mode TM values are 0.8°C (1.6°F) lower than rectal temperatures. In women who menstruate, the A.M. temperature is generally lower in the 2 weeks before ovulation; it then rises by about 0.6°C (1°F) with ovulation and remains at that level until menses occur. Seasonal variation in body temperature has been described but may reflect a metabolic change and is not common. Body temperature is elevated in the postprandial state, but this elevation does not represent fever. Pregnancy and endocrinologic dysfunction also affect body temperature. The daily temperature variation appears to be fixed in early childhood; in contrast, elderly individuals can exhibit a reduced ability to develop fever, with only a modest fever even in severe infections.
FEVER VERSUS HYPERTHERMIA FEVER Fever is an elevation of body temperature that exceeds the normal daily variation and occurs in conjunction with an increase in the hypothalamic set point -- for example, from 37°C to 39°C. This shift of the set point from "normothermic" to febrile levels very much resembles the resetting of the home thermostat to a higher level in order to raise the ambient temperature in a room. Once the hypothalamic set point is raised, neurons in the vasomotor center are activated and vasoconstriction commences. The individual first notices vasoconstriction in the hands and feet. Shunting of blood away from the periphery to the internal organs essentially decreases heat loss from the skin, and the person feels cold. For most fevers, body temperature increases by 1 to 2°C. Shivering, which increases heat production from the muscles, may begin at this time; however, shivering is not required if heat conservation mechanisms raise blood temperature sufficiently. Heat production from the liver also increases. In humans, behavioral instincts (e.g., putting on more clothing or bedding) lead to a reduction of exposed surfaces, which helps raise body temperature. The processes of heat conservation (vasoconstriction) and heat production (shivering and increased metabolic activity) continue until the temperature of the blood bathing the hypothalamic neurons matches the new thermostat setting. Once that point is reached, the hypothalamus maintains the temperature at the febrile level by the same mechanisms of heat balance that are operative in the afebrile state. When the hypothalamic set point is again reset downward (due to either a reduction in the concentration of pyrogens or the use of antipyretics), the processes of heat loss through vasodilation and sweating are initiated. Behavioral changes triggered at this time include the removal of insulating clothing or bedding. Loss of heat by sweating and vasodilation continues until the blood temperature at the hypothalamic level matches the lower setting. A fever of>41.5°C (106.7°F) is called hyperpyrexia. This extraordinarily high fever can develop in patients with severe infections but most commonly occurs in patients with central nervous system hemorrhages. In the preantibiotic era, fever due to a variety of infectious diseases rarely exceeded 106°F, and there has been speculation that this natural "thermal ceiling" is mediated by neuropeptides functioning as central antipyretics. In some rare cases, the hypothalamic set point is elevated as a result of local trauma, hemorrhage, tumor, or intrinsic hypothalamic malfunction. The term hypothalamic fever is sometimes used to describe elevated temperature caused by abnormal hypothalamic function. However, most patients with hypothalamic damage have subnormal, not supranormal, body temperatures. These patients do not respond properly to mild environmental temperature changes. For example, when exposed to only mildly cold conditions, their core temperature falls quickly rather than over the normal period of a few hours. In the very few patients in whom elevated core temperature is suspected to be due to hypothalamic damage, diagnosis depends on the demonstration of other abnormalities in hypothalamic function, such as the production of hypothalamic releasing factors, abnormal response to cold, and absence of circadian temperature and
hormonal rhythms. HYPERTHERMIA Hyperthermia is characterized by an unchanged (normothermic) setting of the thermoregulatory center in conjunction with an uncontrolled increase in body temperature that exceeds the body's ability to lose heat. Exogenous heat exposure and endogenous heat production are two mechanisms by which hyperthermia can result in dangerously high internal temperatures. Excessive heat production can easily cause hyperthermia despite physiologic and behavioral control of body temperature. For example, over-insulating clothing can result in an elevated core temperature, and work or exercise in hot environments can produce heat faster than peripheral mechanisms can lose it. Although most patients with elevated body temperature have fever, there are a few circumstances in which elevated temperature represents not fever but hyperthermia (Table 17-1). Heat stroke, caused by thermoregulatory failure in association with a warm environment, may be categorized as exertional or nonexertional. Exertional heat stroke typically occurs in younger individuals exercising at ambient temperatures and/or humidities that are higher than normal. Even in normal individuals, dehydration or the use of common medications (e.g., over-the-counter antihistamines with anticholinergic side effects) may help to precipitate exertional heat stroke. Nonexertional or classic heat stroke typically occurs in elderly individuals, particularly during heat waves. For example, in Chicago in July 1995, 465 deaths were certified as heat related. The elderly, the bedridden, persons taking anticholinergic or antiparkinsonian drugs or diuretics, and individuals confined to poorly ventilated and non-air-conditioned environments are most susceptible. Drug-induced hyperthermia has become increasingly common as a result of the increased use of prescription psychotropic drugs and illicit drugs. Drug-induced hyperthermia may be caused by monoamine oxidase inhibitors, tricyclic antidepressants, and amphetamines and by the illicit use of phencyclidine, lysergic acid diethylamide (LSD), or cocaine. Malignant hyperthermia occurs in individuals with an inherited abnormality of skeletal-muscle sarcoplasmic reticulum that causes a rapid increase in intracellular calcium levels in response to halothane and other inhalational anesthetics or to succinylcholine. Elevated temperature, increased muscle metabolism, rigidity, rhabdomyolysis, acidosis, and cardiovascular instability develop rapidly. This condition is often fatal. The neuroleptic malignant syndrome can occur with phenothiazines and other drugs such as haloperidol and is characterized by muscle rigidity, autonomic dysregulation, and hyperthermia. This disorder appears to be caused by the inhibition of central dopamine receptors in the hypothalamus, which results in increased heat generation and decreased heat dissipation. Thyrotoxicosis and pheochromocytoma can also cause increased thermogenesis. It is important to distinguish between fever and hyperthermia since hyperthermia can be rapidly fatal and characteristically does not respond to antipyretics. However, there is no rapid way to make this distinction. Hyperthermia is often diagnosed on the basis of the
events immediately preceding the elevation of core temperature -- e.g., heat exposure or treatment with drugs that interfere with thermoregulation. However, in addition to the clinical history of the patient, the physical aspects of some forms of hyperthermia may alert the clinician. For example, in patients with heat stroke syndromes and in those taking drugs that block sweating, the skin is hot but dry. Moreover, antipyretics do not reduce the elevated temperature in hyperthermia, whereas in fever -- and even in hyperpyrexia -- adequate doses of either aspirin or acetaminophen usually result in some decrease in body temperature. PYROGENS The term pyrogen is used to describe any substance that causes fever. Exogenous pyrogens are derived from outside the patient; most are microbial products, microbial toxins, or whole microorganisms. The classic example of an exogenous pyrogen is the lipopolysaccharide endotoxin produced by all gram-negative bacteria. Endotoxins are potent not only as pyrogens but also as inducers of various pathologic changes in gram-negative infections. Another group of potent bacterial pyrogens is produced by gram-positive organisms and includes the enterotoxins of Staphylococcus aureus and the group A and B streptococcal toxins, also called superantigens. One staphylococcal toxin of clinical importance is the toxic shock syndrome toxin associated with isolates of S. aureus from patients with toxic shock syndrome. Like the endotoxins of gram-negative bacteria, the toxins produced by staphylococci and streptococci cause fever in experimental animals when injected intravenously at concentrations of 5 mm in diameter with a more rounded configuration. Wheals (urticaria, hives) are papules or plaques that are pale pink and may appear annular (ringlike) as they enlarge; classic (nonvasculitic) wheals are transient, lasting only 24 to 48 h in any defined area. Vesicles (5 mm) are circumscribed, elevated lesions containing fluid. Pustules are raised lesions containing purulent exudate; vesicular processes such as varicella or herpes simplex may evolve to pustules. Nonpalpable purpura is a flat lesion that is due to bleeding into the skin; if 3 mm, they are termed ecchymoses. Palpable purpura is a raised lesion that is due to inflammation of the vessel wall (vasculitis) with subsequent hemorrhage. An ulcer is a defect in the skin extending at least into the upper layer of the dermis, and an eschar (tache noire) is a necrotic lesion covered with a black crust. Other pertinent features of rashes include their configuration (i.e., annular or target), the arrangement of their lesions, and their distribution (i.e., central or peripheral).*For further discussion, see Chaps. 55 and 57. CLASSIFICATION OF RASH This chapter reviews rashes that reflect systemic disease but does not include localized skin eruptions (i.e., cellulitis, impetigo) that may also be associated with fever (Chap. 128). Rashes are classified herein on the basis of the morphology and distribution of lesions. For practical purposes, this classification system is based on the most typical disease presentations. However, morphology may vary as rashes evolve, and the presentation of diseases with rashes is subject to many variations (Chap. 57). For instance, the classic petechial rash of Rocky Mountain spotted fever (RMSF) may initially consist of blanchable erythematous macules distributed peripherally; at times,
the rash associated with RMSF may not be predominantly acral, or a rash may not develop at all. Diseases with fever and rash may be classified by type of eruption: centrally distributed maculopapular, peripheral, confluent desquamative erythematous, vesiculobullous, urticarial, nodular, purpuric, ulcerated, or eschars (Table 18-1). For a more detailed discussion of each disease associated with a rash, the reader is referred to the chapter dealing with that specific disease. (Reference chapters and color plates are cited in the text and listed in Table 18-1.) Centrally Distributed Maculopapular Eruptions Centrally distributed rashes, in which lesions are primarily truncal, are the most common type of eruption. The rash of measles (rubeola) starts at the hairline 2 to 3 days into the illness and moves down the body, sparing the palms and soles (Chap. 194). It begins as discrete erythematous lesions, which become confluent as the rash spreads. Koplik's spots (1- to 2-mm white or bluish lesions with an erythematous halo on the buccal mucosa) are pathognomonic for measles and are generally seen during the first 2 days of symptoms. They should not be confused with Fordyce's spots (ectopic sebaceous glands), which have no erythematous halos and are found in the mouth of healthy individuals. Koplik's spots may briefly overlap with the measles exanthem. German measles (rubella) also spreads from the hairline downward; unlike that of measles, however, the rash of rubella tends to clear from originally affected areas as it migrates and may be pruritic (Chap. 195). Forchheimer spots (palatal petechiae) may develop but are nonspecific since they also develop in mononucleosis (Chap. 184) and scarlet fever (Chap. 140). Postauricular and suboccipital adenopathy and arthritis are common among adults with German measles. Exposure of pregnant women to ill individuals should be avoided, as rubella causes severe congenital abnormalities. Numerous strains of enteroviruses (Chap. 193), primarily echoviruses and coxsackieviruses, cause nonspecific syndromes of fever and eruptions that may mimic rubella or measles. Patients with infectious mononucleosis caused by Epstein-Barr virus or with primary infection caused by HIV (Chap. 309) may exhibit pharyngitis, lymphadenopathy, and a nonspecific maculopapular exanthem. The rash of erythema infectiosum (fifth disease), which is caused by human parvovirus B19, primarily affects children 3 to 12 years old; it develops after fever has resolved as a bright blanchable erythema on the cheeks ("slapped cheeks") with perioral pallor (Chap. 187). A more diffuse rash (often pruritic) appears the next day on the trunk and extremities and then rapidly develops into a lacy reticular eruption that may wax and wane (especially with temperature change) over 3 weeks. Adults with fifth disease often have arthritis, and fetal hydrops can develop in association with this condition in pregnant women. Exanthem subitum (roseola,Fig. 18-CD2) is most common among children under 3 years of age (Chap. 185). As in erythema infectiosum, the rash usually appears after fever has subsided. It consists of 2- to 3-mm rose-pink macules and papules that rarely coalesce, occur initially on the trunk and sometimes on the extremities (sparing the face), and fade within 2 days.
Though drug reactions have many manifestations, including urticaria, exanthematous drug-induced eruptions (Chap. 59) are most common and are often difficult to distinguish from viral exanthems. Eruptions elicited by drugs are usually more intensely erythematous and pruritic than viral exanthems, but this distinction is not reliable. A history of new medications and an absence of prostration may help to distinguish a drug-related rash from an eruption of another etiology. Rashes may persist for up to 2 weeks after administration of the offending agent is discontinued. Certain populations are more prone than others to drug rashes. Of HIV-infected patients, 50 to 60% develop a rash in response to sulfa drugs; 50 to 100% of patients with mononucleosis due to Epstein-Barr virus develop a rash when given ampicillin. Rickettsial illnesses (Chap. 177) should be considered in the evaluation of individuals with centrally distributed maculopapular eruptions. The usual setting for epidemic typhus is a site of war or natural disaster in which people are exposed to body lice. A diagnosis of recrudescent typhus should be considered in European immigrants to the United States. However, an indigenous form of typhus, presumably transmitted by flying squirrels, has been reported in the southeastern United States. Endemic typhus or leptospirosis (the latter caused by a spirochete;Chap. 174) may be seen in urban environments where rodents proliferate. Outside the United States, other rickettsial diseases cause a spotted-fever syndrome and should be considered in residents of or travelers to endemic areas. Similarly, typhoid fever, a nonrickettsial disease caused by Salmonella typhi (Chap. 156), is usually acquired during travel outside the United States. Some centrally distributed maculopapular eruptions have distinctive features. Erythema chronicum migrans (ECM), the rash of Lyme disease (Chap. 176), typically manifests as singular or multiple annular plaques. Untreated ECM lesions usually fade within a month but may persist for more than a year. Erythema marginatum, the rash of acute rheumatic fever (Chap. 235), has a distinctive pattern of enlarging and shifting transient annular lesions. Collagen vascular diseases may cause fever and rash. Patients with systemic lupus erythematosus (Chap. 311) typically develop a sharply defined, erythematous eruption in a butterfly distribution on the cheeks (malar rash) as well as many other skin manifestations. Still's disease (Chap. 326) manifests as an evanescent salmon-colored rash on the trunk and proximal extremities that coincides with fever spikes. Peripheral Eruptions These rashes are alike in that they are most prominent peripherally or begin in peripheral (acral) areas before spreading centripetally. Early diagnosis and therapy are critical inRMSF(Chap. 177) because of its grave prognosis if untreated. Lesions evolve from macular to petechial, start on the wrists and ankles, spread centripetally, and appear on the palms and soles only later in the disease. The rash of secondary syphilis (Chap. 172), which may be diffuse but is prominent on the palms and soles, should be considered in the differential diagnosis of pityriasis rosea, especially in sexually active patients. Atypical measles (Chap. 194) is seen in individuals contracting measles who received the killed measles vaccine between 1963 and 1967 in the United States and who were not subsequently protected with the live vaccine. Hand-foot-and-mouth disease (Chap. 193) is distinguished by tender vesicles distributed peripherally and in the mouth; outbreaks commonly occur within families. The classic
target lesions of erythema multiforme appear symmetrically on the elbows, knees, palms, and soles. In relatively severe cases, these lesions may spread diffusely and involve mucosal surfaces. Lesions may develop on the hands and feet in endocarditis (Chap. 126). Confluent Desquamative Erythemas These eruptions consist of diffuse erythema frequently followed by desquamation. The eruptions caused by group A Streptococcus or Staphylococcus aureus are toxin mediated. Certain disease features may provide diagnostic clues. Scarlet fever (Chap. 140) usually follows pharyngitis; patients have a facial flush, a "strawberry" tongue, and accentuated petechiae in body folds (Pastia's lines). Kawasaki disease (Chaps. 57 and317) presents in the pediatric population as fissuring of the lips, a strawberry tongue, conjunctivitis, adenopathy, and sometimes cardiac abnormalities. Streptococcal toxic shock syndrome (Chap. 140,Fig. 18-CD3) manifests with hypotension, multiorgan failure, and often a severe group A streptococcal infection (e.g., necrotizing fasciitis, Fig. 18-CD4). Staphylococcal toxic shock syndrome (Chap. 139) also presents with hypotension and multiorgan failure, but usually only S. aureus colonization -- not a severe S. aureus infection -- is documented. Staphylococcal scalded-skin syndrome (Chap. 139) is seen primarily in children and in immunocompromised adults. Generalized erythema is often evident during the prodrome of fever and malaise; profound tenderness of the skin is distinctive. In the exfoliative stage, the skin can be induced to form bullae with light lateral pressure (Nikolsky's sign). In a mild form, a scarlatiniform eruption mimics scarlet fever, but the patient does not exhibit a strawberry tongue or circumoral pallor. In contrast to the staphylococcal scalded-skin syndrome, in which the cleavage plane is superficial in the epidermis, toxic epidermal necrolysis (Chap. 59) (Fig. 18-CD5) involves sloughing of the entire epidermis, resulting in severe disease. Exfoliative erythroderma syndrome (Chaps. 56 and59) is a serious reaction associated with systemic toxicity that is often due to eczema, psoriasis, mycosis fungoides, or a severe drug reaction. Vesiculobullous Eruptions Varicella (Chap. 183) is highly contagious, often occurring in winter or spring. At a given time within a given region of the body, varicella lesions are in different stages of development. In immunocompromised hosts, varicella vesicles may lack the characteristic erythematous base or may appear hemorrhagic. Rickettsialpox (Chap. 177) is often documented in urban settings and is characterized by vesicles. It can be distinguished from varicella by an eschar at the site of the mouse-mite bite and the papule/plaque base of each vesicle. Disseminated Vibrio vulnificus infection (Chap. 159) or ecthyma gangrenosum due to Pseudomonas aeruginosa (Chap. 155) should be considered in immunosuppressed individuals with sepsis and hemorrhagic bullae. Urticarial Eruptions Individuals with classic urticaria ("hives") usually have a hypersensitivity reaction without associated fever. In the presence of fever, urticarial eruptions are usually due to urticarial vasculitis (Chap. 317). Unlike individual lesions of classic urticaria, which last up to 48 h, these lesions may last up to 5 days. Etiologies include serum sickness (often induced by drugs such as penicillins, sulfas, salicylates, or barbiturates), connective-tissue disease (e.g., systemic lupus erythematosus or Sjogren's syndrome), and infection (e.g., with hepatitis B virus, coxsackievirus A9, or parasites). Malignancy may be associated with fever and chronic urticaria (Chap. 57).
Nodular Eruptions In immunocompromised hosts, nodular lesions often represent disseminated infection. Patients with disseminated candidiasis (often due to Candida tropicalis) may have a triad of fever, myalgias, and eruptive nodules (Chap. 205). Disseminated cryptococcosis lesions (Chap. 204) may resemble molluscum contagiosum. Necrosis of nodules should raise the suspicion of aspergillosis (Chap. 206) or mucormycosis (Chap. 207). Erythema nodosum presents with exquisitely tender nodules on the lower extremities. Sweet's syndrome (Chap. 57) should be considered in individuals with multiple nodules and plaques, often so edematous that they give the appearance of vesicles or bullae. Sweet's syndrome may affect either healthy individuals or persons with lymphoproliferative disease. Purpuric Eruptions Acute meningococcemia (Chap. 146) classically presents in children as a petechial eruption, but initial lesions may appear as blanchable macules or urticaria.RMSFshould be considered in the differential diagnosis of acute meningococcemia. Echovirus 9 infection (Chap. 193) may mimic acute meningococcemia; patients should be treated as if they have bacterial sepsis since prompt differentiation of these conditions may be impossible. Large ecchymotic areas of purpura fulminans (Chaps. 124 and146) reflect severe underlying disseminated intravascular coagulation, which may be due to infectious or noninfectious causes. The lesions of chronic meningococcemia (Chap. 146) may have a variety of morphologies, including petechial. Purpuric nodules may develop on the legs and resemble erythema nodosum but lack its exquisite tenderness. Lesions of disseminated gonococcemia (Chap. 147) are distinctive, sparse, countable hemorrhagic pustules, usually located near joints. The lesions of chronic meningococcemia and those of gonococcemia may be indistinguishable in terms of appearance and distribution. Viral hemorrhagic fever (Chaps. 198 and199) should be considered in patients with an appropriate travel history and a petechial rash. Thrombotic thrombocytopenic purpura (Chaps. 57,108, and116) is a noninfectious cause of fever and petechiae. Cutaneous small-vessel vasculitis (leukocytoclastic vasculitis) typically manifests as palpable purpura and has a wide variety of causes (Chap. 57). Eruptions with Ulcers or Eschars The presence of an ulcer or eschar in the setting of a more widespread eruption can provide an important diagnostic clue. For example, the presence of an eschar may suggest the diagnosis of scrub typhus or rickettsialpox in the appropriate setting. In other illnesses (e.g., anthrax), an ulcer or eschar may be the only skin manifestation. (Bibliography omitted in Palm version) Back to Table of Contents
19. APPROACH TO THE ACUTELY ILL INFECTED FEBRILE PATIENT - Tamar F. Barlam, Dennis L. Kasper The physician treating the acutely ill febrile patient must be able to recognize infections that require emergent attention. If such infections are not adequately evaluated and treated at initial presentation, the opportunity to alter an adverse outcome may be lost. In this chapter, the clinical presentations of and approach to patients with relatively common infectious disease emergencies are discussed. These infectious processes are discussed in detail in other chapters.*Noninfectious causes of fever are not covered in this chapter; information on the approach to fever of unknown origin, including that eventually shown to be of noninfectious etiology, is presented in Chap. 125. GENERAL CONSIDERATIONS APPEARANCE A physician must have a consistent approach to acutely ill patients. Even before the history is elicited and a physical examination performed, an immediate assessment of the patient's general appearance yields valuable information. The perceptive physician's subjective sense that a patient is septic or toxic often proves accurate. Visible agitation or anxiety in a febrile patient can be a harbinger of critical illness. HISTORY Presenting symptoms are frequently nonspecific. In addition to a general description of symptoms, it is important to obtain a sense of disease progression. Detailed questions should be asked about the onset and duration of symptoms and about changes in severity or rate of progression over time. Host factors and comorbid conditions may enhance the risk of infection with certain organisms or of a more fulminant course than is usually seen. Lack of splenic function, alcoholism with significant liver disease, intravenous drug use, HIV infection, diabetes, malignancy, and chemotherapy all predispose to specific infections and frequently to increased severity. The patient should be questioned about factors that might help identify a nidus for invasive infection, such as recent upper respiratory tract infections, influenza, or varicella; prior trauma; disruption of cutaneous barriers due to lacerations, burns, surgery, or decubiti; and the presence of foreign bodies, such as nasal packing after rhinoplasty, barrier contraceptives, tampons, arteriovenous fistulas, or prosthetic joints. Travel, contact with pets or other animals, or activities that might result in tick exposure can lead to diagnoses that would not otherwise be considered. Recent dietary intake, medication use, social contact with ill individuals, vaccination history, and menstrual history may be relevant. A review of systems should focus on any neurologic signs or sensorium alterations, rashes or skin lesions, and focal pain or tenderness and should also include a general review of respiratory, gastrointestinal, or genitourinary symptoms. It is especially important to determine the duration and progression of these symptoms in order to gain an appreciation of the pace and urgency of the process. PHYSICAL EXAMINATION A complete physical examination should be performed, with special attention to some
areas that are sometimes given short shrift in routine examinations. Assessment of the patient's general appearance and vital signs, skin and soft tissue examination, and the neurologic evaluation are of particular importance. The patient may appear either anxious and agitated or lethargic and apathetic. Fever is usually present, although the elderly and compromised hosts, such as those who are uremic or cirrhotic and patients who are taking glucocorticoids or nonsteroidal anti-inflammatory agents, may be afebrile despite serious underlying infection. Measurement of blood pressure, heart rate, and respiratory rate helps determine the degree of hemodynamic and metabolic compromise. The patient's airway must be evaluated to rule out the risk of obstruction from an invasive oropharyngeal infection. The etiologic diagnosis may become evident in the context of a thorough skin examination. Petechial rashes are typically seen with meningococcemia or Rocky Mountain spotted fever (RMSF); erythroderma is usual with toxic shock syndrome (TSS) and drug fever. The soft tissue and muscle examination is critical. Areas of erythema or duskiness, edema, and tenderness may indicate underlying necrotizing fasciitis, myositis, or myonecrosis. The neurologic examination must include a careful assessment of mental status for signs of early encephalopathy. Evidence of nuchal rigidity or focal neurologic findings should be sought. Focal findings, depressed mental status, or papilledema should be evaluated by brain imaging prior to lumbar puncture, which, in this setting, could initiate herniation. SPECIFIC PRESENTATIONS For most infections, there is time for careful evaluation, diagnostic testing, and consultation with other physicians. However, the infections considered below according to common clinical presentation can have rapidly catastrophic outcomes, and their immediate recognition can be life-saving. Recommended therapeutic regimens are presented in Table 19-1. SEPSIS WITHOUT AN OBVIOUS FOCUS OF PRIMARY INFECTION These patients initially have a brief prodrome of nonspecific symptoms and signs that progresses quickly to hemodynamic instability with hypotension, tachycardia, tachypnea, or respiratory distress. A patient may display altered mental status. Disseminated intravascular coagulation (DIC) with clinical evidence of a hemorrhagic diathesis is a poor prognostic sign. Septic Shock (See also Chap. 124) Patients with bacteremia leading to septic shock may have a primary site of infection (e.g., pneumonia, pyelonephritis, or cholangitis) that is not evident initially. Elderly patients with comorbid conditions, hosts compromised by malignancy and neutropenia, or patients who have recently undergone a surgical procedure or hospitalization are at increased risk for an adverse outcome. Gram-negative bacteremia with organisms such as Pseudomonas aeruginosa, Aeromonas hydrophila, or Escherichia coli and gram-positive infection with organisms such as Staphylococcus aureus or group A streptococci can present as intractable hypotension and multiorgan failure. Treatment can usually be initiated empirically on the basis of the presentation (Table 124-3).
Overwhelming Infection in Asplenic Patients (See alsoChap. 124) Patients without splenic function are at risk for overwhelming bacterial sepsis. Asplenic patients succumb to sepsis at 600 times the rate of the general population; 50 to 70% of cases occur within the first 2 years after splenectomy, with a mortality rate of up to 80%. However, in the asplenic individual, an increased risk of overwhelming sepsis continues throughout life. In asplenia, encapsulated bacteria cause the majority of infections, and adults are at lower risk than children because they are more likely to have antibody to these organisms. Streptococcus pneumoniae infection is most common, but the risk of infection with Haemophilus influenzae or Neisseria meningitidis is also high. Severe clinical manifestations of infections due to E. coli, S. aureus, group B streptococci, P. aeruginosa, Capnocytophaga, Babesia, and Plasmodium have been described. Babesiosis (See alsoChap. 214) A history of recent travel to endemic areas should raise the possibility of infection with Babesia. Between 1 and 4 weeks after a tick bite, the patient experiences chills, fatigue, anorexia, myalgia, arthralgia, nausea, and headache; ecchymosis and/or petechiae are occasionally seen. The tick that most commonly transmits Babesia, Ixodes scapularis, also transmits Borrelia burgdorferi (the agent of Lyme disease) and Ehrlichia, and co-infection can occur, resulting in more severe disease. Infection with the European species Babesia divergens is more frequently fulminant than that due to the U.S. species B. microti, causing a febrile syndrome with hemolysis, jaundice, hemoglobinemia, and renal failure and a mortality rate of>50%. Severe babesiosis is especially common in asplenic hosts but does occur in hosts with normal splenic function. Other Sepsis Syndromes Tularemia (Chap. 161) is seen throughout the United States, but primarily in Arkansas, Oklahoma, and Missouri, in association with wild rabbit, tick, and tabanid fly contact. The uncommon typhoidal form can be associated with gram-negative septic shock and a mortality rate of>30%. In the United States, plague (Chap. 162) is found primarily in New Mexico, Arizona, and Colorado after contact with ground squirrels, prairie dogs, or chipmunks. The septic form is particularly rare and is associated with shock, multiorgan failure, and a 30% mortality rate. These rare infections should be considered in the appropriate epidemiologic setting. SEPSIS WITH SKIN MANIFESTATIONS (See also Chap. 18) Maculopapular rashes may reflect early meningococcal or rickettsial disease but are usually associated with nonemergent infections. Exanthems are usually viral. Petechiae Petechial rashes caused by viruses are seldom associated with hypotension or a toxic appearance, although severe measles can be an exception. In other settings, petechial rashes require more urgent attention. Meningococcemia (See alsoChap. 146) Almost three-quarters of patients with bacteremic N. meningitidis infection have a rash. Meningococcemia most often affects young children (i.e., those 6 months to 5 years old, often in daycare). However, sporadic cases and outbreaks occur in schools (grade school through college) and army barracks. Between 10 and 20% of all cases have a fulminant course, with shock,DIC, and multiorgan failure. Of these patients, 50 to 60% die, and survivors often require
extensive debridement or amputation of gangrenous extremities. Patients may exhibit fever, headache, nausea, vomiting, myalgias, change in mental status, and meningismus. However, the rapidly progressive form of disease is not usually associated with meningitis. The rash is initially pink, blanching, and maculopapular, appearing on the trunk and extremities, but then becomes hemorrhagic, forming petechiae. Petechiae are first seen at the ankles, wrists, axillae, mucosal surfaces, and palpebral and bulbar conjunctiva, with subsequent spread to the lower extremities and trunk. A cluster of petechiae may be seen at pressure points, e.g., where a blood pressure cuff has been inflated. In rapidly progressive meningococcemia, the petechial rash quickly becomes purpuric (Plate IID-44) and patients develop DIC. Hypotension with petechiae for 30% overall,>70% in association withTSS, and nearly 100% without surgical intervention. Life-threatening necrotizing fasciitis may also be due to Clostridium perfringens (Chap. 145); in this condition, the patient is extremely toxic and the mortality rate is high. Within 48 h, rapid tissue invasion and systemic toxicity associated with hemolysis and death ensue. The distinction between this entity and clostridial
myonecrosis is made by muscle biopsy. Clostridial Myonecrosis (See alsoChap. 145) Myonecrosis is often associated with trauma or surgery but can be spontaneous. The incubation period is usually 12 to 24 h long, and massive necrotizing gangrene develops within hours of onset. Systemic toxicity, shock, and death can occur within 12 h. The patient's pain and toxic appearance are out of proportion to physical findings. On examination, the patient is febrile, apathetic, tachycardic, and tachypneic and may express a feeling of impending doom. Hypotension and renal failure develop later, and hyperalertness is evident preterminally. The skin over the affected area is bronze-brown, mottled, and edematous. Bullous lesions with serosanguineous drainage and a mousy or sweet odor can be present. Crepitus can occur secondary to gas production in muscle tissue. The mortality rate is>65% with spontaneous myonecrosis, which is often associated with C. septicum and underlying malignancy. The mortality rates associated with trunk and limb infection are 63% and 12%, respectively, and any delay in surgical treatment increases the risk of death. NEUROLOGIC INFECTIONS WITH OR WITHOUT SEPTIC SHOCK Bacterial Meningitis (See also Chap. 372) Bacterial meningitis is one of the most common infectious emergencies involving the central nervous system. Although hosts with cell-mediated immune deficiency, including transplant recipients, diabetic patients, the elderly, and cancer patients treated with certain chemotherapeutic agents, are at particular risk for Listeria monocytogenes meningitis, most cases in adults are due to S. pneumoniae (30 to 50%) and N. meningitidis (10 to 35%). An early presentation of headache, meningismus, and fever is classic but is seen in only half of patients. The elderly can present without fever or meningeal signs despite lethargy and confusion. Cerebral dysfunction is evidenced by confusion, delirium, and lethargy that can progress to coma. The presentation is fulminant, with sepsis and brain edema, in some cases; papilledema at presentation is unusual and suggests another diagnosis (e.g., an intracranial lesion). Focal signs, including cranial nerve palsies (IV, VI, VII), can be seen in 10 to 20% of cases; 50 to 60% of patients have bacteremia. A poor neurologic outcome is associated with coma at any time during the course or with aCSFglucose level of 1500 ms) is a specific finding (85 to 100%) for diagnosis of sinus node dysfunction but has a low sensitivity; continuous electrocardiographic monitoring is usually more effective for diagnosing this abnormality. Prolongation of the HV interval and conduction block below the His bundle indicate that His-Purkinje disease may be responsible for syncope. Although an HV interval>100 ms is abnormal, this finding is not common in patients with syncope, and some patients with shorter intervals are also at risk for AV block. Programmed stimulation for ventricular arrhythmias is most useful in patients who have experienced a myocardial infarction; the sensitivity and specificity of this technique is lower in patients with normal hearts or those with heart disease other than coronary artery disease.
Upright tilt table testing is indicated for recurrent syncope, a single syncopal episode that caused injury, or a single syncopal event in a "high-risk" setting (pilot, commercial vehicle driver, etc.), whether or not there is a history of preexisting heart disease or prior vasovagal episodes. In susceptible patients, upright tilt at an angle between 60 and 80° for 30 to 60 min induces a vasovagal episode. The protocol can be shortened if upright tilt is combined with intravenous administration of drugs that cause venous pooling or increase adrenergic stimulation (isoproterenol, nitroglycerin, edrophonium, or adenosine). The sensitivity and specificity of tilt table testing is difficult to ascertain because of the lack of validated criteria. Moreover, the reflexes responsible for vasovagal syncope can be elicited in most, if not all, individuals given the appropriate stimulus. The reported accuracy of the test ranges from 30 to 80%, depending on the population studied and the techniques used. Whereas the reproducibility of a negative test is 85 to 100%, the reproducibility of a positive tilt table test is only between 62 and 88%. A variety of other tests may be useful to determine the presence of structural heart disease that may cause syncope. The echocardiogram with Doppler examination detects valvular, myocardial, and pericardial abnormalities. The echocardiogram is the "gold standard" for the diagnosis of hypertrophic cardiomyopathy and atrial myxoma. Cardiac cine magnetic resonance (MR) imaging provides an alternative noninvasive modality that may be useful for patients in whom diagnostic-quality echocardiographic images cannot be obtained. This test is also indicated for patients suspected of having arrhythmogenic right ventricular dysplasia or right ventricular outflow tract ventricular tachycardia. Both are associated with right ventricular structural abnormalities that are better visualized on MR imaging than by echocardiogram. Exercise testing may detect ischemia or exercise-induced arrhythmias. In some patients, cardiac catheterization may be necessary to diagnose the presence or severity of coronary artery disease or valvular abnormalities. Ultrafast computed tomographic scan, ventilation-perfusion scan, or pulmonary angiography are indicated in patients in whom syncope may be due to pulmonary embolus. In possible cases of cerebrovascular syncope, a variety of neuroimaging tests may be indicated, including Doppler ultrasound studies of the carotid and vertebrobasilar systems,MRimaging, MR angiography, and x-ray angiography of the cerebral vasculature (Chaps. 358 and361). Electroencephalography is indicated if seizures are suspected. TREATMENT The treatment of syncope is directed toward the underlying cause. This discussion will focus on the treatment of disorders of autonomic control.*Arrythmias are discussed in Chaps. 229 and 230, valvular heart diseases in Chap. 236, and cerebrovascular disorders in Chap. 361. Certain precautions should be taken regardless of the cause of syncope. At the first sign of symptoms, patients should make every effort to avoid injury should they lose consciousness. Patients with frequent episodes, or those who have experienced syncope without warning symptoms should avoid situations in which sudden loss of consciousness might result in injury (e.g., climbing ladders, swimming alone, operating
heavy machinery, driving). Patients should lower their head to the extent possible, and preferably should lie down. Lowering the head by bending at the waist should be avoided because it may further compromise venous return to the heart. When appropriate, family members or other close contacts should be educated as to the problem. This will ensure appropriate therapy and may prevent delivery of inappropriate therapy (chest compressions associated with cardiopulmonary resuscitation) that may inflict trauma. Patients who have lost consciousness should be placed in a position that maximizes cerebral blood flow, offers protection from trauma, and secures the airway. Whenever possible, the patient should be placed supine with the head turned to the side to prevent aspiration and the tongue from blocking the airway. Assessment of the pulse and direct cardiac auscultation may assist in determining if the episode is associated with a bradyarrythmia or tachyarrhythmia. Clothing that fits tightly around the neck or waist should be loosened. Peripheral stimulation, such as by sprinkling cold water on the face, may be helpful. Patients should not be given anything by mouth or be permitted to rise until the sense of physical weakness has passed. Patients with vasovagal syncope should be instructed to avoid situations or stimuli that have caused them to lose consciousness. Episodes associated with intravascular volume depletion may be prevented by salt and fluid loading prior to provocative events.b-Adrenoceptor antagonists, the most widely used agents, mitigate the increase in myocardial contractility that stimulates left ventricular mechanoreceptors and also block central serotonin receptors. Disopyramide, a vagolytic with negative inotropic properties, and another vagolytic, transdermal scopolamine, are used to treat vasovagal syncope. Paroxetine, a serotonin reuptake inhibitor used for depression, appears to be an effective treatment, as are theophylline and ephedrine. Midodrine, ana agonist, has been a first-line agent for some patients. Permanent cardiac pacing is effective for patients with frequent episodes of vasovagal syncope and is indicated for those with prolonged asystole associated with vasovagal episodes. Patients with orthostatic hypotension should be instructed to rise slowly and systematically (supine to seated, seated to standing) from the bed or a chair. Movement of the legs prior to rising facilitates venous return from the lower extremities. Whenever possible, medications that aggravate the problem (vasodilators, diuretics, etc.) should be discontinued. Elevation of the head of the bed [20 to 30 cm (8 to 12 in.)] and use of elastic stockings may help. Therapeutic modalities include devices that prevent lower limb blood pooling, such as an antigravity or g suit or elastic stockings; salt loading; and a variety of pharmacologic agents including sympathomimetic amines, monamine oxidase inhibitors, beta blockers, and levodopa.*The treatment of orthostatic hypotension secondary to central or peripheral disorders of the autonomic nervous system is discussed in Chap. 366. Glossopharyngeal neuralgia is treated with carbamazepine, which is effective for the syncope as well as for the pain. Patients with carotid sinus syndrome should be instructed to avoid clothing and situations that stimulate carotid sinus baroreceptors. Patients should turn their entire body, rather than just their head, to look to one side. Those with intractable syncope due to the cardioinhibitory response to carotid sinus
stimulation should undergo permanent pacemaker implantation. Patients with syncope should be hospitalized when the episode may have resulted from a life-threatening abnormality or if recurrence with significant injury seems likely. These individuals should be admitted to a bed with continuous electrocardiographic monitoring. Patients who are known to have a normal heart and for whom the history strongly suggests vasovagal or situational syncope may be treated as outpatients if the episodes are neither frequent nor severe. DIZZINESS AND VERTIGO Dizziness is a common and often vexing symptom. Patients use the term to encompass a variety of sensations, including those that seem semantically appropriate (e.g., lightheadedness, faintness, spinning, giddiness, etc.) and those that are misleadingly inappropriate, such as mental confusion, blurred vision, headache, or tingling. Moreover, some individuals with gait disorders complain of dizziness despite the absence of vertigo or other abnormal cephalic sensations. The causes include peripheral neuropathy, myelopathy, spasticity, parkinsonian rigidity, and cerebellar ataxia. In this context, the term dizziness is being used to describe disturbed mobility. There may be mild associated lightheadedness, particularly with impaired sensation from the feet or poor vision; this is known as multiple-sensory-defect dizziness and occurs in elderly individuals who complain of dizziness only during ambulation. Decreased position sense (secondary to neuropathy or myelopathy) and poor vision (from cataracts or retinal degeneration) create an overreliance on the aging vestibular apparatus. A less precise but sometimes comforting designation to patients is benign dysequilibrium of aging. Thus, a careful history is necessary to determine exactly what a patient who states, "Doctor, I'm dizzy," is experiencing. After eliminating the misleading symptoms or gait disturbance, "dizziness" usually means either faintness (presyncope) or vertigo (an illusory or hallucinatory sense of movement of the body or environment, most often a feeling of spinning). Operationally, dizziness is classified into three categories: (1) faintness, (2) vertigo, and (3) miscellaneous head sensations. FAINTNESS Prior to an actual faint (syncope), there are often prodromal presyncopal symptoms (faintness) reflecting ischemia to a degree insufficient to impair consciousness (see above). VERTIGO Vertigo is usually due to a disturbance in the vestibular system. The end organs of this system, situated in the bony labyrinths of the inner ears, consist of the three semicircular canals and the otolithic apparatus (utricle and saccule) on each side. The canals transduce angular acceleration, while the otoliths transduce linear acceleration and static gravitational forces, the latter providing a sense of head position in space. The neural output of the end organs is conveyed to the vestibular nuclei in the brainstem via the eighth cranial nerve. The principal projections from the vestibular nuclei are to the nuclei of cranial nerves III, IV, and VI, the spinal cord, the cerebral cortex, and the cerebellum. The vestibuloocular reflex (VOR) serves to maintain visual
stability during head movement and depends on direct projections from the vestibular nuclei to the sixth cranial nerve (abducens) nuclei in the pons and, via the medial longitudinal fasciculus, to the third (oculomotor) and fourth (trochlear) cranial nerve nuclei in the midbrain. These connections account for the nystagmus (to-and-fro oscillation of the eyes) that is an almost invariable accompaniment of vestibular dysfunction. The vestibular nerves and nuclei project to areas of the cerebellum (primarily the flocculus and nodulus) that modulate the VOR. The vestibulospinal pathways assist in the maintenance of postural stability. Projections to the cerebral cortex, via the thalamus, provide conscious awareness of head position and movement. The vestibular system is one of three sensory systems subserving spatial orientation and posture; the other two are the visual system (retina to occipital cortex) and the somatosensory system that conveys peripheral information from skin, joint, and muscle receptors. The three stabilizing systems overlap sufficiently to compensate (partially or completely) for each other's deficiencies. Vertigo may represent either physiologic stimulation or pathologic dysfunction in any of the three systems. Physiologic Vertigo This occurs when (1) the brain is confronted with a mismatch among the three stabilizing sensory systems; (2) the vestibular system is subjected to unfamiliar head movements to which it has never adapted, such as in seasickness; or (3) unusual head/neck positions, such as the extreme extension when painting a ceiling. Intersensory mismatch explains carsickness, height vertigo, and the visual vertigo most commonly experienced during motion picture chase scenes; in the latter, the visual sensation of environmental movement is unaccompanied by concomitant vestibular and somatosensory movement cues. Space sickness, a frequent transient effect of active head movement in the weightless zero-gravity environment, is another example of physiologic vertigo. Pathologic Vertigo This results from lesions of the visual, somatosensory, or vestibular systems. Visual vertigo is caused by new or incorrect spectacles or by the sudden onset of an extraocular muscle paresis with diplopia; in either instance,CNScompensation rapidly counteracts the vertigo. Somatosensory vertigo, rare in isolation, is usually due to a peripheral neuropathy that reduces the sensory input necessary for central compensation when there is dysfunction of the vestibular or visual systems. The most common cause of pathologic vertigo is vestibular dysfunction. The vertigo is frequently accompanied by nausea, jerk nystagmus, postural unsteadiness, and gait ataxia. Since vertigo increases with rapid head movements, patients tend to hold their heads still. Labyrinthine Dysfunction This causes severe rotational or linear vertigo. When rotational, the hallucination of movement, whether of environment or self, is directed away from the side of the lesion. The fast phases of nystagmus beat away from the lesion side, and the tendency to fall is toward the side of the lesion. When the head is straight and immobile, the vestibular end organs generate a tonic resting firing frequency that is equal from the two sides. With any rotational acceleration, the anatomic positions of the semicircular canals on each side necessitate an increased firing rate from one and a commensurate decrease from the other. This change in
neural activity is ultimately projected to the cerebral cortex, where it is summed with inputs from the visual and somatosensory systems to produce the appropriate conscious sense of rotational movement. After cessation of movement, the firing frequencies of the two end organs reverse; the side with the initially increased rate decreases, and the other side increases. A sense of rotation in the opposite direction is experienced; since there is no actual head movement, this hallucinatory sensation is physiologic postrotational vertigo. Any disease state that changes the firing frequency of an end organ, producing unequal neural input to the brainstem and ultimately the cerebral cortex, causes vertigo. The symptom can be conceptualized as the cortex inappropriately interpreting the abnormal neural input from the brainstem as indicating actual head rotation. Transient abnormalities produce short-lived symptoms. With a fixed unilateral deficit, central compensatory mechanisms ultimately diminish the vertigo. Since compensation depends on the plasticity of connections between the vestibular nuclei and the cerebellum, patients with brainstem or cerebellar disease have diminished adaptive capacity, and symptoms may persist indefinitely. Compensation is always inadequate for severe fixed bilateral lesions despite normal cerebellar connections: these patients are permanently symptomatic. Acute unilateral labyrinthine dysfunction is caused by infection, trauma, and ischemia. Often, no specific etiology is uncovered, and the nonspecific terms acute labyrinthitis, acute peripheral vestibulopathy, or vestibular neuritis are used to describe the event. The attacks are brief and leave the patient for some days with a mild positional vertigo. Infection with herpes simplex virus type 1 has been implicated. It is impossible to predict whether a patient recovering from the first bout of vertigo will have recurrent episodes. Acute bilateral labyrinthine dysfunction is usually the result of toxins such as drugs or alcohol. The most common offending drugs are the aminoglycoside antibiotics which damage the fine hair cells of the vestibular end organs and may cause a permanent disorder of equilibrium. Recurrent unilateral labyrinthine dysfunction, in association with signs and symptoms of cochlear disease (progressive hearing loss and tinnitus), is usually due to Meniere's disease (Chap. 29). When auditory manifestations are absent, the term vestibular neuronitis denotes recurrent monosymptomatic vertigo.TIAs of the posterior cerebral circulation (vertebrobasilar insufficiency) very infrequently cause recurrent vertigo without concomitant motor, sensory, visual, cranial nerve, or cerebellar signs. Positional vertigo is precipitated by a recumbent head position, either to the right or to the left. Benign paroxysmal positional (or positioning) vertigo (BPPV) of the posterior semicircular canal is particularly common. Although the condition may be due to head trauma, usually no precipitating factors are identified. It generally abates spontaneously after weeks or months. The vertigo and accompanying nystagmus have a distinct pattern of latency, fatigability, and habituation that differs from the less common central positional vertigo (Table 21-2) due to lesions in and around the fourth ventricle. Moreover, the pattern of nystagmus in posterior canal BPPV is distinctive. The lower eye displays a large-amplitude torsional nystagmus, and the upper eye has a lesser degree of torsion combined with upbeating nystagmus. If the eyes are directed to the
upper ear, the vertical nystagmus in the upper eye increases in amplitude. Vertigo of vestibular nerve origin may occur with diseases that involve the nerve in the petrous bone or the cerebellopontine angle. Except that it is less severe and less frequently paroxysmal, it has many of the characteristics of labyrinthine vertigo. The adjacent auditory division of the eighth cranial nerve also may be affected, which explains the frequent association of vertigo with tinnitus and deafness. The function of the eighth cranial nerve may be disturbed by tumors of the lateral recess (especially schwannomas), less frequently by meningeal inflammation in this region and, rarely, by an abnormal vessel that compresses the nerve. Schwannomas involving the eighth cranial nerve (acoustic neuroma) grow slowly and produce such a gradual reduction of labyrinthine output that central compensatory mechanisms can prevent or minimize the vertigo; auditory symptoms of hearing loss and tinnitus are the most common manifestations. While lesions of the brainstem or cerebellum can cause acute vertigo, associated signs and symptoms usually permit distinction from a labyrinthine etiology (Table 21-3). However, labyrinthine ischemia, presumably due to occlusion of the labyrinthine branch of the internal auditory artery, may be the sole manifestation of vertebrobasilar insufficiency; patients with this syndrome present with the abrupt onset of severe vertigo, nausea and vomiting without tinnitus or hearing loss. Occasionally, an acute lesion of the vestibulocerebellum may present with monosymptomatic vertigo indistinguishable from a labyrinthopathy. Vestibular epilepsy, vertigo secondary to temporal lobe epileptic activity, is rare and almost always intermixed with other epileptic manifestations. Psychogenic vertigo, usually a concomitant of panic attacks or agoraphobia (fear of large open spaces, crowds, or leaving the safety of home), should be suspected in patients so "incapacitated" by their symptoms that they adopt a prolonged housebound status. Most patients with organic vertigo attempt to function despite their discomfort. Organic vertigo is accompanied by nystagmus; a psychogenic etiology is almost certain when nystagmus is absent during a vertiginous episode. Miscellaneous Head Sensations This designation is used, primarily for purposes of initial classification, to describe dizziness that is neither faintness nor vertigo. Cephalic ischemia or vestibular dysfunction may be of such low intensity that the usual symptomatology is not clearly identified. For example, a small decrease in blood pressure or a slight vestibular imbalance may cause sensations different from distinct faintness or vertigo but that may be identified properly during provocative testing techniques. Other causes of dizziness in this category are hyperventilation syndrome, hypoglycemia, and the somatic symptoms of a clinical depression; these patients should have normal neurologic examinations and vestibular function tests. Approach to the Patient The most important diagnostic tool is a careful history focused on the meaning of "dizziness" to the patient. Is it faintness? Is there a sensation of spinning? If either of these is affirmed and the neurologic examination is normal, appropriate investigations for the multiple etiologies of cephalic ischemia or vestibular dysfunction are undertaken.
When the meaning of "dizziness" is uncertain, provocative tests may be helpful. These office procedures simulate either cephalic ischemia or vestibular dysfunction. Cephalic ischemia is obvious if the dizziness is duplicated during maneuvers that produce orthostatic hypotension. Further provocation involves the Valsalva maneuver, which decreases cerebral blood flow and should reproduce ischemic symptoms. The simplest provocative test for vestibular dysfunction is rapid rotation and abrupt cessation of movement in a swivel chair. This always induces vertigo that the patients can compare with their symptomatic dizziness. The intense induced vertigo may be unlike the spontaneous symptoms, but shortly thereafter, when the vertigo has all but subsided, a lightheadedness supervenes that may be identified as "my dizziness." When this occurs, the dizzy patient, originally classified as suffering from "miscellaneous head sensations," is now properly diagnosed as having mild vertigo secondary to a vestibulopathy. Patients with symptoms of positional vertigo should be appropriately tested (Table 21-2); positional testing is more sensitive with special spectacles that preclude visual fixation (Frenzel lenses). A final provocative test, requiring the use of Frenzel lenses, is vigorous head shaking in the horizontal plane for about 10 s. If nystagmus develops after the shaking stops, even in the absence of vertigo, vestibular dysfunction is demonstrated. The maneuver can then be repeated in the vertical plane. If the provocative tests establish the dizziness as a vestibular symptom, an evaluation of vestibular vertigo is undertaken. Evaluation of Patients with Pathologic Vestibular Vertigo The evaluation depends on whether a central etiology is suspected (Table 21-3). If so,MRimaging of the head is mandatory. Such an examination is rarely helpful in cases of recurrent monosymptomatic vertigo with a normal neurologic examination. TypicalBPPVrequires no investigation after the diagnosis is made (Table 21-2). Vestibular function tests serve to (1) demonstrate an abnormality when the distinction between organic and psychogenic is uncertain, (2) establish the side of the abnormality, and (3) distinguish between peripheral and central etiologies. The standard test is electronystagmography (calorics), where warm and cold water (or air) are applied, in a prescribed fashion, to the tympanic membranes, and the slow-phase velocities of the resultant nystagmus from the right and left ears are compared. A velocity decrease from one side indicates hypofunction ("canal paresis"). An inability to induce nystagmus with ice water denotes a "dead labyrinth." Some institutions have the capability of quantitatively determining various aspects of the vestibuloocular reflex using computer-driven rotational chairs and precise oculographic recording of the eye movements. Hyperventilation is the cause of dizziness in many anxious individuals; tingling of the hands and face may be absent. Forced hyperventilation for 1 min is indicated for patients with enigmatic dizziness and normal neurologic examinations. Similarly, depressive symptoms (which patients usually insist are "secondary" to the dizziness) must alert the examiner to a clinical depression as the cause, rather than the effect, of
the dizziness. CNSdisease can produce dizzy sensations of all types. Consequently, a neurologic examination is always required even if the history or provocative tests suggest a cardiac, peripheral vestibular, or psychogenic etiology. Any abnormality on the neurologic examination should prompt appropriate neurodiagnostic studies. TREATMENT Treatment of acute vertigo consists of bed rest and vestibular suppressant drugs such as antihistaminics (meclizine, dimenhydrinate, promethazine), or a tranquilizer with GABA-ergic effects (diazepam). If the vertigo persists beyond a few days, most authorities advise ambulation in an attempt to induce central compensatory mechanisms, despite the short-term discomfort to the patient. Chronic vertigo of labyrinthine origin may be treated with a systematized vestibular rehabilitation program to facilitate central compensation (see alsoTable 21-4). BPPVis often self-limited but, when persistent, responds dramatically to specific repositioning exercise programs designed to empty particulate debris from the posterior semicircular canal. One of these exercises, the Epley procedure, is graphically demonstrated, in four languages, on a website for use in both physician's offices and self-treatment (http://www.charite.de/ch/neuro/vertigo.html). Prophylactic measures to prevent recurrent vertigo are variably effective. Antihistamines are commonly utilized. Meniere's disease may respond to a diuretic or, more effectively, to a very low salt diet (1 g/day). There are a variety of inner ear surgical procedures for refractory Meniere's disease, but these are only rarely necessary. (Bibliography omitted in Palm version) Back to Table of Contents
22. WEAKNESS, MYALGIAS, DISORDERS OF MOVEMENT, AND IMBALANCE Richard K. Olney, Michael J. Aminoff Normal motor function requires integrated muscle activity with appropriate modulation by neuronal activity in the cerebral cortex, basal ganglia, cerebellum, and spinal cord. Symptoms and signs of motor system dysfunction may include weakness, fatigue, myalgias, spasms, cramps, dyskinetic movement, ataxia, imbalance, or disorders in the initiation or planning of movement. WEAKNESS Weakness is a reduction in normal power of one or more muscles. Patients may use the term differently; thus one or more specific examples of weakness should be elicited during the history. Increased fatigability or limitation in function due to pain is often confused with weakness by patients. Increased fatigability is the inability to sustain the performance of an activity that should be normal for a person of the same age, gender, and size. Weakness is described commonly by severity and distribution. Paralysis and the suffix "-plegia" indicate weakness that is so severe that it is complete or nearly complete. "Paresis" refers to weakness that is mild or moderate. The prefix "hemi-" refers to one half of the body, "para-" to both legs, and "quadri-" to all four limbs. Tone is the resistance of a muscle to passive stretch. Central nervous system (CNS) abnormalities that cause weakness generally produce spasticity, an increase in tone due to upper motor neuron disease. Spasticity is velocity-dependent, has a sudden release after reaching a maximum (the "clasp-knife" phenomenon), and predominantly affects antigravity muscles (i.e., upper limb flexors and lower limb extensors). Spasticity is distinct from rigidity and paratonia, two other types of increased tone. Rigidity is increased tone that is present throughout the range of motion (a "lead pipe" or "plastic" stiffness) and affects flexors and extensors equally. In some patients, rigidity has a cogwheel quality that is enhanced by voluntary movement of the contralateral limb (reinforcement). Rigidity occurs with certain extrapyramidal disorders. Paratonia, also referred to as gegenhalten, is increased tone that varies irregularly in a manner that may seem related to the degree of relaxation, is present throughout the range of motion, and affects flexors and extensors equally. Paratonia usually results from disease of the frontal lobes. Weakness with decreased tone (flaccidity) or normal tone occurs with disorders of the motor unit, that is, a single lower motor neuron and all of the muscle fibers it innervates. Three basic patterns of weakness can usually be recognized based on the signs summarized inTable 22-1. One results from upper motor neuron pathology, and the other two from disorders of the motor unit (lower motor neuron and myopathic weakness). Fasciculations and early atrophy help to distinguish lower motor neuron (neurogenic) weakness from myopathic weakness. A fasciculation is a visible or palpable twitch within a single muscle due to the spontaneous discharge of one motor unit. Neurogenic weakness also produces more prominent hypotonia and greater depression of tendon reflexes than myopathic weakness.
PATHOGENESIS Upper Motor Neuron Weakness This pattern of weakness results from disorders that affect the upper motor neurons or their axons in the cerebral cortex, subcortical white matter, internal capsule, brainstem, or spinal cord (Fig. 22-1). Both the pyramidal and bulbospinal pathways contribute to normal strength, tone, coordination, and gait. Upper motor neuron lesions produce weakness through decreased activation of the lower motor neurons. In general, distal muscle groups are affected more severely than proximal ones, and axial movements are spared unless the lesion is severe and bilateral. With corticobulbar involvement, weakness is usually observed only in the lower face and tongue; extraocular, upper facial, pharyngeal, and jaw muscles are almost always spared. With bilateral corticobulbar lesions, pseudobulbar palsy often develops, in which dysarthria, dysphagia, dysphonia, and emotional lability accompany bilateral facial weakness. Spasticity accompanies upper motor neuron weakness but may not be present in the acute phase. Upper motor neuron lesions also affect the ability to perform rapid repetitive movements. Such movements are slow and coarse, but normal rhythmicity is maintained. Finger-nose-finger and heel-knee-shin are performed slowly but adequately. Lower Motor Neuron Weakness This pattern results from disorders of cell bodies of lower motor neurons in the brainstem motor nuclei and the anterior horn of the spinal cord, or from dysfunction of the axons of these neurons as they pass to skeletal muscle (Fig. 22-2). Lower motor weakness is produced by a decrease in the number of motor units that can be activated, through a loss of the a motor neurons or disruption of their connections to muscle. With a decreased number of motor units, fewer muscle fibers are activated with full effort and maximum power is reduced. Loss of g motor neurons does not cause weakness but decreases tension on the muscle spindles. Muscle tone and tendon reflexes depend on g motor neurons, muscle spindles, spindle afferent fibers, and the a motor neurons. A tap on a tendon stretches muscle spindles and activates the primary spindle afferent fibers. These monosynaptically stimulate thea motor neurons in the spinal cord, producing a brief muscle contraction, which is the familiar tendon reflex. When a motor unit becomes diseased, especially in anterior horn cell diseases, it may spontaneously discharge, producing a fasciculation. These isolated small twitches may be seen or felt clinically or recorded by electromyography (EMG) (Chap. 357). When a motor neurons or their axons degenerate, the denervated muscle fibers spontaneously discharge in a manner that cannot be seen or felt but can be recorded with EMG. These small single muscle fiber discharges are called fibrillation potentials. If significant lower motor neuron weakness is present, recruitment of motor units is delayed or reduced, with fewer than normal activated at a given discharge frequency. This contrasts with upper motor neuron weakness, in which a normal number of motor units are activated at a given frequency but in which the maximum discharge frequency is decreased. Myopathic Weakness This pattern of weakness is produced by disorders within the motor unit that affect the muscle fibers or the neuromuscular junctions.
Two types of muscle fibers exist. Type I muscle fibers are rich in mitochondria and oxidative enzymes, produce relatively low force, but have low energy demands that can be supplied by ongoing aerobic metabolism. They produce sustained postural and nonforceful movements. Type II muscle fibers are rich in glycolytic enzymes, can produce relatively high force, but have high energy demands that cannot be supplied for long by ongoing aerobic metabolism. Thus, these units can be activated maximally for only brief periods of time to produce high-force movements. For graded voluntary movements, type I muscle fibers are activated earlier in recruitment. For each muscle fiber, if the nerve terminal releases a normal number of acetylcholine molecules presynaptically and a sufficient number of postsynaptic acetylcholine receptors are opened, the end plate reaches threshold and thereby generates an action potential that spreads across the muscle fiber membrane and into the transverse tubular system. This electrical excitation activates intracellular events that produce an energy-dependent contraction of the muscle fiber (excitation-contraction coupling). Myopathic weakness is produced by a decrease in the number or contractile force of muscle fibers activated within the motor unit. With muscular dystrophies, inflammatory myopathies, or myopathies with muscle fiber necrosis, decreased numbers of muscle fibers survive within many motor units. As demonstrated withEMG, the size of each motor unit action potential is decreased so that motor units must be recruited more rapidly than normal to produce the power necessary for a certain movement. Neuromuscular junction diseases, such as myasthenia gravis, produce weakness in a similar manner, although the loss of muscle fibers within the motor unit is functional rather than actual. Furthermore, the number of muscle fibers activated can vary over time, depending on the state of rest of the neuromuscular junctions. Thus, fatigable weakness is suggestive of myasthenia gravis or another neuromuscular junction disease. Some myopathies produce weakness through loss of contractile force of muscle fibers or through relatively selective involvement of the type II muscle fibers. These may not affect the size of individual motor unit action potentials observed with EMG and are detected by a discrepancy between the electrical activity and force of a muscle. Integrated Movements Most purposeful movements require the integrated coordination of many muscle groups. Consider a simple movement, such as grasping a ball. The primary movement is a flexion of the thumb and fingers of one hand, with opposition of the thumb and little finger. This requires the contraction of several muscles, including flexor digitorum superficialis, flexor digitorum profundus, flexor pollicis longus, flexor pollicis brevis, opponens pollicis, and opponens digiti minimi. These prime movers for this action are called agonists. In order for the grasping to be smooth and forceful, the thumb and finger extensors need to relax at the same rate as the flexors contract. The muscles that act in a directly opposing manner to the agonists are antagonists. A secondary action of the thumb and finger flexors is to flex the wrist; because wrist flexion tends to weaken finger flexion if both occur, activation of wrist extensors assists the grasping movement. Muscles that produce such complementary movements are synergists. Finally, the arm needs to be held in a stable position as the grasp occurs, so that the ball is not knocked away before it is secured. Muscles that stabilize the arm
position are fixators. The coordination of activity by agonists, antagonists, synergists, and fixators is regulated by a three-level hierarchy of motor control. The lowest level of control is mediated through segmental reflexes in the spinal cord. These reflexes facilitate agonists and reciprocally inhibit the antagonists. Spinal segments also control rhythmic patterns of movement that involve more than a single pair of agonists and antagonists. For example, the lumbosacral spinal cord contains the basic programming for cyclical stepping movements that involve the synergistic activation of different muscle groups over time. The intermediate level of control is mediated through the descending bulbospinal pathways, which integrate visual, proprioceptive, and vestibular feedback into the execution of an action. For example, the locomotor center in the midbrain is required to modify the cyclical stepping movements in order that balance be maintained and forward movement occur. The highest level of control is mediated by the cerebral cortex. Superimposition of this highest level of control is necessary for activities such as walking to be goal-directed. Precise movements that are learned and improved through practice are also initiated and controlled by the motor cortex. Although only the agonists are directly activated, during the course of a complex sequence of actions such as playing the piano, the sequential activation of different groups of agonists for each note or chord is a part of the learned motor program. Further, the execution of these actions also involves input from the basal ganglia and cerebellar hemispheres to facilitate agonists, synergists, and fixators and to inhibit undesired antagonists. Apraxia is a disorder of planning and initiating a skilled or learned movement (Chap. 25). Unilateral apraxia of the right hand may be due to a lesion of the left frontal lobe (especially anterior or inferior), the left temporoparietal region (especially the supramarginal gyrus), or their connections. Left body apraxia is produced by lesions of these regions in the right hemisphere or by lesions in the corpus callosum that disconnect the right temporoparietal or frontal regions from those on the left. Bilateral apraxia is often due to bilateral frontal lobe lesions or diffuse bilateral hemispheric disease. Approach to the Patient The mode of onset, distribution, and associated features of weakness should be carefully defined. When there is a discrepancy between the history and physical findings, it is usually because the patient complains of weakness, whereas symptoms are actually due to other causes, such as incoordination or pain limiting effort. Power may be examined in a variety of ways. The patient is asked to push or pull in a specified direction against resistance, and the strength in each muscle group is graded from 0 to 5 by the scale developed by the Medical Research Council (Table 22-2). A second method is indirect testing through observation of task performance such as holding the arms outstretched. This is especially useful in detecting mild, asymmetric upper motor neuron weakness through the observation of a downward drift with pronation of the forearm on one side. A third method is functional testing, which involves quantitation of activities. Common tests include counting the number of times a person can perform a deep-knee bend or step on a stool or chair, or timing the length of time the arms can be held abducted to 90 degrees. When performed serially, functional tests provide useful estimates of changes in the patient's status over time.
Other elements of the motor examination include appraisal of muscular bulk, inspection for fasciculations, and assessment of tone. Fasciculations are most easily determined by observing relaxed limbs that are illuminated from behind, but they can also be palpated as irregular low-amplitude twitches within the muscle. Tone is assessed by passive movement of each limb at its various joints and at several different speeds. In the clinical context of weakness, tone may be spastic or flaccid. The presence of cogwheel rigidity, lead-pipe rigidity, or paratonia suggests a disorder of integrated movements, rather than true weakness. Hemiparesis Hemiparesis results from an upper motor neuron lesion above the midcervical spinal cord; most lesions that produce hemiparesis are located above the foramen magnum. The presence of language disorders, cortical sensory disturbances, cognitive abnormalities, disorders of visual-spatial integration, apraxia, or seizures indicates a cortical lesion. Homonymous visual field defects reflect either a cortical or a subcortical hemispheric lesion. A "pure motor" hemiparesis of the face, arm, and/or leg is due to a small, discrete lesion in the posterior limb of the internal capsule, cerebral peduncle, or upper pons. Some brainstem lesions produce the classic findings of ipsilateral cranial nerve signs and contralateral hemiparesis. These "crossed paralyses" are discussed further inChap. 361. The absence of cranial nerve signs or facial weakness suggests that a hemiparesis is due to a lesion in the high cervical spinal cord, especially if associated with ipsilateral loss of proprioception and contralateral loss of pain and temperature sense (the Brown-Sequard syndrome). However, most spinal cord lesions produce quadriparesis or paraparesis. Acute or episodic hemiparesis usually has a vascular pathogenesis, either ischemia or a primary hemorrhage (Chap. 361). Less commonly, hemorrhage may occur into brain tumors (Chap. 370) or from rupture of normal vessels due to trauma (Chap. 369); the trauma may be trivial in patients who are anticoagulated or elderly. Less likely possibilities include a focal inflammatory lesion from multiple sclerosis (Chap. 371), abscess, or sarcoidosis (Chap. 318). Evaluation begins immediately with a computed tomography (CT) scan of the brain (Fig. 22-3). If CT is normal and an ischemic stroke is unlikely, magnetic resonance imaging (MRI) of the brain or cervical spine may be indicated. Subacute hemiparesis that evolves over days or weeks has a long differential diagnosis. A common cause is subdural hematoma; this readily treatable condition must always be considered, especially in elderly or anticoagulated patients, even in the absence of a history of trauma (Chap. 369). Infectious possibilities include cerebral bacterial abscess (Chap. 372), fungal granuloma or meningitis (Chap. 374), and parasitic infection. Weakness from malignant primary and metastatic neoplasms may evolve over days to weeks (Chap. 370). AIDS (Chap. 309) may present with subacute hemiparesis due to toxoplasmosis or primaryCNSlymphoma. Noninfectious inflammatory processes, such as multiple sclerosis (Chap. 371) or, less commonly, sarcoidosis, are further considerations. If the brainMRI is normal and if cortical and hemispheric signs are not present, MRI of the cervical spine may be required. Chronic hemiparesis that evolves over months is usually due to a neoplasm (Chap. 370), an unruptured arteriovenous malformation (Chap. 361), a chronic subdural
hematoma (Chap. 369), or a degenerative disease (Chaps. 363 to 366). The initial diagnostic test is often an MRIof the brain, especially if the clinical findings suggest brainstem pathology. If MRI of the brain is normal, the possibility of a foramen magnum or high cervical spinal cord lesion should be considered. Paraparesis An intraspinal lesion at or below the upper thoracic spinal cord level is most commonly responsible. A sensory level over the trunk identifies the approximate level of the cord lesion. Paraparesis can also result from lesions at other locations that disturb upper motor neurons (especially parasagittal lesions and hydrocephalus) and lower motor neurons (anterior horn cell disorders, cauda equina syndromes, and occasionally peripheral neuropathies). Acute or episodic paraparesis due to spinal cord disease may be difficult to distinguish from disorders affecting lower motor neurons or cerebral hemispheres. Recurrent episodes of paraparesis are often due to multiple sclerosis or to vascular malformations of the spinal cord. With acute spinal cord disease, the upper motor neuron deficit is usually associated with incontinence and a sensory disturbance of the lower limbs that extends rostrally to a level on the trunk; tone is typically flaccid, and tendon reflexes absent. In such cases, the diagnostic approach starts with an imaging study of the spinal cord (Fig. 22-3). Compressive lesions (particularly epidural tumor, abscess, or hematoma), spinal cord infarction (proprioception is usually spared), an arteriovenous fistula or other vascular anomaly, and transverse myelitis, among other causes may be responsible (Chap. 368). Diseases of the cerebral hemispheres that produce acute paraparesis include anterior cerebral artery ischemia (shoulder shrug also affected), superior sagittal sinus or cortical venous thrombosis, and acute hydrocephalus. If upper motor neuron signs are associated with drowsiness, confusion, seizures, or other hemispheric signs but not a sensory level over the trunk, the diagnostic approach starts with anMRI of the brain. Paraparesis is part of the cauda equina syndrome, which may result from trauma to the low back, a midline disk herniation, or intraspinal tumor; although sphincters are affected, hip flexion is often spared, as is sensation over the anterolateral thighs. Rarely, paraparesis is caused by a rapidly evolving peripheral neuropathy such as Guillain-Barre syndrome or by a myopathy. In such cases, electrophysiologic studies are diagnostically helpful and refocus the subsequent evaluation (Chaps. 378 and381). Subacute or chronic paraparesis with spasticity is caused by upper motor neuron disease. When paraparesis evolves over weeks or months with lower limb sensory loss and sphincter involvement, possible spinal cord disorders include multiple sclerosis, intraparenchymal tumor, chronic spinal cord compression from degenerative disease of the spine, subacute combined degeneration due to vitamin B12deficiency, viral infections (especially human T cell leukemia/lymphoma virus I), and hereditary or other degenerative diseases. Primary progressive multiple sclerosis usually presents in the fourth or fifth decade as progressive paraparesis (Chap. 371). Gliomas of the spinal cord typically produce a progressive myelopathy that is painful (Chap. 370). The clinical approach begins with anMRI of the spinal cord. If the imaging study is normal and spasticity is present, MRI of the brain may be indicated. If hemispheric signs are present, parasagittal meningioma or chronic hydrocephalus is likely and MRI of the brain is the initial test. Progression over months to years is typical of degenerative disorders such as primary lateral sclerosis (Chap. 365) and hereditary disorders such as
familial spastic paraparesis and adrenomyeloneuropathy (Chap. 368). In the rare situations when a chronic paraparesis is due to a lower motor neuron or myopathic etiology, the localization is usually suspected on clinical grounds by the absence of spasticity and confirmed byEMG and nerve conduction tests. Quadriparesis or Generalized Weakness Generalized weakness may be due to disorders of the central nervous system or of the motor unit. Although the terms quadriparesis and generalized weakness are often used interchangeably, quadriparesis is more often chosen when an upper motor neuron cause is suspected and generalized weakness when a disease of the motor unit is likely. Weakness fromCNSdisorders is usually associated with changes in consciousness or cognition, with increased muscle tone and muscle stretch reflexes, and with alterations of sensation. Most neuromuscular causes of intermittent weakness are associated with normal mental function, diminished muscle tone, and hypoactive muscle stretch reflexes. Exceptions are some causes of acute quadriparesis due to upper motor neuron disorders in which transient hypotonia is present. The major causes of intermittent weakness are listed inTable 22-3. A patient with generalized fatigability without objective weakness may have the chronic fatigue syndrome (Chap. 384). Acute Quadriparesis Acute quadriparesis with onset over minutes may result from disorders of upper motor neurons (e.g., anoxia, hypotension, brainstem or cervical cord ischemia, trauma, and systemic metabolic abnormalities) or muscle (electrolyte disturbances, certain inborn errors of muscle energy metabolism, toxins, or periodic paralyses). Onset over hours to weeks may, in addition to the above, be due to lower motor neuron disorders. Guillain-Barre syndrome (Chap. 378) is the most common lower motor neuron weakness that progresses over days to several weeks; the finding of an elevated protein level in the cerebrospinal fluid is helpful but may be absent early in the course. If stupor or coma is present, the evaluation begins with aCT scan of the brain. If upper motor neuron signs are present but the patient is alert, the initial test is usually anMRI of the cervical cord. If weakness is lower motor neuron, myopathic, or uncertain in origin, the clinical approach starts with blood studies for muscle enzymes and electrolytes and anEMG and nerve conduction study. Subacute or Chronic Quadriparesis When quadriparesis due to upper motor neuron disease develops over weeks, months, or years, the distinction between disorders of the cerebral hemispheres, brainstem, and cervical spinal cord is usually possible by clinical criteria alone. The diagnostic approach begins with anMRI of the clinically suspected site of pathology. Lower motor neuron disease usually presents with weakness that is most profound distally, whereas myopathic weakness is typically proximal; the evaluation then begins withEMG and nerve conduction studies. Monoparesis This is usually due to lower motor neuron disease, with or without associated sensory involvement. Upper motor neuron weakness occasionally presents with a monoparesis of distal and nonantigravity muscles. Myopathic weakness is rarely limited to one limb. Acute Monoparesis Distinguishing between upper and lower motor neuron disorders may be difficult clinically because tone and reflexes are frequently decreased in both at presentation. If the weakness is predominantly in distal and nonantigravity muscles and
not associated with sensory impairment or pain, focal cortical ischemia is likely (Chap. 361); in this setting, diagnostic possibilities are similar to those for acute hemiparesis. Sensory loss and pain usually accompany acute lower motor neuron weakness. The distribution of weakness is commonly localized to a single nerve root or peripheral nerve within one limb but occasionally reflects involvement of the brachial or lumbosacral plexus. If lower motor neuron weakness is suspected, or if the pattern of weakness is uncertain, the clinical approach begins with anEMG and nerve conduction study. Subacute or Chronic Monoparesis Weakness with atrophy of one limb that develops over weeks or months is almost always lower motor neuron in origin. If the weakness is associated with numbness, a peripheral nerve or spinal root origin is likely; uncommonly, the brachial or lumbosacral plexus is affected. If numbness is absent, anterior horn cell disease is likely. In either case, an electrodiagnostic study is indicated. If upper rather than lower motor neuron signs are present, a tumor, vascular malformation, or other cortical lesion affecting the precentral gyrus may be responsible. Alternatively, if the leg is affected, a small thoracic cord lesion, often a tumor or multiple sclerosis, may be present. In these situations, the approach begins with an imaging study of the suspicious area. Distal Weakness Involvement of two or four limbs distally suggests lower motor neuron or peripheral nerve disease. Acute distal lower limb weakness occurs occasionally from an acute toxic polyneuropathy or cauda equina syndrome. Distal symmetric weakness usually develops over weeks, months, or years and is due to metabolic, toxic, hereditary, degenerative, or inflammatory diseases of peripheral nerves (Chap. 377). With peripheral nerve disease, weakness is usually less severe than numbness. Anterior horn cell disease may begin distally but is typically asymmetric and is not associated with numbness (Chap. 365). Rarely, myopathies also present with distal weakness (Chap. 381). The first step in evaluation is an electrodiagnostic study (Fig. 22-3). Proximal Weakness Proximal weakness of two or four limbs suggests a disorder of muscle or, less commonly, neuromuscular junction or anterior horn cell. Myopathy often produces symmetric weakness of the pelvic or shoulder girdle muscles (Chap. 381). Diseases of the neuromuscular junction (such as myasthenia gravis) may present with symmetric proximal weakness (Chap. 380), often associated with ptosis, diplopia, or bulbar weakness and fluctuating in severity during the day. Extreme fatigability present in some cases of myasthenia gravis may even suggest episodic weakness, but strength rarely returns fully to normal. The proximal weakness of anterior horn cell disease is most often asymmetric, but may be symmetric if familial (Chap. 365). Numbness does not occur with any of these diseases. The evaluation usually begins with determination of the serum creatine kinase level and electrophysiologic studies. Weakness in a Restricted Distribution In some patients, weakness does not fit any of the above patterns. Examples include weakness limited to the extraocular, hemifacial, bulbar, or respiratory muscles. If unilateral, restricted weakness is usually due to lower motor neuron or peripheral nerve disease, such as in a facial palsy (Chap. 367) or an isolated superior oblique muscle paresis (Chap. 28). Relatively symmetric weakness of extraocular or bulbar muscles is usually due to a myopathy (Chap. 381) or neuromuscular junction disorder (Chap. 380). Bilateral facial palsy with areflexia
suggests Guillain-Barre syndrome (Chap. 378). Worsening of relatively symmetric weakness with fatigue is characteristic of neuromuscular junction disorders (Chap. 380). Asymmetric bulbar weakness is usually due to motor neuron disease. Weakness limited to respiratory muscles is uncommon and is usually due to motor neuron disease, myasthenia gravis, or polymyositis/dermatomyositis (Chap. 382). MYALGIAS, SPASMS, AND CRAMPS Spontaneous or exercise-related discomfort from muscles is usually benign and is rarely caused by a definable neuromuscular disease. However, a number of disorders of the motor system are characteristically painful. Some terms for muscular discomfort or involuntary contractions, such as myalgias, spasms, and cramps, are often used interchangeably by patients but have a more specific meaning to physicians. Other terms, such as aching, heaviness, and stiffness, are less specific. Myalgias are pains that are felt in muscle; the term does not imply an involuntary contraction. Spasms and cramps refer to episodes of involuntary contraction of one or more muscles. Cramps are usually painful, whereas spasms are not necessarily uncomfortable. MYALGIAS Proximal or generalized weakness associated with myalgias is usually due to an inflammatory, metabolic, endocrine, or toxic myopathy (Chap. 381). Spontaneous myalgias not accompanied by objective weakness are often without a clear cause unless associated with a well-defined systemic illness. Myalgias are a common manifestation of fever or infection, especially influenza. Muscle pains and stiffness with elevated serum creatine kinase concentration is common in hypothyroidism, even in patients without objective weakness. Polymyalgia rheumatica (Chap. 317) is characterized by diffuse myalgias and joint stiffness that predominantly affect the pelvic and shoulder girdles in a patient over 50 years of age who has anorexia, mild weight loss, and low-grade fever. Limitation of activity from the myalgias and joint stiffness also leads to disuse atrophy and may give the impression of weakness. However,EMG, serum creatine kinase levels, and muscle biopsy are normal. The erythrocyte sedimentation rate is elevated in most patients, and features of giant-cell arteritis are present in 25%. Diffuse myalgias are common in many rheumatologic diseases, in which the diagnosis and treatment are based on other symptoms and signs. Myalgias are occasionally present in dermatomyositis/polymyositis, but most patients have weakness without significant pain. Fibromyalgia (fibrositis, fibromyositis) is associated with pain and tenderness of muscle and adjacent connective tissue (Chap. 325). Fatigue, insomnia, and depression are often present, but objective weakness, elevation of serum creatine kinase level, or elevation of the erythrocyte sedimentation rate does not occur. The diagnosis is dependent upon identifying characteristic focal "trigger points." Focal Myalgias Focal muscle pain is often traumatic. Rupture of muscle tendons such as the biceps or gastrocnemius muscle may produce visible muscle shortening. Many such tears resolve without surgery but leave an abnormal appearance to the muscle belly. Nontraumatic focal muscle pain is often related to adjacent nonmuscular disorders (e.g., unilateral gastrocnemius pain due to deep venous thrombosis). Rarely, focal muscle pain may be caused by ischemic infarction or bacterial myositis, if acute, or by
neoplasm, parasitic infection, sarcoidosis, or other inflammation or infection, if subacute or chronic. Exertional Myalgias Myalgias following unaccustomed, strenuous physical activity occur in normal individuals are often associated with laboratory evidence for muscle damage, such as an elevation of serum creatine kinase, edema of muscles onMRI, necrosis of muscle fibers on biopsy, and rarely myoglobinuria. Similar symptoms and laboratory abnormalities characterize certain metabolic disorders of muscle, such as carnitine palmitoyl transferase and glycolytic pathway enzyme deficiencies. The association of objective weakness during an episode of myalgias suggests a metabolic muscle disease. The development of an acute contracture (the inability to relax a muscle due to energy depletion) with the myalgias suggests a metabolic muscle disease with a glycolytic enzyme deficiency (Chap. 383). Exertional myalgias with muscle fiber necrosis also occur in muscular dystrophy with partial deficiency of dystrophin, and certain mitochondrial cytopathies (Chap. 383). Exertional myalgias with elevated creatine kinase concentration but without weakness also occur in hypothyroidism, and when confined to the legs may be due to vascular or neurogenic intermittent claudication. Most patients with exertional myalgias and no weakness do not have a definable abnormality. SPASMS AND CRAMPS Involuntary contraction of muscle may occur with disorders of theCNS, lower motor neuron, or muscle. Contractions that originate within the CNS and are associated with upper motor neuron signs are usually referred to as spasms and generally affect the flexors or extensors of one or more limbs. Those that originate within the CNS and are not associated with upper motor neuron signs include movement disorders discussed below, as well as the rare stiff-person syndrome and tetanus. Muscle rigidity from active muscle contraction can occur in the malignant hyperthermia syndrome, usually associated with general anesthesia. In the neuroleptic malignant syndrome, muscle rigidity arises from CNS overactivity and is present in muscle. Involuntary contractions that originate in the lower motor neurons are usually cramps, occasionally tetany, or rarely neuromyotonia. Spasms that originate in muscle or muscle membrane are usually a delayed relaxation after voluntary contraction, either myotonia or rarely a contracture. These conditions may be difficult to distinguish clinically but are often well characterized byEMGstudies. Stiff-Person Syndrome This rare syndrome is characterized by slowly progressive muscle stiffness and superimposed spasms. The stiffness commonly begins in the low back and spreads over months up the spine and into the limbs but not into the jaw. The gait becomes stiff, and there is hyperlordosis of the lumbar spine. Spasms are often produced by startle. Emotional stress tends to worsen the stiffness as well as the frequency and severity of spasms. The spontaneous motor activity disappears during sleep. The syndrome is often associated with diabetes mellitus and can be paraneoplastic, accompanying Hodgkin's lymphoma, small cell cancer of the lung, and breast cancer. Most patients have a serum antibody against glutamic acid decarboxylase, an enzyme responsible for synthesis of the inhibitory neurotransmitter g-aminobutyric acid (GABA). Stiffness results from loss of descending brainstem or segmental spinal inhibitory influences on the lower motor neurons.EMGstudies reveal
continuous motor unit activity that is similar to voluntary effort with preservation of the silent period to muscle stretch. Stiffness and spasms typically respond partially to treatment with baclofen or benzodiazepines. Tetanus This rare hyperexcitable state results from exposure to tetanus toxin in patients infected with Clostridium tetani (Chap. 143). Painful spasms typically begin with jaw closure (trismus) and soon become generalized.EMGstudies reveal continuous motor unit activity that is similar to voluntary effort except for loss of the silent period to muscle stretch. Cramps These are the most common type of involuntary muscle contraction. Cramps are a painful contraction of a single muscle that produces a palpable knot within the muscle for seconds to minutes and is relieved by passive stretch of the muscle or spontaneously.EMGstudies reveal motor unit activity that has too high a discharge frequency to be voluntary. If cramps are associated with weakness, the weakness is almost always lower motor neuron in origin. When strength is normal, no definable condition is usually found, although dehydration, hypothyroidism (Video 330-1), or uremia is occasionally present. If prominent, membrane stabilizing drugs, such as carbamazepine, may provide symptomatic benefit. Tetany Tetany is characterized by contraction of distal muscles of the hands (carpal spasm with extension of interphalangeal joints and adduction and flexion of the metacarpophalangeal joints) and feet (pedal spasm) and is associated with tingling around the mouth and distally in the limbs. Tetany with carpopedal spasms is a common manifestation of hypocalcemia or respiratory alkalosis (even from hyperventilation).EMGstudies reveal single or more often grouped motor unit discharges at low discharge frequency. Neuromyotonia (Isaac's Syndrome) Neuromyotonia is characterized by muscle stiffness at rest that persists during sleep and by delayed relaxation after voluntary effort. Distal limb muscles are usually affected most severely, but all skeletal muscle may be involved. Gait may be stiff, and close inspection of the muscle reveals undulation of the overlying skin due to continuous muscle fiber contractions (myokymia). The continuous muscle fiber activity generates heat, and excessive sweating is common.EMGstudies commonly reveal myokymic discharges, especially in familial cases. Rarely, EMGs record high-frequency neuromyotonic discharges. Autoantibodies against voltage-gated potassium channels have been demonstrated in some cases, and plasma exchange may be effective. Myotonia This is a nonpainful delay in the relaxation of muscle after voluntary activity. Delay in opening the hand after a forceful grip (grip myotonia) is common. These disorders are usually familial and worsen in cold weather.EMGdemonstrates a waxing and waning discharge of individual muscle fibers. Contracture A painful inability to relax a muscle after voluntary activity due to energy depletion characterizes certain metabolic disorders with failure of energy production, such as myophosphorylase deficiency (McArdle's disease).EMGstudies reveal electrical silence.
MOVEMENT DISORDERS Movement disorders are neurologic syndromes in which abnormal movements (or dyskinesias) occur due to a disturbance of fluency and speed of voluntary movement or the presence of unintended extra movements. Because they are so distinct from the pyramidal disorders that cause upper motor neuron weakness, movement disorders are often referred to as extrapyramidal diseases. Hyperkinetic movement disorders are those in which an excessive amount of spontaneous motor activity is seen or in which abnormal involuntary movements occur. Hypokinetic movement disorders are characterized by akinesia or bradykinesia, in which purposeful motor activity is absent or reduced. This is often described as "poverty of movement." PATHOGENESIS Movement disorders result from disease of the basal ganglia, paired subcortical gray matter structures consisting of the caudate and the putamen (which together are called the striatum), the internal and external segments of the globus pallidus, the subthalamic nucleus, and the substantia nigra. The major interconnections and neurotransmitters involved in basal ganglia circuits are illustrated inFig. 22-4A. An understanding of this circuitry can explain, in part, the perturbation that occurs in both the hypo- and hyperkinetic disorders. Parkinson's disease (Video 361-1) (Chap. 363), the prototypic hypokinetic movement disorder, results from a loss of dopaminergic neurons in the substantia nigra pars compacta. This leads to less excitation of striatal neurons that express the D1type of dopamine receptors and less inhibition of D2striatal neurons, both contributing to reduced facilitation of cortically initiated movement (Fig. 22-4B). The resting tremor of Parkinson's disease is less readily explained by this model but may result from effects on cholinergic interneurons in the striatum. Huntington's disease (Chap. 362), a hyperkinetic movement disorder, may be explained by selective loss of D2striatal neurons, resulting in disinhibition of cortically initiated movements without normal feedback control. The pathogenesis of hemiballismus is similar -- a direct lesion of the glutamatergic neurons in the subthalamic nucleus (usually from a stroke) leads to disinhibition of thalamocortical projections. Approach to the Patient An algorithm for the interpretation of abnormal movements is illustrated inFig. 22-5. The initial step is to determine if the movement disorder is due to an excess or a poverty of movement (i.e., a hyperkinetic or a hypokinetic movement disorder). Hyperkinetic Movement Disorders Abnormal involuntary movements are divided into those that are rhythmical and those that are irregular. Those that are rhythmical are termed tremors, with the uncommon exception of palatal and segmental myoclonus. Tremors are divided into three types: rest, postural, and intention tremor. A rest tremor is maximal at rest and becomes less prominent with activity. It is characteristic of parkinsonism, a hypokinetic movement disorder, and is therefore commonly associated with bradykinesia and cogwheel rigidity. A rest tremor that develops acutely is usually due to toxins [such as exposure to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine
(MPTP)] or dopamine blocking drugs (such as phenothiazines). If insidious in onset, the diagnostic approach is the same as for Parkinson's disease (Chap. 363). A postural tremor is maximal while limb posture is actively maintained against gravity; it is lessened by rest and is not markedly enhanced during voluntary movement toward a target. A postural tremor that develops acutely is usually due to toxic or metabolic factors (for example, hyperthyroidism) or stress. The insidious onset of a postural tremor suggests a benign or familial essential tremor (Chap. 363). An intention tremor is most prominent during voluntary movement toward a target and is not present during postural maintenance or at rest. It is a sign of cerebellar disease (Chap. 364). Asterixis, which may superficially resemble a tremor, is an intermittent inhibition of muscle contraction that occurs with metabolic encephalopathy (Chap. 376). This leads, for example, to a momentary and repetitive partial flexion of the wrists during attempted sustained wrist extension. Involuntary movements that are irregular are characterized further by their speed and site of occurrence and by whether they can be suppressed voluntarily. The slowest are athetosis and dystonia. Athetosis is a slow, writhing, sinuous movement that occurs nearly continuously in distal muscles. Dystonia is a slowly varying but nearly continuous deviation of posture about one or more joints; it may occur in a proximal or distal limb or in axial structures. Dystonia is a more sustained deviation of posture than athetosis, although these two phenomena overlap considerably. The further evaluation of athetosis and dystonia are discussed inChap. 363. Among the rapid irregular movements, tics are controlled with voluntary effort, while the others are not. Tics often occur repetitively in a single location but are sometimes multifocal (Chap. 363). Chorea, hemiballismus, and myoclonus are rapid, irregular jerks that cannot be consciously suppressed. Hemiballismus is the most distinctive among them. It is manifest as a sudden and often violent flinging movement of a proximal limb, usually an arm (Chap. 363). Hemiballismus usually develops acutely due to infarction of the contralateral subthalamic nucleus but occasionally develops subacutely or chronically due to other lesions of this nucleus. Chorea is a rapid, jerky, irregular movement that tends to occur in the distal limbs or face but may also occur in proximal limb and axial structures. Acute or subacute onset is usually toxic due to excess levodopa or dopamine-agonist therapy or, less often, neuroleptics, birth control pills, pregnancy (chorea gravidarum), hyperthyroidism, or the antiphospholipid syndrome. In children, it may be associated with rheumatic fever and, in such cases, is referred to as Sydenham's chorea. The gradual onset of chorea is typical of degenerative neurologic diseases, such as Huntington's chorea (Chap. 362). Myoclonus is a rapid, brief, irregular movement that is usually multifocal. Myoclonus can occur spontaneously at rest, in response to sensory stimuli, or with voluntary movements. It is a symptom that occurs in a wide variety of metabolic and neurologic disorders. Posthypoxic intention myoclonus is a special myoclonic syndrome that occurs as a sequel to transient cerebral anoxia. Myoclonus may result from lipid storage disease, encephalitis, Creutzfeldt-Jakob disease, or metabolic encephalopathies due to respiratory failure, chronic renal failure, hepatic failure, or electrolyte imbalance.
Myoclonus is also a feature of certain types of epilepsy, as discussed inChap. 360. Palatal and segmental myoclonus are uncommon rhythmic forms of myoclonus that may resemble tremor; they are caused by structural disease of the brainstem or spinal cord at the level of the abnormal movement. Hypokinetic Movement Disorders These syndromes are manifest as bradykinesia, with a masked, expressionless facial appearance, loss of associated limb movements during walking, and rigid en bloc turning. If bradykinesia is associated only with a rest tremor, cogwheel rigidity, or impairment of postural reflexes (especially with a tendency to fall backwards), Parkinson's disease is likely (Chap. 363). If cognitive, language, upper motor neuron, sensory, or autonomic signs are also present, a multisystem degenerative neurologic disease is present.*These disorders are discussed in Chaps. 363, 364, and 366. IMBALANCE AND DISORDERS OF GAIT Imbalance is the impaired ability to maintain the intended orientation of the body in space. It is generally manifest as difficulty in maintaining an upright posture while standing or walking; a severe imbalance may also affect the ability to maintain posture while seated. Patients with imbalance commonly complain of a feeling of unsteadiness or dysequilibrium. Whereas imbalance and unsteadiness are synonymous, dysequilibrium implies the additional component of impaired spatial orientation even while lying down. Patients with dysequilibrium commonly also experience vertigo, defined as an hallucination of rotatory movement. PATHOGENESIS Imbalance and Limb Ataxia Imbalance results from disorders of the spinal cord (spinocerebellar) or vestibular sensory input, the integration of these inputs in the brainstem or midline cerebellum, or the motor output to the spinal neurons that control axial and proximal muscles. Limb ataxia results from disorders of the spinocerebellar and corticopontocerebellar inputs, the integration of these inputs in the intermediate and lateral cerebellum, or the output to the spinal neurons (via the red nucleus and rubrospinal tract) or to the cortex. These pathways ensure adequate speed, fluency, and integration of limb movements. The lateral cerebellar hemispheres coordinate a complex feedback circuit that modulates cortically initiated limb movement. Sensory ataxia is caused by lesions that affect the peripheral sensory fibers, dorsal root ganglia cells, posterior columns of the spinal cord, lemniscal system in the brainstem, thalamus, or parietal cortex; relevant anatomy is discussed inChap. 23. Impairment of the proprioceptive sensory feedback to the cerebellum, basal ganglia, and cortex produces sensory ataxia. Sensory ataxia results in imbalance and disturbs the fluency and integration of movements that can be partially alleviated by visual feedback. Disorders of Gait Walking is one of the most complicated motor activities. Essentially all structures discussed in this chapter participate in normal walking. Cyclical stepping movements produced by the lumbosacral spinal cord centers are modified by cortical, basal ganglionic, brainstem, and cerebellar influences based on proprioceptive, vestibular, and visual feedback.
Approach to the Patient Examination of coordination, balance, and gait is typically performed at the same time. The finger-nose-finger and the heel-knee-shin maneuvers are observed for signs of incoordination in general and dysmetria in particular. Dysmetria consists of irregular errors in the amplitude and force of limb movements. This is accentuated near the target or point of intention and hence termed intention tremor. The patient is also asked to maintain the arms outstretched against a resistance that is suddenly removed; excessive rebound indicates cerebellar dysfunction. The ability of the patient to rapidly and repetitively tap the hands and feet is assessed for speed and rhythmicity. Errors in rhythm (irregular rate, velocity, or force) indicate dysdiadochokinesia. Slow, coarse, but rhythmical movements indicate upper motor neuron disorders. The patient is asked to demonstrate how to comb the hair or brush the teeth to assess the ability to initiate and execute a simple sequence of activity. Balance is examined by having the patient stand stationary with the feet together. If this position can be maintained, the eyes are closed for 5 to 10 s. Accentuation of sway or actual loss of balance is assessed. If balance is momentarily lost, several trials may be necessary to determine if the loss is consistently in the same direction. Walking along an uncrowded space, such as a hallway, is observed. Symmetry of arm swing and various phases of the gait cycle are observed. Walking is then performed for several steps on the heels, on the toes, and in tandem. Imbalance An algorithm for interpretation of imbalance is presented inFig. 22-6. Cerebellar ataxia results from disorders of the cerebellum or of its afferent inputs or efferent projections. Abnormalities of the midline cerebellar vermis or the flocculonodular lobe produce truncal ataxia which is usually revealed during the process of rising from a chair, assuming the upright stance with the feet together, or performing some other activity while standing. Once a desired position is reached, imbalance may be surprisingly mild. As walking begins, the imbalance recurs. Patients usually learn to lessen the imbalance by walking with the legs widely separated. The imbalance is usually not lateralized and may be accompanied by symmetric nystagmus. Abnormalities of the intermediate and lateral portions of the cerebellum typically produce impaired limb movements rather than truncal ataxia. If involvement is asymmetric, lateralized imbalance is common and usually associated with asymmetric nystagmus. Clinical signs of cerebellar limb ataxia include dysmetria, intention tremor, dysdiadochokinesia, and abnormal rebound. Muscle tone is often modestly reduced; this contributes to the abnormal rebound due to decreased activation of segmental spinal cord reflexes and also to pendular reflexes, i.e., a tendency for a tendon reflex to produce multiple swings to and fro after a single tap.*For further discussion of cerebellar diseases, see Chap. 364. Imbalance with vestibular dysfunction is characterized by a consistent tendency to fall to one side. The patient commonly complains of vertigo rather than imbalance, especially if the onset is acute. Acute vertigo associated with lateralized imbalance but no other neurologic signs is often due to disorders of the semicircular canal (Chap. 21); the presence of other neurologic signs suggests brainstem ischemia (Chap. 361) or multiple sclerosis (Chap. 371). When the vestibular dysfunction is peripheral, positional
nystagmus and vertigo tend to resolve if a provocative position is maintained (extinction) or repeated (habituation). Lateralized imbalance of gradual onset or persisting for more than 2 weeks, accompanied by nystagmus, may result from lesions of the semicircular canal or vestibular nerve, brainstem, or cerebellum. Imbalance with sensory ataxia is characterized by marked worsening when visual feedback is removed. The patient can often assume the upright stance with feet together cautiously with eyes open. With eye closure, balance is rapidly lost (positive Romberg sign) in various directions at random. Sensory examination reveals impairment of proprioception at the toes and ankles, usually associated with an even more prominent abnormality of vibratory perception. Prompt evaluation for vitamin B12deficiency is important, as this disorder is reversible if recognized early (Chap. 368). Depression or absence of reflexes points to peripheral nerve disorders (Chap. 377). Spasticity with extensor plantar responses suggests posterior column and spinal cord disorders (Chap. 368). Rarely, sensory ataxia produces lateralized imbalance. In these cases, the disorder is usually in the parietal lobe or thalamus (Chap. 23), but may also be due to an asymmetric sensory neuropathy (Chap. 377) or posterior column disease (Chap. 368). Sensory limb ataxia is similar to cerebellar limb ataxia but is markedly worse when the eyes are closed. Examination also reveals abnormal proprioception and vibratory perception. The approach focuses on localizing the proprioceptive impairment to the peripheral nerves (Chap. 377), the posterior columns of the spinal cord (Chap. 368), or rarely the parietal lobe. Other forms of imbalance occur, but the fundamental problem is usually a primary disorder of strength, extrapyramidal function, or cortical initiation of movement. Abnormal Gait Each of the disorders discussed in this chapter produces a characteristic gait disturbance. If the neurologic examination is normal except for an abnormal gait, diagnosis may be difficult even for the experienced clinician. Hemiparetic gait characterizes spastic hemiparesis. In its most severe form, an abnormal posture of the limbs is produced by spasticity. The arm is adducted and internally rotated, with flexion of the elbow, wrist, and fingers and with extension of the hip, knee, and ankle. Forward swing of the spastic leg during walking requires abduction and circumduction at the hip, often with contralateral tilt of the trunk to prevent the toes catching on the floor as the leg is advanced. In its mildest form, the affected arm is held in a normal position, but swings less than the normal arm. The affected leg is flexed less than the normal leg during its forward swing and is more externally rotated. A hemiparetic gait is a common residual sign of a stroke (Chap. 361). Paraparetic gait (Video 361-3) is a walking pattern in which both legs are moved in a slow, stiff manner with circumduction, similar to the leg movement in a hemiparetic gait. In many patients, the legs tend to cross with each forward swing ("scissoring"). A paraparetic gait is a common sign of spinal cord disease (Chap. 368) and also occurs in cerebral palsy. Steppage gait is produced by weakness of ankle dorsiflexion. Because of the partial or
complete foot drop, the leg must be lifted higher than usual to avoid catching the toe on the floor during the forward swing of the leg. If unilateral, steppage gait is usually due to L5 radiculopathy, sciatic neuropathy, or peroneal neuropathy (Chap. 377). If bilateral, it is the common result of a distal polyneuropathy or lumbosacral polyradiculopathy (Chap. 377). Waddling gait results from proximal lower limb weakness, most often from myopathy (Chap. 381) but occasionally from neuromuscular junction disease (Chap. 380) or a proximal symmetric spinal muscular atrophy (Chap. 365). With weakness of hip flexion, the trunk is tilted away from the leg that is being moved to lift the hip and provide extra distance between the foot and the floor, and the pelvis is rotated forward to assist with forward motion of the leg. Because pelvic girdle weakness is customarily bilateral, the pelvic lift and rotation alternates from side to side, giving the waddling appearance to the gait. Parkinsonian gait (Video 361-1) is characterized by a forward stoop, with modest flexion at the hips and knees. The arms are flexed at the elbows and adducted at the shoulders, often with a 4- to 6-Hz resting pronation-supination tremor but little other movement, even during walking. Walking is initiated slowly by leaning forward and maintained with short rapid steps, during which the feet shuffle along the floor. The pace tends to accelerate (festination) as the upper body gradually leans further ahead of the feet, whether movement is forward (propulsion) or backward (retropulsion). The postural instability leads to falls (Chap. 363). Apraxic gait (Video 361-4) results from bilateral frontal lobe disease with impaired ability to plan and execute sequential movements. This gait superficially resembles that of parkinsonism, in that the posture is stooped and any steps taken are short and shuffling. However, initiation and maintenance of walking are impaired in a different manner. Each movement that is required for walking can usually be performed, if tested in isolation while sitting or lying. However, when asked to step forward while standing, a long pause often occurs before any attempt is made to flex at the hip and advance, as if the patient is "glued to the ground." Once walking is initiated, it is not maintained, even in an abnormal festinating manner. Rather, after one or several steps are taken, walking is stopped for several seconds or longer. The process is then repeated. Dementia and incontinence may coexist. Choreoathetotic gait is characterized by an intermittent, irregular movement that disrupts the smooth flow of a normal gait. Flexion or extension movements at the hip are common and unpredictable but readily observed as a pelvic lurch (Chap. 363). Cerebellar ataxic gait (Video 361-3) is a broad-based gait disorder in which the speed and length of stride varies irregularly from step to step. With midline cerebellar disease, as in alcoholics, posture is erect but the feet are separated; lower limb ataxia is commonly present as well. Assumption of a particular stance or a change in position may cause instability, yet balance can usually be maintained well with the eyes open or closed. Walking may be rapid, but cadence is irregular. Although patients commonly lack confidence in the stability of their walking, only minimal support is often required for reassurance. With disease of the cerebellar hemispheres, limb ataxia and nystagmus are commonly present as well (Chap. 364).
Sensory ataxic gait may resemble a cerebellar gait, with its broad-based stance and difficulty with change in position. However, although balance may be maintained with the eyes open, loss of visual input through eye closure results in rapid loss of balance with a fall (positive Romberg sign), unless the physician assists the patient. Vestibular gait is one in which the patient consistently tends to fall to one side, whether walking or standing. Cranial nerve examination demonstrates an obviously asymmetric nystagmus. The possibilities of unilateral sensory ataxia and hemiparesis are excluded by the findings of normal proprioception and strength (Chap. 21). Astasia-abasia is a typical hysterical gait disorder. Although the patient usually has normal coordination of leg movements in bed or while sitting, the patient is unable to stand or walk without assistance. If distracted, stationary balance is sometimes maintained and several steps are taken normally, followed by a dramatic demonstration of imbalance with a lunge toward the examiner's arms or a nearby bed. (Bibliography omitted in Palm version) Back to Table of Contents
23. NUMBNESS, TINGLING, AND SENSORY LOSS - Arthur K. Asbury NORMAL SENSATION Normal somatic sensation reflects a continuous day and night monitoring process that occupies considerable moment-to-moment nervous system capacity. Little of this activity reaches consciousness under ordinary conditions. In contrast, disordered sensation, particularly if experienced as painful, is alarming and dominates the sufferer's attention. Abnormalities of sensation, especially if painful, tend to make those suffering seek medical help. The physician must be able to recognize abnormal sensations by how they are described, know their type and likely site of origin, and understand their implications.*For a consideration of pain, see Chap. 12. Positive and Negative Phenomena Abnormal sensory phenomena may be divided into two categories, positive and negative. The prototypical positive phenomenon is tingling (pins-and-needles), and the principal negative phenomenon is numbness. In addition to tingling, positive sensory phenomena include other altered sensations that are often described as pricking, bandlike, lightning-like shooting feelings (lancinations), aching, knifelike, twisting, drawing, pulling, tightening, burning, searing, electrical, or raw feelings. These descriptors are frequently the actual words used by patients. Such sensations may or may not be experienced as painful. Positive phenomena usually result from trains of impulses generated at a site or sites of lowered threshold or heightened excitability along a sensory pathway, either peripheral or central. The nature and severity of an abnormal sensation depend on the number, rate, timing, and distribution of ectopic impulses and the type and function of nervous tissue in which they arise. Because positive phenomena represent excessive activity in sensory pathways, they are not necessarily associated with any sensory deficit (loss) upon examination. Negative phenomena represent loss of sensory function and are characterized by diminished or absent feeling, often experienced as numbness. In contrast to positive phenomena, negative phenomena are accompanied by abnormal findings on sensory examination. In disorders affecting peripheral sensation, it is estimated that at least half the afferent axons innervating a given site are lost or functionless before sensory deficit can be demonstrated by clinical examination. This estimate probably varies according to how rapidly sensory nerve fibers have lost function. If the rate of loss is slow and chronic, lack of cutaneous feeling may be unnoticed by the patient and difficult to demonstrate on examination, even though few sensory fibers are functioning. Rapidly evolving sensory abnormality usually evokes both positive and negative phenomena and is readily recognized by patients. Subclinical degrees of sensory dysfunction not demonstrable on clinical sensory examination may be revealed by sensory nerve conduction studies or somatosensory cerebral evoked potentials (Chap. 357). Sensory symptoms may be either positive or negative, but sensory signs on examination are always a measure of negative phenomena. Terminology Words used to characterize sensory disturbance are descriptive and have been arrived at mainly by convention. Paresthesia and dysesthesia are general terms used to denote sensory symptoms (positive phenomena) and are usually stated in the
plural form. Paresthesias usually refer to tingling or pins-and-needles sensations but may also include a wide variety of other abnormal sensations, excepting pain. Sometimes "paresthesias" carry the implication that the abnormal sensations are perceived without an apparent stimulus. Dysesthesia is a more general term used to subsume all types of abnormal sensations, even painful ones, whether a stimulus is evident or not. While dysesthesias and paresthesias refer to sensations described by patients, another set of terms refers to sensory abnormalities found on examination. These include hypesthesia or hypoesthesia (reduction of cutaneous sensation to a specific type of testing such as pressure, light touch, and warm or cold stimuli); anesthesia (complete absence of skin sensation to the same stimuli plus pinprick); and hypalgesia (referring to reduced pain perception, i.e., nociception, such as the pricking quality elicited by a pin). Hyperesthesia means pain in response to touch. Similarly, allodynia describes the situation in which a nonpainful stimulus, once perceived, is experienced as painful, even excruciating. An example is elicitation of a painful sensation by application of a vibrating tuning fork. Hyperalgesia denotes severe pain in response to a mildly noxious stimulus, and hyperpathia, a broad term, encompasses all the phenomena described by hyperesthesia, allodynia, and hyperalgesia. With hyperpathia, the threshold for a sensory stimulus is increased and the perception is delayed but once felt, is unduly painful. Disorders of deep sensation, arising from muscle spindles, tendons, and joints, affect proprioception (position sense). Manifestations include imbalance (particularly with eyes closed or in the dark), clumsiness of precision movements, and unsteadiness of gait, which are referred to collectively as sensory ataxia (Chap. 22). Other findings on examination usually, but not invariably, include reduced or absent joint position and vibratory sensibility and absent deep tendon reflexes in the affected limbs. Romberg's sign is positive, which means that the patient sways or topples when asked to stand with feet close together and eyes closed. In severe states of deafferentation involving deep sensation, the patient cannot walk or stand unaided or even sit unsupported. Continuous, sometimes wormlike involuntary movements, called pseudoathetosis, of the outstretched hands and fingers occur, particularly with eyes closed. Such patients are severely disabled. Anatomy of Sensation Cutaneous afferent innervation is conveyed by a rich variety of receptors, both naked nerve endings (nociceptors and thermoreceptors) and encapsulated terminals (mechanoreceptors). Each type of receptor has its own set of sensitivities to specific stimuli, size and distinctness of receptive fields, and adaptational qualities. Much of the knowledge about these receptors has come from the development of techniques to study single intact nerve fibers intraneurally in awake unanesthetized human subjects. It is possible not only to record from single nerve fibers, large or small, but also to stimulate single fibers in isolation. A single impulse, whether elicited by a natural stimulus or evoked by electrical microstimulation, in a large myelinated afferent fiber may be both perceived and localized. Afferent fibers of all sizes in peripheral nerve trunks traverse the dorsal roots and enter the dorsal horn of the spinal cord (Fig. 23-1). From there the smaller fibers take a different route to the parietal cortex than the larger fibers. The polysynaptic projections
of the smaller fibers (unmyelinated and small myelinated), which subserve mainly nociception, temperature sensibility, and touch, cross and ascend in the opposite anterior and lateral columns of the spinal cord, through the brainstem, to the ventral posterolateral (VPL) nucleus of the thalamus, and ultimately project to the postcentral gyrus of the parietal cortex (Chap. 12). This is referred to as the spinothalamic pathway, or anterolateral system. The larger fibers, which subserve tactile and position sense and kinesthesia, project rostrally in the posterior column on the same side of the spinal cord and make their first synapse in the gracile or cuneate nuclei of the lower medulla. The second-order neuron decussates and ascends in the medial lemniscus located medially in the medulla and in the tegmentum of the pons and midbrain and synapses in the VPL. The third-order neuron projects to parietal cortex; this large fiber system is referred to as the posterior column-medial lemniscal pathway (lemniscal, for short). Note that although the lemniscal and the anterolateral pathways both project up the spinal cord to the thalamus, it is the (crossed) anterolateral pathway that is referred to as the spinothalamic tract, by convention. Although the fiber types and functions that make up the spinothalamic and lemniscal systems are relatively well known, it has been found that many other fibers, particularly those associated with touch, pressure, and position sense, ascend in a diffusely distributed pattern both ipsilaterally and contralaterally in the anterolateral quadrants of the spinal cord. This explains why an individual with a complete lesion of the posterior columns of the spinal cord may have little sensory deficit on examination. EXAMINATION OF SENSATION The main tasks of the sensory examination are tests of primary sensation. By convention these include the sense of pain, touch, vibration, joint position, and thermal sensation, both hot and cold (Table 23-1). Detailed descriptions of how to perform the various tests of the sensory examination can be found in standard texts (see "Bibliography"). Some general principles pertain. First, the examiner must depend on subjective patient response, particularly when using cutaneous stimuli (pin, touch, vibration, warm or cold). This factor may complicate the interpretation of the sensory examination. Second, with complaints of numbness, patients should be asked to outline on themselves the borders of numb areas. Third, some patients are only partially examinable. In a stuporous patient, sensory examination is reduced to observing the briskness of withdrawal in response to a pinch or other noxious stimulus. Comparison of response on one side of the body to the other is essential. In the alert but uncooperative patient, cutaneous sensation may be unexaminable. However, it is usually possible to get some idea of proprioceptive function by noting the patient's best performance of movements requiring balance and precision. Fourth, sensory examination of a patient who has no neurologic complaints should be abbreviated and may consist of pin, touch, and vibration testing in the hands and feet plus evaluation of stance and gait, including the Romberg maneuver. Evaluation of stance and gait also tests the integrity of motor and cerebellar systems. Primary Sensation (See Table 23-1) The sense of pain is usually tested with a pin, asking the patient to focus on the pricking or unpleasant quality of the stimulus and not just the pressure or touch sensation elicited. Areas of hypalgesia should be mapped by
proceeding radially from the most hypalgesic site (Figs. 23-2 and23-3). Temperature sensation, to both hot and cold, is probably best tested with water flasks filled with water of the desired temperature, using a thermometer to verify the temperature. This is impractical in most settings. An alternative way to test cold sensation is to touch a metal object, such as a tuning fork at room temperature, to the skin. For testing warm temperatures, the tuning fork or other metal object may be held under warm water of the desired temperature and then used. Both cold and warm should be tested because different receptors respond to each. Touch is usually tested with a wisp of cotton or a fine camelhair brush. In general, it is better to avoid testing touch on hairy skin because of the profusion of sensory endings that surround each hair follicle. Joint position testing is a measure of proprioception, one of the most important functions of the sensory system. With the patient keeping eyes closed, joint position is tested in the great toe and in the fingers. If errors are made in recognizing the direction of passive movements of the toe or the finger, more proximal joints should be tested. A test of proximal joint position sense, primarily at the shoulder, is performed by asking the patient to bring the two index fingers together with the arms extended and the eyes closed. Normal individuals should be able to do this quite accurately, with errors of a centimeter or less. The sense of vibration is tested with a tuning fork, preferably a large one that vibrates at 128 Hz. Vibration is usually tested at bony prominences, beginning distally at the malleoli of the ankles, and at the knuckles. If abnormalities are found, more proximal sites can be examined. Vibratory thesholds at the same site in the patient and the examiner can be compared for control purposes. Quantitative Sensory Testing Effective sensory testing devices have been developed over the past two decades. Quantitative sensory testing is particularly useful for serial evaluation of cutaneous sensation in clinical trials. Threshold testing for touch and vibratory and thermal sensation is the most widely used application. Cortical Sensation Cortical sensory testing includes two-point discrimination, touch localization, and bilateral simultaneous stimulation and tests for graphesthesia and stereognosis, to name the most commonly used methods. Abnormalities of these sensory tests, in the presence of normal primary sensation in an alert cooperative patient, signify a lesion of the parietal cortex or thalamocortical projections to the parietal lobe. If primary sensation is altered, these cortical discriminative functions will usually be abnormal, too. Comparisons should always be made between analogous sites on the two sides of the body because the deficit with a specific parietal lesion is likely to be hemilateral. Side-to-side comparisons hold true for all cortical sensory testing. Two-point discrimination is tested by special calipers, the points of which may be set from 2 mm to several centimeters apart and then applied simultaneously to the site to be tested. The pulp of the fingertips is a common site to test; a normal individual can distinguish about 3-mm separation of points there.
Touch localization is usually carried out by light pressure with the examiner's fingertip, asking the patient, whose eyes are closed, to identify the site of touch. It is usual to ask the patient to touch the same site with a fingertip. Bilateral simultaneous stimulation at analogous sites (e.g., the dorsa of both hands) can be carried out to determine whether the perception of touch is extinguished consistently on one side or the other. The phenomenon is referred to as extinction on bilateral simultaneous stimulation. Graphesthesia means the capacity to recognize with eyes closed letters or numbers drawn by the examiner's fingertip on the palm of the hand. Once again, the comparison of one side with the other is of prime importance. Inability to recognize numbers or letters is termed agraphesthesia. Stereognosis refers to the ability to identify common objects by palpation, recognizing their shape, texture, and size. Common standard objects are the best test objects, such as a marble, a paper clip, or coins. Patients with normal stereognosis should be able to distinguish a dime from a penny and a nickel from a quarter without looking. Patients should only be allowed to feel the object with one hand at a time. If they are unable to identify it in one hand, it should be placed in the other for comparison. Individuals unable to identify common objects and coins in one hand who can do so in the other are said to have astereognosis of the abnormal hand. LOCALIZATION OF SENSORY ABNORMALITIES Sensory symptoms and signs can result from lesions at almost any level of the nervous system, including parietal cortex, deep white matter, thalamus, brainstem, spinal cord, spinal root, peripheral nerve, and sensory receptor. Noting the distribution and nature of sensory symptoms and signs is the most important way to localize their source. The extent, configuration, symmetry, quality, and severity are the key observations. Dysesthesias without sensory findings by examination can be difficult to interpret. To illustrate, tingling dysesthesias in an acral distribution (hands and feet) can have more than one interpretation. Distal dysesthesias can be systemic in origin, e.g., secondary to hyperventilation, or can be induced by a medication, such as the diuretic acetazolamide. Distal dysesthesias can also be an early event in an evolving polyneuropathy or can herald a myelopathy, such as with vitamin B12deficiency. Sometimes distal dysesthesias have no definable basis. In contrast, dysesthesias that correspond to a particular peripheral nerve territory denote a lesion of that nerve trunk. For instance, dysesthesias restricted to the fifth digit and the adjacent one-half of the fourth finger on one hand reliably point to disorder of the ulnar nerve, most commonly at the elbow. Nerve and Root In focal nerve trunk lesions severe enough to cause a deficit, sensory abnormalities are readily mapped and generally have discrete boundaries (Figs. 23-2 and23-3). Root lesions, referred to as radicular, are frequently accompanied by deep, aching pain along the course of the related nerve trunk. With compression of a fifth lumbar (L5) or first sacral (S1) root, as may occur with a ruptured intervertebral disc, sciatica is a frequent manifestation. With a lesion affecting a single root, sensory deficit
in the distribution of that root is often minimal or not demonstrable at all. This is because adjacent root territories overlap extensively. Polyneuropathies are generally graded, distal, and symmetric in distribution of deficit (Chap. 377). Dyesthesias begin in the toes and ascend symmetrically, followed by numbness. When dyesthesias reach the knees, they have usually also appeared in the fingertips. The process appears to be nerve length-dependent, and the deficit is often described as "stocking-glove" in type. Although most polyneuropathies are pansensory and affect all modalities of sensation, selective sensory dysfunction according to nerve fiber size may occur. In polyneuropathies that affect small nerve fibers selectively, the hallmark is burning, painful dysesthesias with reduced pinprick and thermal sensation but with sparing of proprioception, motor function, and even deep tendon jerks. Touch is variably involved, but when spared, the sensory pattern is referred to as sensory dissociation. Sensory dissociation patterns can be seen with spinal cord lesions (see below) as well as with small fiber neuropathies. In contrast to small fiber polyneuropathies, large fiber polyneuropathies are characterized by position sense deficit, imbalance, absent tendon jerks, and variable motor dysfunction but preservation of most cutaneous sensation. Dyesthesias, if present at all, tend to be tingling or bandlike. Spinal Cord (SeeChap. 368) If the spinal cord is transected, all sensation is lost below the level of transection. Bladder and bowel function are also lost, as is motor function. Hemisection of the spinal cord produces the Brown-Sequard syndrome, which involves absent pain and temperature sensation on the opposite side below the lesion, and loss of proprioceptive sensation and loss of motor power on the same side below the lesion (seeFigs. 23-1 and368-1). Dissociated sensory deficit patterns (see above) are also a sign of spinothalamic tract involvement in the spinal cord, especially if the deficit is unilateral and has an upper level on the torso. Bilateral spinothalamic tract involvement occurs with lesions affecting the center of the spinal cord, such as happens with expansion of the central canal in syringomyelia. Sensory dissociation is characteristic of syringomyelia. Brainstem Harlequin patterns of sensory disturbance, in which one side of the face and the opposite side of the body are affected, localize to the lateral medulla. Here a small lesion may damage both the ipsilateral descending trigeminal tract and ascending spinothalamic fibers subserving the opposite arm, leg, and hemitorso (see "Lateral medullary syndrome" in Fig. 361-7). In the tegmentum of the pons and midbrain, where the lemniscal and spinothalamic tracts merge, a lesion here causes pansensory loss on the contralateral body. Thalamus Hemisensory disturbance with tingling numbness from head to foot is often thalamic in origin but can also be anterior parietal. If abrupt in onset, the lesion is likely to be due to a small stroke (lacunar infarction), particularly if localized to the thalamus. Occasionally, with lesions affecting theVPL or adjacent white matter, a syndrome of thalamic pain, also called Dejerine-Roussy syndrome, may ensue. This persistent unrelenting hemipainful state is often described in dramatic terms such as "like the flesh is being torn from my limbs" or "as though that side is bathed in acid" (Chap. 12). Cortex With lesions of the parietal lobe, either of the cortex or of subjacent white matter,
the most prominent symptoms are contralateral hemineglect, hemi-inattention, and a tendency not to use the affected hand and arm. Tests of primary sensation may be normal or altered. Anterior parietal infarction may present as a pseudothalamic syndrome with crossed hemilateral loss of primary sensation. Dysesthesias or a sense of numbness may also occur, and rarely a painful state. Focal Sensory Seizures These are generally due to lesions in or near the postcentral gyrus. Symptoms of focal sensory seizures are usually combinations of numbness and tingling, but frequently additional more complex sensations are present, such as a rushing feeling, a sense of warmth, a sense of movement without visible motion, or other unpleasant dysesthesias. Duration of seizures is variable; they may be transient, lasting only seconds, or they may persist for hours. Focal motor features (clonic jerking) may supervene, and seizures can become generalized with loss of consciousness. Likely sites of symptoms are unilaterally in the lips, face, digits, or foot, and symptoms may spread as in a Jacksonian march. On occasion, symptoms may occur in a symmetric bilateral fashion, for instance, in both hands; this results from involvement of the second sensory area (unilaterally) located in the rolandic area at and just above the Sylvian fissure. (Bibliography omitted in Palm version) Back to Table of Contents
24. ACUTE CONFUSIONAL STATES AND COMA - Allan H. Ropper Confusional states and coma are among the most common problems in general medicine. They account for a substantial portion of admissions to emergency wards and are a frequent cause of distress on all hospital services. Because clouding of consciousness and a diminished level of consciousness frequently coexist and result from many of the same diseases, they are presented together here, but from a medical perspective they have different clinical characteristics and physiologic explanations. The basis of consciousness has long been a topic of great interest to psychologists and philosophers and is the subject of a vast literature. Physicians have been mainly concerned with impairments in the level of consciousness (coma, stupor, drowsiness) and with alterations of consciousness, meaning an inability to think coherently, i.e., with accustomed clarity and speed. The latter, broadly termed confusion relates to lessened awareness, perception, apperception (the interpretation of perceptions), thinking, expression in language and action, and all forms of intellection that are dependent on the continuous integration of mental processes. Normal awareness provides a background to our inner mental life, which flows from infancy to death like a "stream of thought," to use William James's metaphor. Self-awareness requires that a person experience these thoughts and be able to reflect and operate upon them. Almost all instances of diminished alertness can be traced to widespread abnormalities of the cerebral hemispheres or to reduced activity of a special thalamocortical alerting system termed the reticular activating system (RAS). The proper functioning of this system, its ascending projections to the cortex, the cortex itself, and corticothalamic connections are required to maintain alertness and coherence of thought. THE CONFUSIONAL STATE Confusion is a mental and behavioral state of reduced comprehension, coherence, and capacity to reason. Inattention, as defined by the inability to sustain uninterrupted thought and actions, and disorientation are its earliest outward signs. As the state of confusion worsens, there are more global mental failings, including impairments of memory, perception, comprehension, problem solving, language, praxis, visuospatial function, and various aspects of emotional behavior that are each attributable to particular regions of the brain. In other instances an apparent confusional state may arise from an isolated deficit in mental function such as an impairment of language (aphasia), loss of memory (amnesia), or lack of appreciation of spatial relations of self and the external environment (agnosia), but the attributes of the problem are then quite different (Chap. 25). Confusion is also a feature of dementia, in which case the chronicity of the process, as in the instance of Alzheimer's disease, distinguishes it from an acute encephalopathy (Chap. 26). The confused patient is usually subdued, not inclined to speak, and is physically inactive. A state of confusion that is accompanied by agitation, hallucinations, tremor, and illusions (misperceptions of environmental sight, sound, or touch) is termed delirium, as typified by delirium tremens from alcohol or drug withdrawal. In psychiatric circles, delirium often refers, albeit imprecisely, to all acute states of confusion with clouding of consciousness and incoherence of thought.
Approach to the Patient Confusion and delirium always signify a disorder of the nervous system. They may be the major manifestation of a head injury; a seizure; drug toxicity (or drug withdrawal); a metabolic disorder resulting from hepatic, renal, pulmonary or cardiac failure; a systemic infection; meningitis or encephalitis; or a chronic dementing disease. The search for these manifold causes begins with a careful history emphasizing the patient's condition before the onset of confusion. The clinical examination should focus on signs of diminished attentiveness, disorientation, and drowsiness and on the presence of localizing neurologic signs. From the clinical data the clinician is directed to the appropriate laboratory tests discussed further on. Often, even after all diagnostic tests are completed, one may still not know the cause of a confusional state. The proper approach is to observe the patient in the hospital for a number of days under stable conditions. New clues may appear or an obscure confusion perhaps related to a medication, may clear up, while other causes such as renal or hepatic failure may worsen and lead to coma. Orientation and memory are tested by asking the patient in a forthright manner the date, inclusive of month, day, year, and day of week; the precise place; and some items of generally acknowledged and universally known information (the names of the President and Vice President, a recent national catastrophe, the state capital). Further probing may be necessary to reveal a defect -- why is the patient in the hospital; what is his or her address, zip code, telephone number, social security number? Problems of increasing complexity may be pursued, but they usually provide little additional information. Attention and coherence of thought can be gauged by the clarity of and speed of responses while the history is being given but are examined more explicitly by having the patient repeat strings of numbers (most adults easily retain seven digits forward and four backward), spell a word such as "world" backwards, and perform serial calculations -- tests of serial subtraction of 3 from 30 or 7 from 100 are useful. It is the inability to sustain coherent mental activity in performing tasks such as these that exposes the most subtle confusional states. Other salient neurologic findings are the level of alertness, which fluctuates if there is drowsiness; indications of focal damage of the cerebrum such as hemiparesis, hemianopia, and aphasia; or adventitious movements of myoclonus or partial convulsions. The language of the confused patient may be disorganized and rambling, even to the extent of incorporating paraphasic words. These features, along with impaired comprehension that is due mainly to inattention, may be mistaken for aphasia. One of the most specific signs of a metabolic encephalopathy is asterixis, which is an arrhythmic flapping tremor that is typically elicited by asking the patient to hold the arms outstretched with the wrists and hands fully extended. After a few seconds, there is a large jerking lapse in the posture of the hand and then a rapid return to the original position. The same movements can be appreciated in any tonically held posture, even of the tongue, and in extreme form the movements may intrude on voluntary limb motion. Bilateral asterixis always signifies a metabolic encephalopathy, e.g., from hepatic failure, hypercapnia, or from drug ingestion, especially with anticonvulsant
medications. Myoclonic jerking and tremor in an awake patient are typical of uremic encephalopathy or the use of antipsychotic drugs such as lithium, phenothiazines, or butyrophenones; myoclonus with coma may also signify anoxic cerebral damage. Confusion in the postoperative period is common but at times so subtle as to escape attention. Cardiac and orthopedic procedures are particularly likely to produce disorientation or delirium in susceptible patients. Often a careful history will reveal that a mild but compensated dementia existed prior to the operation. Medications, particularly those with anticholinergic activity (including meperidine), inadvertent withdrawal from sleeping pills or alcohol, fever, and any of the endogenous metabolic derangements listed above may be responsible, or a stroke may have occurred. Frequently confusion cannot be attributed to any single factor and it clears in several days. In many cases, particularly in the elderly, transient confusion and drowsiness arise with a febrile infection of the urinary tract, lungs, blood, or peritoneum. The term septic encephalopathy is currently used to describe this association, but the mechanism by which infection or inflammation leads to cerebral dysfunction is unknown. Fever can also alter brain function in a way that makes preexisting focal signs worse. Distinguishing dementia from an acute confusional state is a great problem, especially in the elderly, since the two may coexist if a fever, other acute medical problem, or a poorly tolerated medication supervenes in a mildly demented patient, producing a so-called beclouded dementia. The memory loss of dementia brings about a confusional state that varies little in severity from hour to hour and day to day. Poor mental performance is derived mainly from incomplete recollection, inadequate access to names and ideas, and the inability to retain new information, thus affecting orientation and factual knowledge. In contrast to the acute confusional states, attention, alertness, and coherence are preserved until the most advanced stages. Eventually dementia produces a chronic confusion with breakdown of all types of mental performance, and the distinction from an acute encephalopathy depends mainly on the longstanding nature of the condition. Treatment of the confusional state requires that all unnecessary medication be stopped, metabolic alterations be rectified, and infection be treated. Skilled nursing and a quiet room with a window are important. Careful explanations should be given at regular intervals to the family. In the elderly, regular reorientation and active measures to avoid risk factors (sleep deprivation, immobility, and vision and hearing impairments) reduce the number and severity of episodes of delirium in hospitalized patients. COMA AND RELATED DISORDERS OF CONSCIOUSNESS The unnatural situation of reduced alertness and responsiveness represents a continuum that in severest form is called coma, a deep sleeplike state from which the patient cannot be aroused. Stupor defines lesser degrees of unarousability in which the patient can be awakened only by vigorous stimuli, accompanied by motor behavior that leads to avoidance of uncomfortable or aggravating stimuli. Drowsiness, which is familiar to all persons, simulates light sleep and is characterized by easy arousal and the persistence of alertness for brief periods. Drowsiness and stupor are usually attended by some degree of confusion. In clinical practice these terms should be
supplemented by a narrative description of the level of arousal and of the type of responses evoked by various stimuli precisely as observed at the bedside. Such an account is preferable to ambiguous terms such as semicoma or obtundation, the definitions of which differ between physicians. Several other neurologic conditions render patients apparently unresponsive and simulate coma, and certain other subsyndromes of coma must be considered separately because of their special significance. Among the latter, the vegetative state signifies an awake but unresponsive state. Most of these patients were earlier comatose and after a period of days or weeks emerge to an unresponsive state in which their eyelids are open, giving the appearance of wakefulness. Yawning, grunting, swallowing, as well as limb and head movements persist, but there are few, if any, meaningful responses to the external and internal environment -- in essence, an "awake coma." Although respiratory and autonomic functions are retained, the term "vegetative" is nonetheless unfortunate as it is subject to misinterpretation by lay persons. Always there are accompanying signs that indicate extensive damage in both cerebral hemispheres, e.g., decerebrate or decorticate limb posturing and absent responses to visual stimuli (see below). Cardiac arrest and head injuries are the most common causes of the vegetative state (Chaps. 369 and376). The prognosis for regaining mental faculties once the vegetative state has supervened for several months is almost nil hence the term persistent vegetative state. Most instances of dramatic recovery, when investigated carefully, are found to yield to the usual rules for prognosis, but it must be acknowledged that rare instances of awakening to a condition of dementia and paralysis have been documented. Certain other clinical states are prone to be misinterpreted as stupor or coma. Akinetic mutism refers to a partially or fully awake patient who is able to form impressions and think but remains immobile and mute, particularly when unstimulated. The condition may result from damage in the regions of the medial thalamic nuclei, the frontal lobes (particularly situated deeply or on the orbitofrontal surfaces), or from hydrocephalus. The term abulia is used to describe a mental and physical slowness and lack of impulse to activity that is in essence a mild form of akinetic mutism, with the same anatomic origins. Catatonia is a curious hypomobile and mute syndrome associated with a major psychosis. In the typical form patients appear awake with eyes open but make no voluntary or responsive movements, although they blink spontaneously, swallow, and may not appear distressed. As often, the eyes are half-open as if the patient is in a fog or light sleep. There are signs that indicate voluntary attempts to appear less than fully responsive, though it may take some ingenuity on the part of the examiner to demonstrate these. Eyelid elevation is actively resisted, blinking occurs in response to a visual threat, and the eyes move concomitantly with head rotation, all signs belying a brain lesion. It is characteristic but not invariable for the limbs to retain the posture, no matter how bizarre, in which they have been placed by the examiner ("waxy flexibility," or catalepsy.) Upon recovery, such patients have some memory of events that occurred during their catatonic stupor. The appearance is superficially similar to akinetic mutism, but clinical evidence of brain damage is lacking. The locked-in state describes a pseudocoma in which an awake patient has no means of producing speech or volitional limb, face, and pharyngeal movements in order to indicate that he or she is awake, but vertical eye movements and lid elevation remain
unimpaired, thus allowing the patient to signal. Such individuals have written entire treatises using Morse code. Infarction or hemorrhage of the ventral pons, which transects all descending corticospinal and corticobulbar pathways, is the usual cause. A similar awake but deefferented state occurs as a result of total paralysis of the musculature in severe cases of Guillain-Barre syndrome (Chap. 378), critical illness neuropathy (Chap. 376), and pharmacologic neuromuscular blockade. THE ANATOMY AND PHYSIOLOGY OF UNCONSCIOUSNESS To the extent that all complex waking behaviors require the widespread participation of the cerebral cortex, consciousness cannot exist without the activity of these structures. A loosely grouped aggregation of neurons located in the upper brainstem and medial thalamus, theRAS, maintains the cerebral cortex in a state of wakeful consciousness. It follows that the principal causes of coma are (1) lesions that damage a substantial portion of the RAS; (2) destruction of large portions of both cerebral hemispheres; and (3) suppression of thalamocerebral function by drugs, toxins, or by internal metabolic derangements such as hypoglycemia, anoxia, azotemia, or hepatic failure. The classic animal experiments of Moruzzi and Magoun, published in 1949, and subsequent human clinicopathologic observations have established that the regions of the reticular formation that are critical to the maintenance of wakefulness extend from the caudal midbrain to the lower thalamus. A most important practical consideration derives from the anatomic proximity of theRAS to structures that are concerned with pupillary function and eye movements. Pupillary enlargement and loss of vertical and adduction movements of the globes suggest that upper brainstem damage may be the source of coma. Although circumscribed lesions confined to one or both cerebral hemispheres do not affect the brainstem RAS, a large mass on one side of the brain may cause coma by secondarily compressing the upper brainstem and consequently producing abnormalities of the pupils and eye movements (see discussion of transtentorial herniation below). This type of indirect effect is most typical of cerebral hemorrhages and of rapidly expanding tumors within a cerebral hemisphere. In all cases the degree of diminished alertness also relates to the rapidity of evolution and the extent of compression of the RAS. The neurons of theRAS are thought to project rostrally to the cortex primarily via thalamic relay nuclei that in turn exert a tonic influence on the activity of the entire cerebral cortex. The behavioral arousal effected by somesthetic, auditory, and visual stimuli depends upon the rich reciprocal innervation that the RAS receives from these sensory systems. The relays between the RAS and the thalamic and cortical areas utilize a variety of neurotransmittors. Of these, the effect of arousal on acetylcholine and on the biogenic amines has been studied more extensively. Cholinergic fibers connect the midbrain to other areas of the upper brainstem, thalamus, and cortex. Serotonin and norepinephrine also subserve important functions in regulation of the sleep-wake cycle (Chap. 27). Their roles in arousal and coma have not been clearly established, although the alerting effects of amphetamines are likely to be mediated by catecholamine release. Coma Due to Cerebral Mass Lesions and Herniations The cranial cavity is separated into compartments by infoldings of the dura -- the two cerebral hemispheres are
separated by the falx, and the anterior and posterior fossae by the tentorium. Herniation refers to displacement of brain tissue away from a mass and into a compartment that it normally does not occupy. Many of the signs associated with coma, and indeed coma itself, can be attributed to these tissue shifts. Herniation can be transfalcial (displacement of the cingulate gyrus under the falx and across the midline), transtentorial (displacement of the medial temporal lobe into the tentorial opening), and foraminal (downward forcing of the cerebellar tonsils into the foramen magnum; Fig. 24-1). Uncal transtentorial herniation refers to impaction of the anterior medial temporal gyrus (the uncus) into the anterior portion of the tentorial opening. The displaced tissue compresses the third nerve as it traverses the subarachnoid space and results in enlargement of the ipsilateral pupil (putatively because the fibers subserving parasympathetic pupillary function are located peripherally in the nerve). The coma that follows may be due to lateral compression of the midbrain against the opposite tentorial edge by the displaced parahippocampal gyrus (Fig. 24-2). In some cases the lateral displacement causes compression of the opposite cerebral peduncle, producing a Babinski response and hemiparesis contralateral to the original hemiparesis (the Kernohan-Woltman sign). In addition to compressing the upper brainstem, tissue shifts, including herniations, may compress major blood vessels, particularly the anterior and posterior cerebral arteries as they pass over the tentorial reflections, thus producing brain infarctions. The distortions may also entrap portions of the ventricular system, resulting in regional hydrocephalus. Central transtentorial herniation denotes a symmetric downward movement of the upper thalamic region through the tentorial opening. Miotic pupils and drowsiness are the heralding signs. Both temporal and central herniations are thought to cause progressive compression of the brainstem from above: first the midbrain, then the pons, and finally the medulla. The result is a sequential appearance of neurologic signs that corresponds to the affected level. A direct relationship between the various configurations of transtentorial herniations and coma is, at best, tenuous. The orderly progression of signs from midbrain to medulla is often bypassed in catastrophic lesions where all brainstem functions are lost almost simultaneously. It is also clear that displacement of deep brain structures by a mass in any direction, with or without herniation, compresses the region of theRAS and results in coma. Furthermore, drowsiness and stupor typically occur with moderate lateral shifts at the level of the diencephalon (thalami) well before transtentorial or other herniations are evident. Lateral shift is easily quantified on axial images of computed tomography (CT) and magnetic resonance imaging (MRI) scans (Fig. 24-2). In cases of acutely appearing masses, a fairly consistent and simple relationship exists between the degree of horizontal displacement of midline structures and the level consciousness. Specifically, horizontal displacement of the pineal calcification of 3 to 5 mm is generally associated with drowsiness, 6 to 8 mm with stupor, and>9 mm with coma. At the same time, intrusion of the medial temporal lobe into the tentorial opening may be apparent as an obliteration of the cisterns that surround the upper brainstem. Coma and Confusional States Due to Metabolic Disorders A large variety of systemic metabolic abnormalities cause coma by interrupting the delivery of energy
substrates (hypoxia, ischemia, hypoglycemia) or by altering neuronal excitability (drug and alcohol intoxication, anesthesia, and epilepsy). The same metabolic abnormalities that produce coma may in milder form induce widespread cortical dysfunction and an acute confusional state. Thus, in metabolic encephalopathies, clouded consciousness and coma are a continuum. Neuropathologic changes in the various metabolic failures are variable -- very evident in hypoxia-ischemia, manifest as astrocytic changes in hepatic coma, and negligible in renal and other metabolic encephalopathies. Cerebral neurons are fully dependent on cerebral blood flow (CBF) and the related delivery of oxygen and glucose. CBF approximates 75 mL per 100 g/min in gray matter and 30 mL per 100 g/min in white matter (mean = 55 mL per 100 g/min); oxygen consumption is 3.5 mL per 100 g/min, and glucose utilization is 5 mg per 100 g/min. Brain stores of glucose provide energy for approximately 2 min after blood flow is interrupted, and oxygen stores last 8 to 10 s after the cessation of blood flow. Simultaneous hypoxia and ischemia exhaust glucose more rapidly. The electroencephalogram (EEG) rhythm in these circumstances becomes diffusely slowed, typical of metabolic encephalopathies, and as conditions of substrate delivery worsen, eventually all recordable brain electrical activity ceases. In almost all instances of metabolic encephalopathy, the global metabolic activity of the brain is reduced in proportion to the degree of unconsciousness. Conditions such as hyponatremia, hyperosmolarity, hypercapnia, hypercalcemia, and hepatic and renal failure are associated with a variety of alterations in neurons and astrocytes. It should be stated at the outset that the reversible effects of these conditions on the brain are not understood, but they may in different circumstances impair energy supplies, change ion fluxes across neuronal membranes, and cause neurotransmitter abnormalities. For example, the high brain ammonia concentration that is associated with hepatic coma interferes with cerebral energy metabolism and with the Na+, K+ -ATPase pump, increases the number and size of astrocytes, alters nerve cell function, and causes increased concentrations of potentially toxic products of ammonia metabolism; it may also result in abnormalities of neurotransmitters, including possible "false" neurotransmitters that may be active at receptor sites. Apart from hyperammonemia, which of these mechanisms is of critical importance is not clear. The mechanism of the encephalopathy of renal failure is also not known. Unlike ammonia, urea itself does not produce central nervous system (CNS) toxicity. A multifactorial causation has been proposed, including increased permeability of the blood-brain barrier to toxic substances such as organic acids and an increase in brain calcium or cerebrospinal fluid (CSF) phosphate content. Likewise, the basis of confusion and drowsiness that commonly accompanies the septic state has not been clarified. Coma and seizures are a common accompaniment of any large shifts in sodium and water balance. These changes in osmolarity may be the result of a number of systemic medical disorders including diabetic ketoacidosis, the nonketotic hyperosmolar state, and hyponatremia from any cause (e.g., water intoxication, excessive secretion of antidiuretic hormone or atrial natriuretic peptides). The volume of brain water correlates with the level of consciousness in these states, but other factors also play a role. Sodium levels below 125 mmol/L induce confusion, and below 115 mmol/L are associated with coma and convulsions. In hyperosmolar coma the serum osmolarity generally exceeds 350 mosmol/L. As in most other metabolic encephalopathies, the
severity of neurologic change depends to a large degree on the rapidity with which the serum changes occur. Hypercapnia depresses the level of consciousness in proportion to the rise in CO2tension in the blood and depends very much on the rapidity of change. The pathophysiology of other metabolic encephalopathies such as hypercalcemia, hypothyroidism, vitamin B12deficiency, and hypothermia are incompletely understood but must also reflect derangements ofCNSbiochemistry and membrane function. Epileptic Coma Although all metabolic derangements in some way alter neuronal electrophysiologic function, epilepsy is the only primary excitatory disturbance of brain electrical activity that is encountered in clinical practice. Continuous, generalized electrical discharges of the cortex (seizures) are associated with coma even in the absence of epileptic motor activity (convulsions). The self-limited coma that follows seizures, termed the postictal state, may be due to exhaustion of energy reserves or effects of locally toxic molecules that are the byproduct of seizures. The postictal state produces a pattern of continuous, generalized slowing of the backgroundEEGactivity similar to that of other metabolic encephalopathies. Pharmacologic Coma This class of encephalopathy is in large measure reversible and leaves no residual damage providing hypoxia does not supervene. Many drugs and toxins are capable of depressing nervous system function. Some produce coma by affecting both the brainstem nuclei, including theRAS, and the cerebral cortex. The combination of cortical and brainstem signs, which occurs in certain drug overdoses, may lead to an incorrect diagnosis of structural brainstem disease. Approach to the Patient The diagnosis and management of coma depend on knowledge of its main causes (see "Differential Diagnosis," below) and on interpretation of salient clinical signs, notably brainstem reflexes and motor function. Acute respiratory and cardiovascular problems should be attended to prior to neurologic assessment. A complete medical evaluation, except for the vital signs, funduscopy, and examination for nuchal rigidity, may be deferred until the neurologic evaluation has established the severity and nature of coma. History In many cases, the cause of coma is immediately evident (e.g., trauma, cardiac arrest, or known drug ingestion). In the remainder, historic information about the onset of coma is often sparse, but certain historic points are especially useful: (1) the circumstances and rapidity with which neurologic symptoms developed; (2) the details of any immediately preceding medical and neurologic symptoms (confusion, weakness, headache, fever, seizures, dizziness, double vision, or vomiting); (3) the use of medications, illicit drugs, or alcohol; and (4) chronic liver, kidney, lung, heart, or other medical disease. Direct interrogation or telephone calls to family and observers on the scene are an important part of the initial evaluation. Ambulance technicians often provide the most useful information in an enigmatic case. General Physical Examination The temperature, pulse, respiratory rate and pattern, and blood pressure should be measured quickly as the evaluation is getting under way. Fever suggests a systemic infection, bacterial meningitis, or encephalitis; only rarely is it attributable to a brain lesion that has disturbed temperature-regulating centers. A slight elevation in temperature may follow vigorous convulsions. High body temperature, 42 to
44°C, associated with dry skin should arouse the suspicion of heat stroke or anticholinergic drug intoxication. Hypothermia is observed with bodily exposure to lowered environmental temperature; alcoholic, barbiturate, sedative, or phenothiazine intoxication; hypoglycemia; peripheral circulatory failure; or hypothyroidism. Hypothermia itself causes coma only when the temperature is90 mmHg have a significant reduction in morbidity and mortality rate if they receive adequate therapy. These, then, are patients who have hypertension and who should be considered for treatment. The level of systolic pressure is also important in assessing the influence of arterial pressure on cardiovascular morbidity. Some data suggest that it may be more important than diastolic pressure. For example, males with normal diastolic pressures (158 mmHg) have a cardiovascular mortality rate 2.5 times higher than individuals who have similar diastolic pressures but whose systolic pressures clearly are normal (18 mmHg, and overt pulmonary alveolar edema develops at pressures>24 mmHg (Chap. 32). Pulmonary edema impacts cardiac function further by impairing diffusion of oxygen, setting up a vicious cycle. The increase in interstitial and intraalveolar fluid causes a progressive reduction in lung compliance, thereby increasing the work of ventilation while increasing perfusion of poorly ventilated alveoli. In establishing the diagnosis of cardiogenic shock, a history of cardiac disease or of AMIis of value. Associated physical findings include those of hemodynamic instability, peripheral vasoconstriction, and pulmonary and/or systemic venous congestion, as well as findings specific to the underlying cardiac abnormalities. An electrocardiogram may provide evidence of AMI or preexisting cardiac disease. The chest x-ray may show pulmonary edema and cardiomegaly. Transthoracic or transesophageal echocardiograms assist in the diagnosis of structural abnormalities and/or functional impairment of contractility. Serum cardiac markers will support the diagnosis of acute
cardiac injury. Hemodynamic monitoring is usually necessary. Placement of aPAC is helpful and will show a reduced cardiac output and an elevatedPCWP, and direct measurement of right atrial pressure allows calculation of systemic vascular resistance which is elevated. TREATMENT For all forms of cardiogenic shock, preload, afterload, and contractility should be modified using the information provided by thePAC. APCWP of 15 to 20 mmHg should be the initial goal. If the PCWP is excessively elevated, inotropic agents may provide significant reduction. The goal is to increase contractility without significant increases in heart rate. Dopamine and norepinephrine exert both inotropic and vasoconstrictor actions (Chap. 72) that are useful in the presence of persistent hypotension. Dobutamine, a positive inotropic agent with vasodilator properties, may be substituted when arterial pressure has been restored. Pulmonary congestion may be responsive to intravenous furosemide. Patients with an inadequate response to these measures can be supported by using intraaortic balloon counterpulsation to permit recovery of myocardial function. Additional measures to consider in cases of refractory cardiogenic shock include urgent myocardial revascularization in patients withAMI(Chap. 243), correction of anatomic cardiac defects such as rupture of the papillary muscles of the interventricular septum, the placement of ventricular assist devices, and even urgent cardiac transplantation. COMPRESSIVE CARDIOGENIC SHOCK With compression, the heart and surrounding structures are less compliant and, thus, normal filling pressures generate inadequate diastolic filling. Blood or fluid within the poorly distensible pericardial sac may cause tamponade (Chap. 239). Any cause of increased intrathoracic pressure, such as tension pneumothorax, herniation of abdominal viscera through a diaphragmatic hernia, or excessive positive pressure ventilation to support pulmonary function, can also cause compressive cardiogenic shock. Acute right heart failure with a sudden decline in cardiac output can be caused by pulmonary embolism obstructing right ventricular outflow and impairing left ventricular filling. Although initially responsive to increased filling pressures produced by volume expansion, as compression increases, cardiogenic shock occurs. The diagnosis of compressive cardiogenic shock is most frequently based on clinical findings, the chest radiograph, and an echocardiogram. The diagnosis of compressive cardiac shock may be more difficult to establish in the setting of trauma when hypovolemia and cardiac compression are present simultaneously. The classic findings of pericardial tamponade include the triad of hypotension, neck vein distention, and muffled heart sounds (Chap. 239). Pulsus paradoxus, i.e., an inspiratory reduction in systolic pressure >10 mmHg, may also be noted. The diagnosis is confirmed by echocardiography, and treatment consists of immediate pericardiocentesis. A tension pneumothorax produces ipsilateral decreased breath sounds, tracheal deviation away from the affected thorax, and jugular venous distention. Radiographic findings include increased intrathoracic volume, depression of the diaphragm of the affected hemithorax, and shifting of the mediastinum to the contralateral side. Chest decompression must be carried out immediately. Release of air and restoration of normal cardiovascular
dynamics is both diagnostic and therapeutic. SEPTIC SHOCK (See alsoChap. 124) This form of shock is caused by the systemic response to a severe infection. It occurs most frequently in elderly or immunocompromised patients and in those who have undergone an invasive procedure in which bacterial contamination has occurred. Infections of the lung, abdomen, or urinary tract are most common, and approximately half of the patients have bacteremia. Gram-positive and -negative bacteria, viruses, fungi, rickettsiae, and protozoa have all been reported to produce the clinical picture of septic shock, and the overall response is generally independent of the specific type of invading organism. The clinical findings in septic shock are a consequence of the combination of metabolic and circulatory derangements driven by the systemic infection and the release of toxic components of the infectious organisms, e.g., the endotoxin of gram-negative bacteria or the exotoxins and enterotoxins of gram-positive bacteria. Organism toxins lead to the release of cytokines, includingIL-1 andTNF-a, from tissue macrophages. Tissue factor expression and fibrin deposition are increased, and disseminated intravascular coagulation may develop. The inducible form of NO synthase is stimulated, and NO, a powerful vasodilator, is released. Hemodynamic changes in septic shock occur in two characteristic patterns: early, or hyperdynamic, and late, or hypodynamic, septic shock. Hyperdynamic Response In hyperdynamic septic shock, tachycardia is present, the cardiac output is normal, and the systemic vascular resistance is reduced while the pulmonary vascular resistance is elevated. The extremities are usually warm. However, splanchnic vasoconstriction with decreased visceral flow is present. The venous capacitance is increased, which decreases venous return. With volume expansion cardiac output becomes supranormal. Myocardial contractility is depressed in septic shock by mediators including NO,IL-1, and/orTNF-a. Inflammatory mediator-induced processes include increased capillary permeability and continued loss of intravascular volume. In septic shock, in contrast to other types of shock, total oxygen delivery may be increased while oxygen extraction is reduced due to maldistribution of microcirculatory perfusion and impaired utilization. In this setting the presence of a normal mixed venous oxygen saturation is not indicative of adequate peripheral perfusion, and even though the cardiac output may be elevated, it is still inadequate to meet the total metabolic needs. The toxicity of the infectious agents and their byproducts and the subsequent metabolic dysfunction drive the progressive deterioration of cellular and organ function. Acute respiratory distress syndrome, thrombocytopenia, and neutropenia are common complications. Hypodynamic Response As sepsis progresses, vasoconstriction occurs and the cardiac output declines. The patient usually becomes markedly tachypneic, febrile, diaphoretic, and obtunded, with cool, mottled, and often cyanotic extremities. Oliguria, renal failure, and hypothermia develop; there may be striking increases in serum lactate. TREATMENT
Aggressive volume expansion with a crystalloid solution to aPCWP of approximately 15 mmHg and the restoration of arterial oxygenation with inspired oxygen and frequently with mechanical ventilation are the highest priorities. In the presence of hypodynamic septic shock, augmentation of cardiac output may require inotropic support with dopamine or norepinephrine in the presence of hypotension or with dobutamine if arterial pressure is normal. Antibiotics should be administered, either appropriate for the results of cultures or empirical therapy based on the likely source of infection. Surgical debridement or drainage may also be necessary to control the infection. NEUROGENIC SHOCK Interruption of sympathetic vasomotor input after a high cervical spinal cord injury, inadvertent cephalad migration of spinal anesthesia, or severe head injury may result in neurogenic shock. In addition to arteriolar dilatation, venodilation causes pooling in the venous system, which decreases venous return and cardiac output. The extremities are often warm, in contrast to the usual vasoconstriction-induced coolness in hypovolemic or cardiogenic shock. Treatment involves a simultaneous approach to the relative hypovolemia and to the loss of vasomotor tone. Large volumes of fluid may be required to restore normal hemodynamics. Once hemorrhage has been ruled out, norepinephrine may be necessary to augment vascular resistance. HYPOADRENAL SHOCK (See alsoChap. 331) The normal host response to the stress of illness, operation, or trauma requires that the adrenal glands hypersecrete cortisol in excess of that normally required. Hypoadrenal shock occurs in settings in which unrecognized adrenal insufficiency complicates the host response to the stress induced by acute illness or major surgery. Adrenocortical insufficiency may occur as a consequence of the chronic administration of high doses of exogenous glucocorticoids. Recent studies have shown that prolonged stays in a critical state in an intensive care setting may also induce a relative hypoadrenal state. Other, less common causes include adrenal insufficiency secondary to idiopathic atrophy, tuberculosis, metastatic disease, bilateral hemorrhage, and amyloidosis. The shock produced by adrenal insufficiency is characterized by reductions in systemic vascular resistance, hypovolemia, and reduced cardiac output. The diagnosis of adrenal insufficiency may be established by means of anACTHstimulation test (Chap. 331). TREATMENT In the hemodynamically unstable patient, dexamethasone sodium phosphate, 4 mg, should be given intravenously. This agent is preferred because unlike hydrocortisone it does not interfere with theACTHstimulation test. If the diagnosis of adrenal insufficiency has been established, hydrocortisone, 100 mg every 6 to 8 h, can be given and tapered to a maintenance level as the patient achieves hemodynamic stability. Simultaneous volume resuscitation and pressor support is required. ADJUNCTIVE THERAPIES As described above, the sympathomimetic amines dobutamine, dopamine, and norepinephrine are widely used in the treatment of all forms of shock. The clinical
pharmacology of these agents is described in Chap. 72. POSITIONING Positioning of the patient may be a valuable adjunct in the initial treatment of hypovolemic shock. Elevating the foot of the bed (i.e., placing it on "shock blocks") and assumption of the Trendelenburg position without flexion at the knees are effective but may increase work of breathing and risk for aspiration. Simply elevating both legs may be the optimal approach. PNEUMATIC ANTISHOCK GARMENT (PASG) The PASG and the military antishock trousers (MAST) are inflatable external compression devices that can be wrapped around the legs and abdomen and have been widely used in the prehospital setting as a means of providing temporary support of central hemodynamics in shock. They cause an increase in systemic vascular resistance and blood pressure by arterial compression, without causing a significant change in cardiac output. While the use of PASG has been recommended in noncardiogenic forms of shock, the most appropriate use appears to be as a means to tamponade bleeding and augment hemostasis. Inflation of the suit provides splinting of fractures of the pelvis and lower extremities and arrests hemorrhage from fractures. REWARMING Hypothermia is a potential adverse consequence of massive volume resuscitation. The infusion of large volumes of refrigerated blood products and room-temperature crystalloid solutions can rapidly drop core temperatures if fluid is not run through warming devices. Hypothermia may depress cardiac contractility and thereby further impair cardiac output and oxygen delivery. Hypothermia, particularly temperatures 25 g protein per liter is unusual in uncomplicated cirrhosis but is consistent with tuberculous peritonitis or neoplasm. Cloudy fluid with a predominance of polymorphonuclear cells and a positive Gram's stain are characteristic of bacterial peritonitis; if most cells are lymphocytes, tuberculosis should be suspected. The complete examination of each fluid is most important, for occasionally only one finding may be abnormal. For example, if the fluid is a typical transudate but contains>250 white blood cells per microliter, the finding should be recognized as atypical for cirrhosis and should warrant a search for tumor or infection. This is especially true in the evaluation of cirrhotic ascites where
occult peritoneal infection may be present with only minor elevations in the white blood cell count of the peritoneal fluid (300 to 500 cells per microliter). Since Gram's stain of the fluid may be negative in a high proportion of such cases, careful culture of the peritoneal fluid is mandatory. Bedside innoculation of blood culture flasks with ascitic fluid results in a dramatically increased incidence of positive cultures when bacterial infection is present (90 versus 40% positivity with conventional cultures done by the laboratory). Direct visualization of the peritoneum (laparoscopy) may disclose peritoneal deposits of tumor, tuberculosis, or metastatic disease of the liver. Biopsies are taken under direct vision, often adding to the diagnostic accuracy of the procedure. Chylous ascites refers to a turbid, milky, or creamy peritoneal fluid due to the presence of thoracic or intestinal lymph. Such a fluid shows Sudan-staining fat globules microscopically and an increased triglyceride content by chemical examination. Opaque milky fluid usually has a triglyceride concentration of>1000 mg/dL. A turbid fluid due to leukocytes or tumor cells may be confused with chylous fluid (pseudochylous), and it is often helpful to carry out alkalinization and ether extraction of the specimen. Alkali tend to dissolve cellular proteins and thereby reduce turbidity; ether extraction leads to clearing if the turbidity of the fluid is due to lipid. Chylous ascites is most often the result of lymphatic obstruction from trauma, tumor, tuberculosis, filariasis (Chap. 221), or congenital abnormalities. It also may be seen in the nephrotic syndrome. Rarely, ascitic fluid may be mucinous in character, suggesting either pseudomyxoma peritonei (Chap. 289) or rarely a colloid carcinoma of the stomach or colon with peritoneal implants. On occasion, ascites may develop as a seemingly isolated finding in the absence of a clinically evident underlying disease. Then, a careful analysis of ascitic fluid may indicate the direction the evaluation should take. A useful framework for the workup starts with an analysis of whether the fluid is classified as a high (transudate) or low (exudate) gradient fluid. High gradient (transudative) ascites of unclear etiology is most often due to occult cirrhosis, right-sided venous hypertension raising hepatic sinusoidal pressure, or hypoalbuminemic states such as nephrosis or protein-losing enteropathy. Cirrhosis with well-preserved liver function (normal albumin) resulting in ascites invariably is associated with significant portal hypertension (Chap. 298). Evaluation should include liver function tests, liver-spleen scan, or other hepatic imaging procedure (i.e., CT or ultrasound) to detect nodular changes in the liver or a colloid shift of isotope to suggest portal hypertension. On occasion, a wedged hepatic venous pressure can be useful to document portal hypertension. Finally, if clinically indicated, a liver biopsy will confirm the diagnosis of cirrhosis and perhaps suggest its etiology. Other etiologies may result in hepatic venous congestion and resultant ascites. Right-sided cardiac valvular disease and particularly constrictive pericarditis should raise a high index of suspicion and may require cardiac imaging and cardiac catheterization for definitive diagnosis. Hepatic vein thrombosis is evaluated by visualizing the hepatic veins with imaging techniques (Doppler ultrasound, angiography, CT scans, magnetic resonance imaging) to demonstrate obliteration, thrombosis, or obstruction by tumor. Uncommonly, transudative ascites may be associated with benign tumors of the ovary, particularly fibroma (Meigs' syndrome) with ascites and hydrothorax. Low gradient (exudative) ascites should initiate an evaluation for primary peritoneal
processes, most importantly infection and tumor. Routine bacteriologic culture of ascitic fluid often yields a specific organism causing infectious peritonitis. Tuberculous peritonitis (Table 46-1) is best diagnosed by peritoneal biopsy, either percutaneously or via laparoscopy. Histologic examination invariably shows granulomata that may contain acid-fast bacilli. Since cultures of peritoneal fluid and biopsies for tuberculosis may require 6 weeks, characteristic histology with appropriate stains allows antituberculosis therapy to be started promptly. Similarly, the diagnosis of peritoneal seeding by tumor can usually be made by cytologic analysis of peritoneal fluid or by peritoneal biopsy if cytology is negative. Appropriate diagnostic studies can then be undertaken to determine the nature and site of the primary tumor. Pancreatic ascites (Table 46-1) is invariably associated with an extravasation of pancreatic fluid from the pancreatic ductal system, most commonly from a leaking pseudocyst. Ultrasound or CT examination of the pancreas followed by visualization of the pancreatic duct by direct cannulation [viz., endoscopic retrograde cholangiopancreatography (ERCP)] usually discloses the site of leakage and permits resective surgery to be carried out. An analysis of the physiologic and metabolic factors involved in the production of ascites (detailed inChap. 298), coupled with a complete evaluation of the nature of the ascitic fluid, invariably discloses the etiology of the ascites and permits appropriate therapy to be instituted. ACKNOWLEDGEMENT Dr. Kurt J. Isselbacher was the co-author of this chapter in previous editions. (Bibliography omitted in Palm version) Back to Table of Contents (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 7 -ALTERATIONS IN RENAL AND URINARY TRACT FUNCTION 47. AZOTEMIA AND URINARY ABNORMALITIES - Bradley M. Denker, Barry M. Brenner Body homeostasis is maintained predominantly through the cellular processes that together comprise normal kidney function. Disturbances to any of these functions can lead to a constellation of abnormalities that may be detrimental to survival. The clinical manifestations of these diseases will depend upon the pathophysiology of the renal injury and will often be initially identified as a complex of symptoms, abnormal physical findings, and laboratory changes that will allow the identification of specific syndromes. These renal syndromes (summarized in Table 47-1) may arise as the consequence of a systemic illness or can occur as a primary renal disease. Nephrologic syndromes usually consist of several elements that reflect the underlying pathologic processes and the duration of the disease and typically include one or more of the following features: (1) disturbances in urine volume (oliguria, anuria, polyuria); (2) abnormalities of urine sediment [red blood cells (RBC); white blood cells, casts, and crystals]; (3) abnormal excretion of serum proteins (proteinuria); (4) reduction in glomerular filtration rate (GFR) (azotemia); (5) presence of hypertension and/or expanded total body volume (edema); (6) electrolyte abnormalities, or (7) in some syndromes, fever/pain. The combination of these findings should permit identification of one of the major nephrologic syndromes (Table 47-1) and will allow the differential diagnoses to be narrowed and the appropriate diagnostic evaluation and therapeutic course to be determined. Each of these syndromes and their associated diseases are discussed in more detail in subsequent chapters. This chapter will focus on several aspects of renal abnormalities that are critically important to distinguishing these processes: (1) reduction in GFR leading to azotemia, (2) alterations of the urinary sediment and/or protein excretion, and (3) abnormalities of urinary volume. AZOTEMIA ASSESSMENT OF GLOMERULAR FILTRATION RATE Monitoring the GFRis important in both the hospital and outpatient settings, and several different methodologies are available (discussed below). In most acute clinical circumstances a measured GFR is not available, and it is necessary to estimate the GFR from the serum creatinine level in order to provide appropriate doses of drugs that are excreted into the urine. Serum creatinine is the most widely used marker for GFR and is related directly to the urine creatinine excretion and inversely to the serum creatinine (UCr/PCr). Based upon this relationship and some important caveats (discussed below), the GFR will fall proportionately with the increase in PCr. Failure to account for GFR reductions in drug dosing can lead to significant morbidity and mortality from drug toxicities (e.g., digoxin, aminoglycosides). In the outpatient setting, serial determinations of GFR are helpful for following the progression of chronic renal insufficiency, but again, the serum creatinine is often used as a surrogate for GFR (although much less accurate; see below). In patients with chronic progressive renal insufficiency there is an approximately linear relationship between 1/PCrand time. The slope of this line will remain constant for an individual patient, and when values are obtained that do not fall on this line, an investigation for a superimposed acute process
(e.g., volume depletion, drug reaction) should be initiated. It should be emphasized that the signs and symptoms of uremia will develop at significantly different levels of serum creatinine depending upon the patient (size, age, and sex), the underlying renal disease, existence of concurrent diseases, and true GFR. In general, patients do not develop symptomatic uremia until renal insufficiency is usually quite severe (GFR< 15 mL/min) and in some patients it does not occur until the GFR < 5 mL/min. A reducedGFRleads to retention of nitrogenous waste products (azotemia) such as serum urea nitrogen and creatinine. Azotemia may result from reduced renal perfusion, intrinsic renal disease, or postrenal processes (ureteral obstruction; see below and Fig. 47-1). Precise determination of GFR is problematic as both commonly used markers (urea and creatinine) have characteristics that affect their accuracy as markers of clearance. Urea clearance is generally an underestimate of GFR because of tubule urea reabsorption and may be as low as one-half of GFR measured by other techniques. Creatinine is a small, freely filtered solute that varies little from day to day (since it is derived from muscle metabolism of creatine). However, serum creatinine can increase acutely from dietary ingestion of cooked meat. Creatinine can be secreted by the proximal tubule through an organic cation pathway. There are many clinical settings where a creatinine clearance is not available, and decisions concerning drug dosing must be made based on the serum creatinine. A formula that allows an estimate of creatinine clearance in men that accounts for age-related decreases in GFR, body weight, and sex has been derived by Cockcroft-Gault:
This value should be multiplied 0.85 for women, since a lower fraction of the body weight is composed of muscle. The gradual loss of muscle from chronic illness, chronic use of glucocorticoids, or malnutrition can mask significant changes inGFR with small or imperceptible changes in serum creatinine. More accurate determinations of GFR are available using inulin clearance or radionuclide-labeled markers such as125I-iothalamate or EDTA. These methods are highly accurate due to precise quantitation and the absence of any renal reabsorption/secretion and should be used to follow GFR in patients in whom creatinine is not likely to be a reliable indicator (patients with decreased muscle mass secondary to age, malnutrition, concurrent illnesses). Approach to the Patient Once it has been established that GFRis reduced, the physician must decide if this represents acute or chronic renal failure. The clinical situation, history, and laboratory data often make this an easy distinction. However, the laboratory abnormalities characteristic of chronic renal failure, including anemia, hypocalcemia, and hyperphosphatemia, are often also present in patients presenting with acute renal failure. Radiographic evidence of renal osteodystrophy (Chap. 270) would be seen only in chronic renal failure but is a very late finding, and these patients are usually on dialysis. The urinalysis and renal ultrasound can occasionally facilitate distinguishing acute from chronic renal failure. An approach to the evaluation of azotemic patients is shown in Fig. 47-1. Patients with advanced chronic renal insufficiency often have some proteinuria, nonconcentrated urine (isosthenuria), and small kidneys on ultrasound
characterized by increased echogenicity and cortical thinning. Treatment should be directed toward slowing the progression of renal disease and providing symptomatic relief for edema, acidosis, anemia, and hyperphosphatemia, as discussed inChap. 270. Acute renal failure (Chap. 269) can result from processes affecting renal blood flow (prerenal azotemia), intrinsic renal diseases (affecting vessels, glomeruli, or tubules), or postrenal processes (obstruction to urine flow in ureters, bladder, or urethra) (Chap. 281). Prerenal Failure Decreased renal perfusion accounts for 40 to 80% of acute renal failure and, if appropriately treated, is readily reversible. The etiologies of prerenal azotemia include any cause of decreased circulating blood volume including volume loss (gastrointestinal hemorrhage, burns, diarrhea, diuretics), volume sequestration (pancreatitis, peritonitis, rhabdomyolysis), or decreased effective circulating volume (cardiogenic shock, sepsis). Renal perfusion can also be affected by reductions in cardiac output from peripheral vasodilatation (sepsis, drugs) or profound renal vasoconstriction [severe heart failure, hepatorenal syndrome, drugs (such as nonsteroidal anti-inflammatory drugs (NSAIDs)]. True, or "effective," hypovolemia leads to a fall in mean arterial pressure, which in turn triggers a series of neural and humoral responses that include activation of the sympathetic nervous and renin-angiotensin-aldosterone systems and ADH release.GFR is maintained by prostaglandin-mediated relaxation of afferent arterioles and angiotensin II-mediated constriction of efferent arterioles. Once the mean arterial pressure falls below 80 mmHg, there is a steep decline in GFR. Blockade of prostaglandin production byNSAIDscan result in severe vasoconstriction and acute renal failure under these circumstances. Angiotensin-converting enzyme (ACE) inhibitors decrease efferent arteriolar tone and can decrease glomerular capillary perfusion pressure. Patients on NSAIDs and/or ACE inhibitors are most susceptible to hemodynamically mediated acute renal failure when blood volume is reduced for any reason. Patients with renal artery stenosis are dependent upon efferent arteriolar vasoconstriction for maintenance of glomerular filtration pressure and are particularly susceptible to precipitous decline inGFR when given ACE inhibitors. Prolonged renal hypoperfusion can lead to acute tubular necrosis (ATN; an intrinsic renal disease discussed below). The urinalysis and urinary electrolytes can be useful in distinguishing prerenal azotemia from ATN (Table 47-2). The urine of patients with prerenal azotemia can be predicted from the stimulatory actions of norepinephrine, angiotensin II, ADH, and low tubule fluid flow on salt and water reabsorption from the urine. In prerenal conditions the tubules are intact, leading to a concentrated urine (>500 mosm), avid Na retention (urine Na concentration three RBCs/HPF on three urinalyses, or single urinalysis with >100 RBCs, or gross hematuria) identified significant renal or urologic lesions in 9.1% of over 1000 patients. Even patients who are chronically anticoagulated should be investigated as outlined in Fig. 47-2. The suspicion for urogenital neoplasms in patients with isolated painless hematuria (nondysmorphic RBCs) increases with age. Neoplasms are rare in the pediatric population, and isolated hematuria is more likely to be "idiopathic" or associated with a congenital anomaly. Hematuria with pyuria and bacteriuria is typical of infection and should be treated with antibiotics after appropriate cultures. Acute cystitis or urethritis in women can cause gross hematuria. Hypercalciuria and hyperuricosuria are also risk factors for unexplained isolated hematuria in both children and adults. In some of these patients (50 to 60%), reducing calcium and uric acid excretion through dietary interventions can eliminate the microscopic hematuria.
Isolated microscopic hematuria can be a manifestation of glomerular diseases. TheRBCs of glomerular origin are often dysmorphic when examined by phase-contrast microscopy. Irregular shapes of RBCs may also occur due to pH and osmolarity changes found in the distal tubule. There is, however, significant observer variability in detecting dysmorphic RBCs, especially if a phase-contrast microscope is not available. The most common etiologies of isolated glomerular hematuria are IgA nephropathy, hereditary nephritis, and thin basement membrane disease. IgA nephropathy and hereditary nephritis can have episodic gross hematuria. A family history of renal failure is often present in patients with hereditary nephritis, and patients with thin basement membrane disease often have other family members with microscopic hematuria. A renal biopsy is needed for the definitive diagnosis of these disorders, which are discussed in more detail in Chap. 275. Hematuria with dysmorphic RBCs, RBC casts, and protein excretion >500 mg/d is virtually diagnostic of glomerulonephritis. RBC casts form as RBCs that enter the tubular fluid become trapped in a cylindrical mold of gelled Tamm-Horsfall protein. Even in the absence of azotemia, these patients should undergo serologic evaluation and renal biopsy as outlined in Fig. 47-2. Isolated pyuria is unusual since inflammatory reactions in the kidney or collecting system are also associated with hematuria. The presence of bacteria suggests infection, and white blood cell casts with bacteria are indicative of pyelonephritis. White blood cells and/or white blood cell casts may also be seen in tubulointerstitial processes such as interstitial nephritis, systemic lupus erythematosus, and transplant rejection. In chronic renal diseases, degenerated cellular casts called waxy casts can be seen in the urine. Broad casts are thought to arise in the dilated tubules of enlarged nephrons that have undergone compensatory hypertrophy in response to reduced renal mass (i.e., chronic renal failure). A mixture of broad casts typically seen with chronic renal failure together with cellular casts andRBCs may be seen in smoldering processes such as chronic glomerulonephritis with active glomerulitis. ABNORMALITIES OF URINE VOLUME The volume of urine produced varies depending upon the fluid intake, renal function, and physiologic demands of the individual. See "Azotemia," above, for discussion of decreased (oliguria) or absent urine production (anuria).*The physiology of water formation and renal water conservation are discussed in Chap. 268. POLYURIA By history, it is often difficult for patients to distinguish urinary frequency (often of small volumes) from polyuria, and a 24-h urine collection is needed for evaluation (Fig. 47-5). It is necessary to determine if the polyuria represents a solute or water diuresis and if the diuresis is appropriate for the clinical circumstances. The average person excretes between 600 and 800 mosmol of solutes per day, primarily as urea and electrolytes. The urine osmolality can help distinguish a solute from water diuresis. If the urine output is >3 L/d (arbitrarily defined as polyuria) and the urine is dilute (3 L/d and urine osmolality is>300 mosmol/L, then a
solute diuresis is clearly present and a search for the responsible solute(s) is mandatory. Excessive filtration of a poorly reabsorbed solute such as glucose, mannitol, or urea can depress reabsorption of NaCl and water in the proximal tubule and lead to enhanced excretion in the urine. Poorly controlled diabetes mellitus is the most common cause of a solute diuresis, leading to volume depletion and serum hypertonicity. Since the urine Na concentration is less than that of blood, more water than Na is lost, causing hypernatremia and hypertonicity. Common iatrogenic solute diuresis occurs from mannitol administration, radiocontrast media, and high-protein feedings (enterally or parenterally), leading to increased urea production and excretion. Less commonly, excessive Na loss may occur from cystic renal diseases, Bartter's syndrome, or during the course of a tubulointerstitial process (such as resolvingATN). In these so-called salt-wasting disorders, the tubule damage results in direct impairment of Na reabsorption and indirectly reduces the responsiveness of the tubule to aldosterone. Usually, the Na losses are mild, and the obligatory urine output is less than 2 L/d (resolving ATN and postobstructive diuresis are exceptions and may be associated with significant natriuresis and polyuria.) Formation of large volumes of dilute urine represent polydipsic states or diabetes insipidus. Primary polydipsia can result from habit, psychiatric disorders, neurologic lesions, or medications. During deliberate polydipsia, extracellular fluid volume is normal or expanded and vasopressin levels are reduced because serum osmolality tends to be near the lower limits of normal. Central diabetes insipidus may be idiopathic in origin or secondary to a variety of hypothalamic conditions including posthypophysectomy or trauma or neoplastic, inflammatory, vascular, or infectious hypothalamic diseases. Idiopathic central diabetes insipidus is associated with selective destruction of the vasopressin-secreting neurons in the supraoptic and paraventricular nuclei and can be inherited as an autosomal dominant trait or occur spontaneously. Nephrogenic diabetes insipidus can occur in a variety of clinical situations as summarized inFig. 47-5. A plasma vasopressin level is recommended as the best method for distinguishing between central and nephrogenic diabetes insipidus. Alternatively, a water deprivation test plus exogenous vasopressin may also distinguish primary polydipsia from central and nephrogenic diabetes insipidus.*For a detailed discussion, see Chap. 329. (Bibliography omitted in Palm version) Back to Table of Contents (Bibliography omitted in Palm version) Back to Table of Contents
48. INCONTINENCE AND LOWER URINARY TRACT SYMPTOMS - Philippe E. Zimmern, John D. McConnell PHYSIOLOGY OF VOIDING Normal bladder filling depends on unique elastic properties of the bladder wall that allow it to increase in volume at a pressure lower than that of the bladder neck and urethra (otherwise incontinence would occur). Despite provocative maneuvers such as coughing, voluntary bladder contractions do not occur. Emptying is dependent on the integrity of a complex neuromuscular network that causes relaxation of the urethral sphincter a few milliseconds before the onset of the detrusor (bladder muscle) contraction. With normal, sustained detrusor contraction, the bladder empties completely. A bladder that can fill and empty in this manner has a normal detrusor muscle and is described as stable according to conventional terminology. Since the voluntary control of micturition depends on the neural connections between the cerebral cortex and the brainstem, disruption of these pathways (brain tumor, stroke, head trauma, Parkinson's disease) impairs the ability to suppress and control bladder contractions. A bladder contraction without voluntary effort characterizes an unstable bladder. Bladder or detrusor instability of neurologic origin is termed detrusor hyperreflexia. Conversely, the detrusor muscle that cannot contract during voiding is called noncontractile; underactivity of the detrusor due to a lesion of the sacral cord or pelvic nerves is termed detrusor areflexia. Contrary to common belief, the center that controls normal micturition is not in the spinal cord but in the brainstem. Proper coordination (synergia) between the detrusor and urethral sphincters requires an intact neural (autonomic and somatic nervous systems) communication between bladder and urethra. Injury to the upper spinal cord, for example, can cause dyssynergia between bladder and urethra that results in urge incontinence, residual urine retention, bladder wall changes (trabeculation and fibrosis), and possibly renal insufficiency. A simple way to classify voiding dysfunction is to determine whether it is primarily a storage failure or an emptying failure by asking two questions: Is the voiding dysfunction due to the bladder or outlet (bladder neck or urethra) (failure to store)? Is there neurologic dysfunction (failure to empty)? Bladder storage and emptying problems may coexist in the same individual and can cause similar lower urinary tract symptoms (LUTS). LOWER URINARY TRACT SYMPTOMS IN MEN The most common cause of LUTSin men of middle age and older is prostatic hyperplasia, which causes obstruction to urine flow by encroachment on the urethral lumen (Chap. 95). Histologically, 50 to 80% of the prostatic volume is composed of stromal tissue (smooth muscle), while the remainder is glandular. The transitional zone,
which is responsible for benign prostatic growth, comprises 10 to 15% of the prostate at the end of puberty but increases in volume after age 40. However, prostatic enlargement is not always accompanied by symptoms because the direction of growth can be outward, so that little change may occur in urine flow. Alternatively, men with early histologic evidence of prostatic hyperplasia can experience significant voiding symptoms. In this circumstance, increased tone of the prostatic smooth muscle and enhanced prostatic tension within a nondistensible capsule can cause obstruction. In response to obstruction, the bladder smooth-muscle cells hypertrophy to generate the higher pressures necessary for voiding, and the increase in bladder muscle mass leads to reduced elasticity, or compliance, and decreased bladder capacity. Detrusor dysfunction from bladder outlet obstruction can cause any combination of theLUTSdescribed above. When the obstruction progresses, infiltration of extracellular matrix between the smooth-muscle bundles of the bladder wall can result in a hypocontractile or acontractile bladder (bladder failure). Other complications such as urinary tract infections or bladder stones secondary to the large postvoid residuals (stasis) and upper tract damage (hydronephrosis, reflux) can develop during the course of the obstructive process. Although prostatic hyperplasia is the most common cause of bladder outlet obstruction in men, other sources of obstruction include prostate cancer, urethral stricture, and lack of proper sphincteric relaxation (neurologic cause). Nonobstructive causes ofLUTSinclude diabetic neuropathy, which can affect the parasympathetic nerves of the bladder. Decreased sensation of bladder fullness leads to incomplete emptying and overdistention of the bladder and, in turn, to increased frequency and nocturia due to bladder overflow; these symptoms are frequently made worse by the polydipsia/polyuria of diabetes mellitus. At times, storage symptoms can be caused by other neurologic causes such as stroke, multiple sclerosis, or Parkinson's disease. The International Prostate Symptom Score (IPSS) is used to assess the severity ofLUTS: Decreased force of stream -- over the past month how often have you had a weak urinary stream? Intermittency -- over the past month how often have you found you stopped and started again several times when urinating? Incomplete emptying -- over the past month how often have you had a sensation of not emptying your bladder completely after finishing urination? Straining -- over the past month how often have you had to push or strain to begin urination?TheIPSSalso assesses the impact of storage symptoms: Frequency -- over the past month how often have you had to urinate again within 2 h after urinating? Urgency -- over the past month how often have you found it difficult to postpone urination?
Nocturia -- over the past month how many times did you typically get up to urinate between going to bed and getting up in the morning? (Range: none to five or more times.)Except for nocturia, the answers range from 0 (not at all) to 5 (almost always). A total score of 23 suggest significant bladder outlet obstruction. Because similar symptoms can result from neurologic causes, theIPSSquestionnaire cannot be used to make the diagnosis of prostatic hyperplasia but is useful only as an index of severity and of the response to treatment. LOWER URINARY TRACT SYMPTOMS IN WOMEN Urethral obstruction is an uncommon cause ofLUTS in women. A careful bimanual examination and passage of a urethral catheter are sufficient to exclude urethral stenosis, which is usually secondary to prior instrumentation or operative procedures, and urethral cancer. Urinary tract infection (cystitis) is more prevalent in women and must be excluded by urinalysis. Multiple sclerosis should be considered in middle-aged women presenting with frequency, urgency, or incontinence. In addition to many of the same disorders that produce voiding symptoms in men, estrogen deficiency, frequency-urgency syndrome, and interstitial cystitis (IC) with minimal pain must be considered. Cystocele and pelvic prolapse can cause urinary frequency secondary to impairment of bladder emptying. EVALUATION Men and women withLUTS and concomitant neurologic disease should undergo a complete urodynamic evaluation. In the absence of neurologic disease, men with LUTS most commonly have prostatic hyperplasia. However, it is necessary to exclude prostate cancer, especially if there is a positive family history, an abnormal prostate examination, or an elevated level of prostate-specific antigen (PSA). In both sexes bladder cancer can also cause storage symptoms and is suggested by microscopic hematuria and/or abnormal urine cytology. Usually, a detailed genitourinary history, a symptom assessment, a careful neurologic examination including rectal examination and assessment of the bulbocavernosus reflex, measurements of urine flow and postvoid residual urine volume (by bladder ultrasound), and limited laboratory evaluation (urinalysis, urine culture, PSA levels, urine cytology, urea/creatinine levels, as indicated) should be sufficient to direct therapy. More complex investigations of the lower urinary tract (cystoscopy, voiding cystography, urodynamics) and upper urinary tract (pyelogram or ultrasonography) are sometimes indicated.*For therapy of BPH, see Chap. 95. INCONTINENCE Incontinence is a condition where involuntary loss of urine is objectively demonstrated and is a social or hygienic problem. A common variant, stress incontinence, denotes involuntary loss of urine with physical exercise (coughing, sneezing, sports, sexual activity). Urge incontinence is an involuntary loss of urine associated with a strong desire to void, and overflow incontinence is an involuntary loss of urine when the
elevation of intravesical pressure with bladder overfilling or distention exceeds the maximal urethral pressure. Loss of urine through channels other than the urethra is rare (ectopic ureter, fistulae) but causes total or continuous incontinence. INCONTINENCE IN WOMEN Among noninstitutionalized women 60 years of age and older, 25 to 30% have urinary incontinence daily or weekly, and approximately half of institutionalized women are incontinent more than once a day. The annual cost of caring for incontinent persons is very high and, if not well managed, can be associated with complications such as decubitus ulcers. Stress urinary incontinence (SUI) is secondary to urethral hypermobility or, less commonly ( [Na+ + K+], the urine ammonium level is appropriately increased, suggesting an extrarenal cause of the acidosis. Loss of functioning renal parenchyma by progressive renal disease leads to hyperchloremic acidosis when the glomerular filtration rate (GFR) is between 20 and 50 mL/min and to uremic acidosis with a highAG when the GFR falls to 20 mmol/L. Since HCO3-is not reabsorbed normally in the proximal tubule, therapy with NaHCO3will enhance renal potassium wasting and hypokalemia. The typical findings in classic distalRTA(type 1 RTA) (Chap. 276) include hypokalemia, hyperchloremic acidosis, low urinary NH4+excretion (positiveUAG, low urine [NH4+]), and inappropriately high urine pH (pH > 5.5). Such patients are unable to acidify the urine below a pH of 5.5. Most patients have hypocitraturia and hypercalciuria, so that nephrolithiasis, nephrocalcinosis, and bone disease are common. In type 4 RTA, hyperkalemia is disproportionate to the reduction inGFRbecause of coexisting dysfunction of potassium and acid secretion. Urinary ammonium excretion is invariably depressed, and renal function may be compromised, for example, due to diabetic nephropathy, amyloidosis, or tubulointerstital disease.*See Chap. 276 for the pathophysiology, diagnosis, and treatment of RTA.
Hyporeninemic Hypoaldosteronism (See also Chap. 331) This condition typically causes hyperchloremic metabolic acidosis, most commonly in older adults with diabetes mellitus or tubulointerstitial disease and renal insufficiency. Patients usually have mild to moderate renal insufficiency and acidosis, with elevation in serum [K+] (5.2 to 6.0 mmol/L), concurrent hypertension, and congestive heart failure. Both the metabolic acidosis and the hyperkalemia are out of proportion to impairment in GFR. Nonsteroidal anti-inflammatory drugs -- trimethoprim, pentamidine, and ACE-inhibitors -- can also cause hyperkalemia with hyperchloremic metabolic acidosis in patients with renal insufficiency (Table 50-3). METABOLIC ALKALOSIS Metabolic alkalosis is manifested by an elevated arterial pH, an increase in the serum [HCO3-], and an increase in PaCO2 as a result of compensatory alveolar hypoventilation. It is often accompanied by hypochloremia and hypokalemia. The patient with a high [HCO3-] and a low [Cl-] has either metabolic alkalosis or chronic respiratory acidosis. As shown inTable 50-1, the PaCO2increases 6 mmHg for each 10-mmol/L increase in the [HCO3-] above normal. Stated differently, in the range of [HCO3-] from 10 to 40 mmol/L, the predicted PaCO2 is approximately equal to the [HCO3-] + 15. The arterial pH establishes the diagnosis, since it is increased in metabolic alkalosis and decreased or normal in respiratory acidosis. Metabolic alkalosis frequently occurs in association with other disorders such as respiratory acidosis or alkalosis or metabolic acidosis. PATHOGENESIS Metabolic alkalosis occurs as a result of net gain of [HCO3-] or loss of nonvolatile acid (usually HCl by vomiting) from the extracellular fluid. Since it is unusual for alkali to be added to the body, the disorder involves a generative stage, in which the loss of acid usually causes alkalosis, and a maintenance stage, in which the kidneys fail to compensate by excreting HCO3-because of volume contraction, a lowGFR, or depletion of Cl- or K+ . Under normal circumstances, the kidneys have an impressive capacity to excrete HCO3-. Continuation of metabolic alkalosis represents a failure of the kidneys to eliminate HCO3-in the usual manner. For HCO3-to be added to the extracellular fluid, it must be administered exogenously or synthesized endogenously, in part or entirely by the kidneys. The kidneys will retain, rather than excrete, the excess alkali and maintain the alkalosis if (1) volume deficiency, chloride deficiency, and K+ deficiency exist in combination with a reducedGFR, which augments distal tubule H+ secretion; or (2) hypokalemia exists because of autonomous hyperaldosteronism. In the first example, alkalosis is corrected by administration of NaCl and KCl, while in the latter it is necessary to repair the alkalosis by pharmacologic or surgical intervention, not with saline administration. DIFFERENTIAL DIAGNOSIS To establish the cause of metabolic alkalosis (Table 50-4), it is necessary to assess the status of the extracellular fluid volume (ECFV), the recumbent and upright blood pressure, the serum [K + ], and the renin-aldosterone system. For example, the presence
of chronic hypertension and chronic hypokalemia in an alkalotic patient suggests either mineralocorticoid excess or that the hypertensive patient is receiving diuretics. Low plasma renin activity and normal urine [Na+ ] and [Cl-] in a patient who is not taking diuretics indicate a primary mineralocorticoid excess syndrome. The combination of hypokalemia and alkalosis in a normotensive, nonedematous patient can be due to Bartter's or Gitelman's syndrome, magnesium deficiency, vomiting, exogenous alkali, or diuretic ingestion. Determination of urine electrolytes (especially the urine [Cl-]) and screening of the urine for diuretics may be helpful. If the urine is alkaline, with an elevated [Na+] and [K+ ] but low [Cl-], the diagnosis is usually either vomiting (overt or surreptitious) or alkali ingestion. If the urine is relatively acid and has low concentrations of Na+ , K+, and Cl-, the most likely possibilities are prior vomiting, the posthypercapnic state, or prior diuretic ingestion. If, on the other hand, neither the urine sodium, potassium, nor chloride concentrations are depressed, magnesium deficiency, Bartter's or Gitelman's syndrome, or current diuretic ingestion should be considered. Bartter's syndrome is distinguished from Gitelman's syndrome because of hypocalciuria and hypomagnesemia in the latter disorder. The genetic and molecular basis of these two disorders has been elucidated recently (Chap. 276). Alkali Administration Chronic administration of alkali to individuals with normal renal function rarely, if ever causes alkalosis. However, in patients with coexistent hemodynamic disturbances, alkalosis can develop because the normal capacity to excrete HCO3-may be exceeded or there may be enhanced reabsorption of HCO3-. Such patients include those who receive oral or intravenous HCO3-, acetate loads (parenteral hyperalimentation solutions), citrate loads (transfusions), or antacids plus cation-exchange resins (aluminum hydroxide and sodium polystyrene sulfonate). METABOLIC ALKALOSIS ASSOCIATED WITHECFVCONTRACTION, K+DEPLETION, AND SECONDARY HYPERRENINEMIC HYPERALDOSTERONISM Gastrointestinal Origin Gastrointestinal loss of H+ from vomiting or gastric aspiration results in retention of HCO3-. The loss of fluid and NaCl in vomitus or nasogastric suction results in contraction of theECFV and an increase in the secretion of renin and aldosterone. Volume contraction causes a reduction inGFR and an enhanced capacity of the renal tubule to reabsorb HCO3-. During active vomiting, there is continued addition of HCO3-to plasma in exchange for Cl-, and the plasma [HCO3-] exceeds the reabsorptive capacity of the proximal tubule. The excess NaHCO3reaches the distal tubule, where secretion is enhanced by an aldosterone and the delivery of the poorly reabsorbed anion, HCO3-. Because of contraction of the ECFV and hypochloremia, Clis avidly conserved by the kidney. Correction of the contracted ECFV with NaCl and repair of K+deficits corrects the acid-base disorder. Renal Origin Diuretics (See alsoChap. 232) Drugs that induce chloruresis, such as thiazides and loop diuretics (furosemide, bumetanide, torsemide, and ethracrynic acid), acutely diminish theECFVwithout altering the total body bicarbonate content. The serum [HCO3-] increases. The chronic administration of diuretics tends to generate an alkalosis by increasing distal salt delivery, so that K+ and H+secretion are stimulated. The alkalosis is maintained by persistence of the contraction of the ECFV, secondary
hyperaldosteronism, K+ deficiency, and the direct effect of the diuretic (as long as diuretic administration continues). Repair of the alkalosis is achieved by providing isotonic saline to correct the ECFV deficit. Bartter's Syndrome and Gitelman's Syndrome See Chap. 276. Nonreabsorbable Anions and Magnesium Deficiency Administration of large quantities of nonreabsorbable anions, such as penicillin or carbenicillin, can enhance distal acidification and K+secretion by increasing the transepithelial potential difference (lumen negative). Mg2+deficiency results in hypokalemic alkalosis by enhancing distal acidification through stimulation of renin and hence aldosterone secretion. Potassium Depletion Chronic K+depletion may cause metabolic alkalosis by increasing urinary acid excretion. Both NH4+production and absorption are enhanced and HCO3-reabsorption is stimulated. Chronic K+ deficiency upregulates the renal H+ , K+ -ATPase to increase K+absorption at the expense of enhanced H+secretion. Alkalosis associated with severe K+depletion is resistant to salt administration, but repair of the K+ deficiency corrects the alkalosis. After Treatment of Lactic Acidosis or Ketoacidosis When an underlying stimulus for the generation of lactic acid or ketoacid is removed rapidly, as with repair of circulatory insufficiency or with insulin therapy, the lactate or ketones are metabolized to yield an equivalent amount of HCO3-. Other sources of new HCO3-are additive with the original amount generated by organic anion metabolism to create a surfeit of HCO3-. Such sources include (1) new HCO3-added to the blood by the kidneys as a result of enhanced acid excretion during the preexisting period of acidosis, and (2) alkali therapy during the treatment phase of the acidosis. Acidosis-induced contraction of theECFV and K+deficiency act to sustain the alkalosis. Posthypercapnia Prolonged CO2retention with chronic respiratory acidosis enhances renal HCO3-absorption and the generation of new HCO3-(increased net acid excretion). If the PaCO2 is returned to normal, metabolic alkalosis results from the persistently elevated [HCO3-]. Alkalosis develops if the elevated PaCO2 is abruptly returned toward normal by a change in mechanically controlled ventilation. AssociatedECFVcontraction does not allow complete repair of the alkalosis by correction of the PaCO2alone, and alkalosis persists until Cl- supplementation is provided. METABOLIC ALKALOSIS ASSOCIATED WITH ECFV EXPANSION, HYPERTENSION, AND HYPERALDOSTERONISM Mineralocorticoid administration or excess production [primary aldosteronism of Cushing's syndrome and adrenal cortical enzyme defects (Chap. 331)] increases net acid excretion and may result in metabolic alkalosis, which may be worsened by associated K+deficiency.ECFVexpansion from salt retention causes hypertension and antagonizes the reduction in GFRand/or increases tubule acidification induced by aldosterone and by K+ deficiency. The kaliuresis persists and causes continued K+ depletion with polydipsia, inability to concentrate the urine, and polyuria. Increased aldosterone levels may be the result of autonomous primary adrenal overproduction or of secondary aldosterone release due to renal overproduction of renin. In both
situations, the normal feedback of ECFV on net aldosterone production is disrupted, and hypertension from volume retention can result. Liddle's syndrome (Chap. 276) results from increased activity of collecting duct Na+channel (ENaC) and is a rare inherited disorder associated with hypertension due to volume expansion manifested as hypokalemic alkalosis and normal aldosterone levels. Symptoms With metabolic alkalosis, changes in central and peripheral nervous system function are similar to those of hypocalcemia (Chap. 340); symptoms include mental confusion, obtundation, and a predisposition to seizures, paresthesia, muscular cramping, tetany, aggravation of arrhythmias, and hypoxemia in chronic obstructive pulmonary disease. Related electrolyte abnormalities include hypokalemia and hypophosphatemia. TREATMENT This is primarily directed at correcting the underlying stimulus for HCO3-generation. If primary aldosteronism is present, correction of the underlying cause will reverse the alkalosis. [H+ ] loss by the stomach or kidneys can be mitigated by the use of H2receptor blockers, H+, K+ -ATPase inhibitors, or the discontinuation of diuretics. The second aspect of treatment is to remove the factors that sustain HCO 3-reabsorption, such asECFVcontraction or K+deficiency. Although K+deficits should be repaired, NaCl therapy is usually sufficient to reverse the alkalosis if ECFV contraction is present, as indicated by a low urine [Cl-]. If associated conditions preclude infusion of saline, renal HCO3-loss can be accelerated by administration of acetazolamide, a carbonic anhydrase inhibitor, which is usually effective in patients with adequate renal function but can worsen K+losses. Dilute hydrochloric acid (0.1 N HCl) is also effective but can cause hemolysis. Alternatively, acidification can also be achieved with oral NH4Cl, which should be avoided in the presence of liver disease. Hemodialysis against a dialysate low in [HCO3-] and high in [Cl-] can be effective when renal function is impaired. RESPIRATORY ACIDOSIS Respiratory acidosis can be due to severe pulmonary disease, respiratory muscle fatigue, or abnormalities in ventilatory control and is recognized by an increase in PaCO2and decrease in pH (Table 50-5). In acute respiratory acidosis, there is an immediate compensatory elevation (due to cellular buffering mechanisms) in HCO3-, which increases 1 mmol/L for every 10-mmHg increase in PaCO2. In chronic respiratory acidosis (>24 h), renal adaptation increases the [HCO3-] by 4 mmol/L for every 10-mmHg increase in PaCO2. The serum HCO3-usually does not increase above 38 mmol/L. The clinical features vary according to the severity and duration of the respiratory acidosis, the underlying disease, and whether there is accompanying hypoxemia. A rapid increase in PaCO2may cause anxiety, dyspnea, confusion, psychosis, and hallucinations and may progress to coma. Lesser degrees of dysfunction in chronic hypercapnia include sleep disturbances, loss of memory, daytime somnolence,
personality changes, impairment of coordination, and motor disturbances such as tremor, myoclonic jerks, and asterixis. Headaches and other signs that mimic raised intracranial pressure, such as papilledema, abnormal reflexes, and focal muscle weakness, are due to vasoconstriction secondary to loss of the vasodilator effects of CO2. Depression of the respiratory center by a variety of drugs, injury, or disease can produce respiratory acidosis. This may occur acutely with general anesthetics, sedatives, and head trauma or chronically with sedatives, alcohol, intracranial tumors, and the syndromes of sleep-disordered breathing, including the primary alveolar and obesity-hypoventilation syndromes (Chaps. 263 and264). Abnormalities or disease in the motor neurons, neuromuscular junction, and skeletal muscle can cause hypoventilation via respiratory muscle fatigue. Mechanical ventilation, when not properly adjusted and supervised, may result in respiratory acidosis, particularly if CO2production suddenly rises (because of fever, agitation, sepsis, or overfeeding) or alveolar ventilation falls because of worsening pulmonary function. High levels of positive end-expiratory pressure in the presence of reduced cardiac output may cause hypercapnia as a result of large increases in alveolar dead space (Chap. 266). Permissive hypercapnia is being used with increasing frequency because of studies suggesting lower mortality rates than with conventional mechanical ventilation, especially with severe central nervous system or heart disease. Although the potential beneficial effects of permissive hypercapnia may be mitigated by correction of the acidemia, it seems prudent, nevertheless, to keep the pH in the range of 7.2 to 7.3 by administration of NaHCO3. Acute hypercapnia follows sudden occlusion of the upper airway or generalized bronchospasm as in severe asthma, anaphylaxis, inhalational burn, or toxin injury. Chronic hypercapnia and respiratory acidosis occur in end-stage obstructive lung disease. Restrictive disorders involving both the chest wall and the lungs can cause respiratory acidosis because the high metabolic cost of respiration causes ventilatory muscle fatigue. Advanced stages of intrapulmonary and extrapulmonary restrictive defects present as chronic respiratory acidosis. The diagnosis of respiratory acidosis requires, by definition, the measurement of PaCO2and arterial pH. A detailed history and physical examination often indicate the cause. Pulmonary function studies (Chap. 250), including spirometry, diffusion capacity for carbon monoxide, lung volumes, and arterial PaCO2and O2saturation, usually make it possible to determine if respiratory acidosis is secondary to lung disease. The workup for nonpulmonary causes should include a detailed drug history, measurement of hematocrit, and assessment of upper airway, chest wall, pleura, and neuromuscular function. TREATMENT The management of respiratory acidosis depends on its severity and rate of onset. Acute respiratory acidosis can be life-threatening, and measures to reverse the underlying cause should be undertaken simultaneously with restoration of adequate alveolar ventilation. This may necessitate tracheal intubation and assisted mechanical ventilation. Oxygen administration should be titrated carefully in patients with severe
obstructive pulmonary disease and chronic CO2retention who are breathing spontaneously (Chap. 258). When oxygen is used injudiciously, these patients may experience progression of the respiratory acidosis. Aggressive and rapid correction of hypercapnia should be avoided, because the falling PaCO2may provoke the same complications noted with acute respiratory alkalosis (i.e., cardiac arrhythmias, reduced cerebral perfusion, and seizures). The PaCO2should be lowered gradually in chronic respiratory acidosis, aiming to restore the PaCO2 to baseline levels and to provide sufficient Cl- and K+ to enhance the renal excretion of HCO3-. Chronic respiratory acidosis is frequently difficult to correct, but measures aimed at improving lung function (Chap. 258) can help some patients and forestall further deterioration in most. RESPIRATORY ALKALOSIS Alveolar hyperventilation decreases PaCO2and increases the HCO3-/PaCO2ratio, thus increasing pH (Table 50-5). Nonbicarbonate cellular buffers respond by consuming HCO3-. Hypocapnia develops when a sufficiently strong ventilatory stimulus causes CO2output in the lungs to exceed its metabolic production by tissues. Plasma pH and [HCO3-] appear to vary proportionately with PaCO2over a range from 40 to 15 mmHg. The relationship between arterial [H+] concentration and PaCO2 is about 0.7 mmol/L per mmHg (or 0.01 pH unit/mmHg), and that for plasma [HCO3-] is 0.2 mmol/L per mmHg. Hypocapnia sustained longer than 2 to 6 h is further compensated by a decrease in renal ammonium and titrable acid excretion and a reduction in filtered HCO3-reabsorption. Full renal adaptation to respiratory alkalosis may take several days and requires normal volume status and renal function. The kidneys appear to respond directly to the lowered PaCO2rather than to alkalosis per se. In chronic respiratory alkalosis a 1-mmHg fall in PaCO2causes a 0.4- to 0.5-mmol/L drop in [HCO3-] and a 0.3-mmol/L fall (or 0.003 rise in pH) in [H+]. The effects of respiratory alkalosis vary according to duration and severity but are primarily those of the underlying disease. Reduced cerebral blood flow as a consequence of a rapid decline in PaCO2may cause dizziness, mental confusion, and seizures, even in the absence of hypoxemia. The cardiovascular effects of acute hypocapnia in the conscious human are generally minimal, but in the anesthetized or mechanically ventilated patient, cardiac output and blood pressure may fall because of the depressant effects of anesthesia and positive-pressure ventilation on heart rate, systemic resistance, and venous return. Cardiac arrhythmias may occur in patients with heart disease as a result of changes in oxygen unloading by blood from a left shift in the hemoglobin-oxygen dissociation curve (Bohr effect). Acute respiratory alkalosis causes intracellular shifts of Na+ , K+, and PO4-and reduces free [Ca2+] by increasing the protein-bound fraction. Hypocapnia-induced hypokalemia is usually minor. Chronic respiratory alkalosis is the most common acid-base disturbance in critically ill patients and, when severe, portends a poor prognosis. Many cardiopulmonary disorders manifest respiratory alkalosis in their early to intermediate stages, and the finding of normocapnia and hypoxemia in a patient with hyperventilation may herald the onset of rapid respiratory failure and should prompt an assessment to determine if the patient is becoming fatigued. Respiratory alkalosis is common during mechanical ventilation.
The hyperventilation syndrome may be disabling. Paresthesia, circumoral numbness, chest wall tightness or pain, dizziness, inability to take an adequate breath, and, rarely, tetany may themselves be sufficiently stressful to perpetuate the disorder. Arterial blood-gas analysis demonstrates an acute or chronic respiratory alkalosis, often with hypocapnia in the range of 15 to 30 mmHg and no hypoxemia. Central nervous system diseases or injury can produce several patterns of hyperventilation and sustained PaCO2levels of 20 to 30 mmHg. Hyperthyroidism, high caloric loads, and exercise raise the basal metabolic rate, but ventilation usually rises in proportion so that arterial blood gases are unchanged and respiratory alkalosis does not develop. Salicylates are the most common cause of drug-induced respiratory alkalosis as a result of direct stimulation of the medullary chemoreceptor (Chap. 396). The methylxanthines, theophylline, and aminophylline stimulate ventilation and increase the ventilatory response to CO2. Progesterone increases ventilation and lowers arterial PaCO2 by as much as 5 to 10 mmHg. Therefore, chronic respiratory alkalosis is a common feature of pregnancy. Respiratory alkalosis is also prominent in liver failure, and the severity correlates with the degree of hepatic insufficiency. Respiratory alkalosis is often an early finding in gram-negative septicemia, before fever, hypoxemia, or hypotension develop. The diagnosis of respiratory alkalosis depends on measurement of arterial pH and PaCO2. The plasma [K+] is often reduced and the [Cl-] increased. In the acute phase, respiratory alkalosis is not associated with increased renal HCO3-excretion, but within hours net acid excretion is reduced. In general, the HCO3-concentration falls by 2.0 mmol/L for each 10-mmHg decrease in PaCO2. Chronic hypocapnia reduces the serum [HCO3-] by 5.0 mmol/L for each 10-mmHg decrease in PaCO2. It is unusual to observe a plasma HCO3-< 12 mmol/L as a result of a pure respiratory alkalosis. When a diagnosis of respiratory alkalosis is made, its cause should be investigated. The diagnosis of hyperventilation syndrome is made by exclusion. In difficult cases, it may be important to rule out other conditions such as pulmonary embolism, coronary artery disease, and hyperthyroidism. TREATMENT The management of respiratory alkalosis is directed toward alleviation of the underlying disorder. If respiratory alkalosis complicates ventilator management, changes in dead space, tidal volume, and frequency can minimize the hypocapnia. Patients with the hyperventilation syndrome may benefit from reassurance, rebreathing from a paper bag during symptomatic attacks, and attention to underlying psychological stress. Antidepressants and sedatives are not recommended. b-Adrenergic blockers may ameliorate peripheral manifestations of the hyperadrenergic state. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 8 -ALTERATIONS IN SEXUAL FUNCTION AND REPRODUCTION 51. ERECTILE DYSFUNCTION - Kevin T. McVary Erectile dysfunction (ED) affects 10 to 25% of middle-aged and elderly men. Demographic changes, the popularity of newer treatments, and greater acceptance of ED by patients and society have led to increased diagnosis and associated health care expenditures for the management of this common disorder. Impairment of erectile function has a profound impact on the well-being of affected men. Because many patients are reluctant to initiate discussion of sexual function, the physician should address this topic directly to elicit a history of ED. PHYSIOLOGIC CONTROL OF ERECTION AND MALE SEXUAL FUNCTION Normal male sexual function requires (1) an intact libido, (2) the ability to achieve and maintain penile erection, (3) ejaculation, and (4) detumescence. Libido refers to sexual desire and is influenced by a variety of visual, olfactory, tactile, auditory, imaginative and hormonal stimuli. Sex steroids, particularly testosterone, act to increase libido. Libido can be diminished by hormonal or psychiatric disorders or by medications. The major anatomic structures of the penis that are involved in erectile function include the three corpora, which consist of the paired cavernosa and a single spongiosum that encloses the urethra. A collagenous sheath, called the tunica albuginea, individually surrounds each corpora. The micro-architecture of the corpora is composed of a mass of smooth muscle (trabecula) which contains a network of endothelial-lined vessels (lacunar spaces). Penile tumescence leading to erection depends on the increased flow of blood into the lacunar network after complete relaxation of the arteries and corporal smooth muscle. Subsequent compression of the trabecular smooth muscle against the fibroelastic tunica albuginea causes a passive closure of the emissary veins and accumulation of blood in the corpora. In the presence of a full erection and a competent valve mechanism, the corpora become noncompressible cylinders from which blood does not escape. The central nervous system exerts an important influence by either stimulating or antagonizing spinal pathways that mediate erectile function and ejaculation. The erectile response is mediated by a combination of central (psychogenic) and peripheral (reflexogenic) innervation. Sensory nerves that originate from receptors in the penile skin and glans converge to form the dorsal nerve of the penis, which travels to the S2-S4 dorsal root ganglia via the pudendal nerve. Parasympathetic nerve fibers to the penis arise from neurons in the intermediolateral columns of S2-S4 sacral spinal segments. Sympathetic innervation originates from the T-11 to the L-2 spinal segments and descends through the hypogastric plexus. Neural input to smooth muscle tone is crucial to the initiation and maintenance of an erection. There is also an intricate interaction between the corporal smooth muscle cell and its overlying endothelial cell lining (Fig. 51-1A). Nitric oxide, which induces vascular relaxation, promotes erection and is opposed by endothelin-1 (ET-1), which mediates vascular contraction. Nitric oxide is synthesized from L-arginine by nitric oxide synthase,
and is released from the nonadrenergic, noncholinergic (NANC) autonomic nerve supply to act postjunctionally on smooth muscle cells. Nitric oxide increases the production of cyclic 3¢,5¢-guanosine monophosphate (cyclic GMP), which interacts with protein kinase G and decreases intracellular calcium, causing relaxation of the smooth muscle (Fig. 51-1B). Cyclic GMP is gradually broken down by phosphodiesterase type 5 (PDE-5). Inhibitors of PDE-5, such as the oral medication sildenafil, maintain erections by reducing the breakdown of cyclic GMP. However, if nitric oxide is not produced at some level, the addition of PDE-5 inhibitor is not effective, as the drug facilitates but does not initiate the initial enzyme cascade. In addition to nitric oxide, vasoactive prostaglandins (PGE1, PGF2a) are synthesized within the cavernosal tissue and increase cyclic AMP levels, also leading to relaxation of cavernosal smooth muscle cells. Ejaculation is stimulated by the sympathetic nervous system, which results in contraction of the epididymis, vas deferens, seminal vesicles, and prostate, causing seminal fluid to enter the urethra. Seminal fluid emission is followed by rhythmic contractions of the bulbocavernosus and ischiocavernosus muscles, leading to ejaculation. Premature ejaculation is usually related to anxiety or a learned behavior and is amenable to behavioral therapy or treatment with medications such as selective serotonin reuptake inhibitors (SSRIs). Retrograde ejaculation results when the internal urethral sphincter does not close, and it may occur in men with diabetes or after surgery involving the bladder neck. Detumescence is mediated by released norepinephrine from the sympathetic nerves, release of endothelin from the vascular surface, and contraction of smooth muscle induced by activation of postsynaptic a-adrenergic receptors. These events increase venous outflow and restore the flaccid state. Venous leak can cause premature detumescence and is thought to be caused by insufficient relaxation of the corporal smooth muscle rather than a specific anatomic defect. Priapism refers to a persistent and painful erection and may be associated with sickle cell anemia, hypercoagulable states, spinal cord injury, or injection of vasodilator agents into the penis. ERECTILE DYSFUNCTION EPIDEMIOLOGY In the Massachusetts Male Aging Study (MMAS), a community-based survey of men between the ages of 40 and 70, 52% of responders reported some degree ofED. Complete ED occurred in 10% of respondents, moderate ED occurred in 25%, and minimal ED in 17%. The incidence of moderate or severe ED more than doubled between the ages of 40 and 70. In the National Health and Social Life Survey (NHSLS), which was a nationally representative sample of men and women age 18 to 59 years, 10% of men reported being unable to maintain an erection (corresponding to the proportion of men in the MMAS reporting severe ED). Incidence was highest among men in the 50 to 59 age group (21%) and among men who were poor (14%), divorced (14%), and less educated (13%). The incidence ofED is also higher among men with certain medical disorders. In theMMAS, ED correlated with the presence of diabetes mellitus, heart disease, hypertension, and decreased HDL levels. Medications used to treat diabetes or
cardiovascular disease are additional risk factors (see below). There is a higher incidence of ED among men who have undergone radiation or surgery for cancer of the prostate and in those with a lower spinal cord injury. Psychological causes of ED include depression and anger. TheNHSLSfound a higher incidence of ED among men who reported fair-to-poor health or experienced stress from unemployment or other causes. ED is not considered a normal part of the aging process. Nonetheless, it is associated with certain physiologic and psychological changes related to age. PATHOPHYSIOLOGY EDmay result from three basic mechanisms: (1) failure to initiate (psychogenic, endocrinologic, or neurogenic); (2) failure to fill (arteriogenic); or (3) failure to store (venoocclusive dysfunction) adequate blood volume within the lacunar network. The inability to initiate an erection may have psychogenic, endocrinologic, or neurogenic etiologies. These categories are not mutually exclusive, and multiple factors contribute to ED in many patients. For example, diminished filling pressure can lead secondarily to venous leak. Psychogenic factor frequently co-exist with other etiologic factors and should be considered in all cases. Diabetic, atherosclerotic, and drug-related causes account for>80% of cases of ED in older men. Vasculogenic The most frequent organic cause ofED is a disturbance of blood flow to and from the penis. Atherosclerotic or traumatic arterial disease can decrease flow to the lacunar spaces, resulting in decreased rigidity and an increased time to full erection. Excessive outflow through the veins, despite adequate inflow, may also contribute to ED. In this case, the achieved perfusion pressures cannot compensate for the unrestricted outflow needed to ensure adequate erection. This situation may be due to insufficient relaxation of trabecular smooth muscle and may occur in anxious individuals with excessive adrenergic tone or in those with damaged parasympathetic outflow. Structural alterations to the fibroelastic components of the corpora may cause a loss of compliance and an inability to compress the tunical veins. This condition may result from aging, increased cross-leaking of collagen fibers induced by nonenzymatic glycosylation, hypoxia, or altered synthesis of collagen associated with hypercholesterolemia. Fibroelastic structures can also be damaged by surgery, radiation, or trauma to the penis. Neurogenic Disorders that affect the sacral spinal cord or the autonomic fibers to the penis preclude nervous system relaxation of penile smooth muscle, thus leading toED. In patients with spinal cord injury, the degree of ED depends on the completeness and level of the lesion. Patients with incomplete lesions or injuries to the upper part of the spinal cord are more likely to retain erectile capabilities than those with complete lesions or injuries to the lower part. Although 75% of patients with spinal cord injuries have some erectile capability, only 25% have erections sufficient for penetration. Other neurologic disorders commonly associated with ED include multiple sclerosis and peripheral neuropathy. The latter is often due to either diabetes or alcoholism. Pelvic surgery may cause ED through disruption of the autonomic nerve supply. Endocrinologic Androgens increase libido, but their exact role in erectile function remains unclear. Individuals with castrate levels of testosterone can achieve erections from visual or sexual stimuli. Nonetheless, normal levels of testosterone appear to be
important for erectile function, particularly in older males. Androgen replacement therapy can improve depressed erectile function when it is secondary to hypogonadism; it is not useful for ED when endogenous testosterone levels are normal. Increased prolactin may decrease libido by suppressing gonadotropin-releasing hormone (GnRH), and it also leads to decreased testosterone levels. Treatment of hyperprolactinemia with dopamine agonists can restore libido and testosterone. Diabetic EDoccurs in 35 to 75% of men with diabetes mellitus. Pathologic mechanisms are primarily related to diabetes-associated vascular and neurologic complications. Diabetic macrovascular complications are mainly related to age, whereas microvascular complications correlate with the duration of diabetes and the degree of glycemic control (Chap. 333). Individuals with diabetes also have reduced amounts of nitric oxide synthase in both endothelial and neural tissues. Psychogenic Two mechanisms contribute to the inhibition of erections in psychogenicED. First, psychogenic stimuli to the sacral cord may inhibit reflexogenic responses, thereby blocking activation of vasodilator outflow to the penis. Second, excess sympathetic stimulation in an anxious man may increase penile smooth muscle tone. The most common causes of psychogenic ED are performance anxiety, depression, relationship conflict, loss of attraction, sexual inhibition, conflicts over sexual preference, sexual abuse in childhood, and fear of pregnancy or sexually transmitted disease. Almost all patients with ED, even when it has a clear-cut organic basis, develop a psychogenic component as a reaction to ED. Medication-Related Medication-inducedED(Table 51-1) is estimated to occur in 25% of men seen in general medical outpatient clinics. Among the antihypertensive agents, the thiazide diuretics and beta blockers have been implicated most frequently. Calcium channel blockers and angiotensin-converting enzyme inhibitors are less frequently cited. These drugs may act directly at the corporal level (e.g., calcium channel blockers) or indirectly by reducing pelvic blood pressure, which is important in the development of penile rigidity. Alpha adrenergic blockers are less likely to cause ED. Estrogens,GnRHagonists, H2antagonists, and spironolactone cause ED by suppressing gonadotropin production or by blocking androgen action. Antidepressant and antipsychotic agents -- particularly neuroleptics, tricyclics, andSSRIs -- are associated with erectile, ejaculatory, orgasmic, and sexual desire difficulties. Digoxin induces ED via blockade of the Na+ ,K+-ATPase pump, resulting in a net increase in intracellular calcium and increased corporal smooth muscle tone. Although many medications can causeED, patients frequently have concomitant risk factors that confound the clinical picture. If there is a strong association between the institution of a drug and the onset of ED, alternative medications should be considered. Otherwise, it is often practical to treat the ED without attempting multiple changes in medications, as it may be difficult to establish a causal role for the drug. CLINICAL EVALUATION A good physician-patient relationship helps to unravel the possible causes ofED, many of which require discussion of personal and sometimes embarrassing topics. For this reason, a primary care provider is often ideally suited to initiate the evaluation. A
complete medical and sexual history should be taken in an effort to assess whether the cause of ED is organic, psychogenic, or multifactorial (Fig. 51-2). Initial questions should focus on the onset of symptoms, the presence and duration of partial erections, and the progression of ED. A history of nocturnal or early morning erections is useful for distinguishing physiologic from psychogenic ED. Nocturnal erections occur during rapid eye movement (REM) sleep and require intact neurologic and circulatory systems. Organic causes of ED are generally characterized by a gradual and persistent change in rigidity or the inability to sustain nocturnal, coital, or self-stimulated erections. The patient should also be questioned about the presence of penile curvature or pain with coitus. It is also important to address libido, as decreased sexual drive and ED are sometimes the earliest signs of endocrine abnormalities (e.g., increased prolactin, decreased testosterone levels). It is useful to ask whether the problem is confined to coitus with one or other partners; ED arises not uncommonly in association with new or extramarital sexual relationships. Situational ED, as opposed to consistent ED, suggests psychogenic causes. Ejaculation is much less commonly affected than erection, but questions should be asked about whether ejaculation is normal, premature, delayed, or absent. Relevant risk factors should be identified, such as diabetes mellitus, coronary artery disease, lipid disorders, hypertension, peripheral vascular disease, smoking, alcoholism, and endocrine or neurologic disorders. The patient's surgical history should be explored with an emphasis on bowel, bladder, prostate, or vascular procedures. A complete drug history is also important, as medications constitute a major source of reversible ED. Social changes that may precipitate ED are also crucial to the evaluation, including health worries, spousal death, divorce, relationship difficulties, and financial concerns. The physical examination is an essential element in the assessment ofED. Signs of hypertension as well as evidence of thyroid, hepatic, hematologic, cardiovascular, or renal diseases should be sought. An assessment should be made of the endocrine and vascular systems, the external genitalia, and the prostate gland. The penis should be carefully palpated along the corpora to detect fibrotic plaques. Reduced testicular size and loss of secondary sexual characteristics are suggestive of hypogonadism. Neurologic examination should include assessment of anal sphincter tone, the bulbocavernosus reflex, and testing for peripheral neuropathy. Selected laboratory testing is recommended in all cases. Although hyperprolactinemia is uncommon, a serum prolactin level should be measured, as decreased libido and/or erectile dysfunction may be the presenting symptoms of a prolactinoma or other mass lesions of the sella (Chap. 328). The serum testosterone level should be measured and, if low, gonadotropins should be measured to determine whether hypogonadism is primary (testicular) or secondary (hypothalamic-pituitary) in origin (Chap. 335). Serum chemistries, CBC, and lipid profiles may be of value, if not performed recently, as they can yield evidence of anemia, diabetes, hyperlipidemia, or other systemic diseases associated withED. Determination of serum PSA should be conducted according to recommended clinical guidelines (Chap. 95). Additional diagnostic testing is rarely necessary in the evaluation of ED. However, in selected patients, specialized testing may provide insight into pathologic mechanisms of ED and aid in the selection of treatment options. Optional specialized testing includes: (1) studies of nocturnal penile tumescence and rigidity; (2) vascular testing (in-office
injection of vasoactive substances, penile Doppler ultrasound, penile angiography, dynamic infusion cavernosography/cavernosometry); (3) neurologic testing (biothesiometry-graded vibratory perception; somatosensory evoked potentials); and (4) psychological diagnostic tests. The information potentially gained from these procedures must be balanced against their invasiveness and cost. TREATMENT Patient Education Patient and partner education is essential in the treatment ofED. In goal-directed therapy, education facilitates understanding of the disease, results of the tests, and selection of treatment. Discussion of treatment options helps to clarify how treatment is best offered, and to stratify first- and second-line therapies. Patients with high-risk lifestyle issues, such as smoking, alcohol abuse, or recreational drug use, should be counseled on the role these factors play in the development of ED. Oral Agents Sildenafil is the only approved and effective oral agent for the treatment ofED. Sildenafil has markedly improved the management of ED because it is effective for the treatment of a broad range of causes of ED, including psychogenic, diabetic, vasculogenic, post-radical prostatectomy (nerve-sparing procedures), and spinal cord injury. Sildenafil is a selective and potent inhibitor of PDE-5, the predominant phosphodiesterase isoform found in the penis. It is administered in doses of 25, 50, or 100 mg, and enhances erections after sexual stimulation. The onset of action is approximately 60 to 90 min. Reduced initial doses should be considered for patients who are elderly, have renal insufficiency, or are taking medications that inhibit the CYP3A4 metabolic pathway in the liver (e.g., erythromycin, cimetidine, ketoconazole, and, possibly, itraconazole and mibefradil), as they may increase the serum concentration of sildenafil. The drug does not affect ejaculation, orgasm, or sexual drive. Side effects associated with sildenafil include headaches (19%), facial flushing (9%), dyspepsia (6%) and nasal congestion (4%). Approximately 7% of men may experience transient altered color vision (blue halo effect). Sildenafil is contraindicated in men receiving nitrate therapy for cardiovascular disease, including agents delivered by oral, sublingual, transnasal, or topical routes. These agents can potentiate its hypotensive effect and may result in profound shock. Likewise, amyl/butyl nitrates (poppers) may have a fatal synergistic effect on blood pressure. Sildenafil should also be avoided in patients with congestive heart failure and cardiomyopathy because of the risk of vascular collapse. Because sexual activity leads to an increase in physiologic expenditure [5 to 6 metabolic equivalents (METS)], physicians have been advised to exercise caution in prescribing any drug for sexual activity to those with active coronary disease, heart failure, borderline hypotension, hypovolemia, and to those on complex antihypertensive regimens. Androgen Therapy Testosterone replacement is used to treat both primary and secondary causes of hypogonadism (Chap. 335). Androgen supplementation in the setting of normal testosterone is rarely efficacious and is discouraged. Methods of androgen replacement include parenteral administration of long-acting testosterone esters (enanthate and cypionate), oral preparations (17a-alkylated derivatives), and transdermal patches (Chap. 335). The long-acting 17b-hydroxy esters of testosterone are the safest, most cost-effective, and practical preparations available. The administration of 200 to 300 mg intramuscularly every 2 to 3 weeks provides a practical
option but is far from an ideal physiologic replacement. Oral androgen preparations have the potential for hepatotoxicity and should be avoided. Transdermal delivery of testosterone more closely mimics physiologic testosterone levels, but it is unclear whether this translates into improved sexual function. Because testosterone gradually decreases into the hypogonadal range by 24 hours, patches need to replaced daily. Testosterone therapy is contraindicated in men with androgen-sensitive cancers and may be inappropriate for men with bladder neck obstruction. It is generally advisable to measure PSA before giving androgen. Hepatic function should be tested before and during testosterone therapy. Vacuum Constriction Devices Vacuum constriction devices (VCD) are a well-established, noninvasive therapy. They are a reasonable treatment alternative for select patients who cannot take sildenafil or do not desire other interventions. VCD draw venous blood into the penis and use a constriction ring to restrict venous return and maintain tumescence. Adverse events with VCD include pain, numbness, bruising, and altered ejaculation. Additionally, many patients complain that the devices are cumbersome and that the induced erections have a non-physiologic appearance. Intraurethral Alprostadil If a patient fails to respond to oral agents, a reasonable next choice is intraurethral or self-injection or vasoactive substances. Intraurethral prostaglandin E1(alprostadil), in the form of a semisolid pellet (doses of 125 to 1000 ug), is delivered with an applicator. Approximately 65% of men receiving intraurethral alprostadil respond with an erection when tested in the office, but only 50% of those achieve successful coitus at home. Intraurethral insertion is associated with a markedly reduced incidence of priapism in comparison to intracavernosal injection. Intracavernosal Self-Injection Injection of synthetic formulations of alprostadil is effective in 70 to 80% of patients withED, but discontinuation rates are high because of the invasive nature of administration. Doses range between 1 and 40 ug. Injection therapy is contraindicated in men with a history of hypersensitivity to the drug and in men at risk for priapism (hypercoagulable states, sickle cell disease). Side effects include local adverse events, prolonged erections, pain, and fibrosis with chronic use. Various combinations of alprostadil, phentolamine, and/or papaverine are sometimes used. Surgery A less frequently used form of therapy forEDinvolves the surgical implantation of a semi-rigid or inflatable penile prosthesis. These surgical treatments are invasive, associated with potential complications, and generally reserved for treatment of refractory ED. Despite their high cost and invasiveness, penile prostheses are associated with high rates of patient satisfaction. Sex Therapy A course of sex therapy may be useful for addressing specific interpersonal factors that may affect sexual functioning. Sex therapy generally consists of in-session discussion and at-home exercises specific to the person and the relationship. It is preferable if therapy includes both partners, provided the patient is involved in an ongoing relationship. (Bibliography omitted in Palm version)
Back to Table of Contents
52. DISTURBANCES OF MENSTRUATION AND OTHER COMMON GYNECOLOGIC COMPLAINTS IN WOMEN - Bruce R. Carr, Karen D. Bradshaw Complaints related to the female reproductive tract can be categorized as disorders of menstruation, pelvic pain, disturbances in sexual function, or infertility. However, a single disorder, e.g., leiomyoma of the uterus, can present with symptoms referable to any one or more of these categories. Furthermore, sexual dysfunction can interdigitate with other problems in several ways. On the one hand, in women with complaints related to other reproductive tract functions, the underlying problem may actually be severe sexual dysfunction or marital conflict. Alternatively, women with severe organic disorders of the pelvis, e.g., pelvic inflammatory disease or endometriosis, may present with sexual dysfunction such as dyspareunia that in fact is only a minor manifestation of the underlying disease. Since normal reproductive function depends on the integrated action of the central nervous system, the endocrine glands, and the reproductive organs, menstrual cycle abnormalities, sexual dysfunction, and infertility may be the result of systemic and psychological disorders as well as of primary defects in the endocrine and reproductive organs. The endocrine and physiologic control -- normal and abnormal -- of puberty, reproductive life, and menopause are discussed in Chap. 336. The focus of this chapter is on the initial evaluation of women with disturbances of the reproductive tract. DISTURBANCES IN MENSTRUATION Disorders of menstruation can be divided into abnormal uterine bleeding and amenorrhea. Abnormal Uterine Bleeding The menstrual cycle is defined as the interval between the onset of one bleeding episode and the onset of the next. In normal women the cycle averages 28± 3 days, the mean duration of menstrual flow is 4 ± 2 days, and the average blood loss is 35 to 80 mL. Between menarche and menopause most women experience one or more episodes of abnormal uterine bleeding, here defined as any bleeding pattern outside the normal ranges of frequency, duration, and/or amount of blood loss. The decision to evaluate a patient depends on the severity and frequency of the abnormal bleeding pattern. When vaginal bleeding occurs, it should first be determined whether the blood is derived from the uterine endometrium. Rectal, bladder, cervical, and vaginal sources of bleeding must be excluded. Once the bleeding is established to be uterine in origin, a pregnancy-related disorder (such as threatened or incomplete abortion or ectopic pregnancy) must be ruled out by physical examination and appropriate laboratory tests. It should also be remembered that uterine bleeding may also be the initial or principal manifestation of a generalized bleeding diathesis. The remaining causes of abnormal uterine bleeding can be divided into those associated with ovulatory or anovulatory cycles. Ovulatory Cycles Menstrual bleeding with ovulatory cycles is spontaneous, regular in onset, predictable in duration and amount of flow, and frequently associated with discomfort; it is the consequence of progesterone withdrawal at the end of the luteal
(postovulatory) phase and requires prior estrogen priming of the endometrium during the follicular (preovulatory) phase of the cycle. When deviations from an established pattern of menstrual flow occur but the cycles are still regular, the usual cause is disease of the outflow tract. For example, regular, prolonged, excessive bleeding episodes can result from abnormalities of the uterus such as submucous leiomyomas, adenomyosis, or endometrial polyps. On the other hand, cyclic, predictable menstruation characterized by spotting or light bleeding suggests obstruction of the outflow tract as with uterine synechiae or scarring of the cervix. Intermittent bleeding between cyclic ovulatory menses is often due to cervical or endometrial lesions. Anovulatory Cycles Uterine bleeding that is irregular in occurrence, unpredictable as to amount and duration of flow, and usually painless is called dysfunctional or anovulatory uterine bleeding. This type of bleeding is the result of a failure of normal follicular maturation with consequent anovulation and may be either transient or chronic. Transient disruption of ovulatory cycles occurs most often in the early menarcheal years, during the perimenopausal period, or as the consequence of a variety of stresses and intercurrent illnesses. Persistent dysfunctional uterine bleeding during the reproductive years can occur in several organic diseases that affect ovarian function and is most often due to estrogen breakthrough bleeding. Estrogen breakthrough bleeding occurs when estrogen stimulation of the endometrium is continuous and is not interrupted by cyclic progesterone withdrawal, as can occur in polycystic ovarian disease. Amenorrhea Amenorrhea is defined either as failure of menarche by age 16, regardless of the presence or absence of secondary sexual characteristics, or as the absence of menstruation for 6 months in a woman with previous periodic menses. Amenorrhea in a woman who has never menstruated is termed primary amenorrhea; cessation of menses is termed secondary amenorrhea. Because some disorders can cause both primary and secondary amenorrhea, we prefer a functional classification based on the nature of the underlying defect, namely, anatomic defects of the outflow tract (uterus, cervix, or vagina), ovarian failure, and chronic anovulation. Anatomic defects of the outflow tract include congenital defects of the vagina, imperforate hymen, transverse vaginal septa, cervical stenosis, intrauterine adhesions (synechiae), absence of the vagina or uterus, and uterine maldevelopment. The diagnosis of an anatomic defect is usually made by physical examination but may be confirmed by demonstrating failure of bleeding following administration of estrogen plus a progestogen for 21 days. Pelvic ultrasonography, magnetic resonance imaging, hysterosalpingogram, or hysteroscopy may be helpful in defining the defect. Causes of ovarian failure include gonadal dysgenesis, deficiency of 17a-hydroxylase, resistant ovary syndrome, and premature ovarian failure. Ovarian failure encompasses disorders in which the ovary is deficient in germ cells and those in which the germ cells are resistant to follicle-stimulating hormone (FSH). The diagnosis of ovarian failure as the cause of amenorrhea is confirmed by an elevated plasma FSH level. Women with chronic anovulation fail to ovulate spontaneously but have the capability of ovulating with appropriate therapy. In some women with chronic anovulation, total estrogen production is adequate, but it is not secreted in a cyclic fashion. In others,
estrogen production is deficient. Women who have adequate estrogen production and demonstrate withdrawal bleeding after progestogen challenge often have polycystic ovarian disease (seeFig. 336-8). Other causes include hormone-secreting ovarian and adrenal tumors. Women with deficient or absent estrogen production, and therefore with absence of withdrawal bleeding after progestogen administration, usually have hypogonadotropic hypogonadism due to organic or functional disorders of the pituitary or central nervous system such as brain tumors, pituitary tumors (especially prolactin-secreting adenomas), primary hypopituitarism, or Sheehan's syndrome. PELVIC PAIN Pelvic pain may originate in the pelvis or be referred from another region of the body. A pelvic source is suggested by the history (e.g., dysmenorrhea and dyspareunia) and physical findings, but a high index of suspicion must be entertained for extrapelvic disorders that refer to the pelvis, such as appendicitis, diverticulitis, cholecystitis, intestinal obstruction, and urinary tract infections (Chap. 14). "Physiologic" Pelvic Pain Pain Associated with Ovulation ("Mittelschmerz") Many women experience low abdominal discomfort with ovulation, typically a dull aching pain at midcycle in one lower quadrant lasting from minutes to hours. It is rarely severe or incapacitating. The pain may result from peritoneal irritation by follicular fluid released into the peritoneal cavity at ovulation. The onset at midcycle and short duration of pain suggest this diagnosis. Premenstrual or Menstrual Pain In normal ovulatory women, somatic symptoms during the few days prior to menses may be insignificant or disabling. Such symptoms include edema, breast engorgement, and abdominal bloating or discomfort. A symptom complex of cyclic irritability, depression, and lethargy is known as the premenstrual syndrome (PMS). PMS appears to be caused by changes in gonadal steroid levels. Although there is no consensus about therapy, randomized, controlled trials suggest significant improvement with the daily use of serotonin-reuptake inhibitors. Severe or incapacitating uterine cramping during ovulatory menses and in the absence of demonstrable disorders of the pelvis is termed primary dysmenorrhea. Primary dysmenorrhea is caused by prostaglandin-induced uterine ischemia and is treated with prostaglandin synthetase inhibitors and/or oral contraceptive agents. Pelvic Pain due to Organic Causes Severe dysmenorrhea associated with disease of the pelvis is termed secondary dysmenorrhea. Organic causes of pelvic pain can be classified as (1) uterine, (2) adnexal, (3) vulvar or vaginal, and (4) pregnancy-associated. Uterine Pain Pain of uterine etiology is often chronic and continuous and increases in intensity during menstruation and intercourse. Causes include leiomyomas of the uterus (particularly submucous and degenerating leiomyomas), adenomyosis, and cervical stenosis. Infections of the uterus associated with intrauterine manipulation following
dilatation and curettage or with the insertion of intrauterine devices can also cause pelvic pain (Chap. 336). Pelvic pain due to endometrial or cervical cancer is usually a late manifestation (Chap. 336). Adnexal Pain The most common cause of pain in the adnexae (fallopian tubes and ovaries) is infection (Chap. 133). Acute salpingo-oophoritis presents as low abdominal pain, fever, and chills; begins a few days after a menstrual period; and is usually due to chlamydial or gonococcal disease with or without a superimposed pyogenic infection. Chronic pelvic inflammatory disease results from either a single episode or multiple episodes of infection and may present as infertility associated with chronic pelvic pain that increases in intensity with menses and intercourse. On physical examination, cervical motion tenderness, adnexal tenderness, and adnexal thickening and/or masses may be present. Pelvic inflammatory disease may become a surgical emergency if peritonitis results from rupture of a tuboovarian abscess. Ovarian cysts or neoplasms may cause pelvic pain that becomes more severe with torsion or rupture of the mass, and ectopic pregnancy must be considered in the differential diagnosis (see below). Endometriosis involving fallopian tubes, ovaries, or peritoneum may cause both chronic low abdominal pain and infertility; the magnitude of tissue involvement does not always correlate with the severity of symptoms. Endometriosis pain typically increases with menstruation and, if the posterior ligaments of the uterus are involved, with intercourse. Vulvar or Vaginal Pain Pain in these areas is most often due to infectious vaginitis caused by Monilia, Trichomonas, or bacteria and is characteristically associated with vaginal discharge and pruritus. Herpetic vulvitis, other dermatologic conditions of the vulva, condyloma acuminatum, and cysts or abscesses of Bartholin's glands may also cause vulvar pain. Pregnancy-Associated Disorders Pregnancy must be considered in the differential diagnosis of pelvic pain during the reproductive years. Threatened abortion or incomplete abortion often presents with uterine cramping, bleeding, or passage of tissue following a period of amenorrhea. Ectopic pregnancy may be insidious in presentation or result in abrupt intraperitoneal hemorrhage and maternal death. Evaluation of Pelvic Pain The evaluation of pelvic pain requires a careful history and pelvic examination. This often leads to the correct diagnosis and institution of appropriate treatment. If the pain is severe and the diagnosis is unclear, the workup should follow that outlined for the acute abdomen (Chap. 14). A culdocentesis may be indicated if a ruptured ectopic pregnancy is suspected. If there is a question of an adnexal mass or if the patient is so obese as to preclude a thorough pelvic examination, abdominal or vaginal sonography may be useful. Serial human chorionic gonadotropin (hCG) measurements may help in establishing a diagnosis of tubal pregnancy and are useful in determining if an intrauterine pregnancy is viable. Finally, diagnostic laparoscopy and laparotomy may be indicated with pain of undetermined etiology. SEXUAL DYSFUNCTION Some women with sexual dysfunction describe minor complaints related to the reproductive tract as a means of bringing sexual problems to the attention of the physician. Alternatively, sexual dysfunction may be thought to be the cause of low
abdominal discomfort or dyspareunia when the actual etiology is organic. However, more and more women seek medical advice because of sexual problems that interface in provenance between medicine, psychiatry, and sociology. The normal sexual response begins with sexual arousal, which causes genital vasocongestion that results in vaginal lubrication in preparation for intromission. The lubrication is due to the formation of a transudate in the vagina and in conjunction with genital congestion produces the so-called orgasmic platform prior to orgasm. Sexual stimuli (visual, tactile, auditory, and olfactory) as well as healthy vaginal tissue are prerequisites for genital vasocongestion and vaginal lubrication. During the second stage of the sexual response, involuntary contractions of the muscles of the pelvis result in a pleasurable cortical sensory phenomenon known as orgasm. Direct or indirect stimulation of the clitoris is important in the production of the female orgasm. In simple terms, sexual dysfunction can be due to interference with the arousal or orgasmic phases of the sexual response. Either disorder can be due to an organic or functional cause or both. Illnesses that impair neurologic function such as diabetes mellitus or multiple sclerosis can prevent normal sexual arousal. Local pelvic diseases such as vaginitis, endometriosis, and salpingo-oophoritis may preclude normal sexual response because of resulting dyspareunia. Debilitating systemic diseases such as cancer and cardiovascular diseases may inhibit normal sexual response indirectly. More commonly, failure of a normal sexual response is due to psychological factors that impair sexual arousal. Such problems include misinformation, e.g., the perception of sexual satisfaction as bad, or feelings of guilt about previous psychologically traumatic events such as incest, rape, or unwanted pregnancy. In addition, women who have had previous hysterectomy or mastectomy may perceive themselves as "incomplete." Stresses such as anxiety, depression, fatigue, and marital or interpersonal conflicts may lead to failure of the vasocongestive response and prevent normal vaginal lubrication. Women with such experiences may be unable to achieve normal sexual response unless they receive professional counseling. Such problems are approached by attempting to identify and reduce the causative stresses. Failure to achieve orgasm is a specific form of sexual dysfunction. In the absence of orgasm many women enjoy sexual encounters to variable degrees because of the pleasure derived from closeness in a cherished relationship, particularly with a loving partner. However, for other women sexual relations with rare or absent orgasms are frustrating and unsatisfying. In many instances, failure of orgasm is due to insufficient clitoral stimulation and may be rectified by appropriate counseling and patient education. A specific entity, "vaginismus," painful, involuntary contractions of the musculature surrounding the entrance to the vagina, is a rare cause of dyspareunia. It is a conditioned response to a previous real or imagined frightening or traumatic sexual experience. Treatment is directed to elimination of the conditioned response by progressive vaginal dilation by the patient in conjunction with marital therapy. REPRODUCTION
Infertility is discussed in detail in Chap. 54. The approach to infertile couples always involves evaluation of both the man and woman. The history should address the frequency of intercourse, the sexual responses of both, the use of contraceptives or lubricants, prior pregnancies, interval to conception and outcome of pregnancy, previous or past medical illnesses, and all medications taken. Male-associated factors account for a third of infertility problems. Therefore, one of the first procedures in the workup of infertile couples should be a semen analysis. The initial evaluation of the woman includes documentation of normal ovulatory cycles. A history of regular, cyclic, predictable, spontaneous menses usually indicates ovulatory cycles, which may be confirmed by basal body temperature graphs, properly timed endometrial biopsies, or plasma progesterone measurements during the luteal phase of the cycle. Also, the diagnosis of luteal-phase dysfunction (low progesterone secretion during the luteal phase) can be established by these methods. Transvaginal ultrasonography is useful for evaluating follicular development. The most common cause of infertility in women is tubal disease, usually due to infection (pelvic inflammatory disease) or endometriosis. Tubal disease can be evaluated by obtaining a hysterosalpingogram or by diagnostic laparoscopy. Tubal diseases can usually be treated by laparoscopic tuboplasty and lysis of adhesions. In many instances of infertility, it is now possible to use assisted reproductive technologies including in vitro fertilization and embryo transfer, gamete intrafallopian tube transfer, transfer of cryopreserved ova and embryos, donor oocytes or donor sperm, and ovarian hyperstimulation with clomiphene citrate or gonadotropins followed by intrauterine insemination. The desire for contraception is also a frequent cause for women to seek medical treatment or evaluation. The most widely used methods for fertility control include (1) rhythm and withdrawal techniques, (2) barrier methods, (3) intrauterine devices, (4) oral steroid contraceptives, (5) sterilization, and (6) abortion.*These methods and their complications are discussed in Chap. 54. (Bibliography omitted in Palm version) Back to Table of Contents
53. HIRSUTISM AND VIRILIZATION - David A. Ehrmann Hirsutism, defined as excessive male-pattern hair growth, affects approximately 10% of women of reproductive age. Hirsutism may be mild, essentially representing a variation of normal hair growth, or rarely it may be the harbinger of a serious underlying condition. It is often idiopathic but may be caused by several conditions associated with androgen excess, such as polycystic ovarian syndrome (PCOS) or congenital adrenal hyperplasia (CAH) (Table 53-1). Cutaneous manifestations commonly associated with hirsutism include acne and male-pattern balding (androgenic alopecia). Virilization, on the other hand, refers to the state in which androgen levels are sufficiently high to cause additional signs and symptoms such as deepening of the voice, breast atrophy, increased muscle bulk, clitoromegaly, and increased libido; virilization is an ominous sign that suggests the possibility of an ovarian or adrenal neoplasm. HAIR FOLLICLE GROWTH AND DIFFERENTIATION Hair can be categorized as either vellus (fine, soft, and not pigmented) or terminal (long, coarse, and pigmented). The number of hair follicles does not change over an individual's lifetime, but the follicle size and type of hair can change in response to numerous factors, particularly androgens. Androgens are necessary for terminal hair and sebaceous gland development and mediate differentiation of pilosebaceous units (PSUs) into either a terminal hair follicle or a sebaceous gland. In the former case, androgens transform the vellus hair into a terminal hair; in the latter, the sebaceous component proliferates and the hair remains vellus. There are three phases in the cycle of hair growth: (1) anagen (growth phase), (2) catagen (involution phase), and (3) telogen (rest phase). Depending on the body site, hormonal regulation may play an important role in the hair growth cycle. For example, the eyebrows, eyelashes, and vellus hairs are androgen-insensitive, whereas the axillary and pubic areas are sensitive to low doses of androgens. Hair growth on the face, chest, upper abdomen, and back requires greater levels of androgens and is therefore more characteristic of the pattern typically seen in males. Androgen excess in women leads to increased hair growth in most androgen-sensitive sites but will manifest with loss of hair in the scalp region, in part by reducing the time hairs spend in anagen phase. Although androgen excess underlies most cases of hirsutism, there is only a modest correlation between androgen levels and the quantity of hair growth. This is due to the fact that hair growth from the follicle depends on local factors and variability in end-organ sensitivity, as well as circulating androgen concentrations. Genetic factors and ethnic background also influence hair growth. In general, dark-haired individuals tend to be more hirsute than blonde or fair individuals. Asians and Native Americans have relatively sparse hair in regions sensitive to high androgen levels, whereas people of Mediterranean descent are more hirsute. For these reasons, family history and ethnic background are important considerations when assessing the etiology and severity of hirsutism. CLINICAL ASSESSMENT
Historic elements relevant to the assessment of hirsutism include the age of onset and rate of progression of hair growth and associated symptoms or signs (e.g., acne). Depending on the cause, excess hair growth is typically first noted during the second and third decades. The growth is usually slow but progressive. Sudden development and rapid progression of hirsutism suggests the possibility of an androgen-secreting neoplasm, in which case findings of virilization may also be present. The age of onset of menstrual cycles (menarche) and the pattern of the menstrual cycle should be ascertained; irregular cycles from the time of menarche onward are more likely to result from ovarian rather than adrenal androgen excess. Associated symptoms such as galactorrhea should prompt evaluation for hyperprolactinemia (Chap. 328) and possibly hypothyroidism (Chap. 330). Hypertension, striae, easy bruising, centripetal weight gain, and weakness suggest hypercortisolism (Cushing's syndrome; Chap. 331). Rarely, patients with growth hormone excess (i.e., acromegaly) will present with hirsutism. Use of medications such as phenytoin, minoxidil, or cyclosporine may be associated with androgen-independent causes of excess hair growth (i.e., hypertrichosis). A family history of infertility and/or hirsutism may indicate disorders such as nonclassic congenital adrenal hyperplasia (CAH), a disorder particularly common in Ashkenazi Jews, among others (Chap. 331). Physical examination should include measurement of height, weight, and calculation of body mass index (BMI). A BMI >25 kg/m2 is indicative of excess weight for height, and values>30 kg/m2are often seen in association with hirsutism. Notation should be made of blood pressure. Cutaneous signs sometimes associated with androgen excess and insulin resistance include acanthosis nigricans and skin tags. An objective clinical assessment of hair distribution and quantity is central to the evaluation in any woman presenting with hirsutism. This assessment permits the distinction between hirsutism and hypertrichosis and provides a baseline reference point to gauge the response to treatment. Hypertrichosis refers to the excessive growth of androgen-independent hair which is vellus, prominent in nonsexual areas, and most commonly familial or caused by metabolic disorders (e.g., thyroid disturbances, anorexia nervosa) or medications (e.g., phenytoin, minoxidil or cyclosporine). A simple and commonly used method to grade hair growth is the modified scale of Ferriman and Gallwey (Fig. 53-1), where each of nine androgen-sensitive sites is graded from 0 to 4. Approximately 95% of Caucasian women have a score below 8 on this scale; thus, it is normal for most women to have some hair growth in androgen-sensitive sites. Scores above 8 suggest an excess of androgen-mediated hair growth, a finding that should be assessed further by hormonal evaluation (see below). In racial/ethnic groups that are less likely to manifest hirsutism (e.g., Asian women), additional cutaneous evidence of androgen excess should be sought, including pustular acne or thinning hair. HORMONAL EVALUATION Androgens are secreted by both the ovaries and adrenal glands in response to their respective tropic hormones, luteinizing hormone (LH) and adrenocorticotropic hormone (ACTH). The principal circulating steroids involved in the etiology of hirsutism are
androstenedione, dehydroepiandrosterone (DHEA) and its sulfated form (DHEAS), and testosterone. The ovaries and adrenal glands normally contribute about equally to testosterone production. Further, approximately half of the total testosterone originates from direct glandular secretion, and the remainder is derived from the peripheral conversion of androstenedione and DHEA (Chap. 335). Although it is the most important circulating androgen, testosterone is, in effect, the penultimate androgen in mediating hirsutism; it is converted to the more potent dihydrotestosterone (DHT) by the enzyme 5a-reductase, which is located in the pilosebaceous unit. DHT has a higher affinity for, and slower dissociation from, the androgen receptor. The local production of DHT allows it to serve as the primary mediator of androgen action at the level of the pilosebaceous unit. There are two isoenzymes of 5a-reductase: type 2 is found in the prostate gland and in hair follicles, whereas type 1 is primarily found in sebaceous glands. One approach to testing for hyperandrogenemia is depicted in Fig. 53-2. This involves measuring blood levels of testosterone andDHEAS. It is also important to measure the level of free (or unbound) testosterone, because it is the fraction of testosterone that is not bound to its carrier protein, sex-hormone binding globulin (SHBG), that is biologically available. Hyperinsulinemia and/or androgen excess decrease hepatic production of SHBG, often resulting in levels of total testosterone within the high-normal range at a time when the free hormone is substantially elevated. Because adrenal androgens are readily suppressed by low doses of glucocorticoids, the dexamethasone androgen-suppression test may broadly distinguish ovarian from adrenal androgen overproduction. A blood sample is obtained before and after administering dexamethasone (0.5 mg orally every 6 h for 4 days). An adrenal source is suggested by suppression of plasma free testosterone into the normal range; incomplete suppression suggests ovarian androgen excess. A baseline plasma total testosterone level >12 nmol/L (>3.5 ng/mL) usually indicates a virilizing tumor, whereas a level>7 nmol/L (>2 ng/mL) is suggestive. A basalDHEASlevel >18.5 umol/L (>7000 ug/L) suggests an adrenal tumor. Although DHEAS has been proposed as a "marker" of predominant adrenal androgen excess, it is not unusual to find modest elevations in DHEAS among women withPCOS. Computed tomography (CT) or magnetic resonance imaging (MRI) should be used to localize an adrenal mass, and ultrasound will usually suffice to identify an ovarian mass, if clinical evaluation and hormonal levels suggest these possibilities. PCOSis the most common cause of ovarian androgen excess (Chap. 336). However, the increased ratio ofLH to follicle-stimulating hormone that is often seen in carefully studied patients with PCOS may not be exhibited in up to half of these women due to the pulsatility of gonadotropins. If performed, ultrasound shows enlarged ovaries and/or increased stroma in many women with PCOS. However, polycystic ovaries may also be found in women without clinical or laboratory features of PCOS. Therefore, polycystic ovaries are a relatively insensitive and nonspecific finding for the diagnosis of ovarian hyperandrogenism. Though it is not widely used, gonadotropin-releasing hormone agonist testing can be used to make a specific diagnosis of ovarian hyperandrogenism. A peak 17-hydroxyprogesterone level ³7.8 nmol/L (³2.6 ug/L), after the administration of 100 ug nafarelin (or 10 ug/kg leuprolide) subcutaneously, is virtually diagnostic of
ovarian hyperandrogenism. NonclassicCAH is most commonly due to 21-hydroxylase deficiency but can also be caused by autosomal recessive defects in other steroidogenic enzymes necessary for adrenal corticosteroid synthesis (Chap. 331). Because of the enzyme defect, the adrenal gland cannot secrete glucocorticoids efficiently (especially cortisol). This results in diminished negative feedback inhibition ofACTH, leading to compensatory hyperplasia of the adrenal cortex and accumulation of steroid precursors proximal to the enzyme defect. These precursors are subsequently converted to androgen. Deficiency of 21-hydroxylase can be reliably excluded by determining a morning 17-hydroxyprogesterone level 0.3°C (>0.6°F) for 10 days]. An endometrial biopsy to exclude luteal phase insufficiency is no longer considered an essential part of the infertility workup for most patients. Even in the presence of ovulatory cycles, evaluation of ovarian reserve is recommended for women over 35 by measurement ofFSH on day 3 of the cycle or in response to clomiphene, an estrogen antagonist (see below). An FSH level 14% normal forms (using strict Kruger criteria), whereas low fertilization is seen with1.5 cm (see "Hyperpigmentation") and multiple Lisch nodules, are seen in von Recklinghausen's disease (NF type I). Lisch nodules are 1-mm yellow-brown spots within the iris that are best observed with slit-lamp examination. Additional manifestations include axillary
freckling and peripheral andCNStumors (Chap. 370). In some patients the neurofibromas are localized and unilateral, whereas in others they are limited to the CNS. Angiofibromas are firm, pink to skin-colored papules that measure from 3 mm to several centimeters in diameter. When they are located on the central cheeks (adenoma sebaceum) or multiple fibromas are seen around the nails, the patient has tuberous sclerosis. It is an autosomal disorder due to mutations in two different genes, and the associated findings are discussed in the section on ash leaf spots as well as inChap. 370. Multiple facial angiofibromas have also been observed in patients with multiple endocrine neoplasia (MEN) syndrome, type 1. Neuromas (benign proliferations of nerve fibers) are also firm, skin-colored papules. They are more commonly found at sites of amputation and as rudimentary supernumerary digits. However, when there are multiple neuromas on the eyelids, lips, distal tongue, and/or oral mucosa, the patient should be investigated for other signs of theMENsyndrome, type 2b. Associated findings include marfanoid habitus, protuberant lips, intestinal ganglioneuromas, and medullary thyroid carcinoma (>75% of patients;Chap. 339). Adnexal tumors are derived from pluripotential cells of the epidermis that can differentiate toward hair, sebaceous, apocrine, or eccrine glands or remain undifferentiated. Basal cell epitheliomas (BCEs) are examples of adnexal tumors that have little or no evidence of differentiation. Clinically, they are translucent papules with rolled borders, telangiectasias, and central erosion. BCEs commonly arise in sun-damaged skin of the head and neck. When a patient has multiple BCEs, especially prior to age 30, the possibility of the basal cell nevus syndrome should be raised. It is inherited as an autosomal dominant trait and is associated with jaw cysts, palmar and plantar pits, frontal bossing, medulloblastomas and calcification of the falx cerebri and diaphragma sellae. Tricholemmomas are also skin-colored adnexal tumors but differentiate toward hair follicles and can have a wartlike appearance. The presence of multiple tricholemmomas on the face and cobblestoning of the oral mucosa points to the diagnosis of Cowden's disease (multiple hamartoma syndrome) due to mutations in the PTEN gene. Internal organ involvement (in decreasing order of frequency) includes fibrocystic disease and carcinoma of the breast, adenomas and carcinomas of the thyroid, and gastrointestinal polyposis. Keratoses of the palms, soles, and dorsa of the hands are also seen. Pink Lesions The cutaneous lesions associated with primary systemic amyloidosis are pink in color and translucent. Common locations are the face, especially the periorbital and perioral regions, and flexural areas. On biopsy, homogeneous deposits of amyloid are seen in the dermis and in the walls of blood vessels; the latter lead to an increase in vessel wall fragility. As a result, petechiae and purpura develop in clinically normal skin as well as in lesional skin following minor trauma, hence the term "pinch purpura." Amyloid deposits are also seen in the striated muscle of the tongue and result in macroglossia. Even though specific mucocutaneous lesions are rarely seen in secondary amyloidosis and are present in only about 30% of the patients with primary amyloidosis, a rapid
diagnosis of systemic amyloidosis can be made by an examination of abdominal subcutaneous fat. By special staining, deposits are seen around blood vessels or individual fat cells in 40 to 50% of patients. There are also three forms of amyloidosis that are limited to the skin and that should not be construed as cutaneous lesions of systemic amyloidosis. They are macular amyloidosis (upper back), lichenoid amyloidosis (usually lower extremities), and nodular amyloidosis. In macular and lichenoid amyloidosis, the deposits are composed of altered epidermal keratin. Recently, macular and lichenoid amyloidosis have been associated withMENsyndrome, type 2a. Patients with multicentric reticulohistiocytosis also have pink-colored papules and nodules on the face and mucous membranes as well as on the extensor surface of the hands and forearms. They have a polyarthritis that can mimic rheumatoid arthritis clinically. On histologic examination, the papules have characteristic giant cells that are not seen in biopsies of rheumatoid nodules. Pink to skin-colored papules that are firm, 2 to 5 mm in diameter, and often in a linear arrangement are seen in patients with papular mucinosis. This disease is also referred to as lichen myxedematosus or scleromyxedema. The latter name comes from the brawny induration of the face and extremities that may accompany the papular eruption. Biopsy specimens of the papules show localized mucin deposition, and serum protein electrophoresis demonstrates a monoclonal spike of IgG, usually with a l light chain. Yellow Lesions Several systemic disorders are characterized by yellow-colored cutaneous papules or plaques -- hyperlipidemia (xanthomas), gout (tophi), diabetes (necrobiosis lipoidica), pseudoxanthoma elasticum, and Torre syndrome (sebaceous tumors). Eruptive xanthomas are the most common form of xanthomas, and are associated with hypertriglyceridemia (types I, III, IV, and V). Crops of yellow papules with erythematous halos occur primarily on the extensor surfaces of the extremities and the buttocks, and they spontaneously involute with a fall in serum triglycerides. Increasedb-lipoproteins (primarily types II and III) result in one or more of the following types of xanthoma: xanthelasma, tendon xanthomas, and plane xanthomas. Xanthelasma are found on the eyelids, whereas tendon xanthomas are frequently associated with the Achilles and extensor finger tendons; plane xanthomas are flat and favor the palmar creases, face, upper trunk, and scars. Tuberous xanthomas are frequently associated with hypertriglyceridemia, but they are also seen in patients with hypercholesterolemia (type II) and are found most frequently over the large joints or hand. Biopsy specimens of xanthomas show collections of lipid-containing macrophages (foam cells). Patients with several disorders, including biliary cirrhosis, can have a secondary form of hyperlipidemia with associated tuberous and planar xanthomas. However, patients with myeloma have normolipemic flat xanthomas. This latter form of xanthoma may be³12 cm in diameter and is most frequently seen on the upper trunk or side of the neck. It is also important to note that the most common setting for eruptive xanthomas is uncontrolled diabetes mellitus. The least specific sign for hyperlipidemia is xanthelasma, because at least 50% of the patients with this finding have normal lipid profiles. In tophaceous gout there are deposits of monosodium urate in the skin around the joints, particularly those of the hands and feet. Additional sites of tophi formation include
the helix of the ear and the olecranon and prepatellar bursae. The lesions are firm, yellow in color, and occasionally discharge a chalky material. Their size varies from 1 mm to 7 cm, and the diagnosis can be established by polarization of the aspirated contents of a lesion. Lesions of necrobiosis lipoidica are found primarily on the shins (90%), and patients can have diabetes mellitus or develop it subsequently. Characteristic findings include a central yellow color, atrophy (transparency), telangiectasias, and an erythematous border. Ulcerations can also develop within the plaques. Biopsy specimens show necrobiosis of collagen, granulomatous inflammation, and obliterative endarteritis. In pseudoxanthoma elasticum (PXE) there is an abnormal deposition of calcium on the elastic fibers of the skin, eye, and blood vessels. In the skin, the flexural areas such as the neck, axillae, antecubital fossae, and inguinal area are the primary sites of involvement. Yellow papules coalesce to form reticulated plaques that have an appearance similar to that of plucked chicken skin. In severely affected skin, hanging, redundant folds develop. Some patients have a more subtle macular form of the disease, and careful inspection is required. Biopsy specimens of involved skin show swollen and irregularly clumped elastic fibers with deposits of calcium. In the eye, the calcium deposits in Bruch's membrane lead to angioid streaks and choroiditis; in the arteries of the heart, kidney, gastrointestinal tract, and extremities, the deposits lead to angina, hypertension, gastrointestinal bleeding, and claudication, respectively. Long-term administration of D-penicillamine can lead to PXE-like skin changes as well as elastic fiber alterations in internal organs. Adnexal tumors that have differentiated toward sebaceous glands include sebaceous adenoma, sebaceous epithelioma, sebaceous carcinoma, and sebaceous hyperplasia. Except for sebaceous hyperplasia, which is commonly seen on the face, these tumors are fairly rare. Patients with Torre syndrome have sebaceous adenomas, and in the majority of cases there are multiple such tumors. These patients can also have sebaceous carcinomas and sebaceous hyperplasia as well as keratoacanthomas. The internal manifestations of Torre syndrome include multiple carcinomas of the gastrointestinal tract (primarily colon) as well as cancers of the larynx, genitourinary tract, and endometrium. Some patients also have a strong family history of cancer. Red Lesions Cutaneous lesions that are red in color have a wide variety of etiologies; in an attempt to simplify their identification, they will be subdivided into papules, papules/plaques, and subcutaneous nodules. Common red papules include arthropod bites and cherry hemangiomas; the latter are small, bright-red, dome-shaped papules that represent benign proliferation of capillaries. In patients with AIDS, the development of multiple red hemangioma-like lesions points to bacillary angiomatosis, and biopsy specimens show clusters of bacilli that stain positive with the Warthin-Starry stain; the pathogens have been identified as Bartonella henselae and B. quintana. Disseminated visceral disease is seen primarily in immunocompromised hosts but can occur in immunocompetent individuals. Multiple angiokeratomas are seen in Fabry's disease, an X-linked recessive lysosomal storage disease that is due to a deficiency of a-galactosidase A. The lesions are red to red-blue in color and can be quite small in size (1 to 3 mm), with the most common location being the lower trunk. Associated findings include chronic renal failure,
peripheral neuropathy, and corneal opacities (cornea verticillata). Electron photomicrographs of angiokeratomas and clinically normal skin demonstrate lamellar lipid deposits in fibroblasts, pericytes, and endothelial cells that are diagnostic of this disease. Widespread acute eruptions of erythematous papules are discussed in the section on exanthems. There are several infectious diseases that present as erythematous papules or nodules in a sporotrichoid pattern, that is, in a linear arrangement along the lymphatic channels. The two most common etiologies are Sporothrix schenckii (sporotrichosis) and M. marinum (atypical mycobacteria). The organisms are introduced as a result of trauma, and a primary inoculation site is often seen in addition to the lymphatic nodules. Additional causes include Nocardia, Leishmania, and other dimorphic fungi; culture of lesional tissue will aid in the diagnosis. The diseases that are characterized by erythematous plaques with scale are reviewed in the papulosquamous section, and the various forms of dermatitis are discussed in the section on erythroderma. Additional disorders in the differential diagnosis of red papules/plaques include erysipelas, polymorphous light eruption (PMLE), lymphocytoma cutis, cutaneous lupus, lymphoma cutis, and leukemia cutis. The first three diseases represent primary cutaneous disorders. PMLE is characterized by erythematous papules and plaques in a primarily sun-exposed distribution -- dorsum of the hand, extensor forearm, and face. Lesions follow exposure to UV-B and/or UV-A, and in northern latitudes PMLE is most severe in the late spring and early summer. A process referred to as "hardening" occurs with continued UV exposure, and the eruption fades, but in temperate climates it will recur in the spring. PMLE must be differentiated from cutaneous lupus, and this is accomplished by histologic examination and direct immunofluorescence of the lesions. Lymphocytoma cutis (pseudolymphoma) is a benign polyclonal proliferation of lymphocytes in the skin that presents as infiltrated pink-red to red-purple papules and plaques; it must be distinguished from lymphoma cutis. Several types of red plaques are seen in patients with systemic lupus, including (1) erythematous urticarial plaques across the cheeks and nose in the classic butterfly rash; (2) erythematous discoid lesions with fine or "carpet-tack" scale, telangiectasias, central hypopigmentation, peripheral hyperpigmentation, follicular plugging, and atrophy located on the face, scalp, external ears, arms, and upper trunk; and (3) psoriasiform or annular lesions of subacute lupus with hypopigmented centers located on the face, extensor arms, and upper trunk. Additional cutaneous findings include (1) a violaceous flush on the face and V of the neck; (2) urticarial vasculitis (see "Urticaria"); (3) lupus panniculitis (see below); (4) diffuse alopecia; (5) alopecia secondary to discoid lesions; (6) periungual telangiectasias and erythema; (7) erythema multiforme-like lesions that may become bullous; and (8) distal ulcerations secondary to Raynaud's phenomenon, vasculitis, or livedoid vasculitis. Patients with only discoid lesions usually have the form of lupus that is limited to the skin. However, 2 to 10% of these patients eventually develop systemic lupus. Direct immunofluorescence of involved skin shows deposits of IgG or IgM and C3 in a granular distribution along the dermal-epidermal junction. In lymphoma cutis there is a proliferation of malignant lymphocytes or histiocytes in the skin, and the clinical appearance resembles that of lymphocytoma cutis -- infiltrated pink-red to red-purple papules and plaques. Lymphoma cutis can occur anywhere on
the surface of the skin, whereas the sites of predilection for lymphocytomas include the malar ridge, tip of the nose, and earlobes. Patients with non-Hodgkin's lymphomas have specific cutaneous lesions more often than those with Hodgkin's disease, and occasionally, the skin nodules precede the development of extracutaneous non-Hodgkin's lymphoma or represent the only site of involvement. Arcuate lesions are sometimes seen in lymphoma and lymphocytoma cutis as well as in CTCL. Leukemia cutis has the same appearance as lymphoma cutis, and specific lesions are seen more commonly in monocytic leukemias than in lymphocytic or granulocytic leukemias. Cutaneous chloromas (granulocytic sarcomas) may precede the appearance of circulating blasts in acute nonlymphocytic leukemia and, as such, represent a form of aleukemic leukemia cutis. Common causes of erythematous subcutaneous nodules include inflamed epidermoid inclusion cysts, acne cysts, and furuncles. Panniculitis, an inflammation of the fat, also presents as subcutaneous nodules and is frequently a sign of systemic disease. There are several forms of panniculitis, including erythema nodosum, erythema induratum, lupus profundus, lipomembranous lipodermatosclerosis,a 1-antitrypsin deficiency, facticial, and fat necrosis secondary to pancreatic disease. Except for erythema nodosum, these lesions may break down and ulcerate or heal with a scar. The shin is the most common location for the nodules of erythema nodosum, whereas the calf is the most common location for lesions of erythema induratum. In erythema nodosum the nodules are initially red but then develop a blue color as they resolve. Patients with erythema nodosum but no underlying systemic illness can still have fever, malaise, leukocytosis, arthralgias, and/or arthritis. However, the possibility of an underlying illness should be excluded, and the most common associations are streptococcal infections, upper respiratory infections, sarcoidosis, and inflammatory bowel disease. The less common associations include tuberculosis, histoplasmosis, coccidioidomycosis, psittacosis, drugs (oral contraceptives, sulfonamides, aspartame, bromides, iodides), cat-scratch fever, and infections with Yersinia, Salmonella, and Chlamydia. In some patients, erythema induratum/nodular vasculitis is an idiopathic disease; however, in approximately 25 to 70% of patients, polymerase chain reaction (PCR) analysis will demonstrate M. tuberculosis complex DNA. The lesions of lupus profundus are found primarily on the upper arms and buttocks (sites of abundant fat) and are seen in both the cutaneous and systemic forms of lupus. The overlying skin may be normal, erythematous, or have the changes of discoid lupus. The subcutaneous fat necrosis that is associated with pancreatic disease is presumably secondary to circulating lipases and is seen in patients with pancreatic carcinoma as well as in patients with acute and chronic pancreatitis. In this disorder there may be an associated arthritis, fever, and inflammation of visceral fat. Histologic examination of deep incisional biopsy specimens will aid in the diagnosis of the particular type of panniculitis. Subcutaneous erythematous nodules are also seen in cutaneous polyarteritis nodosa (PAN) and as a manifestation of systemic vasculitis, e.g., systemic PAN, allergic granulomatosis, or Wegener's granulomatosis (Chap. 317). Cutaneous PAN presents with painful subcutaneous nodules and ulcers within a red-purple, netlike pattern of livedo reticularis. The latter is due to slowed blood flow through the superficial horizontal venous plexus. The majority of lesions are found on the lower extremity, and while
arthralgias and myalgias may accompany cutaneous PAN, there is no evidence of systemic involvement. In both the cutaneous and systemic forms of vasculitis, skin biopsy specimens of the associated nodules will show the changes characteristic of a vasculitis; the size of the vessel involved will depend on the particular disease. Red-Brown Lesions The cutaneous lesions in sarcoidosis (Chap. 318) are classically red to red-brown in color, and with diascopy (pressure with a glass slide) a yellow-brown residual color is observed that is secondary to the granulomatous infiltrate. The waxy papules and plaques may be found anywhere on the skin, but the face is the most common location. Usually there are no surface changes, but occasionally the lesions will have scale. Biopsy specimens of the papules show "naked" granulomas in the dermis, i.e., granulomas surrounded by a minimal number of lymphocytes. Other cutaneous findings in sarcoidosis include annular lesions with an atrophic or scaly center, papules within scars, hypopigmented macules and papules, alopecia, acquired ichthyosis, erythema nodosum, and lupus pernio (see below). Additional physical findings are peripheral lymphadenopathy and parotid and lacrimal gland enlargement. When there is cutaneous involvement of the hands, radiographs will often show lytic lesions in the underlying bone. The differential diagnosis of sarcoidosis includes foreign-body granulomas produced by chemicals such as beryllium and zirconium, late secondary syphilis, and lupus vulgaris. Lupus vulgaris is a form of cutaneous tuberculosis that is seen in previously infected and sensitized individuals. There is often underlying active tuberculosis elsewhere, usually in the lungs or lymph nodes. At least 90% of the lesions occur in the head and neck area and are red-brown plaques with a yellow-brown color on diascopy. Secondary scarring and squamous cell carcinomas can develop within the plaques. Cultures orPCRanalysis of the lesions should be done because it is rare for the acid-fast stain to show bacilli within the dermal granulomas. Sweet's syndrome is characterized by red to red-brown plaques and nodules that are frequently painful and occur primarily on the head, neck, and upper extremities. The patients also have fever, neutrophilia, and a dense dermal infiltrate of neutrophils in the lesions. In approximately 10% of the patients there is an associated malignancy, most commonly acute nonlymphocytic leukemia. Sweet's syndrome has also been reported with lymphoma, chronic leukemia, myeloma, myelodysplastic syndromes, and solid tumors (primarily of the genitourinary tract). The differential diagnosis includes neutrophilic eccrine hidradenitis and atypical forms of pyoderma gangrenosum. Extracutaneous sites of involvement include joints, muscles, eye, kidney (proteinuria, occasionally glomerulonephritis), and lung (neutrophilic infiltrates). The idiopathic form of Sweet's syndrome is seen more often in women, following a respiratory tract infection. A generalized distribution of red-brown macules and papules is seen in the form of mastocytosis known as urticaria pigmentosa (Chap. 310). Each lesion represents a collection of mast cells in the dermis, with hyperpigmentation of the overlying epidermis. Stimuli such as rubbing cause these mast cells to degranulate, and this leads to the formation of localized urticaria (Darier's sign). Additional symptoms can result from mast cell degranulation and include headache, flushing, diarrhea, and pruritus. Mast cells also infiltrate various organs such as the liver, spleen, and gastrointestinal tract in up to
30 to 50% of patients with urticaria pigmentosa, and accumulations of mast cells in the bones may produce either osteosclerotic or osteolytic shadows on radiographs. In the majority of these patients, however, the internal involvement remains fairly static. A subtype of chronic leukocytoclastic vasculitis, erythema elevatum diutinum (EED), also presents with papules that are red-brown in color. The papules coalesce into plaques on the extensor surfaces of knees, elbows, and the small joints of the hand. Flares of EED have been associated with streptococcal infections. Blue Lesions Lesions that are blue in color are the result of either vascular ectasias and tumors or melanin pigment in the dermis. Venous lakes (ectasias) are compressible dark-blue lesions that are found commonly in the head and neck region. Venous malformations are also compressible blue papules and nodules that can occur anywhere on the body, including the oral mucosa. When they are multiple rather than single congenital lesions, the patient may have the blue rubber bleb syndrome or Mafucci's syndrome. Patients with the blue rubber bleb syndrome also have vascular anomalies of the gastrointestinal tract that may bleed, whereas patients with Mafucci's syndrome have associated dyschondroplasia and osteochondromas. Blue nevi (moles) are seen when there are collections of pigment-producing nevus cells in the dermis. These benign papular lesions are dome-shaped and occur most commonly on the dorsum of the hand or foot. Violaceous Lesions Violaceous papules and plaques are seen in lupus pernio, lymphoma cutis, and cutaneous lupus. Lupus pernio is a particular type of sarcoidosis that involves the tip of the nose and the earlobes, with lesions that are violaceous in color rather than red-brown. This form of sarcoidosis is associated with involvement of the upper respiratory tract. The plaques of lymphoma cutis and cutaneous lupus may be red or violaceous in color and were discussed above. Purple Lesions Purple-colored papules and plaques are seen in vascular tumors, such as Kaposi's sarcoma (Chap. 309) and angiosarcoma, and when there is extravasation of red blood cells into the skin in association with inflammation, as in palpable purpura (see "Purpura"). Patients with congenital or acquiredAVfistulas and venous hypertension can develop purple papules on the lower extremities that can resemble Kaposi's sarcoma clinically and histologically; this condition is referred to as pseudo-Kaposi sarcoma (acral angiodermatitis). Angiosarcoma is found most commonly on the scalp and face of elderly patients or within areas of chronic lymphedema and presents as purple papules and plaques. In the head and neck region the tumor often extends beyond the clinically defined borders and may be accompanied by facial edema. Brown and Black Lesions Brown- and black-colored papules are reviewed in "Hyperpigmentation." Cutaneous Metastases These are discussed last because they can have a wide range of colors. Most commonly they present as either firm, skin-colored subcutaneous nodules or firm, red to red-brown papulonodules. The lesions of lymphoma cutis range from pink-red to plum in color, whereas metastatic melanoma can be pink, blue, or black in color. Cutaneous metastases develop from hematogenous or lymphatic spread and are most often due to the following primary carcinomas: in men, lung, colon, melanoma,
and oral cavity; and in women, breast, colon, and lung. These metastatic lesions may be the initial presentation of the carcinoma, especially when the primary site is the lung, kidney, or ovary. PURPURA (Table 57-16) Purpura are seen when there is an extravasation of red blood cells into the dermis, and as a result, the lesions do not blanch with pressure. This is in contrast to those erythematous or violet-colored lesions that are due to localized vasodilatation -- they do blanch with pressure. Purpura (³3 mm) and petechiae (£2 mm) are divided into two major groups, palpable and nonpalpable. The most frequent causes of nonpalpable petechiae and purpura are primary cutaneous disorders such as trauma, solar purpura, and capillaritis. Less common causes are steroid purpura and livedoid vasculitis (see "Ulcers"). Solar purpura are seen primarily on the extensor forearms, while glucocorticoid purpura secondary to potent topical steroids or endogenous or exogenous Cushing's syndrome can be more widespread. In both cases there is alteration of the supporting connective tissue that surrounds the dermal blood vessels. In contrast, the petechiae that result from capillaritis are found primarily on the lower extremities. In capillaritis there is an extravasation of erythrocytes as a result of perivascular lymphocytic inflammation. The petechiae are bright red, 1 to 2 mm in size, and scattered within annular or coin-shaped yellow-brown macules. The yellow-brown color is caused by hemosiderin deposits within the dermis. Systemic causes of nonpalpable purpura fall into several categories, and those secondary to clotting disturbances and vascular fragility will be discussed first. The former group includes thrombocytopenia (Chap. 116), abnormal platelet function as is seen in uremia, and clotting factor defects. The initial site of presentation for thrombocytopenia-induced petechiae is the distal lower extremity. Capillary fragility leads to nonpalpable purpura in patients with systemic amyloidosis (see "Papulonodular Skin Lesions"), disorders of collagen production such as Ehlers-Danlos syndrome, and scurvy. In scurvy there are flattened corkscrew hairs with surrounding hemorrhage on the lower extremities, in addition to gingivitis. Vitamin C is a cofactor for lysyl hydroxylase, an enzyme involved in the posttranslational modification of procollagen that is necessary for cross-link formation. In contrast to the previous group of disorders, the purpura seen in the following group of diseases are associated with thrombi formation within vessels. It is important to note that these thrombi are demonstrable in skin biopsy specimens. This group of disorders includes disseminated intravascular coagulation (DIC), monoclonal cryoglobulinemia, thrombotic thrombocytopenic purpura, and reactions to warfarin. DIC is triggered by several types of infection (gram-negative, gram-positive, viral, and rickettsial) as well as by tissue injury and neoplasms. Widespread purpura and hemorrhagic infarcts of the distal extremities are seen. Similar lesions are found in purpura fulminans, which is a form of DIC associated with fever and hypotension that occurs more commonly in children following an infectious illness such as varicella, scarlet fever, or an upper respiratory tract infection. In both disorders, hemorrhagic bullae can develop in involved skin. Monoclonal cryoglobulinemia is associated with multiple myeloma, Waldenstrom's
macroglobulinemia, lymphocytic leukemia, and lymphoma. Purpura, primarily of the lower extremities, and hemorrhagic infarcts of the fingers and toes are seen in these patients. Exacerbations of disease activity can follow cold exposure or an increase in serum viscosity. Biopsy specimens show precipitates of the cryoglobulin within dermal vessels. Similar deposits have been found in the lung, brain, and renal glomeruli. Patients with thrombotic thrombocytopenic purpura can also have hemorrhagic infarcts as a result of intravascular thromboses. Additional signs include thrombocytopenic purpura, fever, and microangiopathic hemolytic anemia (Chap. 108). Administration of warfarin can result in painful areas of erythema that become purpuric and then necrotic with an adherent black eschar. This reaction is seen more often in women and in areas with abundant subcutaneous fat -- breasts, abdomen, buttocks, thighs, and calves. The erythema and purpura develop between the third and tenth day of therapy, most likely as a result of a transient imbalance in the levels of anticoagulant and procoagulant vitamin K-dependent factors. Continued therapy does not exacerbate preexisting lesions, and patients with an inherited or acquired deficiency of protein C are at increased risk for this particular reaction as well as for purpura fulminans. Purpura secondary to cholesterol emboli are usually seen on the lower extremities of patients with atherosclerotic vascular disease. They often follow anticoagulant therapy or an invasive vascular procedure such as an arteriogram but also occur spontaneously from disintegration of atheromatous plaques. Associated findings include livedo reticularis, gangrene, cyanosis, subcutaneous nodules, and ischemic ulcerations. Multiple step sections of the biopsy specimen may be necessary to demonstrate the cholesterol clefts with the vessels. Petechiae are also an important sign of fat embolism and occur primarily on the upper body 2 to 3 days after a major injury. By using special fixatives, the emboli can be demonstrated in biopsy specimens of the petechiae. Emboli of tumor or thrombus are seen in patients with atrial myxomas and marantic endocarditis. In the Gardner-Diamond syndrome (autoerythrocyte sensitivity), female patients develop large ecchymoses within areas of painful, warm erythema. An episode of significant trauma frequently precedes the onset of this syndrome. Intradermal injections of autologous erythrocytes or phosphatidyl serine derived from the red cell membrane can reproduce the lesions in some patients; however, there are instances where a reaction is seen at an injection site of the forearm but not in the midback region. The latter has led some observers to view Gardner-Diamond syndrome as a cutaneous manifestation of severe emotional stress. Waldenstrom's hypergammaglobulinemic purpura is a chronic disorder characterized by petechiae on the lower extremities. There are circulating complexes of IgG-anti-IgG molecules, and exacerbations are associated with prolonged standing or walking. Palpable purpura are further subdivided into vasculitic and embolic. In the group of vasculitic disorders, leukocytoclastic vasculitis (LCV), also known as allergic vasculitis, is the one most commonly associated with palpable purpura (Chap. 317). Henoch-Schonlein purpura is a subtype of acute LCV that is seen primarily in children and adolescents following an upper respiratory infection. The majority of lesions are found on the lower extremities and buttocks. Systemic manifestations include fever, arthralgias (primarily of the knees and ankles), abdominal pain, gastrointestinal
bleeding, and nephritis. Direct immunofluorescence examination shows deposits of IgA within dermal blood vessel walls. In polyarteritis nodosa, specific cutaneous lesions result from a vasculitis of arterial vessels rather than postcapillary venules as in LCV. The arteritis leads to ischemia of the skin, and this explains the irregular outline of the purpura (see below). Several types of infectious emboli can give rise to palpable purpura. These embolic lesions are usually irregular in outline as opposed to the lesions of leukocytoclastic vasculitis, which are circular in outline. The irregular outline is indicative of a cutaneous infarct, and the size corresponds to the area of skin that received its blood supply from that particular arteriole or artery. The palpable purpura inLCV are circular because the erythrocytes simply diffuse out evenly from the postcapillary venules as a result of inflammation. Infectious emboli are most commonly due to gram-negative cocci (meningococcus, gonococcus), gram-negative rods (Enterobacteriaceae), and gram-positive cocci (staphylococcus). Additional causes include Rickettsia and, in immunocompromised patients, Candida and opportunistic fungi. The embolic lesions in acute meningococcemia are found primarily on the trunk, lower extremities, and sites of pressure, and a gunmetal-gray color often develops within them. Their size varies from 1 mm to several centimeters, and the organisms can be cultured from the lesions. Associated findings include a preceding upper respiratory tract infection, fever, meningitis,DIC, and, in some patients, a deficiency of the terminal components of complement. In disseminated gonococcal infection (arthritis-dermatitis syndrome), a small number of papules and vesicopustules with central purpura or hemorrhagic necrosis are found over the joints of the distal extremities. Additional symptoms include arthralgias, tenosynovitis, and fever. To establish the diagnosis, a Gram stain of these lesions should be performed. Rocky mountain spotted fever is a tick-borne disease that is caused by R. rickettsii. A several-day history of fever, chills, severe headache, and photophobia precedes the onset of the cutaneous eruption. The initial lesions are erythematous macules and papules on the wrists, ankles, palms, and soles. With time, the lesions spread centripetally and become purpuric. Lesions of ecthyma gangrenosum begin as edematous, erythematous papules or plaques and then develop central purpura and necrosis. Bullae formation also occurs in these lesions, and they are frequently found in the girdle region. The organism that is classically associated with ecthyma gangrenosum is Pseudomonas aeruginosa, but other gram-negative rods such as Klebsiella, Escherichia coli, and Serratia can produce similar lesions. In immunocompromised hosts, the list of potential pathogens is expanded to include Candida and opportunistic fungi. ULCERS The approach to the patient with a cutaneous ulcer, is outlined inTable 57-17.*Peripheral vascular diseases of the extremities are reviewed in Chap. 248, as is Raynaud's phenomenon. Livedoid vasculitis (atrophie blanche) represents a combination of a vasculopathy with intravascular thrombosis. Purpuric lesions and livedo reticularis are found in association with painful ulcerations of the lower extremities. These ulcers are often slow to heal, but
when they do, irregularly shaped white scars are formed. The majority of cases are secondary to venous hypertension, but possible underlying illnesses include cryofibrinogenemia and disorders of hypercoagulability, e.g., the antiphospholipid syndrome (Chap. 117). In pyoderma gangrenosum, the border of the ulcers has a characteristic appearance of an undermined necrotic bluish edge and a peripheral erythematous halo. The ulcers often begin as pustules that then expand rather rapidly to a size as large as 20 cm. Although these lesions are most commonly found on the lower extremities, they can arise anywhere on the surface of the body, including sites of trauma (pathergy). An estimated 30 to 50% of cases are idiopathic, and the most common associated disorders are ulcerative colitis and Crohn's disease. Less commonly, it is associated with chronic active hepatitis, seropositive rheumatoid arthritis, acute and chronic granulocytic leukemia, polycythemia vera, and myeloma. Additional findings in these patients, even those with idiopathic disease, are cutaneous anergy and a benign monoclonal gammopathy. Because the histology of pyoderma gangrenosum is nonspecific, the diagnosis is made clinically by excluding less common causes of similar-appearing ulcers such as necrotizing vasculitis, Meleney's ulcer (synergistic infection at a site of trauma or surgery), dimorphic fungi, cutaneous amebiasis, spider bites, and facticial. In the myeloproliferative disorders, the ulcers may be more superficial with a pustulobullous border, and these lesions provide a connection between classic pyoderma gangrenosum and acute febrile neutrophilic dermatosis (Sweet's syndrome). FEVER AND RASH The major considerations in a patient with a fever and a rash are inflammatory diseases versus infectious diseases. In the hospital setting, the most common scenario is a patient who has a drug rash plus a fever secondary to an underlying infection. However, it should be emphasized that a drug reaction can lead to both a cutaneous eruption and a fever ("drug fever"). Additional inflammatory diseases that are often associated with a fever include pustular psoriasis, erythroderma, and Sweet's syndrome. Lyme disease, secondary syphilis, and viral and bacterial exanthems (see "Exanthems") are examples of infectious diseases that produce a rash and a fever. Lastly, it is important to determine whether or not the cutaneous lesions represent septic emboli (see "Purpura"). Such lesions usually have evidence of ischemia in the form of purpura, necrosis, or impending necrosis (gunmetal-gray color). In the patient with thrombocytopenia, however, purpura can be seen in inflammatory reactions such as morbilliform drug eruptions and infectious lesions. (Bibliography omitted in Palm version) Back to Table of Contents
58. IMMUNOLOGICALLY MEDIATED SKIN DISEASES - Kim B. Yancey, Thomas J. Lawley A number of immunologically mediated skin diseases and cutaneous manifestations of immunologically mediated systemic disorders are now recognized as distinct entities with relatively consistent clinical, histologic, and immunopathologic findings. Many of these disorders are due to autoimmune mechanisms. Clinically, they are characterized by morbidity (pain, pruritus, disfigurement) and in some instances by mortality (largely due to loss of epidermal barrier function and/or secondary infection). The major features of the more common immunologically mediated skin diseases are summarized in this chapter (Table 58-1). PEMPHIGUS VULGARIS Pemphigus vulgaris (PV) is a blistering skin disease seen predominantly in elderly patients. Patients with PV have an increased incidence of the HLA-DR4 and -DRw6 serologically defined haplotypes. This disorder is characterized by the loss of cohesion between epidermal cells (a process termed acantholysis) with the resultant formation of intraepidermal blisters. Clinical lesions of PV typically consist of flaccid blisters on either normal-appearing or erythematous skin. These blisters rupture easily, leaving denuded areas that may crust and enlarge peripherally (Plate IIE-69). Substantial portions of the body surface may be denuded in severe cases. Manual pressure to the skin of these patients may elicit the separation of the epidermis (Nikolsky's sign). This finding, while characteristic of PV, is not specific to this disorder and is also seen in toxic epidermal necrolysis, Stevens-Johnson syndrome, and a few other skin diseases. Lesions in PV typically present on the oral mucosa, scalp, face, neck, axilla, and trunk. In half or more of patients, lesions begin in the mouth; approximately 90% of patients have oromucosal involvement at some time during the course of their disease. Involvement of other mucosal surfaces (e.g., pharyngeal, laryngeal, esophageal, conjunctival, vulval, or rectal) can occur in severe disease. Pruritus may be a feature of early pemphigus lesions; extensive denudation may be associated with severe pain. Lesions usually heal without scarring, except at sites complicated by secondary infection or mechanically induced dermal wounds. Nonetheless, postinflammatory hyperpigmentation is usually present at sites of healed lesions for some time. Biopsies of early lesions demonstrate intraepidermal vesicle formation secondary to loss of cohesion between epidermal cells (i.e., acantholytic blisters). Blister cavities contain acantholytic epidermal cells, which appear as round homogeneous cells containing hyperchromatic nuclei. Basal keratinocytes remain attached to the epidermal basement membrane, hence blister formation is within the suprabasal portion of the epidermis. Lesional skin may contain focal collections of intraepidermal eosinophils within blister cavities; dermal alterations are slight, often limited to an eosinophil-predominant leukocytic infiltrate. Direct immunofluorescence microscopy of lesional or intact patient skin shows deposits of IgG on the surface of keratinocytes; in contrast, deposits of complement components are typically found in lesional but not uninvolved skin. Deposits of IgG on keratinocytes are derived from circulating autoantibodies directed against cell-surface antigens. Circulating autoantibodies can be demonstrated in 80 to 90% ofPVpatients by indirect immunofluorescence microscopy; monkey esophagus is the optimal substrate for demonstration of these autoantibodies. Patients with PV have
IgG autoantibodies directed against desmogleins (Dsgs), transmembrane desmosomal glycoproteins that belong to the cadherin supergene family of calcium-dependent adhesion molecules. While Dsg3 is specifically recognized by PV autoantibodies, approximately 50% of PV sera also contain IgG against Dsg1. Most patients with early PV and only mucosal involvement have only anti-Dsg3 autoantibodies, whereas most patients with advanced disease (i.e., involvement of skin and mucosa) have both anti-Dsg3 and anti-Dsg1 autoantibodies. Recent studies have shown that the anti-Dsg autoantibody profile in these patients' sera as well as the tissue distribution of Dsg3 and Dsg1 determine the site of blister formation in patients with pemphigus. Experimental studies have also shown that these autoantibodies are pathogenic (i.e., responsible for blister formation) and that their titer correlates with disease activity. PVcan be life-threatening. Prior to the availability of glucocorticoids, the mortality ranged from 60 to 90%; the current mortality is approximately 5%. Common causes of morbidity and mortality are infection and complications of treatment with glucocorticoids. Bad prognostic factors include advanced age, widespread involvement, and the requirement for high doses of glucocorticoids (with or without other immunosuppressive agents) for control of disease. The course of PV in individual patients is variable and difficult to predict. Some patients achieve remission (40% of patients in some series), but others may require long-term treatment or succumb to complications of their disease or its treatment. The mainstay of treatment is systemic glucocorticoids. Patients with moderate to severe disease are usually started on prednisone, 60 to 80 mg/d. If new lesions continue to appear after 1 to 2 weeks of treatment, the dose should be increased. Many regimens combine an immunosuppressive agent with systemic glucocorticoids for control of PV. The two most frequently used are either azathioprine (1 mg/kg per day) or cyclophosphamide (1 mg/kg per day). It is important to bring severe or progressive disease under control quickly to lessen the severity and/or duration of this disorder. PEMPHIGUS FOLIACEUS Pemphigus foliaceus (PF) is distinguished fromPV by several features. In PF, acantholytic blisters are located high within the epidermis, usually just beneath the stratum corneum. Hence PF is a more superficial blistering disease than PV. The distribution of lesions in the two disorders is much the same, except that in PF mucous membrane lesions are very rare. Patients with PF rarely demonstrate intact blisters but rather exhibit shallow erosions associated with erythema, scale, and crust formation. Mild cases of PF resemble severe seborrheic dermatitis; severe PF may cause extensive exfoliation. Sun exposure (ultraviolet irradiation) may be an aggravating factor. A blistering skin disease endemic to south central Brazil known as fogo selvagem, or Brazilian pemphigus, is clinically, histologically, and immunopathologically indistinguishable from PF. Patients withPF have immunopathologic features in common withPV. Specifically, direct immunofluorescence microscopy of perilesional skin demonstrates IgG on the surface of keratinocytes. As in PV, patients with PF frequently have circulating IgG autoantibodies against keratinocyte cell surface antigens. Guinea pig esophagus is the optimal substrate for indirect immunofluorescence microscopy studies of sera from patients with PF. In PF, autoantibodies are directed against Dsg1, a 160-kDa desmosomal cadherin.
As noted for PV, the autoantibody profile in patients with PF (i.e., anti-Dsg1) and the normal tissue distribution of this autoantigen (i.e., low expression in oral mucosa) is thought to account for the distribution of lesions in this disease. Although pemphigus has been associated with several autoimmune diseases, its association with thymoma and/or myasthenia gravis is particularly notable. To date, more than 30 cases of thymoma and/or myasthenia gravis have been reported in association with pemphigus, usually withPF. Patients may also develop pemphigus as a consequence of drug exposure. The most frequently implicated agent is penicillamine; other offenders include captopril, rifampin, piroxicam, penicillin, and phenobarbital. Drug-induced pemphigus usually resembles PF rather thanPV; autoantibodies in these patients have the same antigenic specificity as they do in other pemphigus patients. In most patients, lesions resolve following discontinuation of the drug; however, some patients require treatment with systemic glucocorticoids and/or immunosuppressive agents. PFis generally a far less severe disease thanPV and carries a better prognosis. Localized disease can be treated conservatively with topical or intralesional glucocorticoids; more active cases can usually be controlled with systemic glucocorticoids. PARANEOPLASTIC PEMPHIGUS Paraneoplastic pemphigus (PNP) is an autoimmune acantholytic mucocutaneous disease associated with an occult or confirmed neoplasm. Patients with PNP typically show painful mucosal erosive lesions in association with pruritic papulosquamous eruptions that often progress to blisters. Palm and sole involvement is common in these patients and raises the possibility that prior reports of neoplasia-associated erythema multiforme actually may have represented unrecognized cases of PNP. Biopsies of lesional skin from these patients show varying combinations of acantholysis, keratinocyte necrosis, and vacuolar-interface dermatitis. Direct immunofluorescence microscopy of patient skin shows deposits of IgG and complement on the surface of keratinocytes and (variably) similar immunoreactants in the epidermal basement membrane zone. Patients with PNP have IgG autoantibodies against cytoplasmic proteins that are members of the plakin family (e.g., desmoplakins I and II, bullous pemphigoid antigen 1, envoplakin, periplakin, and plectin) and cell-surface proteins that are members of the cadherin family (e.g., Dsg3 and Dsg1). Because immunoadsorption of anti-Dsg3 IgG is sufficient to eliminate the ability of PNP sera to induce blisters in an experimental passive transfer animal model, these particular autoantibodies are thought to play a key pathogenic role in blister formation in these patients. Although PNP is generally resistant to conventional therapies (i.e., those used to treat PV), patients may improve (or even remit) following resection of underlying neoplasms. The predominant neoplasms associated with this disorder are non-Hodgkin's lymphoma, chronic lymphocytic leukemia, Castleman's disease, thymoma, and spindle cell tumors. BULLOUS PEMPHIGOID Bullous pemphigoid (BP) is an autoimmune subepidermal blistering disease usually
seen in the elderly. Lesions typically consist of tense blisters on either normal-appearing or erythematous skin (Plate IIE-72). The lesions are usually distributed over the lower abdomen, groin, and flexor surface of the extremities; oral mucosal lesions are found in 10 to 40% of patients. Pruritus may be nonexistent or severe. As lesions evolve, tense blisters tend to rupture and be replaced by flaccid lesions or erosions with or without surmounting crust. Nontraumatized blisters heal without scarring. The major histocompatibility complex class II allele HLA-DQb1*0301 is prevalent in patients with BP. Despite isolated reports, several studies have shown that patients with BP do not have an increased incidence of malignancy in comparison with appropriately age- and sex-matched controls. While biopsies of early lesional skin demonstrate subepidermal blisters, the histologic features depend on the character of the particular lesion. Lesions on normal-appearing skin generally show a sparse perivascular leukocytic infiltrate with some eosinophils; conversely, biopsies of inflammatory lesions typically show an eosinophil-rich infiltrate within the papillary dermis at sites of vesicle formation and in perivascular areas. In addition to eosinophils, cell-rich lesions also contain mononuclear cells and neutrophils. It is not always possible to distinguishBP from other subepidermal blistering diseases by routine histologic techniques. Immunopathologic studies have broadened our understanding of this disease and aided its diagnosis. Direct immunofluorescence microscopy of normal-appearing perilesional skin shows linear deposits of IgG and/or C3 in the epidermal basement membrane. The sera of approximately 70% of these patients contain circulating IgG autoantibodies that bind the epidermal basement membrane of normal human skin in indirect immunofluorescence microscopy. An even higher percentage of patients shows reactivity to the epidermal side of 1 M NaCl split skin [an alternative immunofluorescence microscopy test substrate that is commonly used to distinguish circulating IgG anti-basement membrane autoantibodies in patients with BPfrom those in patients with similar, yet different, subepidermal blistering diseases (e.g., epidermolysis bullosa acquisita, see below)]. No correlation exists between the titer of these autoantibodies and disease activity. In BP, circulating autoantibodies recognize 230- and (in approximately 70% of BP patients) 180-kDa hemidesmosome-associated proteins in basal keratinocytes [i.e., bullous pemphigoid antigen (BPAG)1 and BPAG2, respectively]. Autoantibodies are thought to develop against these antigens (more specifically, initially against BPAG2), deposit in situ, and activate complement that subsequently produces dermal mast cell degranulation and granulocyte-rich infiltrates that cause tissue damage and blister formation. BPmay persist for months to years, with exacerbations or remissions. Although extensive involvement may result in widespread erosions and compromise cutaneous integrity, the mortality rate is low even in the absence of treatment. Nonetheless, deaths may occur in elderly and/or debilitated patients. The mainstay of treatment is systemic glucocorticoids. Patients with local or minimal disease can sometimes be controlled with topical glucocorticoids alone; patients with more extensive lesions generally respond to systemic glucocorticoids either alone or in combination with immunosuppressive agents. Patients will usually respond to prednisone, 40 to 60 mg/d. In some instances, azathioprine (1 mg/kg per day) or cyclophosphamide (1 mg/kg per day) are necessary adjuncts.
PEMPHIGOID GESTATIONIS Pemphigoid gestationis (PG), also known as herpes gestationis, is a rare, nonviral, subepidermal blistering disease of pregnancy and the puerperium. PG may begin during any trimester of pregnancy or present shortly after delivery. Lesions are usually distributed over the abdomen, trunk, and extremities; mucous membrane lesions are rare. Skin lesions in these patients may be quite polymorphic and consist of erythematous urticarial papules and plaques, vesiculopapules, and/or frank bullae. Lesions are almost always very pruritic. Severe exacerbations of PG frequently occur after delivery, typically within 24 to 48 h. PG tends to recur in subsequent pregnancies, often beginning earlier during such gestations. Brief flare-ups of disease may occur with resumption of menses and may develop in patients later exposed to oral contraceptives. Occasionally, infants of affected mothers demonstrate transient skin lesions. Biopsies of early lesional skin show teardrop-shaped subepidermal vesicles forming in dermal papillae in association with an eosinophil-rich leukocytic infiltrate. Differentiation ofPG from other subepidermal bullous diseases by light microscopy is often difficult. However, direct immunofluorescence microscopy of perilesional skin from PG patients reveals the immunopathologic hallmark of this disorder -- linear deposits of C3 in the epidermal basement membrane zone. These deposits develop as a consequence of complement activation produced by low titer IgG anti-basement membrane zone autoantibodies. Recent studies have shown that the majority of PG sera contain autoantibodies that recognizeBPAG2, the same 180-kDa hemidesmosome-associated protein that is targeted by autoantibodies in roughly 70% of patients withBP -- a subepidermal bullous disease that resembles PG morphologically, histologically, and immunopathologically. The goals of therapy in patients withPG are to prevent the development of new lesions, relieve intense pruritus, and care for erosions at sites of blister formation. Most patients require treatment with moderate doses of daily glucocorticoids (i.e., 20 to 40 mg of prednisone) at some point in their course. Mild cases (or brief flare-ups) may be controlled by vigorous use of potent topical glucocorticoids. Although PG was once thought to be associated with an increased risk of fetal morbidity and mortality, the best evidence now suggests that these infants may only be at increased risk of being slightly premature or "small for dates." Current evidence suggests that there is no difference in the incidence of uncomplicated live births in PG patients treated with systemic glucocorticoids and in those managed more conservatively. If systemic glucocorticoids are administered, newborns are at risk for development of reversible adrenal insufficiency. DERMATITIS HERPETIFORMIS Dermatitis herpetiformis (DH) is an intensely pruritic, papulovesicular skin disease characterized by lesions symmetrically distributed over extensor surfaces (i.e., elbows, knees, buttocks, back, scalp, and posterior neck) (Plate IIE-68). The primary lesion in this disorder is a papule, papulovesicle, or urticarial plaque. Because pruritus is prominent, patients may present with excoriations and crusted papules but no observable primary lesions. Patients sometimes report that their pruritus has a
distinctive burning or stinging component; the onset of such local symptoms reliably heralds the development of distinct clinical lesions 12 to 24 h later. Almost all DH patients have an associated, usually subclinical, gluten-sensitive enteropathy (Chap. 286), and more than 90% express the HLA-B8/DRw3 and HLA-DQw2 haplotypes. DH may present at any age, including childhood; onset in the second to fourth decades is most common. The disease is typically chronic. Biopsy of early lesional skin reveals neutrophil-rich infiltrates within dermal papillae. Neutrophils, fibrin, edema, and microvesicle formation at these sites are characteristic of early disease. Older lesions may demonstrate nonspecific features of a subepidermal bulla or an excoriated papule. Because the clinical and histologic features of this disease can be variable and resemble other subepidermal blistering disorders, the diagnosis is confirmed by direct immunofluorescence microscopy of normal-appearing perilesional skin. Such studies demonstrate granular deposits of IgA (with or without complement components) in the papillary dermis and along the epidermal basement membrane zone. IgA deposits in the skin are unaffected by control of disease with medication; however, these immunoreactants may diminish in intensity or disappear in patients maintained for long periods on a strict gluten-free diet (see below). Patients with granular deposits of IgA in their epidermal basement membrane zone typically do not have circulating IgA anti-basement membrane autoantibodies and should be distinguished from individuals with linear IgA deposits at this site (see below). Although mostDHpatients do not report overt gastrointestinal symptoms or laboratory evidence of malabsorption, biopsies of small bowel usually reveal blunting of intestinal villi and a lymphocytic infiltrate in the lamina propria. As is true for patients with celiac disease, this gastrointestinal abnormality can be reversed by a gluten-free diet. Moreover, if maintained, this diet alone may control the skin disease and eventuate in clearance of IgA deposits from these patients' epidermal basement membrane zone. Subsequent gluten exposure in such patients alters the morphology of their small bowel, elicits a flare-up of their skin disease, and is associated with the reappearance of IgA in their epidermal basement membrane zone. Additional evidence that DH develops as a consequence of dietary gluten exposure is the demonstration of IgA anti-endomysial antibodies in these patients' sera (as found in the sera of patients with ordinary gluten-sensitive enteropathy). Recent studies have shown that such autoantibodies are directed against tissue transglutaminase. Patients with DH also have an increased incidence of thyroid abnormalities, achlorhydria, atrophic gastritis, and antigastric parietal cell antibodies. These associations likely relate to the high frequency of the HLA-B8/DRw3 haplotype in these patients, since this marker is commonly linked to autoimmune disorders. The mainstay of treatment of DH is dapsone, a sulfone. Patients respond rapidly (24 to 48 h) to dapsone but require careful pretreatment evaluation and close follow-up to ensure that complications are avoided or controlled. All patients on more than 100 mg/d dapsone will have some hemolysis and methemoglobinemia. These are expected pharmacologic side effects of this agent. Gluten restriction can control DH and lessen dapsone requirements; this diet must rigidly exclude gluten to be of maximal benefit. Many months of dietary restriction may be necessary before a beneficial result is achieved. Good dietary counselling by a trained dietitian is essential. LINEAR IGA DISEASE
Linear IgA disease, once considered a variant form of dermatitis herpetiformis, is actually a separate and distinct entity. Clinically, these patients may resemble patients with typical cases ofDH,BP, or other subepidermal blistering diseases. Lesions typically consist of papulovesicles, bullae, and/or urticarial plaques, predominantly on extensor (as seen in "classic" DH), central, or flexural sites. Oral mucosal involvement occurs in some patients. Severe pruritus resembles that in patients with DH. Patients with linear IgA disease do not have an increased frequency of the HLA-B8/DRw3 haplotype or an associated enteropathy and hence are not candidates for a gluten-free diet. The histologic alterations in early lesions may be virtually indistinguishable from those inDH. However, direct immunofluorescence microscopy of normal-appearing perilesional skin reveals linear deposits of IgA (and often C3) in the epidermal basement membrane zone. Most patients with linear IgA disease demonstrate circulating IgA anti-basement membrane autoantibodies against epitopes in the extracellular domain ofBPAG2, a transmembrane protein found in hemidesmosomes of basal keratinocytes. These patients generally respond to treatment with dapsone, 50 to 150 mg/d. EPIDERMOLYSIS BULLOSA ACQUISITA EBAis a rare, noninherited, polymorphic, subepidermal blistering disease. (The inherited form is discussed in Chap. 351.) Patients with classic or noninflammatory EBA have blisters on noninflamed skin, atrophic scars, milia, nail dystrophy, and oral lesions. Because lesions generally occur at sites exposed to minor trauma, classic EBA is considered to be a mechanobullous disease. Other patients with EBA have widespread inflammatory, scarring, bullous lesions and oromucosal involvement that resembles severeBP. Some patients present with an inflammatory bullous disease that evolves into the classic noninflammatory form of this disorder. In general, EBA is chronic; associations with multiple myeloma, amyloidosis, inflammatory bowel disease, and diabetes mellitus have been reported. The HLA-DR2 haplotype is found with increased frequency in these patients. The histology of lesional skin varies depending on the character of the lesion being studied. Noninflammatory bullae show subepidermal blisters with a sparse leukocytic infiltrate and resemble those in patients with porphyria cutanea tarda. Inflammatory lesions consist of a subepidermal blister and neutrophil-rich leukocytic infiltrates in the superficial dermis.EBApatients have continuous deposits of IgG (and frequently C3 as well as other complement components) in a linear pattern within the epidermal basement membrane zone. Ultrastructurally, these immunoreactants are found in the sublamina densa region in association with anchoring fibrils, wheat stack-like structures that extend from the lamina densa into the underlying papillary dermis. Approximately 25 to 50% of EBA patients have circulating IgG anti-basement membrane autoantibodies directed against type VII collagen -- the collagen species that comprises anchoring fibrils. Such IgG autoantibodies bind the dermal side of 1 M NaCl split skin (in contrast to IgG autoantibodies in patients withBP that bind either epidermal or both sides of this indirect immunofluorescence microscopy test substrate). Treatment ofEBA is generally unsatisfactory. Some patients with inflammatory EBA may respond to systemic glucocorticoids, either alone or in combination with immunosuppressive agents. Other patients (especially those with neutrophil-rich
inflammatory lesions) may respond to dapsone. The chronic, noninflammatory form of this disease is largely resistant to treatment, although some patients may respond to cyclosporine. CICATRICIAL PEMPHIGOID Cicatricial pemphigoid (CP) is a rare, acquired, subepithelial blistering disease characterized by erosive lesions of mucous membranes and skin that result in scarring of at least some sites of involvement. Immunopathologically, perilesional mucosa and skin of patients with CP demonstrate in situ deposits of immunoreactants in epithelial basement membranes. Common sites of involvement include the oral mucosa (especially the gingiva) and conjunctiva; other sites that may be affected include the nasopharyngeal, laryngeal, esophageal, urogenital, and rectal mucosa. Skin lesions (present in about one-third of patients) tend to predominate on the scalp, face, and upper trunk and generally consist of a few scattered erosions or tense blisters on an erythematous or urticarial base. CP is typically a chronic and progressive disorder. Serious complications may arise as a consequence of ocular, laryngeal, esophageal, or urogenital lesions. Erosive conjunctivitis may result in shortened fornices, symblephara, ankyloblepharon, entropion, corneal opacities, and (in severe cases) blindness. Similarly, erosive lesions of the larynx may cause hoarseness, pain, and tissue loss that if unrecognized and untreated may eventuate in complete destruction of the airway. Esophageal lesions may result in stenosis and/or strictures that may place patients at risk for aspiration. Strictures may also complicate urogenital involvement. Biopsies of lesional tissue generally demonstrate subepithelial vesiculobullae and a mononuclear leukocytic infiltrate. Neutrophils and eosinophils may be seen in biopsies of early lesions; older lesions may demonstrate a scant leukocytic infiltrate and fibrosis. Direct immunofluorescence microscopy of perilesional tissue typically demonstrates deposits of IgG, IgA, and/or C3 in these patients' epithelial basement membranes. Because many of these patients show no evidence of circulating anti-basement membrane autoantibodies, testing of perilesional skin is important diagnostically. AlthoughCP was once thought to be a single nosologic entity, it is now largely regarded as a disease phenotype that may develop as a consequence of an autoimmune reaction against a variety of different molecules in epithelial basement membranes (e.g.,BPAG2, laminin 5, type VII collagen, and other antigens yet to be completely defined). Treatment of CP is largely dependent upon sites of involvement. Due to potentially severe complications, ocular, laryngeal, esophageal, and/or urogenital involvement require aggressive systemic treatment with dapsone, prednisone, or the latter in combination with another immunosuppressive agent (e.g., azathioprine or cyclophosphamide). Less threatening forms of the disease may be managed with topical or intralesional glucocorticoids. AUTOIMMUNE SYSTEMIC DISEASES WITH PROMINENT CUTANEOUS FEATURES DERMATOMYOSITIS The cutaneous manifestations of dermatomyositis (Chap. 382) are often distinctive but at times may resemble those of systemic lupus erythematosus (SLE) (Chap. 311), scleroderma (Chap. 313), or other overlapping connective tissue diseases (Chap. 313).
The extent and severity of cutaneous disease may or may not correlate with the extent and severity of the myositis. Patients with severe muscle involvement may have relatively minor skin changes, whereas patients with marked skin involvement may have mild muscle disease. The cutaneous manifestations of dermatomyositis are similar whether the disease appears in childhood or old age, except that calcification of subcutaneous tissue is a common late sequela in childhood dermatomyositis. The cutaneous signs of dermatomyositis may precede or follow the development of myositis by weeks to years. Cases lacking muscle involvement (i.e., dermatomyositis sine myositis) have also been reported. The most common manifestation is a purple-red discoloration of the upper eyelids, sometimes associated with scaling ("heliotrope" erythema; Plate IIE-63) and periorbital edema. Erythema on the cheeks and nose in a "butterfly" distribution may resemble the eruption inSLE. Erythematous or violaceous scaling patches are common on the upper anterior chest, posterior neck, scalp, and the extensor surfaces of the arms, legs, and hands. Erythema and scaling may be particularly prominent over the elbows, knees, and the dorsal interphalangeal joints. Approximately one-third of patients have violaceous, flat-topped papules over the dorsal interphalangeal joints that are pathognomonic of dermatomyositis (Gottron's sign or Gottron's papules; Plate IIE-65). These lesions can be contrasted with the erythema and scaling on the dorsum of the fingers in some patients with SLE, which spares the skin over the interphalangeal joints. Periungual telangiectasia may be prominent, and a lacy or reticulated erythema may be associated with fine scaling on the extensor surfaces of the thighs and upper arms. Other patients, particularly those with long-standing disease, develop areas of hypopigmentation, hyperpigmentation, mild atrophy, and telangiectasia known as poikiloderma vasculare atrophicans. Poikiloderma is rare in both SLE and scleroderma and thus can serve as a clinical sign that distinguishes dermatomyositis from these two diseases. Cutaneous changes may be similar in scleroderma and dermatomyositis and may include thickening and binding down of the skin of the hands (sclerodactyly) as well as Raynaud's phenomenon. However, the presence of severe muscle disease, Gottron's papules, heliotrope erythema, and poikiloderma serve to distinguish patients with dermatomyositis. Skin biopsy of erythematous, scaling lesions of dermatomyositis may reveal only mild nonspecific inflammation but sometimes may show changes indistinguishable from those found in SLE, including epidermal atrophy, hydropic degeneration of basal keratinocytes, edema of the upper dermis, and a mild mononuclear cell infiltrate. Direct immunofluorescence microscopy of lesional skin is usually negative, although granular deposits of immunoglobulin(s) and complement in the epidermal basement membrane zone have been described in some patients. Treatment should be directed at the systemic disease. In the few instances where adjunctive cutaneous therapy is desirable, topical glucocorticoids are sometimes useful. These patients should avoid exposure to ultraviolet irradiation and use photoprotective measures such as sunscreens. LUPUS ERYTHEMATOSUS The cutaneous manifestations of lupus erythematosus (LE) (Chap. 311) can be divided into acute, subacute, and chronic (i.e., discoid LE) types. Acute cutaneous LE is characterized by erythema of the nose and malar eminences in a "butterfly" distribution (Plate IIE-61). The erythema is often sudden in onset, accompanied by edema and fine scale, and correlated with systemic involvement. Patients may have widespread
involvement of the face as well as erythema and scaling of the extensor surfaces of the extremities and upper chest. These acute lesions, while sometimes evanescent, usually last for days and are often associated with exacerbations of systemic disease. Skin biopsy of acute lesions may show only a sparse dermal infiltrate of mononuclear cells and dermal edema. In some instances, cellular infiltrates around blood vessels and hair follicles are notable, as is hydropic degeneration of basal cells of the epidermis. Direct immunofluorescence microscopy of lesional skin frequently reveals deposits of immunoglobulin(s) and complement in the epidermal basement membrane zone. Treatment is aimed at control of systemic disease; photoprotection in this, as well as in other forms of LE, is very important. Subacute cutaneous lupus erythematosus (SCLE) is characterized by a widespread photosensitive, nonscarring eruption. About half of these patients haveSLE in which severe renal and central nervous system involvement is uncommon. SCLE may present as a papulosquamous eruption that resembles psoriasis or annular lesions that resemble those seen in erythema multiforme. In the papulosquamous form, discrete erythematous papules arise on the back, chest, shoulders, extensor surfaces of the arms, and the dorsum of the hands; lesions are uncommon on the face, flexor surfaces of the arms, and below the waist. The slightly scaling papules tend to merge into large plaques, some with a reticulate appearance. The annular form involves the same areas and presents with erythematous papules that evolve into oval, circular, or polycyclic lesions. The lesions of SCLE are more widespread but have less tendency for scarring than do lesions of discoidLE. Skin biopsy reveals a dense mononuclear cell infiltrate around hair follicles and blood vessels in the superficial dermis, combined with hydropic degeneration of basal cells in the epidermis. Direct immunofluorescence microscopy of lesional skin reveals deposits of immunoglobulin(s) in the epidermal basement membrane zone in about half these cases. A particulate pattern of IgG deposition around basal keratinocytes has recently been associated with SCLE. Most SCLE patients have anti-Ro antibodies. Local therapy is usually unsuccessful, and most patients require treatment with aminoquinoline antimalarials. Low-dose therapy with oral glucocorticoids is sometimes necessary; photoprotective measures against both ultraviolet B and A wavelengths are very important. Discoid lupus erythematosus (DLE) is characterized by discrete lesions, most often on the face, scalp, or external ears. The lesions are erythematous papules or plaques with a thick, adherent scale that occludes hair follicles (follicular plugging). When the scale is removed, its underside will show small excrescences that correlate with the openings of hair follicles and is termed a "carpet tack" appearance. This finding is relatively specific for DLE. Long-standing lesions develop central atrophy, scarring, and hypopigmentation but frequently have erythematous, sometimes raised borders at the periphery (Plate IIE-62). These lesions persist for years and tend to expand slowly. Only 5 to 10% of patients with DLE meet the American Rheumatism Association criteria forSLE. However, typical discoid lesions are frequently seen in patients with SLE. Biopsy of DLE lesions shows hyperkeratosis, follicular plugging, and atrophy of the epidermis. The dermal-epidermal junction reveals hydropic degeneration of basal keratinocytes, and a mononuclear cell infiltrate surrounding hair follicles and blood vessels. Direct immunofluorescence microscopy demonstrates immunoglobulin(s) and complement deposits at the basement membrane zone in about 90% of cases. Treatment is focused on control of local cutaneous disease and consists mainly of photoprotection and topical
or intralesional glucocorticoids. If local therapy is ineffective, use of aminoquinoline antimalarials may be indicated. SCLERODERMA AND MORPHEA The skin changes of scleroderma (Chap. 313) usually begin on the hands, feet, and face, with episodes of recurrent nonpitting edema. Sclerosis of the skin begins distally on the fingers (sclerodactyly) and spreads proximally, usually accompanied by resorption of bone of the fingertips, which may have punched out ulcers, stellate scars, or areas of hemorrhage (Plate IIE-66). The fingers may actually shrink in size and become sausage-shaped, and since the fingernails are usually unaffected, the nails may curve over the end of the fingertips. Periungual telangiectasias are usually present, but periungual erythema is rare. In advanced cases, the extremities show contractures and calcinosis cutis. Face involvement includes a smooth, unwrinkled brow, taut skin over the nose, shrinkage of tissue around the mouth, and perioral radial furrowing (Plate IIE-64). Matlike telangiectasias are often present, particularly on the face and hands. Involved skin feels indurated, smooth, and bound to underlying structures; hyperpigmentation and hypopigmentation are also often present. Raynaud's phenomenon, i.e., cold-induced blanching, cyanosis, and reactive hyperemia, is present in almost all patients and can precede development of scleroderma by many years. The combination of calcinosis cutis, Raynaud's phenomenon, esophageal dysmotility, sclerodactyly, and telangiectasia has been termed the CREST syndrome. Anticentromere antibodies have been reported in a very high percentage of patients with the CREST syndrome but in only a small minority of patients with scleroderma. Skin biopsy reveals thickening of the dermis and homogenization of collagen bundles. Direct immunofluorescence microscopy of lesional skin is usually negative. Morphea, which has been called localized scleroderma, is characterized by localized thickening and sclerosis of skin, usually affecting young adults or children. Morphea begins as erythematous or flesh-colored plaques that become sclerotic, develop central hypopigmentation, and demonstrate an erythematous border. In most cases, patients have one or a few lesions, and the disease is termed localized morphea. In some patients, widespread cutaneous lesions may occur, without systemic involvement. This form is called generalized morphea. Most patients with morphea do not have autoantibodies. Skin biopsy of morphea is indistinguishable from that of scleroderma. Linear scleroderma is a limited form of disease that presents in a linear, bandlike distribution and tends to involve deep as well as superficial layers of skin. Scleroderma and morphea are usually quite resistant to therapy. For this reason, physical therapy to prevent joint contractures and to maintain function is employed and is often helpful. Diffuse fasciitis with eosinophilia is a clinical entity that can sometimes be confused with scleroderma. There is usually the sudden onset of swelling, induration, and erythema of the extremities frequently following significant physical exertion. The proximal portions of extremities (arms, forearms, thighs, legs) are more often involved than are the hands and feet. While the skin is indurated, it is usually not bound down as in scleroderma; contractures may occur early secondary to fascial involvement. The latter may also cause muscle groups to be separated (i.e., the "groove sign") and veins to appear depressed (i.e., sunken veins). These skin findings are accompanied by peripheral blood eosinophilia, increased erythrocyte sedimentation rate, and sometimes
hypergammaglobulinemia. Deep biopsy of affected areas of skin reveals inflammation and thickening of the deep fascia overlying muscle. An inflammatory infiltrate composed of eosinophils and mononuclear cells is usually found. Patients with eosinophilic fasciitis appear to be at increased risk to develop bone marrow failure or other hematologic abnormalities. While the ultimate course of eosinophilic fasciitis is uncertain, many patients respond favorably to treatment with prednisone in doses ranging from 40 to 60 mg/d. The eosinophilia-myalgia syndrome, a disorder reported in epidemic numbers in 1989 and linked to ingestion of L-tryptophan manufactured by a single company in Japan, is a multisystem disorder characterized by debilitating myalgias and absolute eosinophilia in association with varying combinations of arthralgias, pulmonary symptoms, and peripheral edema. In a later phase (i.e., 3 to 6 months after initial symptoms), these patients often develop localized sclerodermatous skin changes, weight loss, and/or neuropathy (Chap. 313). The precise cause of this syndrome, which may resemble other sclerotic skin conditions, is unknown. However, the implicated lots of L-tryptophan contained the contaminant 1,1-ethylidene bis[tryptophan]. This contaminant may be pathogenic or a marker for another substance that provokes the disorder. (Bibliography omitted in Palm version) Back to Table of Contents
59. CUTANEOUS DRUG REACTIONS - Robert S. Stern, Olivier M. Chosidow, Bruce U. Wintroub Cutaneous reactions are among the most frequent adverse reactions to drugs. Prompt recognition of these reactions, drug withdrawal, and appropriate therapeutic interventions can minimize toxicity. This chapter focuses on adverse cutaneous reactions to drugs other than topical agents and reviews the incidence, patterns, and pathogenesis of cutaneous reactions to drugs and other therapeutic agents. USE OF PRESCRIPTION DRUGS IN THE UNITED STATES More than 1.5 billion prescriptions for 60,000 drug products, which include over 2000 different active agents, are dispensed each year in the United States. Hospital inpatients alone annually receive about 120 million courses of drug therapy, and half of adult Americans receive prescription drugs on a regular outpatient basis. Many additional patients use over-the-counter medicines that may cause adverse cutaneous reactions. INCIDENCE OF CUTANEOUS REACTIONS Although adverse drug reactions are common, it is difficult to ascertain their incidence, seriousness, and ultimate health effects. Available information comes from evaluations of hospitalized patients, epidemiologic surveys, premarketing studies, and voluntary reporting, most notably to the U.S. Food and Drug Administration's Medwatch System. None of these efforts provides comprehensive comparable data on the risk of cutaneous reactions associated with most medicines. In one study about 2% of medical inpatients had skin reactions consisting of rash, urticaria, or pruritus during hospitalization. The overall reaction rate per course of drug therapy was about 3:1000. Among inpatients, penicillins, sulfonamides, and blood products accounted for two-thirds of cutaneous reactions. Among outpatients, reaction rates for many antibiotics were comparable to those observed in inpatients. Fluoroquinolones are notable causes of cutaneous reactions not observed in earlier studies. Reaction rates for selected commonly used antibiotics are summarized in Table 59-1. Most cutaneous reactions occur within 2 weeks of exposure to a drug. The risk of allergic reactions does not vary greatly with age or sex. Among outpatients, the risk of a reaction to an antibiotic was comparable for first and subsequent courses of a given drug. The distribution of morphologic patterns of drug eruptions cared for within a Finnish hospital dermatology department with a special interest in fixed drug eruptions included exanthematous reactions (32%), urticaria and/or angioedema (20%), fixed drug eruptions (34%), erythema multiforme (2%), Stevens-Johnson syndrome (SJS; 1%), exfoliative dermatitis (1%), and photosensitivity reactions (3%). Other studies suggest that about 80% of all cutaneous reactions are morbilliform or erythematous, 10 to 15% are urticaria or angioedema, and all other types of reactions are relatively rare. The relative risk ofSJS and toxic epidermal necrolysis (TEN), perhaps the most important severe cutaneous reactions, has been quantified in an international case control study and case series. Sulfonamide antibiotics, allopurinol, amine antiepileptic
drugs (phenytoin and carbamazepine), and lamotrigine (a new antiepileptic) are associated with the highest risk of these reactions. PATHOGENESIS OF DRUG REACTIONS Untoward cutaneous responses to drugs can arise as a result of immunologic or nonimmunologic mechanisms. Immunologic reactions require activation of host immunologic pathways and are designated drug allergy. Drug reactions occurring through nonimmunologic mechanisms may be due to activation of effector pathways, overdosage, cumulative toxicity, side effects, ecologic disturbance, interactions between drugs, metabolic alterations, exacerbation of preexisting dermatologic conditions, or inherited protein or enzyme deficiencies. It is often not possible to specify the responsible drug or pathogenic mechanism because the skin responds to a variety of stimuli through a limited number of reaction patterns. The mechanism of many drug reactions is unknown. IMMUNOLOGIC DRUG REACTIONS Drugs frequently elicit an immune response, but only a small number of individuals experience clinical hypersensitivity reactions. For example, most patients exposed to penicillin develop demonstrable antibodies to penicillin but do not manifest drug reactions when exposed to penicillin. Multiple factors determine the capacity of a drug to elicit an immune response, including the molecular characteristics of the drug and host effects. Increases in molecular size and complexity are associated with increased immunogenicity, and macromolecular drugs such as protein or peptide hormones are highly antigenic. Most drugs are small organic molecules > > >IgE-Dependent Reactions IgE-dependent drug reactions are usually manifest in the
skin and gastrointestinal, respiratory, and cardiovascular systems (Chap. 310). Primary symptoms and signs include pruritus, urticaria, nausea, vomiting, cramps, bronchospasm, and laryngeal edema and, on occasion, anaphylactic shock with hypotension and death. Immediate reactions may occur within minutes of drug exposure, and accelerated reactions occur hours or days after drug administration. Accelerated reactions are usually urticarial and may include laryngeal edema. Penicillin and related drugs are the most frequent causes of IgE-dependent reactions. Release of chemical mediators such as histamine, adenosine, leukotrienes, prostaglandins, platelet-activating factor, enzymes, and proteoglycans from sensitized tissue, mast cells, or circulating basophilic leukocytes results in vasodilation and edema. Release is triggered when polyvalent drug protein conjugates cross-link IgE molecules fixed to sensitized cells. The clinical manifestations are determined by interaction of the released chemical mediator with its target organ, i.e., skin, respiratory, gastrointestinal, and/or cardiovascular systems. Certain routes of administration favor different clinical patterns (i.e., oral route: gastrointestinal effects; intravenous route: circulatory effects). Immune-Complex-Dependent Reactions Serum sickness is produced by circulating immune complexes and is characterized by fever, arthritis, nephritis, neuritis, edema, and an urticarial, papular, or purpuric rash (Chap. 317). The syndrome requires an antigen that remains in the circulation for prolonged periods so that when antibody is synthesized, circulating antigen-antibody complexes are formed. Serum sickness was first described following administration of foreign sera, but drugs are now the usual cause. Drugs that produce serum sickness include the penicillins, sulfonamides, thiouracils, cholecystographic dyes, phenytoin, aminosalicylic acid, heparin, and antilymphocyte globulin. Cephalosporin administration in febrile children is associated with a high risk of a clinically similar reaction, but the mechanism of this reaction is unknown. In classic serum sickness, symptoms develop 6 days or more after exposure to a drug, the latent period representing the time needed to synthesize antibody. The antibodies responsible for immune-complex-dependent drug reactions are largely of the IgG or IgM class. Vasculitis, a relatively rare cutaneous complication of drugs, may also be a result of immune complex deposition (Chap. 317). Cytotoxicity and Delayed Hypersensitivity Cytotoxicity and delayed hypersensitivity mechanisms may be important in the etiology of morbilliform exanthema, hypersensitivity syndrome,SJS, orTEN, but this is not proven. Systemic manifestations occur frequently. The nature of the antigen leading to cytotoxic reactions is unknown, but it is likely that different T lymphocyte populations are activated. T H1 type cells will lead to the production of interleukin (IL)-2 and interferon (IFN)-g and subsequent activation of cytotoxic T cells. In early lesions of morbilliform exanthema or TEN, histopathologic studies have shown expression of HLA-DR and intercellular adhesion molecule (ICAM-1) by keratinocytes, CD4 cells (in the dermis), and CD8 T cells (in the epidermis) and apoptosis of keratinocytes (facilitated by tumor necrosis factora secretion and fas-ligand expression). TH2 type cells produce cytokines such as IL-5, which may be involved in hypersensitivity syndrome (see below). NONIMMUNOLOGIC DRUG REACTIONS Nonimmunologic mechanisms are responsible for the majority of drug reactions; however, only the most important mechanisms will be discussed.
Nonimmunologic Activation of Effector Pathways Drug reactions may result from nonimmunologic activation of effector pathways by three mechanisms: First, drugs may release mediators directly from mast cells and basophils and present as anaphylaxis, urticaria, and/or angioedema. Urticarial anaphylactic reactions induced by opiates, polymyxin B, tubocurarine, radiocontrast media, and dextrans may occur by this mechanism. Second, drugs may activate complement in the absence of antibody. This is an additional mechanism through which radiocontrast media may act. Third, drugs such as aspirin and other nonsteroidal anti-inflammatory agents (NSAIDs) may alter pathways of arachidonic acid metabolism and induce urticaria. Phototoxicity Phototoxic reactions may be drug-induced or may occur in metabolic disorders in which a photosensitizing chemical is overproduced. A phototoxic reaction occurs when enough chromophore (drug or metabolic product) absorbs sufficient radiation to cause a reaction or interaction with target tissue. Drug-induced phototoxic reactions can occur on first exposure. The incidence of phototoxicity is a direct function of the concentration of sensitizer and the amount of light of the appropriate wavelengths. At least three distinct photochemical mechanisms have been described: (1) the reaction between the excited state of a phototoxic molecule and a biologic target may cause formation of a covalent photoaddition product, (2) the phototoxic molecule may form stable photoproducts that are toxic to biologic substrates, and (3) radiation of a phototoxic molecule may result in transfer of energy to oxygen molecules and cause formation of toxic oxygen species, such as singlet oxygen superoxide anion, or hydroxyl radicals. Interaction of these reactive species with biologic targets produces photooxidized molecules. Phototoxic injury is usually manifest as a sunburn-like reaction. Exacerbation of Preexisting Diseases A variety of agents can exacerbate preexisting diseases. For example, lithium can exacerbate acne and psoriasis in a dose-dependent manner. Beta-blocking agents andIFN-amay induce psoriasis. Withdrawal of glucocorticoids can exacerbate psoriasis or atopic dermatitis. Inherited Enzyme or Protein Deficiencies Specific genetically determined defects in the ability of an individual to detoxify toxic reactive drug metabolites may predispose such individuals to the development of severe drug reactions, especially hypersensitivity syndrome, and perhapsTENassociated with use of sulfonamides and anticonvulsants. Alterations of Immunologic Status Alterations in patients' immunologic status may also modify the risk of cutaneous reactions. Bone marrow transplant patients, HIV-infected persons, and persons with Epstein-Barr virus infection are at higher risk of developing cutaneous reactions to drugs. Skin reactions to trimethoprim-sulfamethoxazole are seen in about a third of HIV-infected users of this drug, but desensitization can be accomplished. Dapsone, trimethoprim alone, and amoxicillin-clavulanate are also frequent causes of drug eruptions in HIV-infected patients. The advent of highly active antiretroviral therapy (HAART) may have decreased the risk of cutaneous reactions in HIV patients (Chap. 309). A CLINICAL CLASSIFICATION OF CUTANEOUS DRUG REACTIONS
URTICARIA/ANGIOEDEMA Urticaria is a skin reaction characterized by pruritic, red wheals. Lesions may vary from a small point to a large area. Individual lesions rarely last more than 24 h. When deep dermal and subcutaneous tissues are also swollen, this reaction is known as angioedema. Angioedema may involve mucous membranes and may be part of a life-threatening anaphylactic reaction. Urticarial lesions, along with pruritus and morbilliform (or maculopapular) eruptions, are among the most frequent types of cutaneous reactions to drugs. Drug-induced urticaria may be caused by three mechanisms: an IgE-dependent mechanism, circulating immune complexes (serum sickness), and nonimmunologic activation of effector pathways. IgE-dependent urticarial reactions usually occur within 36 h but can occur within minutes. Reactions occurring within minutes to hours of drug exposure are termed immediate reactions, whereas those that occur 12 to 36 h after drug exposure are designated accelerated reactions. Immune-complex-induced urticaria associated with serum sickness usually occurs from 6 to 12 days after first exposure. In this syndrome, the urticarial eruption may be accompanied by fever, hematuria, arthralgias, hepatic dysfunction, and neurologic symptoms. Certain drugs, such asNSAIDs, angiotensin-converting enzyme (ACE) inhibitors, and radiographic dyes, may induce urticarial reactions, angioedema, and anaphylaxis in the absence of drug-specific antibody. Although ACE inhibitors, aspirin, penicillin, and blood products are the most frequent causes of urticarial eruptions, urticaria has been observed in association with nearly all drugs. Drugs also may cause chronic urticaria, which lasts more than 6 weeks. Aspirin frequently exacerbates this problem. The treatment of urticaria or angioedema depends on the severity of the reaction and the rate at which it is evolving. In severe cases, especially with respiratory or cardiovascular compromise, epinephrine is the mainstay of therapy, but its effect is reduced in patients using beta blockers. For more seriously affected patients, treatment with systemic glucocorticoids, sometimes intravenously administered, are helpful. In addition to drug withdrawal, for patients with only cutaneous symptoms and without symptoms of angioedema or anaphylaxis, oral antihistamines are usually sufficient. PHOTOSENSITIVITY ERUPTIONS Photosensitivity eruptions are usually most marked in sun-exposed areas but may extend to sun-protected areas. Phototoxic reactions are more common with some drugs. Photoallergic reactions to systemically administered drugs are very rare. Phototoxic reactions usually resemble sunburn and can occur with the first exposure to a drug. Their severity depends on the tissue level of the drug, the extent of exposure to light, and the efficiency of the photosensitizer (Chap. 60). Orally administered phototoxic drugs include many fluoroquinolones, chlorpromazine, tetracycline, thiazides, and at least twoNSAIDs(benoxaprofen and piroxicam). The majority of the common phototoxic drugs have action spectrums in the long-wave ultraviolet A (UV-A) range. Phototoxic reactions abate with removal of either the drug or ultraviolet radiation. Because UV-A and visible light, which trigger these reactions, are
not easily absorbed by nonopaque sunscreens and are transmitted through window glass, these reactions may be difficult to block. Photosensitivity reactions are treated by avoiding exposure to ultraviolet light (sunlight) and treating the reaction as one would a sunburn. Rarely, individuals develop persistent reactivity to light, necessitating long-term avoidance of sun exposure. PIGMENTATION CHANGES Drugs may cause a variety of pigmentary changes in the skin. Some drugs stimulate melanocytic activity and increase pigmentation. Drug deposition can also lead to pigmentation; this phenomenon occurs with heavy metals. Phenothiazines may be deposited in the skin and cause a slate-gray color. Antimalarial drugs may cause a slate-gray or yellow pigmentation. Long term minocycline use may cause slate-gray hyperpigmentation, especially in areas of chronic inflammation. Inorganic arsenic, once used to treat psoriasis, is associated with diffuse macular pigmentation. Other heavy metals that cause pigmentary changes include silver, gold, bismuth, and mercury. Long-term use of phenytoin can produce a chloasma-like pigmentation in women. Certain cytostatic agents can also cause pigmentary changes. Histologic examination is often diagnostic for drug deposition diseases. Zidovudine (AZT) is a frequent cause of pigmentation, especially of the nails (Chap. 309). Nicotinic acid in large doses may cause brown pigmentation, and oral contraceptives may produce chloasma. In addition, amiodarone may cause violaceous hyperpigmentation that is increased in sun-exposed skin. Drugs such as heavy metals, copper antimalarial and arsenical agents, and ACTH also may discolor oral mucosa. VASCULITIS Cutaneous necrotizing vasculitis often presents as palpable purpuric lesions that may be generalized or limited to the lower extremities or other dependent areas (Chap. 317). Urticarial lesions, ulcers, and hemorrhagic blisters also occur. Vasculitis may involve other organs, including the liver, kidney, brain, and joints. Drugs are only one cause of vasculitis, with infection and collagen vascular disease responsible for the majority of cases. Propylthiouracil induces a cutaneous vasculitis that is accompanied by leukopenia and splenomegaly. Direct immunofluorescent changes in these lesions suggest immune-complex deposition. Drugs implicated in vasculitic eruptions include allopurinol, thiazides, sulfonamides, penicillin, and someNSAIDs. HYPERSENSITIVITY SYNDROME Initially described with phenytoin, hypersensitivity syndrome presents as an erythematous eruption that may become purpuric and is accompanied by many of the following features: fever, facial and periorbital edema, tender generalized lymphadenopathy, leukocytosis (often with atypical lymphocytes and eosinophils), hepatitis, and sometimes nephritis or pneumonitis. The cutaneous reaction usually begins 1 to 6 weeks after phenytoin is begun and usually resolves with drug cessation,
but symptoms, especially hepatitis, may persist. The eruption recurs with rechallenge, and cross-reactions among aromatic anticonvulsants, including phenytoin, carbamazepine, and barbiturates, are frequent. With phenytoin, an increased risk of this syndrome is associated with an inherited deficiency of epoxide hydrolase, an enzyme required for metabolism of a toxic intermediate arene oxide that is formed during metabolism of phenytoin by the cytochrome P450 system. Other drugs causing this syndrome include lamotrigine, dapsone, allopurinol, sulfonamides, minocycline, and sulfones. Systemic glucocorticoids (prednisone, 0.5 to 1.0 mg/kg) seem to reduce symptoms. Mortality as high as 10% has been reported. WARFARIN NECROSIS OF THE SKIN This rare reaction occurs usually between the third and tenth days of therapy with warfarin derivatives, usually in women. Lesions are sharply demarcated, erythematous, indurated, and purpuric and may resolve or progress to form large, irregular, hemorrhagic bullae with eventual necrosis and slow-healing eschar formation. Development of the syndrome is unrelated to drug dose or underlying condition. Favored sites are breasts, thighs, and buttocks. The course is not altered by discontinuation of the drug after onset of the eruption. Similar reactions have been associated with heparin. Warfarin reactions are associated with protein C deficiency. Protein C is a vitamin K-dependent protein with a shorter half-life than other clotting proteins and is in part responsible for control of fibrinolysis. Since warfarin inhibits synthesis of vitamin K-dependent coagulation factors, warfarin anticoagulation in heterozygotes for protein C deficiency causes a precipitous fall in circulating levels of protein C, permitting hypercoagulability and thrombosis in the cutaneous microvasculature, with consequent areas of necrosis. Heparin-induced necrosis may have clinically similar features but is probably due to heparin-induced platelet aggregation with subsequent occlusion of blood vessels. Warfarin-induced cutaneous necrosis is treated with vitamin K and heparin. Vitamin K reverses the effects of warfarin, and heparin acts as an anticoagulant. Treatment with protein C concentrates may also be helpful in individuals with deficiencies of protein C, the predisposing factor for development of these reactions. MORBILLIFORM REACTIONS Morbilliform or maculopapular eruptions are the most common of all drug-induced reactions, often start on the trunk or areas of pressure or trauma, and consist of erythematous macules and papules that are frequently symmetric and may become confluent. Involvement of mucous membranes, palms, and soles is variable; the eruption may be associated with moderate to severe pruritus and fever. The pathogenesis is unclear. A hypersensitivity mechanism has been suggested, although these reactions do not always recur following drug rechallenge. Diagnosis is rarely assisted by laboratory or patch testing; differentiation from viral exanthem is the principal differential diagnostic consideration. Unless the suspect drug is essential it should be discontinued. Occasionally these eruptions may decrease or fade with continued use of the responsible drug.
Morbilliform reactions usually develop within 1 week of initiation of therapy and last 1 to 2 weeks; however, reactions to some drugs, especially penicillin and drugs with long half-lives, may begin more than 2 weeks after therapy has begun and last as long as 2 weeks after therapy has ceased. Morbilliform eruptions are usually treated by discontinuing the suspect medications symptomatically. Oral antihistamines, emollients, and soothing baths are useful for treatment of pruritus. Short courses of potent topical glucocorticoids can reduce inflammation and symptoms and are probably helpful. The beneficial effect of systemic glucocorticoids relative to risk is less clear. FIXED DRUG REACTIONS These reactions are characterized by one or more sharply demarcated, erythematous lesions in which hyperpigmentation results after resolution of the acute inflammation; with rechallenge, the lesion recurs in the same (i.e., "fixed") location. Lesions often involve the lips, hands, legs, face, genitalia, and oral mucosa and cause burning. Most patients have multiple lesions. Patch testing is useful to establish the etiology. Fixed drug eruptions have been associated with phenolphthalein, sulfonamides, tetracyclines, phenylbutazone,NSAIDs, and barbiturates. Although cross-sensitivity appears to occur between different tetracycline compounds, cross-sensitivity was not elicited when different sulfonamide compounds were administered to patients as part of provocation testing. LICHENOID DRUG ERUPTIONS A lichenoid cutaneous reaction, clinically and morphologically indistinguishable from lichen planus, is associated with a variety of drugs and chemicals. Eosinophils are more common when the reaction is drug-induced. Gold and antimalarials are most often associated with this eruption. Antihypertensive agents, including beta blockers and captopril, have also been reported to cause lichenoid reactions. BULLOUS ERUPTIONS Blisters accompany a wide variety of cutaneous reactions, including fixed drug eruptions, severe morbilliform eruptions in dependent areas of the body, and phototoxic reactions.SJS andTEN are the most serious and important bullous reactions to drugs. Nalidixic acid and furosemide cause blistering eruptions indistinguishable from the primary bullous diseases. A pemphigus foliaceus-like eruption is seen with penicillamine. PUSTULAR ERUPTIONS Acute generalized exanthematous pustulosis is often associated with exposure to drugs, most notably antibiotics. Usually beginning on the face or intertriginous areas, small nonfollicular pustules overlying erythematous and edematous skin may coalesce and lead to superficial ulceration. Fever is present and differentiating this eruption fromTEN in its initial stages may be difficult. Acute generalized exanthematous pustulosis often
begins within a few days of initiating drug treatment. ERYTHEMA MULTIFORME Erythema multiforme is an acute, self-limited inflammatory disorder of skin and mucous membranes characterized by distinctive iris or target lesions, usually acrally distributed and often associated with sore throat, mucosal lesions, and malaise. Classic erythema multiforme usually has nondrug causes, most commonly herpes simplex infection, and must be differentiated from trueSJS, which is usually drug related. STEVENS-JOHNSON SYNDROME SJSis a blistering disorder that is usually more severe than erythema multiforme. Initial presentation is often a sore throat, malaise, and fever. Within a few days, in addition to erosions of multiple mucous membranes, small blisters developing on dusky or purpuric macules or atypical target lesions characterize this eruption. Total percent of body surface area blistering and eventual detachment is less than 10%. Overlap SJS/TENshares characteristics of both SJS and TEN, with 10 to 30% of body surface area exhibiting epidermal detachment. TOXIC EPIDERMAL NECROLYSIS TENis the most serious cutaneous drug reaction and may be fatal. Drugs are usually the cause of TEN. Onset is generally acute and is characterized by fever>39°C (102.2°F), blisters or ulcers of multiple mucous membranes, malaise, and epidermal necrosis involving>30% of body surface area. Intestinal and pulmonary involvement is associated with a poor prognosis, as is a greater extent of epidermal detachment and older age. About 30% of affected persons die. Many treatments affecting immune response or cytokines (thalidomide) or apoptosis (intravenous immunoglobulin) have been advocated, but none have been shown to be efficacious in well-controlled trials. In spite of its theoretical potential benefits, thalidomide therapy increases TEN-associated mortality. Supportive treatment in burn units is helpful in reducing morbidity and mortality. DRUGS OF SPECIAL INTEREST PENICILLIN The incidence of cutaneous reactions to penicillin is about 1%. About 85% of cutaneous reactions to penicillin are morbilliform, and about 10% are urticaria or angioedema. IgG, IgM, and IgE antibodies can be produced; IgG and IgM anti-penicillin antibodies play a role in the development of hemolytic anemia, whereas anaphylaxis and serum sickness appear to be due to IgE antibodies in serum. In patients with suspected IgE-mediated reactions to penicillin for whom future treatment is anticipated, accurate tests for sensitization are available. Current practice is to perform skin testing with a commercially available penicilloyl determinant preparation (Pre-pen, Kremers-Urban) and with fresh penicillin and, if possible, with another source
of minor (nonpenicilloyl) determinants such as aged or base-treated penicillin. Antibodies to minor determinants are common in patients experiencing anaphylaxis, but testing with major determinants alone detects most patients at risk for anaphylaxis. About one-fourth of patients with positive history of penicillin allergy have a positive skin test, while 6% (3 to 10%) with no history of penicillin sensitivity demonstrate a positive skin response to penicillin. Administering penicillin to those patients with a positive skin test produces reactions in a high proportion (50 to 100%); conversely, only a few patients (0.5%) with a negative skin test react to the drug, and reactions tend to be mild and to occur late. Since a false-negative skin test may occur during or just after an acute reaction, testing should be performed either prospectively or several months after a suspected reaction. As many as 80% of patients lose anaphylactic sensitivity and IgE antibody after several years. Radioallergosorbent tests and other in vitro tests offer no advantage over properly performed skin testing. Some cross-reactivity between penicillin and nonpenicillin b-lactam antibiotics (e.g., cephalosporins) occurs, but the majority of penicillin-allergic patients will tolerate cephalosporins. Persons who have negative skin tests to penicillin rarely develop reactions to cephalosporins. In the face of a positive clinical history of penicillin reaction, another drug should be chosen. If this is not feasible or prudent (e.g., in a pregnant patient with syphilis or with enterococcal endocarditis), skin testing with penicillin is warranted. If skin tests are negative, cautious administration of penicillin is acceptable, although some recommend desensitization of such patients if the reaction was likely to be IgE-mediated. In those with positive skin tests, desensitization is mandatory if therapeutic use ofb-lactam antibiotics is to be undertaken. Various protocols are available, including oral and parenteral approaches. Oral desensitization appears to have lower risk of serious anaphylactic reactions during desensitization. However, desensitization carries the risk of anaphylaxis regardless of how it is performed. After desensitization, many patients experience non-life-threatening IgE-mediated untoward reactions to penicillin during their course of therapy. Desensitization is not effective in those with exfoliative dermatitis or morbilliform reactions due to penicillin. NONSTEROIDAL ANTI-INFLAMMATORY DRUGS NSAIDs, including aspirin and indomethacin (indometacin), cause two broad categories of allergic-like symptoms in susceptible individuals: (1) approximately 1% of persons experience urticaria or angioedema, and (2) about half as many (0.5%) experience rhinosinusitis and asthma; however, about 10% of adults with asthma and one-third of individuals with nasal polyposis and sinusitis may respond adversely to aspirin. Urticaria/angioedema may be delayed up to 24 h and may occur at any age. The rhinosinusitis-asthma syndrome generally develops within 1 h of drug administration. In young patients, the reaction pattern often begins as watery rhinorrhea, which can be complicated by nasal and sinus infection, and polyposis, bloody discharge, and nasal eosinophilia. In many individuals with this syndrome, asthma that can be life-threatening eventually ensues wheneverNSAIDsare subsequently ingested, and symptoms may persist despite avoidance of these drugs. Proof of the association of symptoms and NSAID use requires either clear-cut history of symptoms following drug ingestion or an oral challenge. For the latter to be performed with relative safety, (1) asthma must be
under good control, (2) the procedure must be conducted in a hospital setting by experienced personnel capable of recognizing and treating acute respiratory responses, and (3) the challenge should begin with very low doses (i.e., not>30 mg) of aspirin and increase every 1 to 2 h in doubling doses as tolerated to 650 mg. While cross-reactivity betweenNSAIDsis common, it is not immunologic, and patients who are sensitive to NSAIDs cannot be identified by assessment of IgE antibody to aspirin, lymphocyte sensitization, or in vitro immunologic testing. RADIOCONTRAST MEDIA Large numbers of patients are exposed to radiocontrast agents. High-osmolality radiocontrast media are about five times more likely to induce urticaria (1%) or anaphylaxis than newer low-osmolality media. Severe reactions are rare with either type of contrast media. About one-third of those with mild reactions to previous exposure rereact on reexposure. In most cases, these reactions are probably not immunologic. Pretreatment with prednisone and diphenhydramine reduces reaction rates. Persons with a reaction to a high-osmolality contrast media should be given low-osmolality media if later contrast studies are required. ANTICONVULSANTS Of the anticonvulsants, the single orally administered agent with the highest risk of severe adverse cutaneous reactions is the antiseizure medicine lamotrigine. Older anticonvulsants, including phenytoin and carbamazepine, are also associated with many types of severe reactions and a high incidence of less severe reactions, particularly in children. In addition toSJS,TEN, and the hypersensitivity syndrome discussed above, the aromatic anticonvulsants can induce a pseudolymphoma syndrome and induce gingival hyperplasia. SULFONAMIDES Sulfonamides have perhaps the highest risk of causing cutaneous eruptions and are the drugs most frequently implicated inSJS andTEN. The combination of sulfamethoxazole and trimethoprim frequently induces adverse cutaneous reactions in patients with AIDS (Chap. 309). Desensitization is often successful in AIDS patients with morbilliform eruptions but is a high-risk procedure in AIDS patients who manifest erythroderma, fever, or a bullous reaction in response to their earlier sulfonamide exposure. AGENTS USED IN CANCER CHEMOTHERAPY Since many agents used in cancer chemotherapy inhibit cell division, rapidly proliferating elements of the skin, including hair, mucous membranes, and appendages, are sensitive to their effects; as a result, stomatitis and alopecia are among the most frequent dose-dependent side effects of chemotherapy. Onychodystrophy (dystrophic changes in nails) is also seen with bleomycin, hydroxyurea (hydroxycarbamide), and 5-fluorouracil. Sterile cellulitis and phlebitis and ulceration of pressure areas occur with many of these agents. Urticaria, angioedema, exfoliative dermatitis, and erythema of the palms and soles have also been seen, as has local and diffuse hyperpigmentation.
GLUCOCORTICOIDS Both systemic and topical glucocorticoids cause a variety of skin changes, including acneiform eruptions, atrophy, striae, and other stigmata of Cushing's syndrome, and in sufficiently high doses can retard wound healing. Patients using glucocorticoids are at higher risk for bacterial, yeast, and fungal skin infections that may be misinterpreted as drug eruptions but are instead drug side effects. CYTOKINE THERAPY Alopecia is a common complication ofIFN-a. Induction or exacerbation of various immune-mediated disorders (psoriasis, lichen planus, lupus erythematosus) has been also reported with this agent. IFN-b injection has been associated with local necrosis of the skin. Granulocyte colony stimulating factor may induce various neutrophilic dermatosis, including Sweet's syndrome, pyoderma gangrenosum, neutrophilic eccrine hidradenitis, and vasculitis, and can exacerbate psoriasis. IL-2 is associated with frequent cutaneous reactions including exanthema, facial edema, xerosis, and pruritus. Cases of pemphigus vulgaris, linear IgA disease, psoriasis, and vitiligo have also been described in association with this drug. ANTIMALARIAL AGENTS Antimalarial agents are used as therapy for several skin diseases, including the skin manifestations of lupus and polymorphous light eruption, but they can also induce cutaneous reactions. Although also used to treat porphyria cutanea tarda at low doses, in patients with asymptomatic porphyria cutanea tarda, higher doses of chloroquine increase porphyrin levels to such an extent that they may exacerbate the disease. Pigmentation disturbances, including black pigmentation of the face, mucous membranes, and pretibial and subungual areas, occur with antimalarials. Quinacrine (mepacrine) causes generalized, cutaneous yellow discoloration. GOLD Chrysotherapy has been associated with a variety of dose-related dermatologic reactions (including maculopapular eruptions), which can develop as long as 2 years after initiation of therapy and require months to resolve. Erythema nodosum, psoriasiform dermatitis, vaginal pruritus, eruptions similar to those of pityriasis rosea, hyperpigmentation, and lichenoid eruptions resembling those seen with antimalarial agents have been reported. After a cutaneous reaction, it is sometimes possible to reinstitute gold therapy at lower doses without recurrence of the dermatitis. DIAGNOSIS OF DRUG REACTIONS Possible causes of an adverse reaction can be assessed as definite, probable, possible, or unlikely based on six variables: (1) previous experience with the drug in the general population, (2) alternative etiologic candidates, (3) timing of events, (4) drug levels or
evidence of overdose, (5) patient reaction to drug discontinuation, and (6) patient reaction to rechallenge. PREVIOUS EXPERIENCE Tables of relative reaction rates are available and are useful to assess the likelihood that a given drug is responsible for a given cutaneous reaction. The specific morphologic pattern of a drug reaction, however, may modify these reaction rates by increasing or decreasing the likelihood that a given drug is responsible for a given reaction. For example, since fixed eruptions due to drugs are more often seen with barbiturates than with penicillin, a fixed drug reaction in a patient taking both types of agents is more likely to be due to the barbiturate, even though penicillins have a higher overall drug reaction rate. ALTERNATIVE ETIOLOGIC CANDIDATES A cutaneous eruption may be due to exacerbation of preexisting disease or to development of new disease unrelated to drugs. For example, a patient with psoriasis may have a flare-up of disease coincidental with administration of penicillin for streptococcal infection; in this case, infection is a more likely cause for the flare-up than drug reaction. TIMING OF EVENTS Most drug reactions of the skin occur within 1 to 2 weeks of initiation of therapy. Hypersensitivity syndrome may occur later (up to 8 weeks) after initiating drug therapy. Fixed drug reactions and generalized exanthematous pustulosis often occur earlier (within 48 h), as do reactions of all types in persons with prior sensitization to that drug or a cross-sensitizing agent. DRUG LEVELS Some cutaneous reactions are dependent on dosage or cumulative toxicity. For example, lichenoid dermatoses due to gold administration appear more often in patients taking high doses. DISCONTINUATION Most adverse cutaneous reactions to drugs remit with discontinuation of the suspected agent. A reaction is considered unlikely to be drug-related if improvement occurs while the drug is continued or if a patient fails to improve after stopping the drug and appropriate therapy. RECHALLENGE Rechallenge provides the most definitive information concerning adverse cutaneous reactions to drugs, since a reaction failing to recur on rechallenge with a drug is unlikely to be due to that agent. Rechallenge is usually impractical, however, because the need to ensure patient safety and comfort outweighs the value of the possible information
derived from rechallenge. Of special importance is the rapid recognition of reactions that may become serious or life-threatening.Table 59-2 lists clinical and laboratory features that, if present, suggest the reaction may be serious.Table 59-3 provides key features of the most serious adverse cutaneous reactions. DIAGNOSIS OF DRUG ALLERGY Tests for IgE responses include in vivo and in vitro methods, but such tests are available for only a limited number of drugs, including penicillins and cephalosporins, some peptide and protein drugs (insulin, xenogeneic sera), and some agents used for general anesthesia. In vivo testing is accomplished by prick puncture and/or by intradermal skin testing. A wheal-and-flare response 2´ 2 mm greater than that seen with a saline control within 20 min is considered indicative of IgE-mediated mast cell degranulation, provided (1) the patient is not dermographic, (2) the drug does not nonspecifically degranulate mast cells, (3) the drug concentration is not high enough to be irritating, and (4) the buffer itself does not cause wheal-and-flare responses. Skin testing with major and minor determinants of penicillins or cephalosporins has proved useful for identifying patients at risk of anaphylactic reactions to these agents. However, skin tests themselves carry a small risk of anaphylaxis. Negative skin tests do not rule out IgE-mediated reactivity, and the risk of anaphylaxis in response to penicillin administration in patients with negative skin tests is about 1%; about two-thirds of patients with a positive skin test and history of a previous adverse reaction to penicillin experience an allergic response on rechallenge. Skin tests may be negative in allergic patients receiving antihistamines or in those whose allergy is to determinants not present in the test reagent. Although less well studied, similar techniques can identify patients who are sensitive to protein drugs and to agents such as gallamine and succinylcholine. Most other drugs are small molecules, and skin testing with them is unreliable. There are no generally available and reliable tests for assessing causality of non-IgE-mediated reactions, except possibly patch tests for assessment of fixed drug reactions. Therefore, diagnosis usually relies on clinical factors rather than test results. (Bibliography omitted in Palm version) Back to Table of Contents
60. PHOTOSENSITIVITY AND OTHER REACTIONS TO LIGHT - David R. Bickers SOLAR RADIATION Sunlight is the most visible and obvious source of comfort in the environment. This natural proclivity for the sun has the beneficial results of warmth and vitamin D synthesis but also can produce pathologic consequences. Few effects of sun exposure beyond those affecting the skin have been identified, but cutaneous exposure to sunlight can evoke immunosuppressive responses and genetic changes that may be relevant to the pathogenesis of nonmelanoma skin cancer and perhaps infections such as herpes simplex. The sun's energy encompasses a broad range from ultrashort highly energetic ionizing radiation (10-2um) to ultralong radiowaves of very low photon energy (107um). Thus, the emission spectrum ranges over nine orders of magnitude, but that reaching the earth's surface is narrow and is limited to components of the ultraviolet (UV), visible light, and portions of the infrared. The cutoff at the short end of the UV is at approximately 290 nm, because stratospheric ozone is formed by ionizing radiation of wavelengths less than 100 nm and absorbs solar energy between 120 and 310 nm, thereby preventing penetration to the earth's surface of the shorter, more energetic, potentially more harmful wavelengths of solar radiation. Indeed, concern about destruction of the ozone layer by chlorofluorocarbons released into the atmosphere has led to international agreements to reduce production of these chemicals. Measurements of solar flux indicate that there is a twentyfold regional variation in the amount of energy at 300 nm that reaches the earth's surface. This variability relates to seasonal effects, the path of sunlight transmission through ozone and air, the altitude (4% increase for each 300 m of elevation), the latitude (increasing intensity with decreasing latitude), and the amount of cloud cover, fog, and pollution. The major components of the photobiologic action spectrum include theUV and visible wavelengths between 290 and 700 nm. In addition, the wavelengths beyond 700 nm in the infrared primarily evoke heat, but warming of the skin may enhance biologic responses to wavelengths in the UV and visible spectrum. TheUVspectrum is arbitrarily divided into three major segments: C, B, and A. This includes the wavelengths between 10 and 400 nm. Ultraviolet C (UV-C) consists of wavelengths between 10 and 290 nm and does not reach the earth because of its absorption by stratospheric ozone. These wavelengths are not a cause of photosensitivity except in occupational settings where artificial sources of this energy are employed -- e.g., for germicidal effects. Ultraviolet B (UV-B) consists of wavelengths between 290 and 320 nm. This portion of the photobiologic action spectrum is the most efficient in producing redness or erythema in human skin and hence is sometimes known as the "sunburn spectrum." Ultraviolet A (UV-A) represents those wavelengths between 320 and 400 nm and is approximately 1000-fold less efficient in producing skin hyperemia than is UV-B. The UV-A has also been divided into two parts known as UV-A 1 (340 to 400 nm) and UV-A 2 (320 to 340 nm). The visible wavelengths between 400 and 700 nm include the familiar white light which
when directed through a prism can be shown to consist of various colors including violet, indigo, blue, green, yellow, orange, and red. The energy possessed by photons in the visible spectrum is not capable of damaging human skin in the absence of a photosensitizing chemical. The absorption of energy is critical to the development of photosensitivity. Thus the absorption spectrum of a molecule is defined as the range of wavelengths absorbed by it, whereas the action spectrum for an effect of incident radiation is defined as the range of wavelengths that evoke the response. Photosensitivity occurs when a photon-absorbing chemical (chromophore) present in the skin absorbs incident energy, becomes excited, and transfers the absorbed energy to various structures or to oxygen. The absorbed energy must be dissipated by processes including heat, fluorescence, and phosphorescence. It is important to emphasize that absorption spectra and action spectra need not be superimposable, but there must be overlap at some point to produce photosensitization. STRUCTURE AND FUNCTION OF SKIN The skin's exposure to sunlight permits the absorption of some wavelengths and the transmission of others. Essentially, human skin is a sandwich of two distinctive compartments, the epidermis and dermis, separated by a basement membrane. The outer epidermis is a stratified squamous epithelium comprising the surface stratum corneum (a protein- and lipid-rich compact membrane), the stratum granulosum, stratum spinosum, and the basal cell layer. The basal cell layer contains a heterogeneous population of cells, a subset of which migrate upward in the process of terminal differentiation that results in the expression of specific keratin genes and the formation of the stratum corneum. Epidermal cells include resident keratinocytes and melanocytes and immigrant cells, including the immunologically active Langerhans cells, lymphocytes, polymorphonuclear leukocytes, monocytes, and macrophages, making the epidermis a major component of the immune system. Branches of sensory nerve endings also reach into this compartment. The second major component of skin is the dermis, which is relatively large and less densely populated with cells that include fibroblasts, endothelial cells within dermal vessels, and mast cells. Tissue macrophages and sparsely distributed inflammatory cells are also present. All these cells exist within an extracellular matrix of collagen, elastin, and glycosaminoglycans. In contrast to the epidermis, rich vascularization of the dermis allows it to play an important role in temperature regulation and in inflammatory responses to skin injury. UV RADIATION (UVR) AND SKIN The epidermis and the dermis contain several chromophores capable of interacting with incident solar energy. These interactions include reflection, refraction, absorption, and transmission. The stratum corneum is a major impediment to the transmission ofUV-B, and less than 10% of incident wavelengths in this region penetrate the basement membrane. Approximately 3% of radiation below 300 nm, 20% of radiation below 360 nm, and 33% of short visible radiation reaches the basal cell layer in untanned human skin. Proteins and nucleic acids absorb intensely in the short UV-B. In contrast, UV-A 1 and 2 penetrate the epidermis efficiently to reach the dermis, where they likely produce
changes in structural and matrix proteins that contribute to the aged appearance of chronically sun-exposed skin, particularly in individuals of light complexion. One of the consequences ofUV-B absorption by DNA is the production of pyrimidine dimers. These structural changes can be repaired by mechanisms that result in their recognition and excision, and the reestablishment of normal base sequences. The efficient repair of these structural aberrations is crucial, since individuals with defective DNA repair are at high risk for the development of cutaneous cancer. For example, patients with xeroderma pigmentosum, an autosomal recessive disorder, are characterized by variably decreased repair of UV-induced photoproducts, and their skin may develop the xerotic appearance of photoaging as well as basal cell and squamous cell carcinomas and melanoma in the first two decades of life. Studies in mice using knockout gene technology have verified the importance of genes regulating these repair pathways in preventing the development of UV-induced cancer. Cutaneous Optics and Chromophores Chromophores are endogenous or exogenous chemical components that can absorb physical energy. Endogenous chromophores of skin are of two types: (1) chemicals that are normally present, including nucleic acids, proteins, lipids, and 7-dehydrocholesterol, the precursor of vitamin D; and (2) chemicals, such as porphyrins, synthesized elsewhere in the body that circulate in the bloodstream and diffuse into the skin. Normally, only trace amounts of porphyrins are present in the skin, but in the diseases known as the porphyrias, increased amounts are released into the circulation and are transported to the skin, where they absorb incident energy both in the Soret band around 400 nm (short visible) and to a lesser extent in the red portion of the visible spectrum (580 to 660 nm). This results in structural damage to the skin that may be manifest as erythema, edema, urticaria, or blister formation (Chap. 346). Acute Effects of Sun Exposure The immediate cutaneous consequences of sun exposure include sunburn and vitamin D synthesis. Sunburn This very common affliction of human skin is caused by exposure toUVR. Generally speaking, the individual's ability to tolerate sunlight is inversely proportional to his or her melanin pigmentation. Melanin is a complex polymer of tyrosine that functions as an efficient neutral-density filter with broad absorbance within theUVportion of the solar spectrum. Melanin is synthesized in specialized epidermal dendritic cells termed melanocytes and is packaged into melanosomes that are transferred via dendritic processes into keratinocytes, where they provide photoprotection. Sun-induced melanogenesis is a consequence of increased tyrosinase activity in melanocytes that in turn may be due to a combination of eicosanoid and endothelin-1 release. Tolerance of sun exposure is a function of the efficiency of the epidermal-melanin unit and can usually be ascertained by asking an individual two questions: (1) Do you burn after sun exposure? and (2) Do you tan after sun exposure? By the answers to these questions, it is usually possible to divide the population into six skin types varying from type I (always burn, never tan) to type VI (never burn, always tan) (Table 60-1). There are two general theories about the pathogenesis of the sunburn response. First, the lag phase in time between skin exposure and the development of visible redness (usually 4 to 12 h) suggests an epidermal chromophore that causes delayed production and/or release of vasoactive mediator(s), or cytokines, that diffuse to the dermal
vasculature to evoke vasodilatation. Indeed,UVRstimulates the release of numerous proinflammatory cytokines and nitric oxide by keratinocytes. Second, it is possible that the small amount of incidentUV-B radiation (10% or less) that penetrates to the dermis can be absorbed directly by endothelial cells in the vasculature, thereby resulting in vasodilatation. The issue remains unresolved. The action spectrum for sunburn erythema includes theUV-B and UV-A regions. Photons in the shorter UV-B are at least 1000-fold more efficient than photons in the longer UV-B and the UV-A in evoking the response. However, UV-A may contribute to sunburn erythema at midday when much more UV-A than UV-B is present. The mechanism of injury remains poorly defined, but the action spectrum forUV-B erythema closely resembles the absorption spectrum for DNA after adjusting for the absorbance of incident energy by the stratum corneum. Apoptotic keratinocytes (so-called sunburn cells) are visible histologically within an hour of exposure and are maximal within 24 h. UV-A is less effective than UV-B in producing sunburn cells. Mast cells may release inflammatory mediators after exposure to UV-B and UV-A. For example, erythema doses of both UV-B and UV-A increase histamine levels in experimentally induced suction blisters of human skin that return to normal after 24 h (before visible erythema has subsided). Prostaglandin E2increases to approximately 150% of control levels after 24 h and then diminishes. Since prostaglandins evoke both pain and redness when injected intradermally, their presence in suction blisters after UV-B exposure suggests a role in UV-B erythema. Age-related declines occur in the amount of inflammatory mediators detectable in human skin after UV-B irradiation. UV-A erythema results in few epidermal sunburn cells, but vascular endothelial injury is greater than with UV-B. In addition, there are increased levels of arachidonic acid and of prostaglandins D2, E2, and I2 that peak within 5 to 9 h and then subside before peak redness occurs. Despite evidence for the role of prostaglandins in both UV-B- and UV-A-irradiated skin, administration of nonsteroidal anti-inflammatory drugs is more effective in reducing erythema evoked by UV-B than by UV-A. UV-B also induces cutaneous matrix-degrading metalloproteinases within hours of exposure. Vitamin D Photochemistry Cutaneous exposure toUV-B causes photolysis of epidermal previtamin D3(7-dehydrocholesterol) to previtamin D3, which then undergoes a temperature-dependent isomerization to form the stable hormone vitamin D 3. This compound then diffuses to the dermal vasculature and circulates systemically where it is converted to the functional hormone 1,25-dihydroxy vitamin D3[1,25(OH)2D3]. Vitamin D metabolites from the circulation or those produced in the skin itself can augment epidermal differentiation signaling. Aging substantially decreases the ability of human skin to produce vitamin D3. This, coupled with the widespread use of sunscreens that filter outUV-B, has led to concern that vitamin D deficiency may become a significant clinical problem in the elderly. Indeed, studies have shown that the use of sunscreens can diminish the production of vitamin D3 in human skin. Chronic Effects of Sun Exposure: Nonmalignant The clinical features of photodamaged sun-exposed skin consist of wrinkling, blotchiness, telangiectasia, and a roughened, irregular, "weather-beaten" appearance. Whether these changes, which some refer to as photoaging or dermatoheliosis, represent accelerated chronologic aging or a separate and distinct process is not clear.
Within chronically sun-exposed epidermis, there is thickening (acanthosis) and morphologic heterogeneity within the basal cell layer. Higher but irregular melanosome content may be present in some keratinocytes, indicating prolonged residence of the cells in the basal cell layer. These structural changes may help to explain the leathery texture and the blotchy discoloration of sun-damaged skin. The dermis is the major site for sun-associated chronic damage, manifest as a massive increase in thickened irregular masses of tangled elastic fibers resulting from enhanced expression of elastin genes. Collagen fibers are also abnormally clumped in the deeper dermis. Fibroblasts are increased in number and show morphologic signs suggesting activation. Degraded mast cells may be present in the dermis, the relevance of which remains unclear. These morphologic changes, both gross and microscopic, are features of chronically sun-exposed skin. The chromophore(s), the action spectra, and the specific biochemical events orchestrating these changes are unknown. Chronic Effects of Sun Exposure: Malignant One of the major known consequences of chronic skin exposure to sunlight is nonmelanoma skin cancer. The two types of nonmelanoma skin cancer are basal cell and squamous cell carcinoma (Chap. 86). There are three major steps for cancer induction: initiation, promotion, and progression. Chronic exposure of animal skin to artificial light sources that mimic solarUVRresults in initiation, a step whereby structural (mutagenic) changes in DNA evoke an irreversible alteration in the target cell (keratinocyte) that begins the tumorigenic process. Exposure to a tumor initiator is believed to be a necessary but not sufficient step in the malignant process, since initiated skin cells not exposed to tumor promoters do not generally develop tumors. The second stage in tumor development is promotion, a multistep process whereby initiated cells are exposed to chemical and physical agents that evoke epigenetic changes that culminate in the clonal expansion of initiated cells and cause the development, over a period of weeks to months, of benign growths known as papillomas. Again, using transgenic animals, the importance ofUVeffects on the expression of additional oncogenes such as fos and jun in developing papillomas has been demonstrated. UV-B is a complete carcinogen, meaning that it can function as both an initiator and a promoter, leading to tumor induction. Incomplete carcinogens can initiate tumorigenesis but require additional skin exposure to tumor promoters to elicit tumors. The prototype tumor promoter is the phorbol ester 12-O-tetradecanoyl phorbol-13-acetate. Tumor promotion usually requires multiple exposures over time to evoke a neoplasm. The final step in the malignant process is the conversion of benign precursors into malignant lesions, a process thought to require additional genetic alterations in already transformed cells. Indeed, ras gene mutations have been detected in a minority of human nonmelanoma skin cancers. Mutations of the tumor suppressor gene p53 also occur in sun-damaged human skin. Sun exposure causes nonmelanoma and melanoma cancers of the skin, although the evidence is far more direct for its role in nonmelanoma (basal cell and squamous cell carcinoma) than in melanoma. Approximately 80% of nonmelanoma skin cancers
develop on exposed body area, including the face, the neck, and the hands. Men of fair complexion who work outdoors are twice as likely as women to develop these types of cancers. Whites of darker complexions (e.g., Hispanics) have one-tenth the risk of developing such cancers as do light-skinned individuals. Blacks are at lowest risk for all forms of skin cancer. Between 600,000 and 800,000 individuals in the United States develop nonmelanoma skin cancer annually, and the lifetime risk for a white individual to develop such a neoplasm is estimated at approximately 15%. A consensus exists that the incidence of nonmelanoma skin cancer in the population is rising, for reasons that are unclear. The relationship of sun exposure to melanoma is less clear-cut, but suggestive evidence supports an association. Melanomas occasionally develop by the teenage years, indicating that the latent period for tumor growth is less than that of nonmelanoma skin cancer. Melanomas are among the most rapidly increasing of all human malignancies (Chap. 86). Epidemiologic studies of immigrants of similar ethnic stock indicate that individuals born in one area or who migrated to the same locale before age 10 have higher age-specific melanoma rates than individuals arriving later. It is thus reasonable to conclude that life in a sunny climate from birth or early childhood increases the risk of melanoma. In general, risk does not correlate with cumulative sun exposure but may relate to sequelae of sun exposure in childhood. Thus, a blistering sunburn is associated with a doubling of melanoma risk at the site of the reaction. Immunologic Effects Exposure to solar radiation influences both local and systemic immune responses.UV-B appears to be most efficient in altering immune responses, likely related to the capacity of such energy to affect antigen presentation in skin by interacting with epidermal Langerhans cells. These bone marrow-derived dendritic cells possess surface markers characteristic of monocytes and macrophages. Following skin exposure to erythema doses of UV-B, Langerhans cells undergo both morphologic and functional changes that result in decreased contact allergic responses when haptens are applied to the irradiated site. This diminished capacity for sensitization is due to the induction of antigen-specific suppressor T lymphocytes. Indeed, while the immunosuppressive effect of irradiation is limited to haptens applied to the irradiated site, the net result is systemic immune suppression to that antigen because of the induction of suppressor T cells. Higher doses of radiation evoke diminished immunologic responses to antigens introduced either epicutaneously or intracutaneously at sites distant from the irradiated site. These suppressed responses are also associated with the induction of antigen-specific suppressor T lymphocytes and may be mediated by as yet undefined factors that are released from epidermal cells at the irradiated site. The implications of this generalized immune suppression in terms of altered susceptibility to cutaneous cancer or to infection remain to be defined. It is known thatUV-induced tumors in murine skin are antigenic and are rapidly rejected when transplanted into normal syngeneic animals. If the tumors are transplanted into animals previously exposed to subcarcinogenic doses of UV-B, they are not rejected and instead grow progressively in the recipients. This failure of irradiated animals to reject the transplanted tumors is due to the development of T suppressor cells that prevent the rejection response. While the mechanism of suppression of tumor rejection
is unknown, such a response might be a critical determinant of cancer risk in human skin. PHOTOSENSITIVITY DISEASES The diagnosis of photosensitivity requires a careful history to define the duration of the signs and symptoms, the length of time between exposure to sunlight and the development of subjective complaints, and visible changes in the skin. The age of onset also can be a helpful clue; for example, the acute photosensitivity of erythropoietic protoporphyria almost always begins in childhood, whereas the chronic photosensitivity of porphyria cutanea tarda typically begins in the fourth and fifth decades. A history of exposure to topical and systemic drugs and chemicals may provide important information. Many classes of drugs can cause photosensitivity on the basis of either phototoxicity or photoallergy. Fragrances such as musk ambrette that were previously present in numerous cosmetic products are also potent photosensitizers. Examination of the skin may also offer important clues. Anatomic areas that are naturally protected from direct sunlight such as the hairy scalp, the upper eyelids, the retroauricular areas, and the infranasal and submental regions may be spared, whereas exposed areas show characteristic features of the pathologic process. These anatomic localization patterns are often helpful but not infallible in making the diagnosis. For example, airborne contact sensitizers that are blown onto the skin may produce dermatitis that can be difficult to distinguish from photosensitivity, despite the fact that such material may trigger skin reactivity in areas shielded from direct sunlight. Many dermatologic conditions may be caused or aggravated by light (Table 60-2). The role of light in evoking these responses may be dependent on genetic abnormalities ranging from well-described defects in DNA repair that occur in xeroderma pigmentosum to the inherited abnormalities in heme synthesis that characterize the porphyrias. In certain photosensitivity diseases, the chromophore has been identified, whereas in the majority, the energy-absorbing agent is unknown. Polymorphous Light Eruption After sunburn, the most common type of photosensitivity disease is polymorphous light eruption, the mechanism of which is unknown. Many affected individuals never seek medical attention because the condition is often transient, becoming manifest each spring with initial sun exposure but then subsiding spontaneously with continuing exposure, a phenomenon known as "hardening." The major manifestations of polymorphous light eruption include pruritic (often intensely so) erythematous papules that may coalesce into plaques on exposed areas of the face and arms or other areas as well, making the distribution spotty and uneven. The diagnosis can be confirmed by skin biopsy and by performing phototest procedures in which skin is exposed to multiple erythema doses ofUV-A and UV-B. The action spectrum for polymorphous light eruption is usually within these portions of the solar spectrum. Treatment of this disease includes the induction of hardening by the cautious administration ofUVlight, either alone or in combination with photosensitizers such as
the psoralens (see below). Phototoxicity and Photoallergy These photosensitivity disorders are related to the topical or systemic administration of drugs and other chemicals. Both reactions require the absorption of energy by a drug or chemical resulting in the production of an excited-state photosensitizer that can transfer its absorbed energy to a bystander molecule or to molecular oxygen, thereby generating tissue-destructive chemical species. Phototoxicity is a nonimmunologic reaction caused by drugs and chemicals, a few of which are listed in Table 60-3. The usual clinical manifestations include erythema resembling a sunburn that quickly desquamates or "peels" within several days. In addition, edema, vesicles, and bullae may occur. Photoallergy is distinct in that the immune system participates in the pathologic process. The excited-state photosensitizer may create highly unstable haptenic free radicals that bind covalently to macromolecules to form a functional antigen capable of evoking a delayed hypersensitivity response. Some of the drugs and chemicals that produce photoallergy are listed in Table 60-4. The clinical manifestations typically differ from those of phototoxicity in that an intensely pruritic eczematous dermatitis tends to predominate and evolves into lichenified, thickened, "leathery" changes in sun-exposed areas. A small subset (perhaps 5 to 10%) of patients with photoallergy may develop a persistent exquisite hypersensitivity to light even when the offending drug or chemical is identified and eliminated. Known as persistent light reaction, this may be incapacitating for years. Some have used the term chronic actinic dermatitis to encompass these chronic hyperresponsive states. Diagnostic confirmation of phototoxicity and photoallergy often can be obtained using phototest procedures. In patients with suspected phototoxicity, determination of the minimal erythema dose (MED) while the patient is exposed to a suspected agent and then repeating the MED after discontinuation of the agent may provide a clue to the causative drug or chemical. Photopatch testing can be performed to confirm the diagnosis of photoallergy. This is a simple variant of ordinary patch testing in which a series of known photoallergens is applied to the skin in duplicate and one set is irradiated with a suberythema dose ofUV-A. Development of eczematous changes at sites exposed to sensitizer and light is a positive result. The characteristic abnormality in patients with persistent light reaction is a diminished threshold to erythema evoked by UV-B. Patients with chronic actinic dermatitis may have a broad spectrum of UV hyperresponsiveness. The management of drug photosensitivity is first and foremost to eliminate exposure to the chemical agents responsible for the reaction and to minimize sun exposure. The acute symptoms of phototoxicity may be ameliorated by cool, moist compresses, topical glucocorticoids, and systemically administered nonsteroidal antiinflammatory agents. In severely affected individuals, a rapidly tapered course of systemic glucocorticoids may be useful. Judicious use of analgesics may be necessary. Photoallergic reactions require a similar management approach. Furthermore, individuals suffering from persistent light reactivity must be meticulously protected
against light exposure. In selected patients in whom chronic systemic high-dose glucocorticoids pose unacceptable risks, it may be necessary to employ cytotoxic agents such as azathioprine or cyclophosphamide. Porphyria The porphyrias (Chap. 346) are a group of diseases that have in common various derangements in the synthesis of heme. Heme is an iron-chelated tetrapyrrole or porphyrin, and the nonmetal chelated porphyrins are potent photosensitizers that absorb light intensely in both the short (400 to 410 nm) and the long (580 to 650 nm) portions of the visible spectrum. Heme cannot be reutilized and must be continuously synthesized, and the two body compartments with the largest capacity for its production are the bone marrow and the liver. Accordingly, the porphyrias originate in one or the other of these organs, with the end result of excessive endogenous production of potent photosensitizing porphyrins. The porphyrins circulate in the bloodstream and diffuse into the skin, where they absorb solar energy, become photoexcited, and evoke cutaneous photosensitivity. The mechanism of porphyrin photosensitization is known to be photodynamic or oxygen-dependent and is mediated by reactive oxygen species such as superoxide anions. Porphyria cutanea tarda is the most common type of human porphyria and is associated with decreased activity of the enzyme uroporphyrinogen decarboxylase associated with a number of gene mutations. There are two basic types of porphyria cutanea tarda: the sporadic or acquired type, generally seen in individuals ingesting ethanol or receiving estrogens; and the inherited type, in which there is autosomal dominant transmission of deficient enzyme activity. Both forms are associated with increased hepatic iron stores. In both types of porphyria cutanea tarda, the predominant feature is a chronic photosensitivity characterized by increased fragility of sun-exposed skin, particularly areas subject to repeated trauma such as the dorsa of the hands, the forearms, the face, and the ears. The predominant skin lesions are vesicles and bullae that rupture, producing moist erosions, often with a hemorrhagic base, that heal slowly with crusting and purplish discoloration of the affected skin. Hypertrichosis, mottled pigmentary change, and scleroderma-like induration are associated features. Biochemical confirmation of the diagnosis can be obtained by measurement of urinary porphyrin excretion, plasma porphyrin assay, and by assay of erythrocyte and/or hepatic uroporphyrinogen decarboxylase. Multiple mutations of the uroporphyrinogen decarboxylase gene have been identified in human populations, including exon skipping and base substitutions. Treatment consists of repeated phlebotomies to diminish the excessive hepatic iron stores and/or intermittent low doses of the antimalarial drugs chloroquine and hydroxychloroquine. Long-term remission of the disease can be achieved if the patient eliminates exposure to porphyrinogenic agents. Erythropoietic protoporphyria originates in the bone marrow and is due to a decrease in the mitochondrial enzyme ferrochelatase secondary to numerous gene mutations. The major clinical features include an acute photosensitivity characterized by subjective burning and stinging of exposed skin that often develops during or just after exposure.
There may be associated skin swelling and, after repeated episodes, a waxlike scarring. The diagnosis is confirmed by demonstration of measurement of free elevated erythrocyte protoporphyrin. Detection of increased plasma protoporphyrin helps to differentiate lead poisoning and iron-deficiency anemia, in both of which elevated erythrocyte protoporphyrin occurs in the absence of cutaneous photosensitivity and of elevated plasma protoporphyrin. Treatment consists of reducing sun exposure and the oral administration of the carotenoidb-carotene, which is an effective scavenger of free radicals. This drug increases tolerance to sun exposure in many affected individuals, although it has no effect on deficient ferrochelatase. An algorithm for the approach to a patient with photosensitivity is illustrated inFig. 60-1. PHOTOPROTECTION Since photosensitivity of the skin results from exposure to sunlight, it follows that avoidance of the sun would eliminate these disorders. Unfortunately, social pressures make this an impractical alternative for most individuals, and this has led to a search for better approaches to photoprotection. Natural photoprotection is provided by structural proteins in the epidermis, particularly keratins and melanin. The amount of melanin and its distribution in cells is genetically regulated, and individuals of darker complexion (skin types IV to VI) are at decreased risk for the development of cutaneous malignancy. Other forms of photoprotection include clothing and sunscreens. Clothing constructed of tightly woven sun-protective fabrics, irrespective of color, affords substantial protection. Wide-brimmed hats, long sleeves, and trousers all reduce direct exposure. Sunscreens are of two major types -- chemical and physical. Chemical sunscreens are chromophores that absorb energy in theUV-B and/or UV-A regions, thereby diminishing photon absorption by the skin (Table 60-5). Sunscreens are rated for their photoprotective effect by their sun protective factor (SPF). The SPF is simply a ratio of the time required to produce sunburn erythema with and without sunscreen application. SPF ratings of 15 or higher provide effective protection against UV-B and, to a lesser extent, UV-A. The major categories of chemical sunscreens include p-aminobenzoic acid and its esters, benzophenones, anthranilates, cinnamates, and salicylates. Physical sunscreens are light-opaque mixtures containing metal particles such as titanium oxide and zinc oxide that scatter light, thereby reducing photon absorption by the skin. In addition to light absorption, a critical determinant of the photoprotective effect of sunscreens is their ability to remain on the skin, a property known as substantivity. In general, the p-aminobenzoic acid esters formulated in moisturizing vehicles provide the greatest substantivity. Photoprotection can also be achieved by limiting the time of exposure during the day. Since the majority of an individual's total lifetime sun exposure may occur by the age of
18, it is important to educate parents and young children about the hazards of sunlight. Simply eliminating exposure at midday will substantially reduce lifetimeUV-B exposure. PHOTOTHERAPY AND PHOTOCHEMOTHERAPY UVRcan also be used therapeutically. The administration ofUV-B alone or in combination with topically applied agents can induce remissions of psoriasis and atopic dermatitis. Photochemotherapy in which topically applied or systemically administered psoralens are combined withUV-A (PUVA) is also effective in treating psoriasis and in the early stages of cutaneous T cell lymphoma and vitiligo. Psoralens are tricyclic furocoumarins that, when intercalated into DNA and exposed to UV-A, form adducts with pyrimidine bases and eventually form DNA cross-links. These structural changes are thought to decrease DNA synthesis and relate to improvement that occurs in psoriasis. The reason that PUVA photochemotherapy is effective in cutaneous T cell lymphoma is not clear. In addition to its effects on DNA, PUVA photochemotherapy also stimulates melanin synthesis, and this provides the rationale for its use in the depigmenting disease vitiligo. Oral 8-methoxypsoralen andUV-A appear to be most effective in this regard, but as many as 100 treatments extending over 12 to 18 months may be required to promote satisfactory repigmentation. The major side effects ofUV-B phototherapy and PUVA photochemotherapy are due to the cumulative effects of photon absorption and include skin dryness, actinic keratoses, and an increased risk of nonmelanoma skin cancer. Despite these risks, the therapeutic index of these modalities is quite acceptable. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 10 -HEMATOLOGIC ALTERATIONS 61. ANEMIA AND POLYCYTHEMIA - John W. Adamson, Dan L. Longo HEMATOPOIESIS AND THE PHYSIOLOGIC BASIS OF RED CELL PRODUCTION Hematopoiesis is the process by which the formed elements of the blood are produced. The process is regulated through a series of steps beginning with the pluripotent hematopoietic stem cell. Stem cells are capable of producing red cells, all classes of granulocytes, monocytes, platelets, and the cells of the immune system. Commitment of the stem cell to the specific cell lineages appears not to be regulated by known exogenous growth factors or cytokines. Rather, stem cells develop into differentiated cell types through incompletely defined molecular events that are intrinsic to the stem cell itself (Chap. 104). Following lineage commitment (or differentiation), hematopoietic progenitor and precursor cells come increasingly under the regulatory influence of growth factors and hormones, such as erythropoietin (EPO) for red cell production. EPO is required for the maintenance of committed erythroid progenitor cells which, in the absence of the hormone, undergo programmed cell death (apoptosis). The regulated process of red cell production is erythropoiesis, and its key elements are illustrated in Fig. 61-1. In the bone marrow, the first morphologically recognizable erythroid precursor is the pronormoblast. This cell can undergo 4 to 5 cell divisions that result in the production of 16 to 32 mature red cells. With increasedEPOproduction, or the administration of EPO as a drug, early progenitor cell numbers are amplified and, in turn, give rise to increased numbers of erythrocytes. The regulation of EPO production itself is linked to O2transport. In mammals, O2 is transported to tissues bound to the hemoglobin contained within circulating red cells. The mature red cell is 8um in diameter, anucleate, discoid in shape, and extremely pliable in order for it to traverse the microcirculation successfully; its membrane integrity is maintained by the intracellular generation of ATP. Normal red cell production results in the daily replacement of 0.8 to 1% of all circulating red cells in the body. The average red cell lives 100 to 120 days. The machinery responsible for red cell production is called the erythron. The erythron is a dynamic organ made up of a rapidly proliferating pool of marrow erythroid precursor cells and a large mass of mature circulating red blood cells. The size of the red cell mass reflects the balance of red cell production and destruction. The physiologic basis of red cell production and destruction provides an understanding of the mechanisms that can lead to anemia. The physiologic regulator of red cell production, the glycoprotein hormoneEPO, is produced and released by peritubular capillary lining cells within the kidney. These cells are highly specialized epithelial-like cells. A small amount of EPO is produced by hepatocytes. The fundamental stimulus for EPO production is the availability of O2 for tissue metabolic needs. Impaired O2delivery to the kidney can result from a decreased red cell mass (anemia), impaired O2loading of the hemoglobin molecule (hypoxemia), or, rarely, impaired blood flow to the kidney (renal artery stenosis). EPO governs the day-to-day production of red cells, and ambient levels of the hormone can be measured in the plasma by sensitive immunoassays -- the normal level being 10 to 25 U/L. When
the hemoglobin concentration falls below 100 to 120 g/L (10 to 12 g/dL), plasma EPO levels increase logarithmically in inverse proportion to the severity of the anemia. In circulation, EPO has a half-clearance time of 6 to 9 h. EPO acts by binding to specific receptors on the surface of marrow erythroid precursors, inducing them to proliferate and to mature. Under the stimulus of EPO, red blood cell production can increase fourto fivefold within a 1- to 2-week period but only in the presence of adequate nutrients, especially iron. The functional capacity of the erythron, therefore, requires normal renal production of EPO, a functioning erythroid marrow, and an adequate supply of substrates for hemoglobin synthesis. A defect in any of these key components can lead to anemia. Generally, anemia is recognized in the laboratory when a patient's hemoglobin level or hematocrit is reduced below an expected value (the normal range). The likelihood and severity of anemia are defined based on the deviation of the patient's hemoglobin/hematocrit from values expected for age- and sex-matched normal subjects. The lower ranges of distribution of hemoglobin/hematocrit values for adult males and females are shown in Fig. 61-2. The hemoglobin concentration in adults has a Gaussian distribution. The mean hematocrit value for adult males is 47% (± SD 7) and that for adult females is 42% (± 5). Any individual hematocrit or hemoglobin value carries with it a likelihood of associated anemia. Thus, a hematocrit of£39% in an adult male or50,000 STSs have been mapped. The goal of achieving a high-resolution physical map of the human genome has essentially been achieved as all of the genome has been cloned into overlapping fragments. The highest resolution physical map will provide the complete DNA sequence of each chromosome in the human genome. STATUS OF DNA SEQUENCING The primary focus of the genome project is to obtain DNA sequence for the entire human genome as well as model organisms. The sequences of E. coli and many other bacteria, S. cerevisiae, C. elegans, and D. melanogaster have already been completed. Sequencing of the laboratory mouse genome is in progress. Although the prospect of determining the complete sequence of the human genome was a daunting prospect several years ago, technical advances in DNA sequencing and bioinformatics have led
to the completion of a draft human sequence in June 2000, well in advance of the original goal of the year 2003. The current standard is to achieve 99.99% (1 error in 10,000 bp) accuracy. This level of accuracy is important for many reasons, including efforts to determine the degree of DNA sequence variation in the population. Comparisons of the DNA sequence from multiple individuals or populations will allow assessments of genetic variance in the human population. Another goal is to develop a complete set of full-length human cDNAs and to define their locations on the physical map. ETHICAL ISSUES Implicit in theHGP is the idea and hope that identifying disease-causing genes can lead to improvements in diagnosis, prognosis, and treatment. It is estimated that most individuals harbor several serious recessive genes. However, completion of the human genome sequence, determination of the association of genetic defects with disease, and studies of genetic variation raise many new issues with implications for the individual and mankind. The controversies concerning the cloning of mammals and the establishment of human embryonic stem cells underscore the relevance of these questions. Moreover, the information gleaned from genotypic results can have quite different impacts, depending on the availability of strategies to modify the course of disease. For example, the identification of mutations that cause multiple endocrine neoplasia (MEN) type 2 or hemochromatosis allows specific interventions for affected family members. On the other hand, at present the identification of an Alzheimer or Huntington disease gene does not alter therapy. Genetic test results can generate anxiety in affected individuals and family members, and there is the possibility of discrimination on the basis of the test results. Most genetic disorders are likely to fall into an intermediate category where the opportunity for prevention or treatment is significant but limited (Chap. 68). For these reasons, the scientific components of the HGP have been paralleled by efforts to examine ethical and legal implications as new issues arise. Many issues raised by the genome project are familiar, in principle, to medical practitioners. For example, an asymptomatic patient with increased low-density lipoprotein (LDL) cholesterol, high blood pressure, or a strong family history of early myocardial infarction, is known to be at increased risk of coronary heart disease. In such cases, it is clear that the identification of risk factors and an appropriate intervention are beneficial. Likewise, patients with phenylketonuria, cystic fibrosis, or sickle cell anemia are often identified as having a genetic disease early in life. These precedents can be helpful for adapting policies that relate to genetic information. We can anticipate similar efforts, whether based on genotypes or other markers of genetic predisposition, to be applied to many disorders. One confounding aspect of the rapid expansion of information is that our ability to make clinical predictions often lags behind genetic advances. For example, when genes that predispose to breast cancer, such as BRCA1, are described, they generate tremendous public interest in the potential to predict disease, but many years of clinical research are still required to rigorously establish genotype and phenotype correlations. Whether related to informed consent, participation in research, or the management of a genetic disorder that affects an individual or their families, there is a great need for more
information about fundamental principles of genetics. The pervasive nature of the role of genetics in medicine makes it imperative for physicians and other health care professionals to become more informed about genetics and to provide advice and counseling in conjunction with trained genetic counselors (Chap. 68). The application of screening and prevention strategies will therefore require intensive patient and physician education, changes in health care financing, and legislation to protect patient's rights. TRANSMISSION OF GENETIC DISEASE ORIGINS AND TYPES OF MUTATIONS A mutation can be defined as any change in the primary nucleotide sequence of DNA regardless of its functional consequences. Some mutations may be lethal, others are less deleterious, and some may confer an evolutionary advantage. Mutations can occur in the germline (sperm or oocytes); these can be transmitted to progeny. Alternatively, mutations can occur during embryogenesis or in somatic tissues. Mutations that occur during development lead to mosaicism, a situation in which tissues are composed of cells with different genetic constitutions. If the germline is mosaic, a mutation can be transmitted to some progeny but not others, which sometimes leads to confusion in assessing the pattern of inheritance. Somatic mutations that do not affect cell survival can sometimes be detected because of variable phenotypic effects in tissues (e.g., pigmented lesions in McCune-Albright syndrome). Other somatic mutations are associated with neoplasia because they confer a growth advantage to cells. Epigenetic events such as altered DNA methylation may also influence gene expression. With the exception of triplet nucleotide repeats, which can expand (see below), mutations are usually stable. Mutations are structurally diverse -- they can involve the entire genome, as in triploidy (one extra set of chromosomes), or gross numerical or structural alterations in chromosomes or individual genes (Chap. 66). Large deletions may affect a portion of a gene or an entire gene, or, if several genes are involved, they may lead to a contiguous gene syndrome. Unequal crossing-over between homologous genes can result in fusion gene mutations, as illustrated by color blindness (Chap. 28). Mutations involving single nucleotides are referred to as point mutations. Substitutions are called transitions if a purine is replaced by another purine base (A« G) or if a pyrimidine is replaced by another pyrimidine (C « T). Changes from a purine to a pyrimidine, or vice versa, are referred to as transversions. If the DNA sequence change occurs in a coding region and alters an amino acid, it is called a missense mutation. Depending on the functional consequences of such a missense mutation, amino acid substitutions in different regions of the protein can lead to distinct phenotypes. Polymorphisms are sequence variations that have a frequency of at least 1%. Usually, they do not result in a perceptible phenotype. Often they consist of single base-pair substitutions that do not alter the protein coding sequence because of the degenerate nature of the genetic code, although it is possible that some might alter mRNA stability, translation, or the amino acid sequence. These types of silent base substitutions andSNPs are encountered frequently during genetic testing and must be distinguished from true mutations that alter protein expression or function. Small nucleotide deletions or insertions cause a shift of the codon reading frame. Most commonly, reading frame
alterations result in an abnormal protein segment of variable length before termination of translation occurs at a stop codon (nonsense mutation). Mutations in intronic sequences or in exon junctions may destroy or create splice donor or splice acceptor sites. Mutations may also be found in the regulatory sequences of genes, resulting in reduced gene transcription. Mutation Rates As noted before, mutations represent an important cause of genetic diversity as well as disease. Mutation rates are difficult to determine in humans because many mutations are silent and because testing is often not adequate to detect the phenotypic consequences. Mutation rates vary in different genes but are estimated to occur at a rate of about 10-10 /bp per cell division. Germline mutation rates (as opposed to somatic mutations) are relevant in the transmission of genetic disease. Because the population of oocytes is established very early in development, only about 20 cell divisions are required for completed oogenesis, whereas spermatogenesis involves about 30 divisions by the time of puberty and 20 cell divisions each year thereafter. Consequently, the probability of acquiring new point mutations is much greater in the male germline than the female germline, in which rates of aneuploidy are increased (Chap. 66). Thus, the incidence of new point mutations in spermatogonia increases with paternal age (e.g., achondrodysplasia, Marfan syndrome, neurofibromatosis). It is estimated that about 1 in 10 sperm carries a new deleterious mutation. The rates for new mutations are calculated most readily for autosomal dominant and X-linked disorders and are ~10-5to 10-6/locus per generation. Because most monogenic diseases are relatively rare, new mutations account for a significant fraction of cases. This is important in the context of genetic counseling, as a new mutation can be transmitted to the affected individual but does not necessarily imply that the parents are at risk to transmit the disease to other children. An exception to this is when the new mutation occurs early in germline development, leading to gonadal mosaicism. Unequal Crossing-Over Normally, DNA recombination in germ cells occurs with remarkable fidelity to maintain the precise junction sites for the exchanged DNA sequences (Fig. 65-2). However, mispairing of homologous sequences leads to unequal crossover, with gene duplication on one of the chromosomes and gene deletion on the other chromosome. A significant fraction of growth hormone (GH) gene deletions, for example, involve unequal crossing-over (Chap. 328). The GH gene is a member of a large gene cluster that includes a growth hormone variant gene as well as several structurally related chorionic somatomammotropin genes and pseudogenes (highly homologous but functionally inactive relatives of a normal gene). Because such gene clusters contain multiple homologous DNA sequences arranged in tandem, they are particularly prone to undergo recombination and, consequently, gene duplication or deletion. On the other hand, duplication of the PMP22 gene as a result of unequal crossing-over results in increased gene dosage and type IA Charcot-Marie-Tooth disease (Chap. 379). Unequal crossing-over resulting in deletion of PMP22 results in a distinct neuropathy called hereditary liability to pressure palsy (Chap. 379). Glucocorticoid-remediable aldosteronism (GRA) is caused by a rearrangement involving the genes that encode aldosterone synthase (CYP11B2) and steroid 11b-hydroxylase (CYP11B1), normally arranged in tandem on chromosome 8q. These two genes are 95% identical, predisposing to gene duplication and deletion by unequal crossing-over. The rearranged gene product contains the regulatory regions of 11b-hydroxylase fused
to the coding sequence of aldosterone synthetase. Consequently, the latter enzyme is expressed in the adrenocorticotropic hormone (ACTH)-dependent zone of the adrenal gland, resulting in overproduction of mineralocorticoids and hypertension (Chap. 331). Gene conversion refers to a nonreciprocal exchange of homologous genetic information; it is probably more common than generally recognized. In human genetics, gene conversion has been used to explain how an internal portion of a gene is replaced by a homologous segment copied from another allele or locus; these genetic alterations may range from a few nucleotides to a few thousand nucleotides. As a result of gene conversion, it is possible for short DNA segments of two chromosomes to be identical, even though these sequences are distinct in the parents. A practical consequence of this phenomenon is that nucleotide substitutions can occur during gene conversion between related genes, often altering the function of the gene. In disease states, gene conversion often involves intergenic exchange of DNA between a gene and a related pseudogene. For example, the 21-hydroxylase gene (CYP21A) is adjacent to a nonfunctional pseudogene. Many of the nucleotide substitutions that are found in the CYP21A gene in patients with congenital adrenal hyperplasia correspond to sequences that are present in the pseudogene, suggesting gene conversion as a mechanism of mutagenesis. In addition, mitotic gene conversion has been suggested as a mechanism to explain revertant mosaicism in which an inherited mutation is "corrected" in certain cells. For example, patients with autosomal recessive generalized atrophic benign epidermolysis bullosa have acquired reverse mutations in one of the two mutated COL17A1 alleles, leading to clinically unaffected patches of skin. Insertions and Deletions Though many instances of insertions and deletions occur as a consequence of unequal crossing-over, there is also evidence for internal duplication, inversion, or deletion of DNA sequences. The fact that certain deletions or insertions appear to occur repeatedly as independent events suggests that specific regions within the DNA sequence predispose to these errors. For example, certain regions of the DMD gene appear to be hot spots for deletions. Errors in DNA Repair Because mutations caused by defects in DNA repair accumulate as somatic cells divide, these types of mutations are particularly important in the context of neoplastic disorders (Chap. 82). Several genetic disorders involving DNA repair enzymes underscore their importance. Patients with xeroderma pigmentosum have defects in DNA damage recognition or in the nucleotide excision and repair pathway (Chap. 86). Exposed skin is dry and pigmented and is extraordinarily sensitive to the mutagenic effects of ultraviolet irradiation. More than 10 different genes have been shown to cause the different forms of xeroderma pigmentosum. This finding is consistent with the earlier classification of this disease into different complementation groups (Table 65-4) in which normal function is rescued by the fusion of cells derived from two different forms of xeroderma pigmentosum. Ataxia telangiectasia causes large telangiectatic lesions of the face, cerebellar ataxia, immunologic defects, and hypersensitivity to ionizing radiation (Chap. 364). The discovery of the ataxia telangiectasia mutated (ATM) gene reveals that it is homologous to genes involved in DNA repair and control of cell cycle checkpoints. Mutations in the ATM gene give rise to defects in meiosis as well as increasing susceptibility to damage from ionizing radiation. Fanconi's anemia is also associated with an increased risk of
multiple acquired genetic abnormalities. It is characterized by diverse congenital anomalies and a strong predisposition to develop aplastic anemia and acute myelogenous leukemia (Chap. 111). Cells from these patients are susceptible to chromosomal breaks caused by a defect in genetic recombination. At least eight different complementation groups have been identified, and several loci and genes associated with Fanconi's anemia have been mapped or cloned (Table 65-4). HNPCCis caused by mutations in one of several different mismatch repair (MMR) genes including MutS homologue 2 (MSH2) and MutL homologue 1 (MLH1) (Chap. 90). These enzymes are involved in the detection of nucleotide mismatches and in the recognition of slipped-strand trinucleotide repeats. Germline mutations in these genes lead to microsatellite instability and a high mutation rate in colon cancer. This syndrome is characterized by autosomal dominant transmission of colon cancer, young age (3 million sequence differences exist between any two unrelated individuals. This sequence variation usually has no significant functional consequence and provides much of the basis for variation in genetic traits. Although many of these sequence variations areSNPs, other variants includeVNTRs or short tandem repeats (STRs). In VNTRs and STRs, the number of times a sequence is repeated is highly variable in the population. Consequently, the probability that sequences will differ on the two homologous chromosomes is high (often>70 to 90%). Most STRs, also called polymorphic microsatellite markers, consist of di-, tri-, or tetranucleotide repeats that can be measured readily usingPCR and primers that reside on either side of the repeat sequences (Fig. 65-8). Many other methods for analyzing polymorphic variation are also available. Historically,RFLPswere used to detect sequence variations that caused changes in the recognition sites for restriction enzymes. This procedure has been largely replaced by the use of STRs. Analyses of SNPs, using DNA chips, provide a promising means for rapid analysis of genetic variation and linkage. In order to identify a chromosomal locus that segregates with a disease, it is necessary to determine the genotype or haplotype of DNA samples from one or several pedigrees. One can then assess whether certain marker alleles cosegregate with the disease. Markers that are closest to the disease gene are less likely to undergo recombination
events and therefore receive a higher linkage score. Linkage is expressed as a lod (logarithm of odds) score -- the ratio of the probability that the disease and marker loci are linked rather than unlinked. Lod scores of +3 (1000:1) are generally accepted as supporting linkage, whereas a score of -2 is consistent with the absence of linkage. An example of the use of linkage analysis is shown in Fig. 65-8. In this case, the gene for the autosomal dominant disorder,MEN-1, is known to be located on chromosome 11q13. Using positional cloning, the MEN1 gene was identified and shown to encode menin, the function of which is poorly understood. However, the transmission of the disorder suggests that menin acts like a tumor-suppressor gene. Affected individuals inherit a mutant form of the MEN1 gene, predisposing them to certain types of tumors (parathyroid, pituitary, pancreatic islet) (Chap. 339). In the tissues that develop a tumor, a "second hit" occurs in the normal copy of the MEN1 gene. This somatic mutation may be a point mutation, a microdeletion, or loss of a chromosomal fragment (detected as loss of heterozygosity, LOH). Within a given family, linkage to the MEN1 gene locus can be assessed without necessarily knowing the specific mutation in the MEN1 gene. Using polymorphicSTRsthat are close to the MEN1 gene, one can assess transmission of the different MEN1 alleles and compare this pattern to development of the disorder to determine which allele is associated with risk of MEN-1. In the pedigree shown, the affected grandfather in generation I carries alleles 3 and 4 on the chromosome with the mutated MEN1 gene and alleles 2 and 2 on his other chromosome 11. Consistent with linkage of the 3/4 genotype to the MEN1 locus, his son in generation II is affected, whereas his daughter (who inherits the 2/2 genotype from her father) is unaffected. In the third generation, transmission of the 3/4 genotype indicates risk of developing MEN-1, assuming that no genetic recombination between the 3/4 alleles and the MEN1 gene has occurred. After a specific mutation in the MEN1 gene is identified within a family, it is possible to track transmission of the mutation itself, thereby eliminating uncertainty caused by recombination. CHROMOSOMAL DISORDERS Chromosomal or cytogenetic disorders are caused by numerical or structural aberrations in chromosomes. Deviations in chromosome number are common causes of abortions, developmental disorders, and malformations.*For discussion of disorders of chromosome number and structure, see Chap. 66. Contiguous Gene Syndromes Large deletions or duplications may affect a portion of a gene, an entire gene, or, if several genes are involved, cause a contiguous gene syndrome. Syndromes associated with chromosomal deletions or duplications have a wide phenotypic spectrum that is dependent on the number of involved gene loci. For example, the cri-du-chat syndrome, one of the most common deletion disorders, is associated with deletions on the short arm of chromosome 5 that vary in size from extremely small deletions within 5p15.2 to the loss of the entire short arm. Because of the variable size of the involved deletions, the phenotype encompasses a spectrum that ranges from severe mental retardation and microcephaly to an isolated catlike cry without morphologic or mental abnormalities. Contiguous gene syndromes have been useful for identifying the location of new disease-causing genes. Because of the variable size of gene deletions in different
patients, a systemic comparison of phenotypes and locations of deletion breakpoints allows positions of particular genes to be mapped within the critical genomic region. MONOGENIC MENDELIAN DISORDERS Monogenic human diseases are frequently referred to as Mendelian disorders because they obey the principles of genetic transmission originally set forth in Gregor Mendel's classic work. The mode of inheritance for a given phenotypic trait or disease is determined by pedigree analysis. All affected and unaffected individuals in the family are recorded in a pedigree using standard symbols (Fig. 65-9). The principles of allelic segregation, and the transmission of alleles from parents to children, are illustrated inFig. 65-10. One dominant (A) allele and one recessive (a) allele can display three Mendelian modes of inheritance: autosomal dominant, autosomal recessive, and X-chromosomal. About 65% of human monogenic disorders are autosomal dominant, 25% are autosomal recessive, and 5% are X-linked (Table 65-5). Genetic testing is now available for many of these disorders and plays an increasingly important role in clinical medicine. Autosomal Dominant Disorders Autosomal dominant disorders assume particular relevance because mutations in a single allele are sufficient to cause the disease. In contrast to recessive disorders, in which disease pathogenesis is relatively straightforward because there is loss of gene function, in dominant disorders there are various disease mechanisms, many of which are unique to the function of the genetic pathway involved. In autosomal dominant disorders, individuals are affected in successive generations; the disease does not occur in the offspring of unaffected individuals. Males and females are affected with equal frequency because the defective gene resides on one of the 22 autosomes (Fig. 65-11A). Autosomal dominant mutations alter one of the two alleles at a given locus. Because the alleles segregate randomly at meiosis, the probability that an offspring will be affected is 50%. Unless there is a new germline mutation, an affected individual has an affected parent. Children with a normal genotype do not transmit the disorder. Due to differences in penetrance or expressivity (see above), the clinical manifestations of autosomal dominant disorders may be variable. Because of these variations, it is sometimes challenging to determine the pattern of inheritance. It should be recognized, however, that some individuals acquire a mutated gene from an unaffected parent. De novo germline mutations occur more frequently during later cell divisions in gametogenesis, explaining why siblings are rarely affected. As noted before, new germline mutations occur more frequently in fathers of advanced age. For example, the average age of fathers with new germline mutations that cause Marfan's syndrome is approximately 37 years, whereas fathers who transmit the disease by inheritance have an average age of about 30 years. Autosomal Recessive Disorders The clinical expression of autosomal recessive disorders is more uniform than in autosomal dominant disorders. Most mutated alleles lead to a complete or partial loss of function. They frequently involve enzymes in metabolic pathways, receptors, or proteins in signaling cascades. Though most recessive disorders are rare, the relatively high frequency of certain recessive disorders,
such as sickle cell anemia, cystic fibrosis, and thalassemia, is partially explained by a selective biologic advantage for the heterozygous state (see below). In an autosomal recessive disease, the affected individual, who can be of either sex, is a homozygote or compound heterozygote for a single-gene defect. With a few important exceptions, autosomal recessive diseases are rare and often occur in the context of parental consanguinity. Though heterozygous carriers of a defective allele are usually clinically normal, they may display subtle differences in phenotype that only become apparent with more precise testing or in the context of certain environmental influences. In sickle cell anemia, for example, heterozygotes are normally asymptomatic. However, in situations of dehydration or diminished oxygen pressure, sickle cell crises can also occur in heterozygotes (Chap. 106). In most instances, an affected individual is the offspring of heterozygous parents. In this situation, there is a 25% chance that the offspring will have a normal genotype, a 50% probability of a heterozygous state, and a 25% risk of homozygosity for the recessive alleles (Fig. 65-11B). In the case of one unaffected heterozygous and one affected homozygous parent, the probability of disease increases to 50% for each child. In this instance, the pedigree analysis mimics an autosomal dominant mode of inheritance (pseudodominance). In contrast to autosomal dominant disorders, new mutations in recessive alleles are rarely manifest because they usually result in an asymptomatic carrier state. X-Linked Disorders Males have only one X chromosome; consequently, a daughter always inherits her father's X chromosome in addition to one of her mother's two X chromosomes. A son inherits the Y chromosome from his father and one maternal X chromosome. Thus, the characteristic features of X-linked inheritance are (1) the absence of father-to-son transmission, and (2) the fact that all daughters of an affected male are obligate carriers of the mutant allele (Fig. 65-11C). The risk of developing disease due to a mutant X-chromosomal gene differs in the two sexes. Because males have only one X chromosome, they are hemizygous for the mutant allele; thus, they are more likely to develop the mutant phenotype, regardless of whether the mutation is dominant or recessive. A female may be either heterozygous or homozygous for the mutant allele, which may be dominant or recessive. The terms X-linked dominant or X-linked recessive are therefore only applicable to expression of the mutant phenotype in women. In addition, the expression of X-chromosomal genes is influenced by X chromosome inactivation (see below). Y-Linked Disorders Only a few genes are known on the Y chromosome. One such gene, the sex-region determining Y factor (SRY), or testis-determining factor (TDF), is crucial for normal male development. Normally there is infrequent exchange of sequences on the Y chromosome with the X chromosome. Because the SRY region is closely adjacent to the pseudoautosomal region, a chromosomal segment on the X and Y chromosomes with a high degree of homology, a crossing-over occasionally involves the SRY region. Translocations can result in XY females with the Y chromosome lacking the SRY gene or XX males harboring the SRY gene on one of the X chromosomes (Chap. 338). Point mutations in the SRY gene may also result in individuals with an XY genotype and an incomplete female phenotype. Most of these mutations occur de novo. Men with oligospermia/azoospermia frequently have microdeletions on the long arm of
the Y chromosome that involve one or more of the azoospermia factor (AZF) genes. EXCEPTIONS TO SIMPLE MENDELIAN INHERITANCE PATTERNS Mitochondrial Disorders Each mitochondrion contains several copies of a circular chromosome. Mitochondrial DNA predominantly encodes transfer RNAs and proteins that are components of the respiratory chain involved in oxidative phosphorylation and ATP generation. The mitochondrial genome is inherited through the maternal line because sperm does not contribute significant cytoplasmic components to the zygote. All children from an affected mother will inherit the disease, but it will not be transmitted from an affected father to his children. During cell replication, the proportion of wild-type and mutant mitochondria can drift; differences in the fraction of defective mitochondria are referred to as heteroplasmia and explain, in part, the phenotypic variability that is common in mitochondrial diseases.*For detailed discussion of mitochondrial disorders, see Chap. 67. Mosaicism Mosaicism refers to the presence of two or more genetically distinct cell lines in the tissues of an individual. It results from a mutation that occurs during embryonic, fetal, or extrauterine development. The developmental stage at which the mutation arises will determine whether germ cells and/or somatic cells are involved. Chromosomal mosaicism results from non-disjunction at an early embryonic mitotic division, leading to the persistence of more than one cell line, as exemplified by some patients with Turner syndrome (Chap. 338). Somatic mosaicism is characterized by a patchy distribution of genetically altered somatic cells. The McCune-Albright syndrome, for example, is caused by activating mutations in the stimulatory G protein a(Gsa) that occur early in development (Chap. 343). The clinical phenotype varies depending on the tissue distribution of the mutation; manifestations include ovarian cysts that secrete sex steroids and cause precocious puberty, polyostotic fibrous dysplasia, cafe-au-lait skin pigmentation, growth hormone-secreting pituitary adenomas, and hypersecreting autonomous thyroid nodules (Chap. 336). X-Inactivation, Imprinting, and Uniparental Disomy According to traditional Mendelian principles, the parental origin of a mutant gene is irrelevant for the expression of the phenotype. Nonetheless, there are important exceptions to this rule. X-inactivation prevents the expression of most genes on one of the two X-chromosomes in every cell of a female. Gene inactivation also occurs on selected chromosomal regions of autosomes. This phenomenon, referred to as genomic imprinting, leads to preferential expression of an allele depending on its parental origin. It is of pathophysiologic importance in disorders where the transmission of disease is dependent on the sex of the transmitting parent and, thus, plays an important role in the expression of certain genetic disorders. Two classic examples are the Prader-Willi syndrome and Angelman syndrome (Chap. 66). Prader-Willi syndrome is characterized by diminished fetal activity, obesity, hypotonia, mental retardation, short stature, and hypogonadotropic hypogonadism. Deletions in the Prader-Willi syndrome occur exclusively on the paternal chromosome 15. In contrast, patients with Angelman syndrome, characterized by mental retardation, seizures, ataxia, and hypotonia, have deletions at the same site of chromosome 15; however, they are located on the maternal chromosome 15. These two syndromes may also result from uniparental disomy. In this case, the syndromes are not caused by deletions on chromosome 15 but
by the inheritance of either two paternal chromosomes (Prader-Willi syndrome), or two maternal chromosomes (Angelman syndrome). Imprinting and the related phenomenon of allelic exclusion may be more common than currently documented, as it is difficult to examine levels of mRNA expression from the maternal and paternal alleles in specific tissues or in individual cells. Genomic imprinting, or uniparental disomy, is involved in the pathogenesis of several other disorders and malignancies (Chap. 66). Hydatidiform mole contains a normal number of diploid chromosomes, but they are all of paternal origin. The opposite situation occurs in ovarian teratomata, with 46 chromosomes of maternal origin. Expression of the imprinted gene for insulin-like growth factor II (IGF-II) is involved in the pathogenesis of the cancer-predisposing Beckwith-Wiedemann syndrome (BWS) (Chap. 81). These children show somatic overgrowth with organomegalies and hemihypertrophy, and they have an increased risk of embryonal malignancies such as Wilm's tumor. Normally only the paternally derived copy of the IGF-II gene is active and the maternal copy is inactive. Imprinting of the IGF-II gene is regulated by H19, which encodes an RNA transcript that is not translated into protein. Disruption or lack of H19 methylation leads to a relaxation of IGF-II imprinting and expression of both alleles. Heritable changes in gene expression not associated with DNA sequence alterations are referred to as epigenetic effects; these changes are increasingly recognized to play a role in human diseases and possibly in aging as well (Chap. 9). Somatic Mutations In many cancer syndromes, there is an inherited predisposition to tumor formation. However, the neoplastic process requires the acquisition of additional somatic mutations (Chap. 81). In retinoblastoma, the tumor develops when both copies of the retinoblastoma (RB) gene are inactivated through two somatic events (sporadic retinoblastoma) or through a somatic loss of the normal allele in an individual with a hereditary defect in the other allele (hereditary retinoblastoma). This "two-hit" model applies to other inherited cancer syndromes such asMEN-1 (Chap. 339) and neurofibromatosis type 2 (Chap. 370). The defective allele is transmitted in a dominant pattern, though tumorigenesis results from a recessive loss of the tumor suppressor gene in an affected tissue. In other instances, the development of cancer typically requires somatic defects in multiple genes, a process termed multistep carcinogenesis (Chap. 82). Nucleotide Repeat Expansion Disorders Several diseases are associated with an increase in the number of nucleotide repeats above a certain threshold (Table 65-6). The repeats are sometimes located within the coding region of the genes, as in Huntington disease or the X-linked form of spinal and bulbar muscular atrophy (SBMA, Kennedy syndrome). In other instances, the repeats probably alter gene regulatory sequences. If an expansion is present, the DNA fragment is unstable and tends to expand further during cell division. The length of the nucleotide repeat often correlates with the severity of the disease. When repeat length increases from one generation to the next, disease manifestations may worsen or be observed at an earlier age; this phenomenon is referred to as anticipation. In Huntington disease, for example, there is a correlation between age of onset and length of the triplet codon expansion (Chap. 362). Anticipation has also been documented in other diseases caused by dynamic mutations in trinucleotide repeats (Table 65-6). The repeat number may also vary in a tissue-specific manner. In myotonic dystrophy, the CTG repeat may be tenfold greater in
muscle tissue than in lymphocytes (Chap. 383). POPULATION GENETICS AND ASSOCIATION STUDIES Overview of Population Genetics In population genetics, the focus changes from alterations in an individual's genome to the distribution pattern of different genotypes of alleles in the population. In a case where there are only two alleles, A and a, the frequency of the genotypes will be p2+ 2pq +q2= 1, with p2corresponding to the frequency of AA, 2pq to the frequency of Aa, and q2to aa. When the frequency of an allele is known, the frequency of the genotype can be calculated. Alternatively, one can determine an allele frequency, if the genotype frequency has been determined. Allele frequencies vary among ethnic groups and geographical regions. For example, heterozygous mutations in the CFTR gene are relatively common in populations of European origin but are rare in the African population. Allele frequencies may vary because certain allelic variants confer a selective advantage. For example, heterozygotes for the sickle cell mutation, which is particularly common in West Africa, are more resistant to malarial infection because the erythrocytes of heterozygotes provide a less favorable environment for Plasmodium parasites. Though homozygosity for the sickle cell gene is associated with severe anemia and sickle crises (Chap. 106), heterozygotes have a higher probability of survival because of the reduced morbidity and mortality from malaria; this phenomenon has led to an increased frequency of the mutant allele. Recessive conditions are more prevalent in geographically isolated populations because of the more restricted gene pool. Allelic Association and Linkage Disequilibrium There are two primary strategies for mapping genes that cause or increase susceptibility to human disease: (1) classic linkage can be performed based on a known genetic model (see above) or, when the model is unknown, by studying pairs of affected relatives; or (2) disease genes can be mapped using allelic association studies (Table 65-7). Allelic association refers to a situation in which the frequency of an allele is significantly increased or decreased in a particular disease. Linkage and association differ in several aspects. Genetic linkage is demonstrable in families or sibships. Association studies, on the other hand, compare a population of affected individuals with a control population. Association studies can be performed as case-control studies that include unrelated affected individuals and matched controls, or as family-based studies that compare the frequencies of alleles transmitted or not transmitted to affected children. Allelic association studies are particularly useful for identifying susceptibility genes in complex diseases. When alleles at two loci occur more frequently in combination than would be predicted (based on known allele frequencies and recombination fractions), they are said to be in linkage disequilibrium. In Fig. 65-12, a mutation, Z, has occurred at a susceptibility locus where the normal allele is Y. The mutation is in close proximity to a genetic polymorphism with allele A or B. With time, the chromosomes carrying the A and Z alleles accumulate and represent 10% of the chromosomes in the population. The fact that the disease susceptibility gene, Z, is found preferentially, or exclusively, in association with the A allele illustrates linkage disequilibrium. Though not all chromosomes carrying the A allele carry the disease gene, the A allele is associated with an increased risk because of its possible association with the Z allele. This model
implies that it may be possible in the future to identify Z directly to provide a more accurate prediction of disease susceptibility. Evidence for linkage disequilibrium can be helpful in mapping disease genes because it suggests that the two loci, in this case A and Z, are tightly linked. POLYGENIC DISEASE AND COMPLEX GENETIC TRAITS Approach to Polygenic and Multifactorial Disease The expression of many common diseases such as cardiovascular disease, hypertension, diabetes, asthma, psychiatric disorders, and certain cancers is determined by genetic background, environmental factors, and lifestyle (Table 65-8). A trait is called polygenic if multiple genes are thought to contribute to the phenotype or multifactorial if multiple genes are assumed to interact with environmental factors. Genetic models for complex traits need to account for genetic heterogeneity and interactions with other genes and the environment. Complex genetic traits may be influenced by modifying genes that are not linked to the main gene involved in the pathogenesis of the trait. This type of gene-gene interaction, or epistasis, plays an important role in polygenic traits that require the simultaneous presence of variations in multiple genes in order to result in a pathologic phenotype. Gene-environment interactions are relevant for many monogenic and polygenic disorders. In phenylketonuria, the phenotypic expression of the disease depends not only on the presence of the mutation in the phenylalanine hydroxylase gene but also on the exposure to the amino acid phenylalanine (Chap. 352). Another example is type 2 diabetes mellitus, in which genetic, nutritional, and lifestyle factors are intimately interrelated in disease pathogenesis (Chap. 333). The identification of genetic variations and environmental factors that either predispose or protect against disease is essential for predicting disease risk, designing preventive strategies, and developing novel therapeutic approaches (Chap. 68). The study of rare monogenic diseases may provide insights into genetic and molecular mechanisms that are subsequently of importance for the understanding of complex diseases. For example, the identification of the insulin promoter factor 1 in maturity-onset of diabetes type 4 was followed by the observation that it also plays a role in the pathogenesis of diabetes mellitus type 2 (Tables 65-1 and65-8). Approach to the Patient Identifying the Disease-Causing Gene Genomic medicine aims to enhance the quality of medical care through the use genotypic analysis (DNA testing) to identify genetic predisposition to disease, to select more specific pharmacotherapy, and to design individualized medical care based on genotype. Genotype can be deduced by analysis of protein (e.g., hemoglobin, apoprotein E), mRNA, or DNA. However, technological advances have made DNA analysis particularly useful because it can be readily applied to all but the largest genes (Fig. 65-13). DNA testing is performed by mutational analysis or linkage studies in individuals at risk for a genetic disorder known to be present in a family. Mass screening programs require tests of high sensitivity and specificity to be cost-effective. Prerequisites for the success of genetic screening programs include the following: that the disorder is potentially serious; that it can be influenced at a presymptomatic stage by changes in behavior, diet, and/or pharmaceutical manipulations; and that the screening does not result in any
harm or discrimination. Screening in Jewish populations for the autosomal recessive neurodegenerative storage disease Tay-Sachs has reduced the number of affected individuals. In contrast, screening for sickle cell trait/disease in African Americans has led to unanticipated problems of discrimination by health insurers and employers. Mass screening programs harbor additional potential problems. For example, screening for the most common genetic alteration in cystic fibrosis, the DF508 mutation with a frequency of ~70% in northern Europe, is feasible and seems to be effective. One has to keep in mind, however, that there is pronounced allelic heterogeneity and that the disease can be caused by >600 other mutations. The search for these less common mutations would substantially increase costs but not the effectiveness of the screening program as a whole. Occupational screening programs aim to detect individuals with increased risk for certain professional activities (e.g.,a 1antitrypsin deficiency and smoke or dust exposure). MUTATIONAL ANALYSES DNA sequence analysis is increasingly used as a diagnostic tool and significantly enhanced diagnostic accuracy. It is used for determining carrier status and for prenatal testing in monogenic disorders (Table 65-5). Certain cancer susceptibility genes, such as BRCA1 and BRCA2, may identify individuals with an increased risk for the development of malignancies. The detection of mutations is an important diagnostic and prognostic tool in leukemias and lymphomas. The demonstration of the presence or absence of mutations is also relevant for the rapidly evolving field of pharmacogenetics, including the identification of differences in drug treatment response or metabolism as a function of genetic background. A general algorithm for the approach to mutational analysis is outlined inFig. 65-13. The importance of a detailed clinical phenotype cannot be overemphasized. This is the step where one should also consider the possibility of genetic heterogeneity and phenocopies. If obvious candidate genes are suggested by the phenotype, they can be analyzed directly. After identification of a mutation, it is essential to demonstrate that it segregates with the phenotype. The functional characterization of novel mutations is labor-intensive and may require analyses in vitro or in transgenic models in order to document the relevance of the genetic alteration. Numerous techniques are available for the detection of mutations (Table 65-9). In a very broad sense, one can distinguish between techniques that allow for screening the absence or presence of known mutations (screening mode) or techniques that definitively characterize mutations. Analyses of large alterations in the genome are possible using cytogenetics, fluorescent in situ hybridization (FISH), and Southern blotting (Chap. 66). More discrete sequence alterations rely heavily on the use of thePCR, which allows rapid gene amplification and analysis. Moreover, PCR makes it possible to perform genetic testing and mutational analysis with small amounts of DNA extracted from leukocytes or even from single cells, buccal cells, or hair roots. Screening for point mutations can be performed by numerous methods (Table 65-9); most are based on the recognition of mismatches between nucleic acid duplexes, electrophoretic separation of single- or double-stranded DNA, or sequencing of DNA fragments amplified by PCR. DNA sequencing can be performed directly on PCR products or on fragments cloned into plasmid vectors amplified in bacterial host cells.
RT-PCRmay be useful to detect absent or reduced levels of mRNA expression due to a mutated allele. Protein truncation tests (PTT) can be used to detect the broad array of mutations that result in premature termination of a polypeptide during its synthesis. The isolated cDNA is transcribed and translated in vitro, and the proteins are analyzed by gel electrophoresis. Comparison of electrophoretic mobility with the wild-type protein allows detection of truncated mutants. The majority of traditional diagnostic methods are gel-based. Novel technologies for the analysis of mutations, genetic mapping, and mRNA expression profiles are in rapid development. DNA chip technologies allow hybridization of DNA or RNA to hundreds of thousands of probes simultaneously. Microarrays are being used clinically for mutational analysis of several human disease genes, as well as for the identification of viral sequence variations. Together with the knowledge gained from theHGP, these technologies provide the foundation to expand from a focus on single genes to analyses at the scale of the genome. ACKNOWLEDGEMENT This chapter reflects the cumulative contributions of many past contributors to Harrison's Principles of Internal Medicine. Most recently, this includes Dr. Joseph L. Goldstein, Dr. Michael S. Brown, Dr. Andrea Ballabio, and Dr. Arthur L. Beaudet. (Bibliography omitted in Palm version) Back to Table of Contents
66. CHROMOSOME DISORDERS - Terry Hassold, Stuart Schwartz In humans, the normal diploid number of chromosomes is 46, consisting of 22 pairs of autosomal chromosomes (numbered 1 to 22 in decreasing size) and one pair of sex chromosomes (XX in females and XY in males). The genome is estimated to contain between 80,000 and 100,000 genes, with the smallest autosome housing between 500 and 1000 genes. Not surprisingly, duplications or deletions of even small chromosome segments have profound consequences on normal gene expression. Deviations in the number or structure of the 46 human chromosomes are astonishingly common, despite severe deleterious consequences. Chromosomal disorders occur in an estimated 10 to 25% of all pregnancies. They are the leading cause of fetal loss and, among pregnancies surviving to term, the leading known cause of birth defects and mental retardation. In recent years, the practice of cytogenetics has shifted from conventional cytogenetic methodology to a union of cytogenetic and molecular techniques. Formerly the province of research laboratories, fluorescence in situ hybridization (FISH) and related molecular cytogenetic technologies have been incorporated into everyday practice in clinical laboratories. As a result, there is an increased appreciation of the importance of "subtle" constitutional cytogenetic abnormalities, such as microdeletions and imprinting disorders, as well as previously recognized translocations and disorders of chromosome number. VISUALIZING CHROMOSOMES CONVENTIONAL CYTOGENETIC ANALYSIS In theory, chromosome preparations can be obtained from any actively dividing tissue by causing the cells to arrest in metaphase, the stage of the cell cycle at which chromosomes are maximally condensed. In practice, only a small number of tissues are used for routine chromosome analysis: amniocytes or chorionic villi for prenatal testing; and blood, bone marrow, or skin fibroblasts for postnatal studies. Samples of blood, bone marrow, and chorionic villi can be processed using short-term culture techniques that yield results in 1 to 3 days. Analysis of other tissue types typically involves long-term tissue culture, requiring 1 to 3 weeks of processing before cytogenetic analysis is possible. Regardless of the culturing technique, cells are processed to recover chromosomes at metaphase or prometaphase and treated chemically or enzymatically to reveal chromosome "bands" (Fig. 66-1). Analysis of the number of chromosomes in the cell, and the distribution of bands on individual chromosomes, allows the identification of numerical or structural abnormalities. This strategy is useful for characterizing the normal chromosome complement and determining the incidence and types of major chromosome abnormalities. Chromosomes are complex structures, consisting of the DNA double helix and chromosome-associated proteins. As for virtually all organisms, each human chromosome contains two specialized structures: a centromere and two telomeres. The
centromere, or primary constriction, divides the chromosome into short (p) and long (q) arms and is responsible for the segregation of chromosomes during cell division. The telomeres, or chromosome ends, "cap" the p and q arms and are important for allowing DNA replication at the ends of the chromosomes. Prior to DNA replication, each chromosome consists of a single chromatid copy of the DNA double helix. After DNA replication and continuing until the time of cell division (including metaphase, when chromosomes are typically visualized), each chromosome consists of two identical sister chromatids (Fig. 66-1). MOLECULAR CYTOGENETICS The introduction ofFISHmethodologies in the late 1980s revolutionized the field of cytogenetics. In principle, FISH is similar to other DNA-DNA hybridization methodologies. The labeled probe DNA and the target DNA (usually metaphase chromosomes) are denatured to become single-stranded and are hybridized together. The probe is labeled with a hapten, such as biotin or digoxigenin, to allow detection with a fluorophore (e.g., FITC or rhodamine). Alternatively, many probes are already labeled with fluorophores and thus can be detected directly. After the hybridization step, the specimen is counter-stained and the preparations are visualized with a fluorescence microscope. Types of FISH Probes A variety of probes are available for use with FISH, including chromosome-specific paints (chromosome libraries), repetitive probes, and single-copy probes. Chromosome libraries were developed initially from flow-sorted individual chromosomes and more recently from monochromosomal human-rodent hybrids. These probes hybridize to sequences that span the entirety of the chromosome from which they are derived and, as a result, they can be used to "paint" individual chromosomes (Fig. 66-2). Repetitive probes recognize amplified DNA sequences present in chromosomes. The most common are a-satellite DNA probes that are complementary to DNA sequences found at the centromeric regions of all human chromosomes. There are also a-satellite probes that hybridize to the centromeric regions of specific chromosomes (Fig. 66-2). A vast number of single-copy probes are now available, both commercially and as a result of the human genome project. These probes can be as small as 1 kb, though normally they are packaged in cosmids (40 kb), bacterial artificial chromosomes (BACs) or P1 clones (100 to 200 kb), or yeast artificial chromosomes (YACs) (1 to 2 Mb). With the advent of the National Cancer Institute BAC initiative, these large DNA fragments will be placed at 1-Mb intervals on every chromosome, each of which can be used forFISHhybridization. Probes for a variety of microdeletion syndromes and for subtelomeric regions of individual chromosomes are commercially available (Fig. 66-2). Applications of FISH The majority of FISH applications involve hybridization of one or two probes of interest as an adjunctive procedure to conventional chromosomal banding techniques. In this regard, FISH can be utilized to identify specific chromosomes, characterize de novo duplications or deletions, and identify and clarify subtle chromosomal rearrangements. Its greatest utilization, however, is in the detection of microdeletions (see below), including those associated with Prader-Willi syndrome
(PWS), Angelman syndrome (AS), William syndrome, velocardiofacial (VCF) and DiGeorge syndromes, Smith-Magenis syndrome, and Miller-Dieker syndrome (MDS) (see below). Though conventional cytogenetic studies can detect some of these microdeletions, initial detection and/or confirmation with FISH is essential. In fact, since appropriate FISH probes have become available, detection of the aforementioned syndromes has increased significantly. In addition to metaphaseFISH, cells can be analyzed at a variety of stages. Interphase analysis, for example, can be used to make a rapid diagnosis in instances where metaphase chromosome preparations are not yet available (e.g., amniotic fluid interphase analysis). Interphase analysis also increases the number of cells available for examination, allows for investigation of nuclear organization, and provides results when cells do not progress to metaphase. One specialized type of interphase analysis involves the application of FISH to paraffin-embedded sections, thereby preserving the architecture of the tissue. The use of interphaseFISH has increased recently, especially for analyses of amniocentesis samples. These studies are performed on uncultured amniotic fluid, typically using DNA probes specific for the chromosomes most commonly identified in trisomies (chromosomes 13, 18, 21, and the X and Y). These studies can be performed rapidly (24 to 72 h) and will ascertain about 60% of the abnormalities detected prenatally. Nevertheless, guidelines from the American College of Medical Genetics suggest that standard cytogenetic analysis be conducted on all specimens following FISH analysis. Another area in which interphase analysis is routinely utilized is cancer cytogenetics (Chap. 81). Many site-specific translocations are associated with specific types of malignancies. For example, there are probes available for both the Abelson (Abl) oncogene and breakpoint cluster region (bcr) involved in chronic myelogenous leukemia (CML). These probes are labeled with rhodamine and FITC, respectively; the fusion of these genes in CML combines the fluorescent colors and appears as a yellow hybridization signal. In addition to standard metaphase and interphaseFISHanalyses, a number of enhanced techniques have been developed for specific types of analysis, including multicolor FISH techniques, reverse painting, comparative genomic hybridization, and fiber FISH. Spectral karyotyping (SKY) andmulticolor FISH (m-FISH) techniques use combinatorially labeled probes that create a unique color for individual chromosomes. In this manner, all of the chromosomes are studied simultaneously, and computer software is used to generate "pseudo-colors" for the individual chromosomes. This technology is useful in the identification of unknown chromosome material (such as markers of duplications) but is most commonly used with the complex rearrangements seen in cancer specimens. Reverse painting is accomplished by either flow-sorting a chromosome of interest or scraping the chromosome off a slide. The DNA from this chromosome (or portion of a chromosome) is extracted, amplified, labeled, and used as aFISHprobe. This probe is then hybridized to a normal metaphase chromosome to identify the origin of the DNA of interest. It is also utilized to identify marker chromosomes or chromosome duplications
of unknown origin. Comparative genomic hybridization (CGH) is a method that can be used when only DNA is available from a specimen of interest. The entire DNA specimen from the sample of interest is labeled in one color (e.g., green), and the normal control DNA specimen in another color (e.g., red). These are mixed in equal amounts and hybridized to normal metaphase chromosomes. The red-to-green ratio is analyzed by a computer program, which determines where the DNA of interest may have gains or loss of material. This technique is useful in the analysis of tumors, particularly in those cases where cytogenetic analysis is not possible. Fiber FISH is a technique in which chromosomes are mechanically stretched, using one of a variety of different methods. FiberFISHprovides a higher resolution of analysis than conventional FISH and more precise information on the chromosomal localization of a specific probe. CYTOGENETIC TESTING IN PRENATAL DIAGNOSIS (See also Chap. 68) The vast majority of prenatal diagnostic studies are performed to rule out a chromosomal abnormality, but cells may also be propagated for biochemical studies or molecular analyses of DNA. Three procedures are used to obtain samples for prenatal diagnosis: amniocentesis, chorionic villus sampling (CVS), and fetal blood sampling. Amniocentesis is the most commonly used procedure and is routinely performed at 15 to 17 weeks of gestation. On some occasions, early amniocentesis at 12 to 14 weeks is done to expedite results, though less fluid is obtained at this time. Early amniocentesis carries a greater risk of spontaneous abortion or fetal injury but provides results at an earlier stage of pregnancy. The vast majority of amniocentesis are performed in the context of advanced maternal age, the best-known correlate of trisomy (see below). Additional reasons for referral for amniocentesis include an abnormal "triple-marker assay" and/or detection of ultrasound abnormalities. In the triple-marker assay, levels of human chorionic gonadotropin, a fetoprotein, and unconjugated estriol in the maternal serum are quantified and used to adjust the maternal age-predicted risk of a trisomy 21 or trisomy 18 fetus. Specific ultrasound abnormalities, when detected at mid-trimester, can also be associated with chromosomal defects. When a nonspecific ultrasound abnormality is present, the estimated risk of a chromosomal defect is approximately 16%. Associations of chromosomal abnormalities and specific types of abnormal ultrasound findings are listed in Table 66-1. Chorionic villus sampling is the second most common procedure for genetic prenatal diagnosis. Because this procedure is routinely performed at about 8 to 10 weeks of gestation, it allows for an earlier detection of abnormalities and a safer pregnancy termination, if desired.CVS is a relatively safe procedure (spontaneous abortions 99%) that systemic concentrations after an oral dose are negligible and no antianginal effect is present. Giving the drug by the sublingual or transdermal routes bypasses the splanchnic organs and allows essentially all of the drug to reach the systemic circulation. For drugs that are efficiently metabolized by the intestinal epithelium and/or liver, i.e., drugs with high extraction ratios in either of these organs, differences in the extent of the first-pass effect between individuals frequently explain variability in drug response. With propranolol, for example, the 15-fold variability in plasma concentrations after the same oral dose results from differences in the individual hepatic extraction ratios reflective of different levels of drug metabolizing activity. Drug administration by nonintravenous routes involves an absorption process characterized by the plasma level increasing to a maximum value at some time after
administration and then declining as the rate of drug elimination exceeds the rate of absorption. Thus, the peak concentration is lower and occurs later than after the same dose given by rapid intravenous injection. The rate of absorption can be an important consideration during the initial period after drug administration, especially for drugs with a narrow therapeutic index -- the ratio of the toxic dose to the therapeutic dose. If absorption is too rapid, then the resulting high concentration may cause adverse effects not observed with a more slowly available formulation. At the other extreme, slow absorption is deliberately designed into "slow-release" or "sustained-release" drug formulations in order to maintain plasma concentrations essentially constant during the dosage interval, because the drug's rate of elimination is offset by an equivalent rate of absorption controlled by formulation factors. Half-Life The organs of elimination can only clear drug from the blood. Thus, the rate at which drug is eliminated from the body is a function of both clearance and the extent to which drug is distributed outside of the vascular compartment. The fraction of total drug in the body that is eliminated in a given time is designated the fractional elimination constant (k).
For example, if the volume of distribution is 10 L and clearance is 1 L/min, then one-tenth of the drug is eliminated per minute. If k is multiplied by the total amount of drug in the body, the actual rate of elimination at any given time can be determined:
This relationship, indicating that the rate of drug elimination is proportional to the drug concentration, describes a first-order, or monoexponential, process. With a few notable exceptions, the elimination of drugs used clinically is first-order. Half-life (t1/2) is the time that it takes for the plasma concentration or amount of drug in the body to decline by 50%. This parameter is related to k as follows:
where 0.693 is the natural logarithm of 2(Co/0.5 Co). Because
then
This is an important relationship since it indicates that the rate of drug elimination, reflected by t1/2, is dependent on both the efficiency of drug removal (Cl) and the drug's volume of distribution (V). When V remains constant, t1/2is a reflection of clearance. Thus, t1/2is shortened when rifampin induces the enzymes responsible for a drug's
hepatic clearance and is lengthened when a drug's renal clearance is impaired in renal failure. However, when there are concomitant alterations in V, as occurs for some drugs in cardiac failure, t1/2is not an accurate measure of Cl or drug dose. DESIGNING DOSAGE REGIMENS Most drugs are administered as part of long-term therapy involving multiple dosing, and it is critical that the dosage regimen be optimized to the individual patient. With some drugs, the desired response, e.g., coagulation or blood pressure, is readily measurable and an individualized dosage regimen can be developed with dosage titration. However, dosage changes should be conservative (90%. If mediastinoscopy is negative, two curative approaches may be used in treating a Pancoast's syndrome tumor. Preoperative irradiation [30 Gy in 10 treatments] is given to the area, followed by an en bloc resection of the tumor and involved chest wall 3 to 6 weeks later. The 3 year survival rate is 42% for epidermoid and 21% for adeno- and large cell carcinomas. The second approach involves radiotherapy alone in curative doses and standard fractionation, which leads to
survival rates similar to those from combined-modality therapy. A meta-analysis of chemotherapy in non-small cell lung cancer used updated data on 9387 individual patients from 52 randomized trials, both published and unpublished, with the main outcome measure being survival. Regimens containing cisplatin were significantly more effective than no treatment. Trials in early-stage disease comparing surgery with surgery plus chemotherapy gave a hazard ratio of 0.87 (13% reduction in risk of death at 5 years) in favor of chemotherapy. Confidence intervals of these data are wide. However, adjuvant chemotherapy is, in general, not considered standard treatment. The most impressive benefits were obtained when chemotherapy was added to radiotherapy for locally advanced disease (stage IIIB and some stage IIIA disease) and when chemotherapy was given preoperatively in a neoadjuvant fashion in stage IIIA disease. Preoperative neoadjuvant chemotherapy is widely used for stage IIIA disease. Preoperative combined modality therapy followed by surgical resection has given promising early results. Whether the surgery adds benefit after chemoradiotherapy has not been defined. Provided the risk/benefit ratio of using chemotherapy is discussed appropriately with patients, such therapy can be given in a noninvestigational setting. For stage IIIA disease, resection followed by postoperative radiation plus chemotherapy for N2 disease, neoadjuvant chemotherapy followed by surgical resection, or neoadjuvant chemoradiotherapy followed by resection are options. For stage IIIB and bulky IIIA disease, neoadjuvant chemotherapy (2 or 3 cycles of a cisplatin-based combination) followed by chest radiation therapy (60 Gy) has improved median survival time from 10 to 14 months and the 5-year survival rate from 7 to 17% compared to results with radiation therapy alone. Administration of radiation and chemotherapy concurrently is being tested; myelotoxicity and esophagitis are increased, but survival improvement is not yet proven. Randomized clinical trials also are needed to evaluate the usefulness of the new agents with activity against non-small cell lung cancer, including the taxanes (paclitaxel and docetaxel), vinorelbine, gemcitabine, and camptothecins (topotecan and CPT-11) in both adjuvant and neoadjuvant settings. Disseminated Non-Small Cell Lung Cancer The 70% of patients who have unresectable non-small cell cancer have a poor prognosis. Patients with performance status scores of 0 (asymptomatic), 1 (symptomatic, fully ambulatory), 2 (in bed 50% of the time), and 4 (bedridden) have median survival times of 34, 25, 17, 8, and 4 weeks, respectively. Standard medical management, the judicious use of pain medications, the appropriate use of radiotherapy, and outpatient chemotherapy form the cornerstone of management. Patients whose primary tumor is causing symptoms such as bronchial obstruction with pneumonitis, hemoptysis, or upper airway or superior vena cava obstruction should have radiotherapy to the primary tumor. The case for prophylactic treatment of the asymptomatic patient is to prevent major symptoms from occurring in the thorax. However, if the patient can be followed closely, it may be appropriate to defer treatment until symptoms develop. Usually a course of 30 to 40 Gy over 2 to 4 weeks is given to the tumor. Radiation therapy provides relief of intrathoracic symptoms with the following frequencies: hemoptysis, 84%; superior vena cava syndrome, 80%; dyspnea, 60%; cough, 60%; atelectasis, 23%; and vocal cord paralysis, 6%. Cardiac tamponade (treated with pericardiocentesis and radiation therapy to the heart), painful bony metastases (with relief in 66%), brain or
spinal cord compression, and brachial plexus involvement may also be palliated with radiotherapy. Usually, with brain metastases and cord compression, dexamethasone (25 to 100 mg/d in four divided doses) is also given and then rapidly tapered to the lowest dosage that relieves symptoms. Brain metastases often are isolated instances of relapse in patients with adenocarcinoma of the lung otherwise controlled by surgery or radiotherapy. However, there is no proven value for prophylactic cranial irradiation or forCT scans of the head in asymptomatic patients. Pleural effusions are common and are usually treated with thoracentesis. If they recur and are symptomatic, chest tube drainage with a sclerosing agent such as intrapleural talc is used. First, the chest cavity is completely drained. Xylocaine 1% is instilled (15 mL), followed by 50 mL normal saline. Then, 10 g sterile talc is dissolved in 100 mL normal saline, and this solution is injected through the chest tube. The chest tube is clamped for 4 h if tolerated, and the patient is rotated onto different sides to distribute the sclerosing agent. The chest tube is removed 24 to 48 h later, after drainage has become slight (usually 6.6 kPa (50 mmHg), and there should be no CO2retention. For patients with limitations in any of these areas, the initial combined-modality therapy or chemotherapy must be modified to prevent undue toxicity. In all patients, these treatments must be coupled with supportive care for infectious, hemorrhagic, and other medical complications. Chemotherapy The combination most widely used is etoposide plus cisplatin or carboplatin, given every 3 weeks on an outpatient basis for 4 to 6 cycles. Another active regimen is etopside, cisplatin, and paclitaxel. Increased dose intensity of chemotherapy adds toxicity without clear survival benefit. Appropriate supportive care (antiemetic therapy, administration of fluid and saline boluses with cisplatin, monitoring of blood counts and blood chemistries, monitoring for signs of bleeding or infection, and, as required, administration of erythropoietin and granulocyte colony-stimulating factor) and adjustment of chemotherapy doses on the basis of nadir granulocyte counts are essential. The initial combination chemotherapy may result in moderate to severe granulocytopenia (e.g., granulocyte counts 50% of esophageal cancers. CLINICAL FEATURES
About 15% of esophageal cancers occur in the upper third of the esophagus (cervical esophagus), 40% in the middle third, and 45% in the lower third. Squamous cell carcinomas and adenocarcinomas of the esophagus cannot be distinguished radiographically or endoscopically. Progressive dysphagia and weight loss of short duration are the initial symptoms in the vast majority of patients. Dysphagia initially occurs with solid foods and gradually progresses to include semisolids and liquids. By the time these symptoms develop, the disease is usually incurable, since difficulty in swallowing does not occur until ³60% of the esophageal circumference is infiltrated with cancer. Dysphagia may be associated with pain on swallowing (odynophagia), pain radiating to the chest and/or back, regurgitation or vomiting, and aspiration pneumonia. The disease most commonly spreads to adjacent and supraclavicular lymph nodes, liver, lungs, and pleura. Tracheoesophageal fistulas may develop as the disease advances, leading to severe suffering. As with other squamous cell carcinomas, hypercalcemia may occur in the absence of osseous metastases, probably from parathormone-related peptide secreted by tumor cells (Chap. 100). DIAGNOSIS Attempts at endoscopic and cytologic screening for carcinoma in patients with Barrett's esophagus, while effective as a means of detecting high-grade dysplasia, have not yet been shown to improve the prognosis in individuals found to have a carcinoma. Routine contrast radiographs effectively identify esophageal lesions large enough to cause symptoms. In contrast to benign esophageal leiomyomas, which result in esophageal narrowing with preservation of a normal mucosal pattern, esophageal carcinomas characteristically cause ragged, ulcerating changes in the mucosa in association with deeper infiltration, producing a picture resembling achalasia. Smaller, potentially resectable tumors are often poorly visualized despite technically adequate esophagograms. Because of this, esophagoscopy should be performed in all patients suspected of having an esophageal abnormality, to visualize the tumor and to obtain histopathologic confirmation of the diagnosis. Because the population of persons at risk for squamous cell carcinoma of the esophagus (i.e., smokers and drinkers) also has a high rate of cancers of the lung and the head and neck region, endoscopic inspection of the larynx, trachea, and bronchi should also be done. A thorough examination of the fundus of the stomach (by retroflexing the endoscope) is imperative as well. Endoscopic biopsies of esophageal tumors fail to recover malignant tissue in one-third of cases because the biopsy forceps cannot penetrate deeply enough through normal mucosa pushed in front of the carcinoma. Cytologic examination of tumor brushings frequently complements standard biopsies and should be performed routinely. The extent of tumor spread to the mediastinum and paraaortic lymph nodes should also be assessed by computed tomography (CT) scans of the chest and abdomen and by endoscopic ultrasound. TREATMENT The prognosis for patients with esophageal carcinoma is poor. Fewer than 5% of patients are alive 5 years after the diagnosis; thus, management focuses on symptom control. Surgical resection of all gross tumor (i.e., total resection) is feasible in only
40%of cases, with residual tumor cells frequently present at the resection margins. Such esophagectomies have been associated with a postoperative mortality rate of ~10% due to anastomotic fistulas, subphrenic abscesses, and respiratory complications. About 20% of patients who survive a total resection live 5 years. The outcome of primary radiation therapy (5500 to 6000 cGy) for squamous cell carcinomas is similar to that of radical surgery, sparing patients perioperative morbidity but often resulting in less satisfactory palliation of obstructive symptoms. The evaluation of chemotherapeutic agents in patients with esophageal carcinoma has been hampered by ambiguity in the definition of "response" (i.e., benefit) and the debilitated physical condition of many treated individuals. Nonetheless, significant reductions in the size of measurable tumor masses have been reported in 15 to 25% of patients given single-agent treatment and in 30 to 60% of patients treated with drug combinations that include cisplatin. Combination chemotherapy and radiation therapy as the initial therapeutic approach, either alone or followed by an attempt at operative resection, may be of benefit. When administered along with radiation therapy, chemotherapy produces a better survival outcome than radiation therapy alone. The use of preoperative chemotherapy and radiation therapy followed by esophageal resection appears to prolong survival as compared with historic controls, but randomized trials have produced inconsistent results. For the incurable, surgically unresectable patient with esophageal cancer, dysphagia, malnutrition, and the management of tracheoesophageal fistulas loom as major issues. Approaches to palliation include repeated endoscopic dilatation, the surgical placement of a gastrostomy or jejunostomy for hydration and feeding, and endoscopic placement of an expansive metal stent to bypass the tumor. Endoscopic fulguration of the obstructing tumor with lasers appears to be the most promising of these techniques. TUMORS OF THE STOMACH GASTRIC ADENOCARCINOMA Incidence and Epidemiology For unclear reasons, the incidence and mortality rates for gastric cancer have decreased markedly during the past 60 years. The mortality rate from gastric cancer in the United States has dropped in men from 28 to 5.0 per 100,000 population, while in women, the rate has decreased from 27 to 2.3 per 100,000. Nonetheless, 21,500 new cases of stomach cancer were diagnosed in the United States and 13,000 Americans died of the disease in 2000. Gastric cancer incidence has decreased worldwide but remains high in Japan, China, Chile, and Ireland. The risk of gastric cancer is greater among lower socioeconomic classes. Migrants from high- to low-incidence nations maintain their susceptibility to gastric cancer, while the risk for their offspring approximates that of the new homeland. These findings suggest that an environmental exposure, probably beginning early in life, is related to the development of gastric cancer, with dietary carcinogens considered the most likely factor(s). Pathology About 85% of stomach cancers are adenocarcinomas, with 15% due to lymphomas and leiomyosarcomas. Gastric adenocarcinomas may be subdivided into two categories: a diffuse type in which cell cohesion is absent, so that individual cells infiltrate and thicken the stomach wall without forming a discrete mass; and an intestinal
type characterized by cohesive neoplastic cells that form glandlike tubular structures. The diffuse carcinomas occur more often in younger patients, develop throughout the stomach (including the cardia), result in a loss of distensibility of the gastric wall (so-called linitis plastica or "leather bottle" appearance), and carry a poorer prognosis. Intestinal-type lesions are frequently ulcerative, more commonly appear in the antrum and lesser curvature of the stomach, and are often preceded by a prolonged precancerous process. While the incidence of diffuse carcinomas is similar in most populations, the intestinal type tends to predominate in the high-risk geographic regions and is less likely to be found in areas where the frequency of gastric cancer is declining. Thus, different etiologic factor(s) may be involved in these two subtypes. In the United States, the distal stomach is the site of origin of ~30% of gastric cancers, ~20% arise in the midportion of the stomach, and ~37% originate in the proximal third of the stomach. The remaining 13% involve the entire stomach. Etiology The long-term ingestion of high concentrations of nitrates in dried, smoked, and salted foods appears to be associated with a higher risk. The nitrates are thought to be converted to carcinogenic nitrites by bacteria (Table 90-2). Such bacteria may be introduced exogenously through the ingestion of partially decayed foods, which are consumed in abundance worldwide by the lower socioeconomic classes. Bacteria such as Helicobacter pylori may also contribute to this effect by causing chronic gastritis, loss of gastric acidity, and bacterial growth in the stomach. Loss of acidity may occur when acid-producing cells of the gastric antrum have been removed surgically to control benign peptic ulcer disease or when achlorhydria, atrophic gastritis, and even pernicious anemia develop in the elderly. Serial endoscopic examinations of the stomach in patients with atrophic gastritis have documented replacement of the usual gastric mucosa by intestinal-type cells. This process of intestinal metaplasia may lead to cellular atypia and eventual neoplasia. Since the declining incidence of gastric cancer in the United States primarily reflects a decline in distal, ulcerating, intestinal-type lesions, it is conceivable that better food preservation and the availability of refrigeration to all socioeconomic classes have decreased the dietary ingestion of exogenous bacteria. Several additional etiologic factors have been associated with gastric carcinoma. Gastric ulcers and adenomatous polyps have occasionally been so linked, but data regarding a cause-and-effect relationship are unconvincing. The inadequate clinical distinction between benign gastric ulcers and small ulcerating carcinomas may, in part, account for this presumed association. The presence of extreme hypertrophy of gastric rugal folds (i.e., Menetrier's disease), giving the impression of polypoid lesions, has been associated with a striking frequency of malignant transformation; such hypertrophy, however, does not represent the presence of true adenomatous polyps. Individuals with blood group A have a higher incidence of gastric cancer than persons with blood group O; this observation may be related to differences in the mucous secretion leading to altered mucosal protection from carcinogens. Duodenal ulcers are not associated with gastric cancer. Clinical Features Gastric cancers, when superficial and surgically curable, usually produce no symptoms. As the tumor becomes more extensive, patients may complain of an insidious upper abdominal discomfort varying in intensity from a vague, postprandial fullness to a severe, steady pain. Anorexia, often with slight nausea, is very common but is not the usual presenting complaint. Weight loss may eventually be
observed, and nausea and vomiting are particularly prominent with tumors of the pylorus; dysphagia may be the major symptom caused by lesions of the cardia. There are no early physical signs. A palpable abdominal mass indicates long-standing growth and predicts regional extension. Gastric carcinomas spread by direct extension through the gastric wall to the perigastric tissues, occasionally adhering to adjacent organs such as the pancreas, colon, or liver. The disease also spreads via lymphatics or by seeding of peritoneal surfaces. Metastases to intraabdominal and supraclavicular lymph nodes occur frequently, as do metastatic nodules to the ovary (Krukenberg's tumor), periumbilical region ("Sister Mary Joseph node") or peritoneal cul-de-sac (Blumer's shelf palpable on rectal or vaginal examination); malignant ascites may also develop. The liver is the most common site for hematogenous spread of tumor. The presence of iron-deficiency anemia in men and of occult blood in the stool in both sexes mandate a search for an occult gastrointestinal tract lesion. A careful assessment is of particular importance in patients with atrophic gastritis or pernicious anemia. Unusual clinical features associated with gastric adenocarcinomas include migratory thrombophlebitis, microangiopathic hemolytic anemia, and acanthosis nigricans. Diagnosis A double-contrast radiographic examination is the simplest diagnostic procedure for the evaluation of a patient with epigastric complaints. The use of double-contrast techniques helps to detect small lesions by improving mucosal detail. The stomach should be distended at some time during every radiographic examination, since decreased distensibility may be the only indication of a diffuse infiltrative carcinoma. Although gastric ulcers can be detected fairly early, distinguishing benign from malignant lesions is difficult. The anatomic location of an ulcer is not in itself an indication of the presence or absence of a cancer. Gastric ulcers that appear benign by radiography present special problems. Some physicians believe that gastroscopy is not mandatory if the radiographic features are typically benign, if complete healing can be visualized by x-ray within 6 weeks, and if a follow-up contrast radiograph obtained several months later shows a normal appearance. However, we recommend gastroscopic biopsy and brush cytology for all patients with a gastric ulcer in order to exclude a malignancy. Malignant gastric ulcers must be recognized before they penetrate into surrounding tissues, because the rate of cure of early lesions limited to the mucosa or submucosa is>80%. Since gastric carcinomas are difficult to distinguish clinically or radiographically from gastric lymphomas, endoscopic biopsies should be made as deep as possible, due to the submucosal location of lymphoid tumors. The staging system for gastric carcinoma is shown in Table 90-3. TREATMENT Complete surgical removal of the tumor with resection of adjacent lymph nodes offers the only chance for cure. However, this is possible in fewer than a third of patients. A subtotal gastrectomy is the treatment of choice for patients with distal carcinomas, while total or near-total gastrectomies are required for more proximal tumors. The inclusion of
extended lymph node dissection to these procedures appears to confer an added risk for complications without enhancing survival. The prognosis following complete surgical resection depends on the degree of tumor penetration into the stomach wall and is adversely influenced by regional lymph node involvement, vascular invasion, and abnormal DNA content (i.e., aneuploidy), characteristics found in the vast majority of American patients. As a result, the probability of survival after 5 years for the 25 to 30% of patients able to undergo complete resection is ~20% for distal tumors and5 cm in diameter and may be palpable on abdominal examination. Bleeding, obstruction, and perforation are common. CANCERS OF THE ANUS Cancers of the anus account for 1 to 2% of the malignant tumors of the large bowel.
Most such lesions arise in the anal canal, the anatomic area extending from the anorectal ring to a zone approximately halfway between the pectinate (or dentate) line and the anal verge. Carcinomas arising proximal to the pectinate line (i.e., in the transitional zone between the glandular mucosa of the rectum and the squamous epithelium of the distal anus) are known as basaloid, cuboidal, or cloacogenic tumors; about one-third of anal cancers have this histologic pattern. Malignancies arising distal to the pectinate line have a squamous cell histology, ulcerate more frequently, and constitute ~55% of anal cancers. The prognosis for patients with basaloid and squamous cell cancers of the anus is identical when corrected for tumor size and the presence or absence of nodal spread. The development of anal cancer is associated with infection by human papillomavirus, the same organism etiologically linked to cervical cancer. The virus is sexually transmitted. The infection may lead to anal warts (condyloma accuminata) which may progress to anal intraepithelial neoplasia and on to squamous cell carcinoma. The risk for anal cancer is increased among homosexual males, presumably related to anal intercourse. Anal cancer risk is increased in both men and women with AIDS, possibly because their immunosuppressed state permits more severe papillomavirus infection. Anal cancers occur most commonly in middle-aged persons and are more frequent in women than men. At diagnosis, patients may experience bleeding, pain, sensation of a perianal mass, and pruritus. Radical surgery (abdominal-perineal resection with lymph node sampling and a permanent colostomy) used to be the treatment of choice for this tumor type. The 5-year survival rate after such a procedure was 55 to 70% in the absence of spread to regional lymph nodes;80% of patients whose initial lesion was 10 cm) and multiple adenomas. Management involves imaging surveillance for small tumors. If the lesion is large (8 to 10 cm), near the surface, and resectable, surgical removal is appropriate. A patient with liver adenoma should stop taking oral contraceptives. Surgical resection may be required for tumors that do not shrink after oral contraceptives are stopped. Pregnancy increases the risk of hemorrhage and should be avoided in women with large adenomas. Patients with multiple large adenomas (e.g., those with glycogen-storage disease) may benefit from liver transplantation. FOCAL NODULAR HYPERPLASIA Focal nodular hyperplasia is a benign tumor often identified incidentally on imaging studies or at laparoscopy done for other reasons. Like hepatic adenomas, it occurs predominantly in women; however, oral contraceptives are not implicated, and hemorrhage and necrosis are rare. The risk of hemorrhage, however, appears to be higher in women taking oral contraceptives. Typically, the lesion is a solid tumor, often in the right lobe, with a fibrous core and stellate projections. The fibrous projections contain atypical hepatocytes, biliary epithelium, Kupffer cells, and inflammatory cells. A technetium scan will usually show a hot spot because of the presence of Kupffer cells. The lesion appears vascular on angiography, and septations may be detectable by angiography, helicalCT scan, and, most reliably, byMRI, but only rarely by ultrasound. Surgery is indicated only for symptomatic lesions.
HEMANGIOMA AND OTHER BENIGN TUMORS Hemangiomas are the most common benign liver tumors, occurring predominantly in women and usually detected incidentally. The prevalence in the general population is in the range of 0.5 to 7.0%. These asymptomatic vascular lesions can be identified byMRI, contrast-enhancedCT, labeled red blood cell nuclide scans, or hepatic angiography. They do not need to be removed unless they are large and are producing a mass effect. Hemorrhage is rare, and malignant change does not occur. Nodular regenerative hyperplasia consists of multiple hepatic nodules resulting from periportal hepatocyte regeneration with surrounding atrophy. It may be associated with an underlying condition such as malignancy or connective tissue disease. Portal hypertension (in the absence of cirrhosis) is the most common clinical manifestation. Other less common benign hepatic lesions include bile duct adenomas and cystadenomas. CARCINOMAS OF THE LIVER HEPATOCELLULAR CARCINOMA Epidemiology and Etiology Primary hepatocellular carcinoma is one of the most common tumors in the world. It is especially prevalent in regions of Asia and sub-Saharan Africa, where the annual incidence is up to 500 cases per 100,000 population. In the United States and western Europe, it is much less common; however, the annual incidence in the United States has increased from 1.4/100,000 in the period 1976 to 1980 to 2.4/100,000 in 1991 to 1995. Hepatocellular carcinoma is up to four times more common in men than in women and usually arises in a cirrhotic liver. The incidence peaks in the fifth to sixth decades of life in western countries but one to two decades earlier in regions of Asia and Africa with a high prevalence of liver carcinoma. The principal reason for the high incidence of hepatocellular carcinoma in parts of Asia and Africa is the frequency of chronic infection with hepatitis B virus (HBV) and hepatitis C virus (HCV). These chronic infections frequently lead to cirrhosis, which itself is an important risk factor for hepatocellular carcinoma (the risk of liver cancer in a cirrhotic liver is ~3% per year); 60 to 90% of these tumors occur in patients with macronodular cirrhosis. Studies in regions of Asia where hepatocellular carcinoma and HBV infection are prevalent have shown that the incidence of this cancer is about 100-fold higher in individuals with evidence of HBV infection than in noninfected controls. In China, the lifetime risk of developing hepatocellular carcinoma in patients with chronic hepatitis B approaches 40%. In patients with HBV infection and hepatocellular carcinoma, HBV DNA may be integrated into host genomic DNA, both in the tumor cells and in adjacent, uninvolved hepatocytes. In addition, modifications of cellular gene expression occur by insertional mutagenesis, chromosomal rearrangements, or the transcriptional transactivating activity of the X and the pre-S2 regions of the HBV genome. HCValso leads to hepatocellular carcinoma. HCV genetic material does not become integrated into host genomic DNA. Therefore, the mechanism of HCV carcinogenesis is unclear. In Europe and Japan, HCV appears to be substantially more prevalent than HBVin cases of hepatocellular carcinoma. Both HBV and HCV can be demonstrated in
some patients, but the clinical course of liver malignancy in these patients does not appear to differ from that when only one virus is implicated. One distinction in high-prevalence areas between hepatocellular carcinoma associated with HBV infection and with HCV infection is in the timing of onset. In Asia, HBV is acquired at birth via perinatal transmission, whereas HCV infection is acquired primarily during adulthood from transfused blood and injections. Correspondingly, the onset of liver carcinoma occurs one to two decades earlier in those with lifelong hepatitis B than in persons with adult-acquired hepatitis C. Retrospective analysis indicates that hepatocellular carcinoma occurs on average approximately 30 years after HCV infection and almost exclusively in patients with cirrhosis. The annual incidence of hepatocellular carcinoma in cirrhotic patients with chronic hepatitis C is 1.5 to 4%. Any agent or factor that contributes to chronic, low-grade liver cell damage and mitosis makes hepatocyte DNA more susceptible to genetic alterations. Thus, as indicated above, chronic liver disease of any type is a risk factor and predisposes to the development of liver cell carcinoma. These conditions include alcoholic liver disease,a1-antitrypsin deficiency, hemochromatosis, and tyrosinemia. In Africa and southern China, aflatoxin B1is an important public health hazard. This mycotoxin appears to induce a very specific mutation at codon 249 in the tumor suppressor gene p53. The loss, inactivation, or mutation of the p53 gene has been implicated in tumorigenesis and is the most common genetic derangement present in human cancers. ThusHBV and aflatoxin B1 have been implicated in the pathogenesis of hepatocellular carcinoma in regions of Africa and southern China where both agents are prevalent. In view of the male predominance of liver cancer, hormonal factors may also play a role. Hepatocellular tumors may occur with long-term androgenic steroid administration, with exposure to thorium dioxide or vinyl chloride (see below), and possibly with exposure to estrogens in the form of oral contraceptives. Clinical and Laboratory Features Cancers of the liver initially may escape clinical recognition because they occur in patients with underlying cirrhosis, and the symptoms and signs may suggest progression of the underlying disease. The most common presenting features are abdominal pain with detection of an abdominal mass in the right upper quadrant. There may be a friction rub or bruit over the liver. Blood-tinged ascites occurs in about 20% of cases. Jaundice is rare, unless there is significant deterioration of liver function or mechanical obstruction of the bile ducts. Serum elevations of alkaline phosphatase and a fetoprotein (AFP) are common (see below). An abnormal type of prothrombin, des-g-carboxy prothrombin, is made and correlates with AFP elevations. A small percentage of patients with hepatocellular carcinoma have a paraneoplastic syndrome; erythrocytosis may result from erythropoietin-like activity produced by the tumor; hypercalcemia may result from secretion of a parathyroid-like hormone. Other manifestations may include hypercholesterolemia, hypoglycemia, acquired porphyria, dysfibrinogenemia, and cryofibrinogenemia. Imaging procedures to detect liver tumors include ultrasound,CT,MRI, hepatic artery angiography (Chap. 282), and technetium scans. Ultrasound is frequently used to
screen high-risk populations and should be the first test if hepatocellular carcinoma is suspected; it is less costly than scans, is relatively sensitive, and can detect most tumors >3 cm. Helical CT and MRI scans are being used with increasing frequency and have higher sensitivities. AFPlevels >500 ug/L are found in about 70 to 80% of patients with hepatocellular carcinoma. Lower levels may be found in patients with large metastases from gastric or colonic tumors and in some patients with acute or chronic hepatitis. High levels of serum AFP (>500 to 1000 ug/L) in an adult with liver disease and without an obvious gastrointestinal tumor strongly suggest hepatocellular carcinoma. A rising level suggests progression of the tumor or recurrence after hepatic resection or therapeutic approaches such as chemotherapy or chemoembolization (see below). Percutaneous liver biopsy can be diagnostic if the sample is taken from an area localized by ultrasound orCT. Because these tumors tend to be vascular, percutaneous biopsies should be done with caution. Cytologic examination of ascitic fluid is invariably negative for tumor cells. Occasionally, laparoscopy or minilaparotomy, to permit liver biopsy under direct vision, may be used. This approach has the additional advantage of sometimes identifying patients who have a localized resectable tumor suitable for partial hepatectomy. TREATMENT Staging of hepatocellular carcinoma is based on tumor size (< or > 50% of the liver), ascites (absent or present), bilirubin ( 3), and albumin (< or >3) to establish Okuda stages I, II, and III. The Okuda system predicts clinical course better than the American Joint Cancer Commission TNM system. The natural history of each stage without treatment is: stage I, 8 months; stage II, 2 months; stage III, less than 1 month. The course of clinically apparent disease is rapid; if untreated, most patients die within 3 to 6 months of diagnosis. When hepatocellular carcinoma is detected very early by serial screening ofAFP and ultrasound, survival is 1 to 2 years after resection. In selected cases, therapy may prolong life. Surgical resection offers the only chance for cure; however, few patients have a resectable tumor at the time of presentation, because of underlying cirrhosis, involvement of both hepatic lobes, or distant metastases (common sites are lung, brain, bone, and adrenal), and the 5-year survival is low. In patients at high risk for the development of hepatocellular carcinoma, screening programs have been initiated to identify small tumors when they are still resectable. Because 20 to 30% of patients with early hepatocellular carcinoma do not have elevated levels of circulating AFP, ultrasonographic screening is recommended as well as AFP determination. In a study in the Far East, persons positive for hepatitis B surface antigen, with or without liver disease, were screened serially; a number of patients with small, subclinical tumors were identified, and surgical resection undertaken. Follow-up observation revealed a 5-year survival rate in this group of 70% and a 10-year survival rate of 50%. These Asian patients, however, were unusual in that they had minimal or no liver disease and their tumors tended to be unifocal or encapsulated. The findings are in contrast to a study in a large population of Italian patients with cirrhosis, associated in most cases with chronicHBVand/orHCVinfections; screening every 3 to 12 months permitted the detection of a 3% annual incidence of
cancer in this cohort but in most cases failed to achieve the goal of early detection of surgically treatable disease. No randomized study has yet shown survival benefit for screening patients at high risk of developing hepatocellular carcinoma. Liver transplantation may be considered as a therapeutic option; tumor recurrence or metastases are the major problems. Patients who have a single lesion£5 cm or three or fewer lesions£3 cm have survival after liver transplantation that is the same as survival after transplantation for nonmalignant liver disease (Chap. 301). Other approaches include (1) hepatic artery embolization and chemotherapy (chemoembolization), (2) alcohol or radio-frequency ablation via ultrasound-guided percutaneous injection, and (3) ultrasound-guided cryoablation. Treatment options for unresectable disease are limited. Randomized trials have not shown a survival advantage after chemoembolization. The liver cannot tolerate high doses of radiation. The disease is not responsive to chemotherapy, including newer agents such as gemcytabine. Investigative immunotherapy and gene therapy techniques have not yet been successful. Based on the presence of hormone receptors on the tumor, tamoxifen has been tested, but without success, and octreotide has had some modest activity. In patients with resectable tumors, polyprenoic acid (a retinoic acid formulation) and intraarterial131I-labeled lipiodol have been reported to reduce the rate of recurrence. Prevention is the preferred strategy. Hepatitis B vaccine can prevent infection and its sequelae, and a reduction in hepatocellular carcinoma has been seen in Taiwan with the introduction of universal vaccination of children. Interferon treatment reduces the incidence of hepatic failure, death, and liver cancer in patients infected withHBV. Treatment with interferon may lower the risk of development of liver cancer in patients with hepatitis C-related cirrhosis (Chap. 297), but additional studies are needed. OTHER MALIGNANT LIVER TUMORS Fibrolamellar carcinoma differs from the typical hepatocellular carcinoma in that it tends to occur in young adults without underlying cirrhosis. This tumor is nonencapsulated but well circumscribed and contains fibrous lamellae; it grows slowly and is associated with a longer survival if treated. Surgical resection has resulted in 5-year survivals >50%; if the lesion is nonresectable, liver transplantation is an option, and the outcome far exceeds that observed in the nonfibrolamellar variety of liver cancer. Hepatoblastoma is a tumor of infancy that typically is associated with very high serumAFPlevels. The lesions are usually solitary, may be resectable, and have a better 5-year survival than that of hepatocellular carcinoma. Angiosarcoma consists of vascular spaces lined by malignant endothelial cells. Etiologic factors include prior exposure to thorium dioxide (Thorotrast), polyvinyl chloride, arsenic, and androgenic anabolic steroids. Epithelioid hemangioendothelioma is of borderline malignancy; most cases are benign, but bone and lung metastases occur. This tumor occurs in early adulthood, presents with right upper quadrant pain, is heterogeneous on sonography, hypodense onCT, and without neovascularity on angiography. Immunohistochemical staining reveals expression of factor VIII antigen. In the absence of extrahepatic metastases, these lesions can be treated by surgical resection or liver transplantation.
METASTATIC TUMORS Metastatic tumors of the liver are common, ranking second only to cirrhosis as a cause of fatal liver disease. In the United States, the incidence of metastatic carcinoma is at least 20 times greater than that of primary carcinoma. At autopsy, hepatic metastases occur in 30 to 50% of patients dying from malignant disease. Pathogenesis The liver is uniquely vulnerable to invasion by tumor cells. Its size, high rate of blood flow, double perfusion by the hepatic artery and portal vein, and its Kupffer cell filtration function combine to make it the next most common site of metastases after the lymph nodes. In addition, local tissue factors or endothelial membrane characteristics appear to enhance metastatic implants. Virtually all types of neoplasms except those primary in the brain may metastasize to the liver. The most common primary tumors are those of the gastrointestinal tract, lung, and breast, as well as melanomas. Less common are metastases from tumors of the thyroid, prostate, and skin. Clinical Features Most patients with metastases to the liver present with symptoms referable only to the primary tumor, and the asymptomatic hepatic involvement is discovered in the course of clinical evaluation. Sometimes hepatic involvement is reflected by nonspecific symptoms of weakness, weight loss, fever, sweating, and loss of appetite. Rarely, features indicating active hepatic disease, especially abdominal pain, hepatomegaly, or ascites, are present. Patients with widespread metastatic liver involvement usually have suggestive clinical signs of cancer and hepatic enlargement. Some have localized induration or tenderness, and, occasionally, a friction rub may be found over tender areas of the liver. Results of liver biochemical tests are often abnormal, but the elevations in marker levels are often only mild and nonspecific. These signs reflect the effects of fever and wasting as well as those of the infiltrating neoplastic process itself. An increase in serum alkaline phosphatase is the most common and frequently the only abnormality. Hypoalbuminemia, anemia, and occasionally a mild elevation of aminotransferase levels may also be found with more widespread disease. Substantially elevated serum levels of carcinoembryonic antigen are usually found when the metastases are from primary malignancies in the gastrointestinal tract, breast, or lung. Diagnosis Evidence of metastatic invasion of the liver should be sought actively in any patient with a primary malignancy, especially of the lung, gastrointestinal tract, or breast, before resection of the primary lesion. An elevated level of alkaline phosphatase or a mass apparent on ultrasound,CT, orMRIexamination of the liver may provide a presumptive diagnosis. Blind percutaneous needle biopsy of the liver will result in a positive diagnosis of metastatic disease in only 60 to 80% of cases with hepatomegaly and elevated alkaline phosphatase levels. Serial sectioning of specimens, two or three repeated biopsies, or cytologic examination of biopsy smears may increase the diagnostic yield by 10 to 15%. The yield is increased when biopsies are directed by ultrasound or CT or obtained during laparoscopy. TREATMENT
Most metastatic carcinomas respond poorly to all forms of treatment, which is usually only palliative. Rarely a single, large metastasis can be removed surgically. Systemic chemotherapy may slow tumor growth and reduce symptoms, but it does not alter the prognosis. Chemoembolization, intrahepatic chemotherapy, and alcohol or radio-frequency ablation may provide palliation. CHOLANGIOCARCINOMA Benign tumors of the extrahepatic bile ducts are extremely rare causes of mechanical biliary obstruction. Most of these are papillomas, adenomas, or cystadenomas and present with obstructive jaundice or hemobilia. Adenocarcinoma of the extrahepatic ducts is more common. There is a slight male preponderance (60%), and the incidence peaks in the fifth to seventh decades. Apparent predisposing factors include (1) some chronic hepatobiliary parasitic infestations, (2) congenital anomalies with ectactic ducts, (3) sclerosing cholangitis and chronic ulcerative colitis, and (4) occupational exposure to possible biliary tract carcinogens (employment in rubber or automotive plants). Cholelithiasis is not clearly a predisposing factor for cholangiocarcinoma. The lesions of cholangiocarcinoma may be diffuse or nodular. Nodular lesions often arise at the bifurcation of the common bile duct (Klatskin tumors) and are usually associated with a collapsed gallbladder, a finding that mandates cholangiography to view proximal hepatic ducts. Patients with cholangiocarcinoma usually present with biliary obstruction, painless jaundice, pruritus, weight loss, and acholic stools. A deep-seated, vaguely localized right upper quadrant pain may be noted. Hepatomegaly and a palpable, distended gallbladder (unless the lesion is high in the duct) are frequent accompanying signs. Fever is unusual unless associated with ascending cholangitis. Because the obstructing process is gradual, the cholangiocarcinoma is often far advanced by the time it presents clinically. The diagnosis is most frequently made by cholangiography following ultrasound demonstration of dilated intrahepatic bile ducts. Any focal strictures of the bile ducts should be considered malignant until proved otherwise. Endoscopic cholangiography permits obtaining specimens for cytology (sensitivity ~60%) and insertion of stents for biliary drainage. Survival of 1 to 2 years is possible in some cases. Perhaps 20% of patients have surgically resectable tumors, but 5-year survival is only 10 to 30%. The high recurrence rate limits the value of liver transplantation. Photodynamic therapy (intravenous hematoporphyrin with cholangioscopically delivered light) has been used with promising early results. CARCINOMA OF THE PAPILLA OF VATER The ampulla of Vater may be involved by extension of tumor arising elsewhere in the duodenum or may itself be the site of origin of a sarcoma, carcinoid tumor, or adenocarcinoma. Papillary adenocarcinomas are associated with slow growth and a more favorable clinical prognosis than diffuse, infiltrative cancers of the ampulla, which are more frequently widely invasive. The presenting clinical manifestation is usually obstructive jaundice. Endoscopic retrograde cannulation of the pancreatic duct is the preferred diagnostic technique when ampullary carcinoma is suspected, because it allows for direct endoscopic inspection and biopsy of the ampulla and for pancreatography to exclude a pancreatic malignancy. Cancer of the papilla is usually
treated by wide surgical excision. Lymph node or other metastases are present at the time of surgery in approximately 20% of cases, and the 5-year survival rate following surgical therapy in this group is only 5 to 10%. In the absence of metastases, radical pancreaticoduodenectomy (the Whipple procedure) is associated with 5-year survival rates as high as 40%. CANCER OF THE GALLBLADDER Most cancers of the gallbladder develop in conjunction with stones rather than polyps. In patients with gallstones, the risk for developing gallbladder cancer, while increased, is still quite low. In one study, gallbladder cancer developed in only 5 of 2583 patients with gallstones followed for a median of 13 years. In the United States, adenocarcinomas make up the vast majority of the estimated 6500 new cases of gallbladder cancer diagnosed each year. The female/male ratio is 4:1, and the mean age at diagnosis is approximately 70 years. The clinical presentation is most often one of unremitting right upper quadrant pain associated with weight loss, jaundice, and a palpable right upper quadrant mass. Cholangitis may supervene. The preoperative diagnosis of the condition has been facilitated by ultrasound andCT. CT is also useful in guiding fine-needle aspiration and biopsy. Once symptoms have appeared, spread of the tumor outside the gallbladder by direct extension or by lymphatic or hematogenous routes is almost invariable. Over 75% of gallbladder carcinomas are unresectable at the time of surgery, the exceptions being tumors discovered incidentally at laparotomy. If the tumor is found by the pathologist, no additional therapy is required. If the tumor is noted by the surgeon on routine cholecystectomy, a second operation is generally performed to resect the adjacent liver, bile duct, and local lymph nodes. Incidental resectable gallbladder tumors have a 50% 5-year survival. The 1-year mortality rate for unresectable disease is about 95%, and 98% of afflicted patients. 28,200 individuals died of pancreatic cancer in 2000, making it the fifth most common cause of cancer-related mortality. The disease is more common in males than in females and in blacks than in whites. It rarely develops before the age of 50. Little is known about the causes of pancreatic cancer. Cigarette smoking is the most consistent risk factor, with the disease being two to three times more common in heavy smokers than in nonsmokers. Whether this association is due to a direct carcinogenic effect of tobacco metabolites on the pancreas or an as yet undefined exposure that occurs more frequently in cigarette smokers is uncertain. Patients with chronic pancreatitis are at increased risk of pancreatic cancer, as are persons with long-standing diabetes mellitus. Obesity is a risk factor for pancreatic cancer; risk is directly related to increased calorie intake. Alcohol abuse or cholelithiasis are not risk factors for pancreatic cancer. Nor is pancreatic cancer associated with coffee consumption. Mutations in K-ras genes have been found in >85% of specimens of human pancreatic cancer. Pancreatic cancer has been associated with mutation of the p16INK4gene located on chromosome 9p21, a gene also implicated in the pathogenesis of malignant melanoma. CLINICAL FEATURES More than 90% of pancreatic cancers are ductal adenocarcinomas, with islet cell tumors constituting the remaining 5 to 10%. Pancreatic cancers occur twice as frequently in the pancreatic head (70% of cases) as in the body (20%) or tail (10%) of the gland. With the exception of jaundice, the initial symptoms associated with pancreatic cancer are often insidious and are usually present for >2 months before the cancer is diagnosed (Table 92-1). Pain and weight loss are present in>75% of patients. The pain typically has a gnawing, visceral quality, occasionally radiating from the epigastrium to the back. Pain is often a more severe problem in lesions arising in the body or tail of the gland, as such tumors may become quite large before being detected. Characteristically, the pain improves somewhat when the patient bends forward. The development of significant pain suggests retroperitoneal invasion and infiltration of the splanchnic nerves, indicating that the primary lesion is advanced and is not surgically resectable. Rarely, such pain may be transient and associated with hyperamylasemia, indicative of acute pancreatitis caused by ductal obstruction by tumor. The weight loss observed in most patients is primarily the result of anorexia, although in the initial period of the disease, subclinical malabsorption may also be a contributing factor. Jaundice due to biliary obstruction is found in >80% of patients having tumors in the pancreatic head and is typically accompanied by dark urine, a claylike appearance of stool, and pruritus. In contrast to the "painless jaundice" sometimes observed in patients having carcinomas of the bile ducts, duodenum, or periampullary regions, most icteric
individuals with ductal carcinomas of the pancreatic head will complain of significant abdominal discomfort. Although the gallbladder is usually enlarged in patients with carcinoma of the head of the pancreas, it is palpable in 80% of cases; in 5 to 15% of patients with proven pancreatic carcinoma, the CT scan shows only generalized pancreatic enlargement suggesting pancreatitis rather than malignancy. False-positive results occur in about 5 to 10% of cases where no tumor was found on laparotomy. Magnetic resonance imaging (MRI) has not been shown to be better than CT in the evaluation of pancreatic lesions. The value of positron emission tomography (PET) has not been defined. When clinical circumstances dictate additional diagnostic evaluation, endoscopic retrograde cholangiopancreatography (ERCP) with endoscopic ultrasonography (EUS) may clarify the cause of ambiguous CT or ultrasound findings. The characteristic findings are stenosis or obstruction of either the pancreatic or the common bile duct; both duct systems are abnormal in over half the cases. Carcinoma and chronic pancreatitis can be difficult to distinguish by ERCP, particularly if both diseases are present. False-negative results with ERCP are infrequent (15% have been associated with this procedure. Curative resection is usually preceded by laparoscopic inspection of the abdomen to confirm absence of occult disease spread to the omentum, peritoneum, or liver, which would preclude curative resection. Although the potential for cure in patients with pancreatic cancer is restricted to the few who are able to undergo a complete surgical resection, the 5-year survival rate following such operations is only 10%. Nonetheless, the procedure is worth attempting, particularly for lesions in the pancreatic head, since ductal carcinomas often cannot be distinguished preoperatively from ampullary, duodenal, and distal bile duct tumors or pancreatic cyst adenocarcinomas, all of which have far higher rates of resectability and cure. Furthermore, patients who undergo resection and eventually experience disease recurrence survive three to four times longer than those whose tumor is not excised, indicating that such operations have a palliative effect. The risk for tumor recurrence is not affected by the type of operative procedure -- i.e., total pancreatectomy versus pancreaticoduodenectomy ("Whipple resection") -- but it is increased by the presence of lymph node metastases or tumor invasion into adjacent viscera. As a rule, pancreaticoduodenectomy or distal pancreatectomy seems preferable to total pancreatectomy because of the retention of exocrine function and avoidance of brittle diabetes. The median survival for patients whose pancreatic cancers are surgically unresectable is 6 months. Management is directed at palliation of symptoms. Ambulatory patients having tumors in the pancreatic head should be considered for surgical diversion of the
biliary system. If jaundice has already developed, therapeutic options include either nonoperative biliary decompression by endoscopic or percutaneous, transhepatic biliary drainage or surgical biliary bypass. External beam radiation in patients with unresectable tumors that have not spread beyond the pancreas does not appear to prolong survival, although a sufficient reduction in tumor size may lead to palliation of pain. However, the addition of chemotherapy with fluorouracil (5-FU) to external beam irradiation has increased the survival time for these patients, perhaps because 5-FU acts as a radiosensitizing agent. In a small patient population, a similar combination of radiation therapy and 5-FU appears to have prolonged the survival and increased the cure rate as compared to a prospectively randomized nontreatment control group of patients who had a complete surgical resection of their pancreatic cancer. This observation needs to be confirmed before it can be accepted. The possibility of administering such chemoradiation therapy at diagnosis and before surgery ("neoadjuvant" treatment), to increase the potential for resectability, is under investigation. Intraoperative radiation therapy has the potential to deliver higher doses of radiation to the tumor while sparing neighboring tissues but does not give better results than external beam treatment. Chemotherapy in the management of patients with widely metastatic pancreatic cancer has been disappointing. Gemcitabine, a deoxycytidine analogue, produces improvement in the quality of life for patients with advanced pancreatic cancer. However, duration of survival is only moderately improved. Newer forms of treatment, including combining gemcitabine with other cytotoxic agents or therapies directed at specific molecular targets, such as K-ras, or p53 are being evaluated. Experimental therapy should constitute the initial treatment for consenting, ambulatory patients.*Pancreatic endocrine tumors are discussed in Chap. 93. (Bibliography omitted in Palm version) Back to Table of Contents
93. ENDOCRINE TUMORS OF THE GASTROINTESTINAL TRACT AND PANCREAS - Robert T. Jensen Gastrointestinal neuroendocrine tumors (NETs) are derived from the diffuse neuroendocrine system of the gastrointestinal (GI) tract, which is composed of amineand acid-producing cells with different hormonal profiles, depending on the site of origin. The tumors they produce can be divided into carcinoid tumors and pancreatic endocrine tumors (PETs). These tumors were originally classified as APUDomas (for amine precursor uptake and decarboxylation), as were pheochromocytomas, melanomas, and medullary thyroid carcinomas because they share certain cytochemical, pathologic, and biologic features (Table 93-1). APUDomas were thought to have a similar embryonic origin from neural crest cells, but the peptide-secreting cells are not of neuroectodermal origin. CLASSIFICATION, PATHOLOGY, AND TUMOR BIOLOGY OF NETS NETs are generally composed of monotonous sheets of small round cells with uniform nuclei; mitoses are uncommon. They can be tentatively identified on routine histology; however, these tumors are principally recognized by their histologic staining patterns due to shared cellular proteins. Historically, silver staining was used, and tumors were classified as showing an argentaffin reaction if they took up and reduced silver or as being argyrophilic if they did not reduce it. Immunocytochemical localization of chromogranins (A,B,C), neuron-specific enolase, or synaptophysin, which are all neuroendocrine cell markers, are now used (Table 93-1). Chromogranin A is the most widely used. Ultrastructurally, these tumors possess electron-dense neurosecretory granules and frequently contain small clear vesicles that correspond to synaptic vesicles of neurons.NETssynthesize numerous peptides, growth factors, and bioactive amines that may be ectopically secreted, giving rise to a specific clinical syndrome (Table 93-2). The diagnosis of the specific syndrome, such as a VIPoma [vasoactive intestinal peptide (VIP)-secreting tumor], requires the clinical features of the disease and cannot be made from the immunocytochemistry results only (Table 93-1). Furthermore, pathologists cannot distinguish between benign and malignant NETs unless metastases or invasion are present. Carcinoid tumors are frequently classified according to their anatomic area of origin (i.e., foregut, midgut, hindgut) because tumors with similar areas of origin share functional manifestations, histochemistry, and secretory products (Table 93-3). Foregut tumors generally have a low serotonin [5-hydroxytryptamine (5-HT)] content, are argentaffin-negative but argyrophilic, occasionally secrete adrenocorticotropic hormone (ACTH) or 5-hydroxytryptophan (5-HTP) causing an atypical carcinoid syndrome (Fig. 93-1), make several hormones, and may metastasize to bone. They uncommonly produce a clinical syndrome due to the secreted products. Midgut carcinoids are argentaffin-positive, have a high serotonin content, most frequently cause the typical carcinoid syndrome when they metastasize (Table 93-3,Fig. 93-1), release serotonin and tachykinins (substance P, neuropeptide K, substance K), rarely secrete 5-HTP or ACTH, and uncommonly metastasize to bone. Hindgut carcinoids (rectum and transverse and descending colon) are argentaffin-negative, often argyrophilic, rarely
contain serotonin or cause the carcinoid syndrome, rarely secrete 5-HTP or ACTH, contain numerous peptides, and may metastasize to bone. PETscan be classified into specific functional or nonfunctional syndromes (Table 93-2). Each of the functional syndromes is associated with symptoms due to the specific hormone released. In contrast, nonfunctional PETs release no products that cause a specific clinical syndrome. "Nonfunctional" is a misnomer in the strict sense because these tumors frequently ectopically secrete a number of peptides [pancreatic polypeptide (PP) chromogranin A, neurotensin]; however, they cause no specific clinical syndrome. The symptoms caused by nonfunctional PETs are entirely due to the tumor per se. Carcinoid tumors can occur in almost any GI tissue (Table 93-3); however, most (70%) originate from one of four sites; bronchus, jejunum/ileum, rectum, or appendix. In the past, carcinoid tumors most frequently occurred in the appendix (i.e., 40%); however, the bronchus/lung is now the most common site (32%). Overall, the GI tract is the most common site for these tumors, comprising 74%, with the respiratory tract a distant second at 25%. The term pancreatic endocrine tumor, although widely used, is also a misnomer, because these tumors can occur either almost exclusively in the pancreas (insulinomas, glucagonomas, nonfunctionalPETs, PETs causing hypercalcemia) or at both pancreatic and extrapancreatic sites [gastrinomas,VIPomas, somatostatinomas, GRFomas (GRF, growth hormone-releasing factor)]. PETs are also called islet cell tumors; however, this term is discouraged because it is not established that many originate from the islets and many can occur at extrapancreatic sites. The exact incidence of carcinoid tumors orPETsvaries according to whether only symptomatic or all tumors are considered. The incidence of clinically significant carcinoids is 7 to 13 cases per million population per year, whereas malignant carcinoids are reported at autopsy in 21 to 84 cases per million population per year. Clinically significant PETs have a prevalence of 10 cases per million population, with insulinomas, gastrinomas, and nonfunctional PETs having an incidence of 0.5 to 2 cases per million population per year (Table 93-2); VIPomas are 2- to 8-fold less common, glucagonomas are 17- to 30-fold less common, and somatostatinomas the least common. In autopsy studies 0.5 to 1.5% of all cases have a PET; however, in fewer than 1 in 1000 cases was a functional tumor present. Both carcinoid tumors andPETscommonly show malignant behavior (Tables 93-2,93-3). With PETs, except for insulinomas in which appendix (35%)> lung/bronchus (27%)> rectum (14%). A number of factors are important in determining survival and the aggressiveness of the tumor (Table 93-4). The presence of liver metastases is the single most important prognostic factor for both carcinoid tumors and PETs. The size of the primary tumor is particularly important in the development of liver metastases. For example, with small-intestinal carcinoids, which are the most frequent cause of the carcinoid syndrome due to metastatic disease in the liver (Table 93-2), metastases
occur in 15 to 25% if the tumor diameter is 75% if it is>2 cm. The size of the primary tumor has also been shown to be an independent predictor of the development of liver metastases for gastrinomas and other PETs. The presence of lymph node metastases, the depth of invasion, various histologic features (differentiation, mitotic rates, growth indices), and flow cytometric results such as the presence of aneuploidy are all important prognostic factors for the development of metastatic disease. The development of the carcinoid syndrome, older age, male gender, the presence of a symptomatic tumor, or greater increases in a number of tumor markers [5-hydroxyindolacetic acid (5-HIAA), neuropeptide K, chromogranin A] also adversely affect prognosis in patients with carcinoid tumors (Table 93-4). With PETs or gastrinomas, a worse prognosis is associated with female gender, overexpression of the Ha-Ras oncogene or p53, the absence of multiple endocrine neoplasia (MEN) type 1, and higher levels of various tumor markers (e.g., chromogranin A, gastrin). A number of genetic disorders are associated with an increased incidence of neuroendocrine tumors (Table 93-5). Each one is caused by a loss of a possible tumor suppressor gene. The most important isMEN-1, an autosomal dominant disorder due to a defect in a 10-exon gene on 11q13, which encodes for a 610-amino-acid nuclear protein, menin (Chap. 339). In patients with MEN-1, 95 to 100% develop hyperparathyroidism due to parathyroid hyperplasia, 80 to 100% develop nonfunctionalPETs, 54 to 80% develop pituitary adenomas, and bronchial carcinoids develop in 8%, thymic carcinoids in 8%, and gastric carcinoids in 13 to 30% of patients with Zollinger-Ellison syndrome. Functional PETs occur in 80% of patients with MEN-1, with 54% developing Zollinger-Ellison syndrome, 21% insulinomas, 3% glucagonomas, and 1%VIPomas. MEN-1 is present in 20 to 25% of all patients with Zollinger-Ellison syndrome, in 4% of those with insulinomas, and in 1 L per day and >3 L per day in 70%. Most patients do not have accompanying steatorrhea (16%), and the increased stool volume is due to increased excretion of sodium and potassium, which, with the anions, accounts for the osmolality of the stool. Patients frequently have hyperglycemia (25 to 50%) and hypercalcemia (25 to 50%). VIPis a 28-amino-acid peptide neurotransmitter, ubiquitously present in the central nervous system and GI tract. Its known actions include stimulation of small-intestinal chloride secretion as well as effects on smooth-muscle contractility, inhibition of acid secretion, and vasodilatory effects which explain most features of the clinical syndrome. In adults 80 to 90% of VIPomas are pancreatic;VIP-secreting pheochromocytomas, intestinal carcinoids, and occasional ganglioneuromas account for the rest. These tumors are usually single; 50 to 75% are in the pancreatic tail and 37 to 68% have hepatic metastases at diagnosis. Diagnosis The diagnosis requires the demonstration of an elevated plasmaVIPlevel and the presence of large-volume diarrhea. A stool volume of350 mmol/day (350 meq/day) of potassium. Because 37 to 68% of adults withVIPomas have metastatic disease in the liver at presentation, a significant number of patients cannot be cured surgically. In these patients, long-acting somatostatin analogues such as octreotide or lanreotide (Fig. 93-2) are the drugs of choice. Octreotide will control the diarrhea in 87% of patients. In nonresponsive patients, the combination of glucocorticoids and octreotide has proved helpful in a few. Other drugs that may be helpful include prednisone (60 to 100 mg/d), clonidine, indomethacin, phenothiazines, loperamide, lidamidine, lithium, proparanolol, and metoclopramide. Treatment of advanced disease with embolization, chemoembolization, and chemotherapy may also be helpful (see below). NONFUNCTIONAL PANCREATIC ENDOCRINE TUMORS NonfunctionalPETs are endocrine tumors that originate in the pancreas and either secrete no products or their secreted products do not cause a specific clinical syndrome. The symptoms are due entirely to the tumor per se. Nonfunctional PETs almost always secrete chromogranin A (90 to 100%), chromogranin B (90 to 100%),PP(58%), a-human chorionic gonadotropin (hCG) (40%), andb-HCG (20%), but none cause a specific syndrome. Patients with nonfunctional PETs usually present late in their disease course with invasive tumors and hepatic metastases (in 64 to 92%), and the tumors are usually large (72% >5 cm). These tumors are usually solitary except in patients withMEN-1, where they are multiple; they occur primarily in the pancreatic head; and though they do not cause a functional syndrome, they synthesize numerous peptides and cannot be distinguished from functional tumors by immunocytomchemistry. The most common symptoms are abdominal pain (30 to 80%), jaundice (20 to 35%), and weight loss, fatigue, or bleeding; 10 to 15% are found incidentally. The average time from the beginning of symptoms to diagnosis is 5 years. Diagnosis The diagnosis is established by histology in a patient with aPETwithout either clinical symptoms or elevated plasma hormone levels of one of the established syndromes (Table 93-2). Even though chromogranin A levels are elevated in almost every patient, this can be found in functional PETs, carcinoids, and other neuroendocrine disorders. PlasmaPP is increased in 22 to 71% of patients and should suggest the diagnosis in a patient with a pancreatic mass because it is usually normal in patients with pancreatic adenocarcinomas. However, elevated plasma PP is not diagnostic of this tumor because it is elevated in a number of other conditions such as chronic renal failure, old age, inflammatory conditions, and diabetes.
TREATMENT Unfortunately, surgical curative resection can be considered in only a minority of patients because 64 to 92% present with metastatic disease. Treatment needs to be directed against the tumor itself using chemotherapy, embolization, chemoembolization, or hormonal therapy (see below). GRFOMAS GRFomas are endocrine tumors that secrete excessive amounts ofGRF that causes acromegaly. The frequency is not known. GRF (also called growth hormone-releasing hormone, GHRH) is a 44-amino-acid peptide, and 25 to 44% ofPETshave GRF immunoreactivity, although excess secretion is uncommon. GRFomas are lung tumors in 47 to 54% of cases, PETs in 29 to 30%, and small-intestinal carcinoids in 8 to 10% and up to 12% occur at other sites. Patients have a mean age of 38 years, and the symptoms are usually due to either acromegaly or the tumor per se. The acromegaly caused by GRFomas is indistinguishable from classic acromegaly. The pituitary abnormality is growth hormone-secreting somatotrope cell hyperplasia rather than a pituitary adenoma. The pancreatic tumors are usually large (>6 cm), and liver metastases are present in 39%. They should be suspected in any patient with acromegaly and an abdominal tumor, in a patient withMEN-1 with acromegaly, or in a patient without a pituitary adenoma with acromegaly or associated with hyperprolactinemia, which occurs in 70% of GRFomas. GRFomas are an uncommon cause of acromegaly. The diagnosis is established by performing plasma assays for GRF and growth hormone. The normal level for GRF is70%. Most men with benign or malignant conditions of the prostate are not diagnosed during their lifetimes. The high prevalence of these diseases, coupled with comorbid conditions and competing causes of death that are frequent in this age group, mandates a careful consideration of the risk/benefit ratio of any proposed intervention. Management is centered on the continual reassessment of the disease as it unfolds in the individual. For the benign proliferative disorders, the symptoms of urinary frequency, infection, and potential for obstruction are counterbalanced by the side effects and complications of medical or surgical therapy. For malignant disease, the risk of developing symptoms or death from cancer is balanced against treatment efficacy and treatment-related morbidity for interventions proposed at different points in the natural history. The incidence and mortality of prostate cancer have declined over the past few years. The decline is not clearly related to any meaningful decrease in the disease or its severity. Instead, the number of cases diagnosed increased dramatically in the early 1990s based on the widespread use of serum prostate-specific antigen (PSA) levels. The test led to the diagnosis of more asymptomatic cancers, some of which may never have produced symptoms -- so-called lead-time bias (Chap. 80). Screening for prostate cancer has not been proven effective in prospective randomized trials. Prostate cancer is the most common cancer diagnosis and the second leading cause of cancer death in men. In 2000, 180,400 cases were diagnosed and 31,900 men died of prostate cancer, down from the peak of 352,000 new cases in 1996. The projected lifetime risk of developing prostate cancer for a 50-year-old man is 42%, of being diagnosed is 9.5%, and of dying from prostate cancer is 2.9%. ANATOMY The prostate is located in the pelvis and is surrounded by the rectum, bladder, dorsal and periprostatic venous complexes, musculature of the pelvic sidewall, the urethral sphincter (responsible for passive urinary control), the pelvic plexus, and cavernous nerves (which innervate the pelvic organs and corpora cavernosa). It is divided into a peripheral zone, a central zone, and a transition zone. The anterior surface is covered by the fibromuscular stroma. Most cancers develop in the peripheral zone, while nonmalignant proliferation occurs predominantly in the transition zone. The functional unit is the glandular acinus, which consists of an epithelial compartment including epithelial, basal, and neuroendocrine cells, and a stromal compartment including fibroblasts and smooth-muscle cells. These compartments are separated by a basement membrane.PSA and prostate-specific acid phosphatase are produced in the epithelial cells. Both stromal and epithelial cells express androgen receptors and depend on androgens for growth. Additional growth regulatory signals occur via paracrine signaling between the two compartments. In cancer, the relationship between stromal and epithelial elements contributes to growth both in the primary and in
metastatic sites. The major circulating androgen in the blood is testosterone, which is converted to dihydrotestosterone, the active form, by 5a-reductase. Changes in prostate size occur during two distinct periods: diffuse enlargement during puberty and in focal regions in the periurethral area after the age of 55. DIAGNOSIS AND SCREENING Symptoms Most cancers are asymptomatic in their early stages. By contrast, benign proliferative disorders may encroach on the urethra early in the clinical course, giving rise to symptoms of outlet obstruction such as hesitancy, intermittent voiding, diminished stream, incomplete emptying, and postvoid leakage. For the patient with symptoms, the history is focused on the urinary tract to identify other causes of voiding dysfunction. For quantification of symptoms, the preferred questionnaire is the self-administered American Urological Association (AUA) Symptom Index in which the symptoms can be classified as mild, moderate, or severe on the basis of seven questions (Table 95-1). This index is useful in planning and in follow-up. Over time, the resistance to the flow of urine reduces the compliance of the detrusor muscle, resulting in nocturia, urgency, and bladder instability and ultimately in urinary retention. The relationship between the signs and symptoms of obstruction and prostate size is not straightforward, and a small gland does not exclude significant blockage. In severe cases the bladder may be palpable on physical examination. Infection, tranquilizing drugs, antihistamines, or alcohol can precipitate urinary retention. Obstructive symptoms are distinct from irritative symptoms such as frequency, dysuria, or urgency, which may occur from infectious, inflammatory, or neoplastic diseases. Conditions that can mimic cancer include acute prostatitis, granulomatous prostatitis, and prostate calculus. Prostatitis usually produces induration and/or pain and is treated with antibiotics. Prostate cancer may manifest in the same manner, and the distinction can only be established histologically, but a biopsy should not be performed before a trial of antibiotics if prostatitis is a possible diagnosis. In cases where the tumor has extended beyond the confines of the gland, symptoms of hematospermia or erectile dysfunction may occur. Prostate cancer may also present with pain secondary to bone metastases, although many patients are asymptomatic despite extensive spread. Less common presentations include myelophthisic disorders, disseminated intravascular coagulation, or spinal cord compression. The proportion of men diagnosed at these late stages has also decreased significantly as a result ofPSA-based detection strategies. Physical Examination The standard evaluation for prostatic diseases includes the digital rectal examination (DRE). It should be performed with careful attention to the size and consistency of the gland, the presence of lesions within the gland, or evidence of extension beyond its confines. Its importance can not be overemphasized. The posterior surfaces of the lateral lobes, where carcinoma begins most often, are easily palpable on DRE. Carcinoma characteristically is hard, nodular, and irregular, but induration may also be due to fibrous areas in a benign hyperplastic background or to calculi. Extraprostatic extension to the seminal vesicles can often be detected by rectal examination, while scrotal and/or lower extremity lymphedema secondary to infiltration of pelvic lymph nodes indicates extensive disease. The need for establishing a histologic diagnosis is based on the finding of an abnormal DRE or an elevated serumPSAlevel.
Prostate-Specific Antigen PSA is a serine protease that is produced by both nonmalignant and malignant epithelial cells. In serum, it circulates as an inactive complex with two protease inhibitors -- a1-antichymotrypsin andb 2-macroglobulin. PSA is prostate specific but not prostate cancer specific and is measured most commonly by radioimmunoassay. The normal range is 0 to 4 ng/mL; ~30% of men with a PSA in the range of 4 to 10 ng/mL and 50% of those with a PSA>10 ng/mL will have cancer. Between 20 and 25% of men with an abnormalDRE have cancer at biopsy, and 20% of men have cancer detected when the PSA is in the normal range. African American men normally have higher PSA levels, even if they do not have prostate cancer. They also have a 50% higher risk of prostate cancer. The reason for the racial differences is not known. Several refinements have been proposed to increase the sensitivity of the test for younger men more likely to die of the disease, while reducing the frequency of diagnosing cancers of low biologic potential in elderly men more likely to die of other causes. These modifications include age-specific reference ranges, using a lower "upper" limit of normal for younger males and higher "upper" limit for older individuals. Prostate-specific antigen density (PSAD) is calculated by dividing the serumPSAlevel by the estimated prostate weight calculated from transrectal ultrasonography (TRUS). It was proposed to correct for the influence of benign prostatic hyperplasia (BPH) on the measured level of PSA. Values0.15 suggest the presence of cancer. PSAD levels also increase with age. PSA velocity is derived from calculations of the rate of change inPSAbefore the diagnosis of cancer was established. Increases of>0.75 ng/mL per year are suggestive of cancer. For a 50-year-old male, an increase from 2.5 to 3.9 ng/mL in a 1-year period would warrant further testing, even though the level is still in the "normal" range. Free and complexed PSA measurements are used to determine which men require a biopsy when the PSA level is in the range of 4 to 10 ng/mL. In cancer, the level of free PSA is lower. Using a 25% threshold of free PSA for patients with levels in the range of 4 to 10 ng/mL, specificity was improved by 20% while maintaining a sensitivity of 95%. Further refinements to increase the specificity of distinguishing benign and malignant conditions involve the determination of the ratios of free to total, complexed to total, and free to complexed PSA. Using normal ranges for free/total PSA of >0.15, for complexed/total PSA of 0.25 improved specificity in one study by 20%. These modifications are designed to reduce the frequency of biopsies in men without cancer. Figure 95-1 illustrates a diagnostic algorithm based on the DREand PSA findings. Transrectal Ultrasonography Most cancers are hypoechoic by ultrasonography. Unfortunately, no single finding on ultrasound permits univeral distinction between cancer and benign conditions and identification of extracapsular disease. Cancers 95% are adenocarcinomas; the remainder include squamous cell tumors, transitional cell tumors, and, rarely, carcinosarcomas. Metastases to the prostate are rare, but in some cases, transitional cell tumors originating in the bladder or colonic lesions may invade the gland directly. In the evaluation for adenocarcinoma, each core is examined for the presence or absence of cancer. When cancer is identified, the extent and grade are assessed and the presence or absence of perineural invasion or extracapsular extension reported. Histologic grade is based most commonly on the Gleason system, in which the dominant and secondary glandular histologic patterns are independently assigned numbers from 1 to 5 (best to least differentiated) and summed to give a total score of 2 to 10 for each tumor. The grading is reproducible and correlates with clinical outcomes. The most poorly differentiated area of tumor (i.e., the area with the highest histologic grade) often determines biologic behavior. STAGING The TNM staging system (Table 95-2) includes categories for cancers identified solely on the basis of an abnormalPSA with no palpable abnormalities onDRE(T1c), those that are palpable but clinically confined to the gland (T2), and those that have extended outside of the gland (T3 and T4). The presence or absence of nodal (N) and distant metastases (M) are also recorded. Clinical staging alone is inaccurate in assessing capsular invasion and the probability of spread to nodal or more distant sites. To refine this assessment, the TNM system has been modified to incorporate the results of imaging studies such as ultrasound or magnetic resonance imaging (MRI) in the assignment of T stage. Computed tomography (CT) scans lack sensitivity and specificity to detect extraprostatic extension and in visualization of lymph nodes. MRI is an improvement, particularly with an endorectal coil and is superior to CT. T1-weighted images demonstrate the periprostatic fat, periprostatic venous plexus, perivesicular tissues, lymph nodes, and bone marrow. T2-weighted images demonstrate the internal architecture of the prostate and seminal vesicles. Most cancers have a low signal, while the normal peripheral zone has a high signal. Nevertheless, MRI lacks sensitivity and specificity. No single test accurately predicts pathologic stage at surgery. Another limitation of the TNM system is that the majority of men are now being diagnosed with T1c or T2 disease. Thus, to refine the prediction of local disease extent, most groups are now using multiplex staging models based on a combination of the findings of theDRE, biopsy, Gleason score, and baselinePSA(Table 95-3). Others are developing models based on the number of cores and the percentage of each core
involved by tumor. This information can be used to assist patients in selecting treatments, although it remains controversial how recommendations should be affected by a particular level of probability of node-positive disease. These same parameters are also being used to assess the probability of cure (Fig. 95-2). Some tumors that have extended beyond the confines of the gland may still be curable, while others that are still organ-confined may not. Because successful surgery removes all prostatic tissues, both benign and malignant, and radiation therapy eliminates only the malignant component, different definitions of cure are needed depending on the modality used. Thus, following surgery, thePSAlevel should become undetectable; following radiation therapy, it should generally fall to 10-year life expectancy. Some physicians consider observation only for patients with low-grade tumors (Gleason score£6) that do not involve more than a small percentage of a single core. Radical Prostatectomy The objective of a radical prostatectomy is the removal of all prostate tissue with a clear margin of resection, preservation of the external sphincter to maintain continence, and sparing of the autonomic nerves in the neurovascular bundle so that potency is retained. The procedure is performed through a retropubic or perineal approach. In contemporary series, hospital stays are short; mortality 10 years. The operation is not justified in men with a life expectancy 10 were 98%, 90%, and 89%, respectively. Nevertheless, many physicians feel that implantation is best reserved for patients with good or intermediate prognostic features. Overall, the procedure is well tolerated, although most patients experience urinary frequency and urgency, which can persist for several months. Incontinence has been seen in 2 to 4% of cases. Higher complication rates are observed in patients who have undergone a priorTURP or who have obstructive symptoms at baseline. Proctitis has been reported in1 year after primary treatment tend to be localized, while those recurring in35 U/mL. However, in postmenopausal women with an asymptomatic pelvic mass and CA-125 levels³65 U/mL, the test has a sensitivity of 97% and a specificity of 78%. Screening In contrast to patients who present with advanced disease, patients with early ovarian cancers (stages I and II) are commonly curable with conventional therapy. Thus, effective screening procedures would improve the cure rate in this disease. Although pelvic examination can occasionally detect early disease, it is a relatively insensitive screening procedure. Transvaginal sonography has replaced the slower and less sensitive abdominal sonography, but significant false-positive results are noted, particularly in premenopausal women. In one study, 67 laparotomies were required to diagnose 1 primary ovarian cancer. Doppler flow imaging coupled with transvaginal ultrasound may improve accuracy and reduce the high rate of false positives. CA-125 has been studied as a screening tool. Unfortunately, half of women with stages I and II ovarian cancer have CA-125 levels 10 to 15 cm have usually already spread into the intraabdominal space. Spread eventually results in intraabdominal carcinomatosis, which leads to bowel and renal obstruction and cachexia. Although most ovarian tumors are epithelial, two other important ovarian tumor types exist -- stromal and germ cell tumors. These tumors are distinct in their cell of origin but also have different clinical presentations and natural histories and are often managed differently (see below). Metastasis to the ovary can occur from breast, colon, gastric, and pancreatic cancers, and the Krukenberg tumor was classically described as bilateral ovarian masses from metastatic mucin-secreting gastrointestinal cancers. Staging and Prognostic Factors Laparotomy is often the primary procedure used to establish the diagnosis. Less invasive studies useful in defining the extent of spread include chest x-rays, abdominal computed tomography scans, and abdominal and pelvic sonography. If the woman has specific gastrointestinal symptoms, a barium enema or gastrointestinal series can be performed. Symptoms of bladder or renal dysfunction can
be evaluated by cystoscopy or intravenous pyelography. A careful staging laparotomy will establish the stage and extent of disease and allow for the cytoreduction of tumor masses in patients with advanced disease. Proper laparotomy requires a vertical incision of sufficient length to ensure adequate examination of the abdominal contents. The presence, amount, and cytology of any ascites fluid should be noted. The primary tumor should be evaluated for rupture, excrescences, or dense adherence. Careful visual and manual inspection of the diaphragm and peritoneal surfaces is required. In addition to total abdominal hysterectomy and bilateral salpingo-oophorectomy, a partial omentectomy should be performed and the paracolic gutters inspected. Pelvic lymph nodes as well as para-aortic nodes in the region of the renal hilus should be biopsied. Since this surgical procedure defines stage, establishes prognosis, and determines the necessity for subsequent therapy, it should be performed by a surgeon with special expertise in ovarian cancer staging. Studies have shown that patients operated upon by gynecologic oncologists were properly staged 97% of the time, compared to 52 and 35% of cases staged by obstetricians/gynecologists and general surgeons, respectively. At the end of staging, 23% of women have stage I disease (cancer confined to the ovary or ovaries); 13% have stage II (disease confined to the true pelvis); 47% have stage III (disease spread into but confined to the abdomen); and 16% have stage IV disease (spread outside the pelvis and abdomen). The 5-year survival correlates with stage of disease: stage I -- 90%, stage II -- 70%, stage III -- 15 to 20%, and stage IV -- 1 to 5% (Table 97-1). Prognosis in ovarian cancer is dependent not only upon stage but on the extent of residual disease and histologic grade. Patients presenting with advanced disease but left without significant residual disease after surgery have a median survival of 39 months, compared to 17 months for those with suboptimal tumor resection. Prognosis of epithelial tumors is also highly influenced by histologic grade but less so by histologic type. In early-stage disease, survival is better in mucinous adenocarcinoma than endometrial and serous types, and clear cell carcinomas have the worst prognosis. Although grading systems differ among pathologists, all grading systems show a better prognosis for well- or moderately differentiated tumors and a poorer prognosis for poorly differentiated histologies. Typical 5-year survivals for patients with all stages of disease are: well differentiated -- 88%, moderately differentiated -- 58%, poorly differentiated -27%. The prognostic significance of pre- and postoperative CA-125 levels is uncertain. Serum levels generally reflect volume of disease, and high levels usually indicate unresectability and a poorer survival. Postoperative levels, if elevated, usually indicate residual disease. Nevertheless, on multivariate analysis, CA-125 is not an independent prognostic factor because of the association with volume of disease. The rate of decline of CA-125 levels during initial therapy or the absolute level after one to three cycles of chemotherapy correlates with prognosis but is not sufficiently accurate to guide individual treatment decisions. Even when the CA-125 level falls to normal after surgery or chemotherapy, "second-look" laparotomy identifies residual disease in 60% of women. Other more quantitative approaches to define prognosis include ploidy analysis and image cytometry (automated analysis of cell morphology); they remain
investigational. Genetic and biologic factors may influence prognosis. Increased tumor levels of p53 are associated with a worse prognosis in advanced disease. Epidermal growth factor receptors in ovarian cancer are associated with a high risk of progression, but the increased expression of HER-2/neu has given conflicting prognostic results, and expression of Mdr-1 has not been of prognostic value. HER-2/neu is being evaluated as a target for antibody therapy. TREATMENT The selection of therapy for patients with epithelial ovarian cancer depends upon the stage, extent of residual tumor, and histologic grade. In general, patients are considered in three separate treatment groups: (1) those with early (stages I and II) ovarian cancer and microscopic or no residual disease; (2) patients with advanced (stage III) disease but minimal residual tumor (15% response rates in patients relapsing after initial combination chemotherapy include gemcitabine, topotecan, ifosfamide, etoposide, and hexamethylmelamine. Intraperitoneal chemotherapy (usually cisplatin) may be used if a small residual volume (3.5 mmol/L) include confusion, lethargy, coma, and death. The cancers associated withHHM are non-small cell lung cancer and cancers of the breast, kidney, head and neck, and bladder. HHM is particularly common in patients with cancers of squamous cell histology. Hypercalcemia is uncommon at presentation (3.50 mmol/L (>14 mg/dL)] with alteration of mental status can be treated with all of the above plus salmon calcitonin, 4 to 8 U/kg, administered intramuscularly or subcutaneously every 12 h. Calcitonin administration will decrease the serum calcium within 24 h, and its hypocalcemic effect can be prolonged in patients with LOH by adding glucocorticoids. If these agents are not effective in reducing the serum calcium, plicamycin and gallium nitrate may be added. HYPONATREMIA OF MALIGNANCY Hyponatremia of malignancy (Na+ level52% in men; >48% in women) that is detected on a routine blood count. Approximately 3% of patients with renal cell cancer, 10% of patients with hepatoma, and 15% of patients with cerebellar hemangioblastomas have erythrocytosis. In most cases the erythrocytosis is asymptomatic. Patients with erythrocytosis due to a renal cell cancer, hepatoma, or central nervous system cancer should have measurement of red cell mass. If the red cell mass is elevated, the serum erythropoietin level should then be measured. Patients with an appropriate cancer, elevated erythropoietin levels, and no other explanation for erythrocytosis (e.g., a hemoglobinopathy that causes increased O2affinity, Chap. 106) have the paraneoplastic syndrome. TREATMENT Successful resection of the cancer usually resolves the erythrocytosis. If the tumor neither can be resected nor treated effectively with radiation therapy or chemotherapy, phlebotomy may control any symptoms related to erythrocytosis. GRANULOCYTOSIS Approximately 30% of patients with solid tumors have granulocytosis (granulocyte count>8000/uL). In about half of patients with granulocytosis and cancer, the granulocytosis has an identifiable nonparaneoplastic etiology (infection, tumor necrosis, glucocorticoid administration, etc.). The other patients have proteins in urine and serum that stimulate the growth of bone marrow cells. Tumors and tumor cell lines from patients with lung, ovarian, and bladder cancers have been documented to produce granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), and/orIL-6. However, the etiology of granulocytosis has not been characterized in most patients. Patients with granulocytosis are nearly all asymptomatic, and the differential white blood cell count does not have a shift to immature forms of neutrophils. Granulocytosis occurs in 40% of patients with lung and gastrointestinal cancers, 20% of patients with breast cancer, 30% of patients with brain tumors and ovarian cancers, and 10% of patients with renal cell carcinoma. Patients with advanced-stage disease are more likely to have granulocytosis than those with early-stage disease. Paraneoplastic granulocytosis does not require treatment. The granulocytosis resolves
when the underlying cancer is successfully treated. THROMBOCYTOSIS Thirty-five percent of patients with thrombocytosis (platelet count>400,000/uL) have an underlying diagnosis of cancer.IL-6, a candidate molecule for the etiology of paraneoplastic thrombocytosis, stimulates the production of platelets in vitro and in vivo. Some patients with cancer and thrombocytosis have elevated levels of IL-6 in plasma. Another candidate molecule is thrombopoietin, a peptide hormone that stimulates megakaryocyte proliferation and platelet production. The etiology of thrombocytosis has not been established in most cases. Patients with thrombocytosis are nearly all asymptomatic. Thrombocytosis is not clearly linked to thrombosis in patients with cancer. Thrombocytosis is present in 40% of patients with lung and gastrointestinal cancers, 20% of patients with breast, endometrial, and ovarian cancers, and 10% of patients with lymphoma. Patients with thrombocytosis are more likely to have advanced-stage disease and have a poorer prognosis than patients without thrombocytosis. Paraneoplastic thrombocytosis does not require treatment. EOSINOPHILIA Eosinophilia is present in ~1% of patients with cancer. Tumors and tumor cell lines from patients with lymphomas or leukemia may produceIL-5, which stimulates eosinophil growth. Activation of IL-5 transcription in lymphomas and leukemias may involve translocation of the long arm of chromosome 5, to which the genes for IL-5 and other cytokines map. Patients with eosinophilia are typically asymptomatic. Eosinophilia is present in 10% of patients with lymphoma, 3% of patients with lung cancer, and occasional patients with cervical, gastrointestinal, renal, and breast cancer. Patients with markedly elevated eosinophil counts (>5000/uL) can develop shortness of breath and wheezing. A chest radiograph may reveal diffuse pulmonary infiltrates from eosinophil infiltration and activation in the lungs. TREATMENT Definitive treatment is directed at the underlying malignancy: tumors should be resected or treated with radiation or chemotherapy. In most patients who develop shortness of breath related to eosinophilia, symptoms resolve with the use of oral or inhaled glucocorticoids. THROMBOPHLEBITIS Deep venous thrombosis and pulmonary embolism are the most common thrombotic conditions in patients with cancer. Migratory or recurrent thrombophlebitis may be the initial manifestation of cancer. Approximately 15% of patients who develop deep venous thrombosis or pulmonary embolism have a diagnosis of cancer (Chap. 117). The coexistence of peripheral venous thrombosis with visceral carcinoma, particularly
pancreatic cancer, is called Trousseau's syndrome. Pathogenesis Patients with cancer are predisposed to thromboembolism because they are often at bedrest or immobilized, and tumors may obstruct or slow blood flow. In addition, clotting may be promoted by release of procoagulants or cytokines from tumor cells or associated inflammatory cells, or by platelet adhesion or aggregation. The specific molecules that mediate the increased risk of thromboembolism have not been identified. Clinical Manifestations Patients with cancer who develop deep venous thrombosis usually develop swelling or pain in the leg, and physical examination reveals tenderness, warmth, and redness. Patients who present with pulmonary embolism develop dyspnea, chest pain, and syncope, and physical examination shows tachycardia, cyanosis, and hypotension. Approximately 5% of patients with no history of cancer who have a diagnosis of deep venous thrombosis or pulmonary embolism will have a diagnosis of cancer within 1 year. The most common cancers associated with thromboembolic episodes include lung, pancreatic, gastrointestinal, breast, ovarian, and genitourinary cancers, lymphomas, and brain tumors. Patients with cancer who undergo surgical procedures requiring general anesthesia have a 20 to 30% risk of deep venous thrombosis. Diagnosis The diagnosis of deep venous thrombosis in patients with cancer is made by impedance plethysmography or bilateral compression ultrasonography of the leg veins. Patients with a noncompressible venous segment have deep venous thrombosis. If compression ultrasonography is normal and a high clinical suspicion exists for deep venous thrombosis, venography should be done to look for a luminal filling defect. Elevation of D-dimer is not as predictive of deep venous thrombosis in patients with cancer as in patients without cancer. Patients with symptoms and signs suggesting a pulmonary embolism should be evaluated with a chest radiograph, electrocardiogram, arterial blood gas analysis, and ventilation-perfusion scan. Patients with mismatched segmental perfusion defects have a pulmonary embolus. Patients with equivocal ventilation-perfusion findings should be evaluated as described above for deep venous thrombosis in their legs. If deep venous thrombosis is detected, they should be anticoagulated. If deep venous thrombosis is not detected, they should be considered for a pulmonary angiogram. Patients without a diagnosis of cancer who present with an initial episode of thrombophlebitis or pulmonary embolus need no additional tests for cancer other than a careful history and physical exam. In light of the many possible primary sites, diagnostic testing in asymptomatic patients is wasteful. However, if the clot is refractory to standard treatment or is in an unusual site, or if the thrombophlebitis is migratory or recurrent, efforts to find an underlying cancer are indicated. TREATMENT Patients with cancer and a diagnosis of deep venous thrombosis or pulmonary embolism should be treated initially with intravenous unfractionated heparin or low molecular weight heparin for at least 5 days and coumadin started within 1 or 2 days.
The coumadin dose should be adjusted so the INR is 2 to 3. Patients with proximal deep venous thrombosis and a relative contraindication to heparin anticoagulation (hemorrhagic brain metastases or pericardial effusion) should be considered for placement of a filter in the inferior vena cava (Greenfield filter) to prevent pulmonary embolism. Coumadin should be administered for 3 to 6 months. Patients with cancer who undergo a major surgical procedure should be considered for heparin prophylaxis or pneumatic boots. Breast cancer patients undergoing chemotherapy and patients with implanted catheters should be considered for prophylaxis (1 mg coumadin per day).*Cutaneous paraneoplastic syndromes are discussed in Chap. 57. Neurologic paraneoplastic syndromes are discussed in Chap. 101. More extensive discussion of functional endocrine tumors is given in Chap. 93. (Bibliography omitted in Palm version) Back to Table of Contents
101. PARANEOPLASTIC NEUROLOGIC SYNDROMES - Muhammad T. Al-Lozi, Alan Pestronk GENERAL PRINCIPLES A paraneoplastic neurologic syndrome (PNNS) is a neurologic disorder that is associated with a neoplasm but lies anatomically remote from it. Paraneoplastic disorders are caused by immune or other mechanisms and are not due to direct effects of the tumor itself, metastases, opportunistic infections, complications of drug or radiation therapy, or malnutrition. Clinical features of a PNNS are often distinctive. Onset can be dramatic, arising subacutely over weeks or even days to produce neurologic symptoms that may be profoundly disabling. PNNSassociated with autoantibodies can be grouped into (1) disorders in which the neoplasm contains a surface antigen or intracellular protein that is the antigenic target, and (2) monoclonal gammopathy syndromes associated with secretion of an antibody by the neoplasm. Each subgroup has typical clinical, pathologic, and immune characteristics (Table 101-1). Some PNNS, including lymphoma-associated motor neuropathy, subacute necrotic myopathy, dermatomyositis, and necrotizing myopathy, have no currently identified antibody or target antigen and are not yet classifiable in this scheme. The temporal relationship of aPNNS to the associated neoplasm is variable. The PNNS may precede or follow the identification of a neoplasm by weeks, months, or occasionally years. The strength of the association between neoplasms and PNNS varies with different syndromes, different neoplasms, and the clinical context. In some PNNS, such as the sensory neuronopathy associated with anti-Hu antibodies, the association with neoplasm is very strong. By contrast, the Lambert-Eaton myasthenic syndrome (LEMS) is associated with neoplasm in approximately 50% of cases only; the relationship is probably stronger in older individuals who have a history of cigarette smoking. Disorders that are clinically identical to most PNNS also occur in the absence of cancer. Nonetheless, the development of a PNNS in a previously healthy individual should in most circumstances prompt a thorough search for its associated neoplasms. PNNS-associated neoplasms vary considerably in terms of malignancy. Tumors associated with subacute necrotic myelopathy are often severe and unresponsive to therapy. In other syndromes, the tumor either remains small or can be effectively treated; in such cases, the long-term prognosis is determined by the effectiveness of management of the paraneoplastic syndrome. Thymoma associated with myasthenia gravis is an example of a tumor in this category. In some PNNS, such as those associated with small cell lung cancer (SCLC) and anti-Hu antibodies, it has been suggested that the presence of the autoantibodies may confer a more favorable prognosis by inhibiting tumor growth. As outlined inTable 101-2, certain syndromes are associated with particular types of tumors, and more than one syndrome may occur with a given neoplasm. For example,SCLCsare associated with a variety ofPNNS, including limbic encephalitis, cerebellar ataxia, opsoclonus-myoclonus, necrotic myelopathy, sensory neuronopathy, autonomic neuropathy, andLEMS.
Prevalence estimates vary with the particular syndrome. Tumors that are most often associated withPNNS are cancers of the lung, stomach, breast, ovary, and colon. Some 30% of patients with thymoma also develop myasthenia gravis as a paraneoplastic syndrome. A poorly characterized neuromyopathy with proximal weakness and distal sensory loss is very common in patients who have lost more than 15% of their body weight. In contrast, most of the well-defined PNNS are rare, with estimated prevalence rates of 15% of baseline weight). The neuropathy is characterized by distal, symmetric sensory loss and paresthesias, which may be painful, and by weakness and muscle wasting, which is especially prominent in the distal legs. Pathologically there is noninflammatory degeneration of axons and mild myelin loss, presumably secondary to the axonopathy. An accompanying myopathy with atrophy of type II muscle fibers may produce proximal muscle weakness. Axonal loss, with low-amplitude sensory and motor amplitudes and normal conduction velocities, is seen on electrophysiologic studies. This neuromyopathy has been described in association with a variety of solid tumors (lung, breast, stomach), lymphoma, and plasma cell dyscrasia. Successful treatment of the neoplasm may result in improvement or stabilization of the neuromyopathy. Other axonal neuropathies may bePNNS, but their associations with neoplasms are less clearly established. Peripheral nerve vasculitis, producing mononeuritis multiplex or
asymmetric sensorimotor polyneuropathy, has been reported with lymphomas or carcinoma of the lung, prostate, kidney, or stomach. Polyneuropathy, presenting with either a subacute mononeuritis multiplex or a slowly progressive distal symmetric sensorimotor polyneuropathy, occurs in approximately 20% of patients with cryoglobulinemia. Guillain-Barre syndrome (Chap. 378) may be associated with Hodgkin's disease; it is characterized by subacute motor weakness; sensory loss, which is often mild in comparison with the motor deficits; areflexia; and a characteristic elevation of spinal fluid protein concentration without pleocytosis. Enteric autonomic neuropathy with anti-Hu antibodies, commonly presenting as intestinal pseudoobstruction, has been described in association withSCLC. NEUROMUSCULAR JUNCTION Lambert-Eaton myasthenic syndrome is a disorder of the presynaptic component of neuromuscular transmission. Common symptoms in LEMS are weakness, fatigue, and dryness of the mouth. Some patients complain of paresthesias, myalgia, or impotence. Weakness is symmetric, proximal, and most prominent in the lower limbs. Strength can decrease with rest and improve with exercise. Ocular (diplopia and ptosis) and bulbar (dysphagia and dysarthria) symptoms may occur in some patients. Respiratory muscle weakness is rare. Tendon reflexes are either diminished or absent at rest but may increase after exercise. About 50% of patients have an associated neoplasm, most commonlySCLC and less often a lymphoproliferative disorder. About 3% of patients with SCLC haveLEMS. Almost all patients with both SCLC and LEMS have a smoking history. LEMS may precede the detection of cancer by 2 to 3 years. The most useful diagnostic test for LEMS is repetitive nerve stimulation, specifically the finding of compound muscle action potential (CMAP) amplitudes that are small at rest but increase by at least 100% after rapid repetitive nerve stimulation (30 to 50 Hz) or maximal muscle contraction sustained for at least 10 s. LEMS is believed to be an autoimmune disorder associated with diminished quantal release of acetylcholine. IgG antibodies directed against P/Q voltage-gated calcium channels (VGCC) in the motor nerve terminal are found in the sera of ~90% of patients with LEMS and in nearly 100% when LEMS is associated with neoplasm. False-positive findings occur with hypergammaglobulinemia, chronic liver disorders, and (in15% is present. Muscle wasting is more prominent than muscle weakness in this syndrome. Inflammatory myopathy, especially dermatomyositis in older females, may occur in association with a variety of neoplasms (Chap. 382). Necrotizing myopathy presents with a subacute onset of weakness that is typically proximal and ranges from mild to severe. Some patients experience myalgia in addition to muscle weakness. The serum creatine kinase levels are very high. Muscle fiber necrosis is the predominant finding in muscle biopsy. Necrotizing myopathy is most commonly seen with adenocarcinoma and non-small cell cancer of the lung but may also be associated with a variety of other neoplasms. The overall prognosis depends on the malignancy of the associated neoplasm. Weakness may improve following tumor resection or glucocorticoid treatment. Chronic proximal myopathies have been described with an IgM M-protein binding to decorin; scleromyxedema with IgG or IgA M-proteins; a rippling muscle disease has been reported to occur with thymoma. Hormone-secreting (ACTH or parathyroid hormone-like) tumors may also be associated with proximal myopathies. (Bibliography omitted in Palm version) Back to Table of Contents
102. ONCOLOGIC EMERGENCIES - Rasim Gucalp, Janice Dutcher Emergencies in patients with cancer may be classified into three groups: pressure or obstruction caused by a space-occupying lesion, metabolic or hormonal problems (paraneoplastic syndromes, Chap. 100), and complications arising from the effects of treatment. STRUCTURAL-OBSTRUCTIVE ONCOLOGIC EMERGENCIES SUPERIOR VENA CAVA SYNDROME Superior vena cava syndrome (SVCS) is the clinical manifestation of superior vena cava (SVC) obstruction, with severe reduction in venous return from the head, neck, and upper extremities. Malignant tumors, such as lung cancer, lymphoma, and metastatic tumors, are responsible for more than 90% of all SVCS cases. Lung cancer, particularly of small-cell and squamous-cell histologies, accounts for approximately 85% of all cases of malignant origin. Metastatic cancers to the mediastinum, such as testicular and breast carcinomas, account for a small proportion of cases. Other causes include benign tumors, aortic aneurysm, thyroid enlargement, thrombosis, and fibrosing mediastinitis caused by prior irradiation or histoplasmosis. Patients withSVCSusually present with neck and facial swelling (especially around the eyes), dyspnea, and cough. Other symptoms include hoarseness, tongue swelling, headaches, nasal congestion, epistaxis, hemoptysis, dysphagia, pain, dizziness, syncope, and lethargy. Bending forward or lying down may aggravate the symptoms. The characteristic physical findings are dilated neck veins, an increased number of collateral veins covering the anterior chest wall, cyanosis, and edema of the face, arms, and chest. More severe cases include proptosis, glossal and laryngeal edema, and obtundation. The clinical picture is milder if the obstruction is located above the azygos vein. The diagnosis ofSVCS is a clinical one. The most significant chest radiographic finding is widening of the superior mediastinum, most commonly on the right side. Pleural effusion occurs in only 25% of patients, often on the right side. However, a normal chest radiograph is still compatible with the diagnosis if other characteristic findings are present. Computed tomography (CT) provides the most reliable view of the mediastinal anatomy. The diagnosis of SVCS requires diminished or absent opacification of central venous structures with prominent collateral venous circulation. Magnetic resonance imaging (MRI) has no advantages over CT. Invasive procedures, including bronchoscopy, percutaneous needle biopsy, mediastinoscopy, and even thoracotomy, can be performed by a skilled clinician without any major risk of bleeding. For patients with a known cancer, a detailed workup usually is not necessary, and appropriate treatment may be started after obtaining a CT scan of the thorax. For those with no history of malignancy, a detailed evaluation is absolutely necessary to rule out benign causes and determine a specific diagnosis to direct the appropriate therapy. TREATMENT The one potentially life-threatening complication of a superior mediastinal mass is
tracheal obstruction. Upper airway obstruction demands emergent therapy. Diuretics with a low salt diet, head elevation, and oxygen may produce temporary symptomatic relief. Radiation therapy is the primary treatment forSVCScaused by non-small cell lung cancer and other metastatic solid tumors. Chemotherapy is effective when the underlying cancer is small cell carcinoma of the lung or lymphoma. Recurrent SVCS occurs in 10 to 30% of patients after initial therapy; it may be palliated with the use of intravascular self-expanding stents (Fig. 102-1). Surgery may provide immediate relief for patients in whom a benign process is the cause. Clinical improvement occurs in most patients, although this improvement may be due to the development of adequate collateral circulation. The mortality associated withSVCSdoes not relate to caval obstruction, but rather to the underlying cause. SVCS and Central Venous Catheters in Adults The use of long-term central venous catheters has become common practice in patients with cancer. Major vessel thrombosis may occur. In these cases, catheter removal should be combined with anticoagulation to prevent embolization. SVCS in this setting, if detected early, can be treated successfully by fibrinolytic therapy without sacrificing the catheter. Warfarin (1 mg/d) reduces the incidence of thrombosis without altering coagulation tests. PERICARDIAL EFFUSION/TAMPONADE Malignant pericardial disease is found at autopsy in 5 to 10% of patients with cancer, most frequently with lung cancer, breast cancer, leukemias, and lymphomas. Cardiac tamponade as the initial presentation of extrathoracic malignancy is rare. The origin is not malignancy in about 50% of cancer patients with symptomatic pericardial disease, but can be related to irradiation, drug-induced pericarditis, hypothyroidism, idiopathic pericarditis, infection, or autoimmune diseases. Two types of radiation pericarditis have been described: an acute inflammatory, effusive pericarditis occurring within months of irradiation, which usually resolves spontaneously, and a chronic effusive pericarditis that may appear up to 20 years after radiotherapy and is accompanied by a thickened pericardium. Most patients with pericardial metastasis are asymptomatic. However, the common symptoms are dyspnea, cough, chest pain, orthopnea, and weakness. Pleural effusion, sinus tachycardia, jugular venous distension, hepatomegaly, peripheral edema, and cyanosis are the most frequent physical findings. Relatively specific diagnostic findings, such as paradoxical pulse, diminished heart sounds, pulsus alternans (pulse waves alternating between those of greater and lesser amplitude with successive beats), and friction rub are less common than with nonmalignant pericardial disease. Chest radiographs and ECG reveal abnormalities in 90% of patients, but half of these abnormalities are nonspecific. Echocardiography is the most helpful diagnostic test. Pericardial fluid may be serous, serosanguineous, or hemorrhagic, and cytologic examination of pericardial fluid is diagnostic in most patients. False negative cytology may occur in patients with lymphoma and mesothelioma. TREATMENT
Pericardiocentesis with or without the introduction of sclerosing agents, the creation of a pericardial window, complete pericardial stripping, cardiac irradiation, or systemic chemotherapy are effective treatments. Acute pericardial tamponade with life-threatening hemodynamic instability requires immediate drainage of fluid. This can be quickly achieved by pericardiocentesis. Alternatively, subxyphoid pericardiotomy can be performed in 45 min under local anesthesia. INTESTINAL OBSTRUCTION Intestinal obstruction and reobstruction are common problems in patients with advanced cancer, particularly colorectal or ovarian carcinoma. However, other cancers, such as lung or breast cancer and melanoma, can metastasize within the abdomen, leading to intestinal obstruction. Typically, obstruction occurs at multiple sites. Intestinal pseudoobstruction is caused by infiltration of the mesentery or bowel muscle by tumor, involvement of the celiac plexus, or paraneoplastic neuropathy in patients with small cell lung cancer. Paraneoplastic neuropathy is associated with IgG antibodies reactive to neurons of the myenteric and submucosal plexuses of the jejunum and stomach. Ovarian cancer can lead either to authentic luminal obstruction or to pseudoobstruction that results when circumferential invasion of a bowel segment arrests the forward progression of peristaltic contractions. The onset of obstruction is usually insidious. Pain is the most common symptom and is usually colicky in nature. Pain can also be due to abdominal distention, tumor masses, or hepatomegaly. Vomiting can be intermittent or continuous. Patients with complete obstruction usually have constipation. Physical examination may reveal abdominal distention with tympany, ascites, visible peristalsis, high-pitched bowel sounds, and tumor masses. Erect plain abdominal films may reveal multiple air-fluid levels and dilation of the small or large bowel. Acute cecal dilation to more than 12 to 14 cm is considered a surgical emergency because of the high likelihood of rupture. The overall prognosis for the patient with cancer who develops intestinal obstruction is poor; median survival is 3 to 4 months. About one-fourth to one-third of patients are found to have intestinal obstruction due to causes other than cancer. Adhesions from previous operations are a common benign cause. Ileus induced by vincristine is another reversible cause. TREATMENT The management of intestinal obstruction in patients with advanced malignancy depends on the extent of the underlying malignancy and the functional status of the major organs. The initial management should include surgical evaluation. Operation is not always successful and may lead to further complications with a substantial mortality rate (10 to 20%). Self-expanding metal stents placed in the gastric outlet, duodenum, proximal jejunum, colon, or rectum may palliate obstructive symptoms at those sites without major surgery. Patients known to have advanced intraabdominal malignancy should receive a prolonged course of conservative management, including nasogastric decompression. Treatment with antiemetics, antispasmodics, and analgesics may allow patients to remain outside the hospital. The somatostatin analogue octreotide may relieve obstructive symptoms through its inhibitory effect on gastrointestinal secretion.
URINARY OBSTRUCTION Urinary obstruction may occur in patients with prostatic or gynecologic malignancies, particularly cervical carcinoma, or metastatic disease from other primary sites. Radiation therapy to pelvic tumors may cause fibrosis and subsequent ureteral obstruction. Bladder outlet obstruction is usually due to prostate and cervical cancers and may lead to bilateral hydronephrosis and renal failure. Flank pain is the most common symptom. Persistent urinary tract infection, persistent proteinuria, or hematuria in patients with a cancer should raise suspicion of ureteral obstruction. Total anuria and/or anuria alternating with polyuria may occur. A slow, continuous rise in the serum creatinine level necessitates immediate evaluation in patients with cancer. Renal ultrasound examination is the safest and cheapest way to identify hydronephrosis. The function of an obstructed kidney can be evaluated by a nuclear scan.CT can be helpful in identifying a retroperitoneal mass or retroperitoneal adenopathy. TREATMENT Obstruction associated with flank pain, sepsis, or fistula formation is an indication for immediate palliative urinary diversion. There are many newer techniques by which internal ureteral stents can be placed under local anesthesia. Percutaneous nephrostomy offers an alternative approach for drainage. In the case of bladder outlet obstruction due to malignancy, a suprapubic cystostomy can be used for urinary drainage. MALIGNANT BILIARY OBSTRUCTION This common clinical problem can be caused by a primary carcinoma arising in the pancreas, ampulla of Vater, bile duct, or liver or by metastatic disease to the periductal lymph nodes or liver parenchyma. The most common metastatic tumors causing biliary obstruction are gastric, colon, breast, and lung cancers. Jaundice, light-colored stools, dark urine, pruritus, and weight loss due to malabsorption are usual symptoms. Pain and secondary infection are uncommon in malignant biliary obstruction. Ultrasound,CT, or percutaneous transhepatic or endoscopic retrograde cholangiography will identify the site and nature of the biliary obstruction. TREATMENT Palliative intervention is indicated only in patients with disabling pruritus resistant to medical treatment, severe malabsorption, or infection. Stenting under radiographic control, surgical bypass, or radiation therapy with or without chemotherapy may alleviate the obstruction. The choice of modality should be based on the site of obstruction (proximal versus distal), the type of tumor (sensitive to radiotherapy, chemotherapy, or neither), and the general condition of the patient. In the absence of pruritus, biliary obstruction may be a largely asymptomatic cause of death. SPINAL CORD COMPRESSION
Spinal cord compression occurs in 5 to 10% of patients with cancer. Epidural tumor is the first manifestation of malignancy in about 10% of patients. The underlying cancer is usually identified during the initial evaluation; lung cancer is most commonly the primary malignancy. Metastatic tumor involves the vertebral column more often than any other part of the bony skeleton. Lung, breast, and prostate cancer are the most frequent offenders. Multiple myeloma also has a high incidence of spine involvement. The thoracic spine is the most common site (70%), followed by the lumbosacral spine (20%) and the cervical spine (10%). Involvement of multiple sites is most frequent in patients with breast and prostatic carcinoma. Cord injury develops when metastases to the vertebral body or pedicle enlarge and compress the underlying dura. Another cause of cord compression is direct extension of a paravertebral lesion through the intervertebral foramen. These cases usually involve a lymphoma, myeloma, or pediatric neoplasm. Parenchymal spinal cord metastasis due to hematogenous spread is rare. The most common initial symptom in patients with spinal cord compression is localized back pain and tenderness due to involvement of vertebrae by tumor. Pain is usually present for days or months before other neurologic findings appear. It is exacerbated by movement and by coughing or sneezing. It can be differentiated from the pain of disk disease by the fact that it worsens when the patient is supine. Radicular pain is less common than localized back pain and usually develops later. Radicular pain in the cervical or lumbosacral areas may be unilateral or bilateral. Radicular pain from the thoracic roots is often bilateral and is described by patients as a feeling of tight, band-like constriction around the thorax and abdomen. Typical cervical radicular pain radiates down the arm; in the lumbar region, the radiation is down the legs. Loss of bowel or bladder control may be the presenting symptom, but usually occurs late in the course. On physical examination, pain induced by straight leg raising, neck flexion, or vertebral percussion may help to determine the level of cord compression. Patients develop numbness and paresthesias in the extremities or trunk. Loss of sensibility to pinprick is as common as loss of sensibility to vibration or position. The upper limit of the zone of sensory loss is often one or two vertebrae below the site of compression. Motor findings include weakness, spasticity, and abnormal muscle stretching. The presence of an extensor plantar reflex reflects significant compression. Deep tendon reflexes may be brisk. Motor and sensory loss usually precede sphincter disturbance. Patients with autonomic dysfunction may present with decreased anal tonus, decreased perineal sensibility, and a distended bladder. The absence of the anal wink reflex or the bulbocavernosus reflex confirms cord (conus or cauda equina) involvement. In doubtful cases, evaluation of post-voiding urinary residual volume can be helpful. A residual volume of more than 150 mL suggests bladder dysfunction. Autonomic dysfunction is an unfavorable prognostic factor. Patients with progressive neurologic symptoms should have frequent neurologic examinations and rapid therapeutic intervention. Patients with cancer who develop back pain should be evaluated for spinal cord compression as quickly as possible (Fig. 102-2). Treatment is more often successful in patients who are ambulatory and still have sphincter control at the time treatment is
initiated. Patients should have a neurologic examination and plain films of the spine. Those whose physical examination suggests cord compression should receive dexamethasone (24 mg intravenously every 6 h), starting immediately. Erosion of the pedicles (the "winking owl" sign) is the earliest radiologic finding of vertebral tumor. Other radiographic changes include increased intrapedicular distance, vertebral destruction, lytic or sclerotic lesions, scalloped vertebral bodies, and vertebral body collapse. Vertebral collapse is not a reliable indicator of the presence of tumor; about 20% of cases of vertebral collapse, particularly those in older patients and postmenopausal women, are due not to cancer but to osteoporosis. Also, a normal appearance on plain films of the spine does not exclude the diagnosis of cancer. The role of bone scans in the detection of cord compression is not clear; this method is sensitive but less specific than spinal radiography. The full-length image of the cord provided byMRI is useful. On T1-weighted images, good contrast is noted between the cord, cerebrospinal fluid, and extradural lesions. Owing to their sensitivity in demonstrating the replacement of bone marrow by tumor, MRI can show which parts of a vertebra are involved by tumor (the body, pedicle, lamina, spinous process). MRI also visualizes intraspinal extradural masses compressing the cord. T2-weighted images are most useful for the demonstration of intramedullary pathology. Gadolinium-enhanced MRI can help to characterize and delineate intramedullary disease. MRI is as good as or better than myelography plus postmyelogramCT in detecting metastatic epidural disease with cord compression. Myelography should be reserved for patients who have poor MR images or who cannot undergo MRI promptly. CT in conjunction with myelography enhances the detection of small areas of spinal destruction. In patients with spinal cord compression and an unknown primary tumor, a simple workup including chest radiography, mammography, measurement of prostate-specific antigen, and abdominalCTusually reveals the underlying malignancy. TREATMENT The treatment of patients with spinal cord compression is aimed at relief of pain and restoration of neurologic function (Fig. 102-2). Radiation therapy plus glucocorticoids is generally the initial treatment of choice for spinal cord compression. Up to 75% of patients treated when still ambulatory remain ambulatory, but only 10% of patients with paraplegia recover walking capacity. Indications for surgical intervention include unknown etiology, failure of radiation therapy, a radioresistant tumor type (e.g., melanoma or renal cell cancer), pathologic fracture dislocation, and rapidly evolving neurologic symptoms. Until recently, laminectomy was the standard operation for metastatic spinal cord compression, although results were poor. At present, laminectomy should be used only for tissue diagnosis and for the removal of posteriorly localized epidural deposits in the absence of vertebral disease. Because most cases of epidural spinal cord compression are due to anterior or anterolateral extradural disease, resection of the anterior vertebral body along with the tumor, followed by spinal stabilization, has achieved good results and low mortality rate. Chemotherapy may have a role in patients with chemosensitive tumors
who have had prior radiation therapy to the same region and who are not candidates for surgery. The histology of the tumor is an important determinant of both recovery and survival. Rapid onset and quick progression are poor prognostic features. INCREASED INTRACRANIAL PRESSURE About 25% of patients with cancer die with intracranial metastases. The cancers that most often metastasize to the brain are lung and breast cancers and melanoma. Brain metastases often occur in the presence of systemic disease, and they frequently cause major symptoms, disability, and early death. The signs and symptoms of a metastatic brain tumor are similar to those of other intracranial expanding lesions: headache, nausea, vomiting, behavioral changes, seizures, and focal, progressive neurologic changes. Occasionally the onset is abrupt, resembling a stroke, with the sudden appearance of headache, nausea, vomiting, and neurologic deficits. This picture is usually due to hemorrhage into the metastasis. Melanoma, germ cell tumors, and renal cell cancers have a particularly high incidence of intracranial bleeding. The tumor mass and surrounding edema may cause obstruction of the circulation of cerebrospinal fluid, with resulting hydrocephalus. Patients with increased intracranial pressure may have papilledema with visual disturbances and neck stiffness. As the mass enlarges, brain tissue may be displaced through the fixed cranial openings, producing various herniation syndromes. CTand MRIare equally effective in the diagnosis of brain metastases. CT with contrast should be used as a screening procedure. The CT scan shows brain metastases as multiple enhancing lesions of various sizes with surrounding areas of low-density edema. If a single lesion or no metastases are visualized by contrast-enhanced CT, MRI of the brain should be performed. Gadolinium-enhanced MRI is more sensitive than CT at revealing small lesions, particularly in the brainstem or cerebellum. TREATMENT If signs and symptoms of brain herniation (particularly headache, drowsiness, and papilledema) are present, the patient should be intubated and hyperventilated to maintain PCO2between 25 and 30 mmHg and should receive infusions of mannitol (1 to 1.5 g/kg) every 6 h. Dexamethasone is the best initial treatment for all symptomatic patients with brain metastases (see above). Patients with multiple lesions should receive whole-brain radiation therapy. Patients with a single brain metastasis and with controlled extracranial disease may be treated with surgical excision followed by whole-brain radiation therapy, especially if they are younger than 60 years. Radioresistant tumors should be resected if possible. Stereotactic radiosurgery is an effective treatment for inaccessible or recurrent lesions. With a gamma knife or linear accelerator, multiple small, well-colimated beams of ionizing radiation destroy lesions seen onMRI. Some patients with increased intracranial pressure associated with hydrocephalus may benefit from shunt placement. NEOPLASTIC MENINGITIS
Tumor involving the leptomeninges is a complication of both primary tumors of the central nervous system (CNS) and tumors that metastasize to the CNS. The incidence is estimated at 3 to 8% of patients with cancer. Melanoma, breast and lung cancer, lymphoma (including AIDS-associated), and acute leukemia are the most common causes. Patients typically present with multifocal neurologic signs and symptoms including headache, gait abnormality, mental changes, nausea, vomiting, seizures, back or radicular pain, and limb weakness. Signs include cranial nerve palsies, extremity weakness, paresthesia, and decreased deep tendon reflexes. Diagnosis is made by demonstrating malignant cells in the cerebrospinal fluid (CSF); however, up to 40% of patients may have false negative CSF cytology. An elevated CSF protein level is nearly always present (except in HLTV-1-associated adult T cell leukemia). Patients with neurologic signs and symptoms consistent with neoplastic meningitis who have a negative CSF cytology but an elevated CSF protein level should have the spinal tap repeated at least three times for repeated cytologic examination before the diagnosis is rejected.MRI may show hydrocephalus or smooth or nodular enhancement of the meninges. The development of neoplastic meningitis usually occurs in the setting of uncontrolled cancer outside theCNS; thus, prognosis is poor (median survival 10 to 12 weeks). However, treatment of the neoplastic meningitis may successfully alleviate symptoms and control the CNS spread. TREATMENT Intrathecal chemotherapy, usually methotrexate, cytarabine, or thiotepa, is delivered by lumbar puncture or by an intraventricular reservoir (Ommaya) three times a week until theCSF is free of malignant cells. Then injections are given twice a week for a month and then weekly for a month. An extended release preparation of cytarabine (Depocyte) has a longer half-life and is more effective than regular formulations. Among solid tumors, breast cancer responds best to therapy. Patients with neoplastic meningitis from either acute leukemia or lymphoma may be cured of theirCNSdisease if the systemic disease can be eliminated. SEIZURES Seizures occurring in a patient with cancer can be caused by the tumor itself, by metabolic disturbances, by radiation injury, by cerebral infarctions, by chemotherapy-related encephalopathies, or byCNSinfections. Metastatic disease to the CNS is the most common cause of seizures in patients with cancer. Seizures are a presenting symptom of CNS metastasis in 6 to 29% of cases. Approximately 10% of patients with CNS metastasis eventually develop seizures. The presence of frontal lesions correlates with early seizures, and the presence of hemispheric symptoms increases the risk for late seizures. Both early and late seizures are uncommon in patients with posterior fossa lesions. Seizures are also common in patients with CNS metastases from melanoma. Very rarely, cytotoxic drugs such as etoposide, busulfan,
and chlorambucil cause seizures. TREATMENT Patients in whom seizures due toCNSmetastases have been demonstrated should receive anticonvulsive treatment with diphenylhydantoin. Prophylactic anticonvulsant therapy is not recommended unless the patient is at a high risk for late seizures. In those patients, serum diphenylhydantoin levels should be monitored closely and the dosage adjusted accordingly. INTRACEREBRAL LEUKOCYTOSTASIS Intracerebral leukocytostasis (Ball's disease) is a potentially fatal complication of acute leukemia (particularly myelogenous leukemia) that can occur when the peripheral blast cell count is greater than 100,000/uL. At such high blast cell counts, blood viscosity is increased and blood flow is slowed, and the primitive leukemic cells are capable of invading through endothelium and causing hemorrhage into the brain. Patients may experience stupor, dizziness, visual disturbances, ataxia, coma, or sudden death. Administration of 600 cGy of whole-brain irradiation can protect against this complication and can be followed by rapid institution of antileukemic therapy. This complication is not a feature of the high white cell counts associated with chronic lymphocytic leukemia or chronic myelogenous leukemia. HEMOPTYSIS Hemoptysis may be caused by nonmalignant conditions, but lung cancer accounts for a large proportion of cases. Up to 20% of patients with lung cancer have hemoptysis some time in their course. Endobronchial metastases from carcinoid tumors, breast, colon, kidney cancer, and melanoma may also cause hemoptysis. The volume of bleeding is often difficult to gauge. Massive hemoptysis is defined as more than 600 mL of blood produced in 48 h. When respiratory difficulty occurs, hemoptysis should be treated emergently. Often patients can tell where the bleeding is occurring. They should be placed bleeding side down, given supplemental oxygen, and subjected to emergency bronchoscopy. If the site of the lesion is detected, either the patient undergoes a definitive surgical procedure or the lesion is treated with a neodymium:yttrium-aluminum-garnet (Nd:YAG) laser. The surgical option is preferred. Bronchial artery embolization may control brisk bleeding in 75 to 90% of patients, permitting the definitive surgical procedure to be done more safely. Embolization without definitive surgery is associated with rebleeding in 20 to 50% of patients. Pulmonary hemorrhage with or without hemoptysis in hematologic malignancies is often associated with fungal infections, particularly Aspergillus sp. After granulocytopenia resolves, the lung infiltrates in aspergillosis may cavitate and cause massive hemoptysis. Thrombocytopenia and coagulation defects should be corrected, if possible. AIRWAY OBSTRUCTION Generally, airway obstruction refers to a blockage at the level of the mainstem bronchi
or above. It may result either from intraluminal tumor growth or from extrinsic compression of the airway. If the obstruction is proximal to the larynx, a tracheostomy may be life-saving. For more distal obstructions, particularly intrinsic lesions incompletely obstucting the airway, bronchoscopy with laser treatment, photodynamic therapy, or stenting can produce immediate relief in most patients. However, radiation therapy (either external-beam irradiation or brachytherapy) given together with glucocorticoids may also open the airway. Symptomatic extrinsic compression may be palliated by stenting. METABOLIC EMERGENCIES HYPERCALCEMIA Hypercalcemia is the most common paraneoplastic syndrome (Chaps. 100 and341), occurring in about 10% of patients with advanced cancer. It is associated most often with cancers of the lung, breast, head and neck, and kidney and with multiple myeloma and some B and T cell lymphomas. Increased release of calcium from bone is the main factor leading to hypercalcemia. Bone resorption is increased dramatically through stimulation of the proliferation and activity of osteoclasts, and bone formation is not stimulated in parallel. The kidney may play an important role through an increase in the reabsorption of calcium in the distal tubule. Parathormone-related protein (PTHrP) produced by tumors has a central role as a mediator of hypercalcemia in cancer. PTHrP shares 80% homology with the first 13 amino acids of parathormone (PTH), which are in the region responsible for binding to the PTH receptor. PTHrP acts via the PTH hormone receptors on osteoblasts and renal tubular cells to stimulate bone resorption and renal calcium conservation, leading to hypercalcemia. Elevated plasma PTHrP levels are also found in most hypercalcemic patients with bone metastases, whose hypercalcemia has traditionally been explained by local osteolysis due to the production of osteolytic factors by tumors. Transforming growth factors, cytokines (interleukins 1 and 6), and other unknown factors could play a contributory role. True "ectopic" PTH production by malignant tumors is rare. In lymphoma, a vitamin D-related product of the tumor may also increase calcium absorption in the gut. The clinical features of hypercalcemia in patients with cancer are nonspecific and include fatigue, anorexia, constipation, polydipsia, muscle weakness, nausea, and vomiting. They may easily be attributed to the malignancy itself or to its treatment. Laboratory assessment should include measurement of serum electrolytes, calcium, phosphate, and albumin. Hypoalbuminemia is common in malignancy and affects the total serum concentration of calcium. If the ionized calcium level cannot be obtained, then the corrected serum calcium concentration should be calculated with the following formula:
Most patients with hypercalcemia of malignancy have obvious evidence of malignancy, and their serumPTHlevels are suppressed. Measurements ofPTHrPand serum 1,25-dihydroxyvitamin D are not indicated. Routine serum chemistry evaluations are not
able to distinguish between malignant and nonmalignant causes of hypercalcemia. TREATMENT Not all patients with moderate to severe hypercalcemia (corrected calcium³12 mg/dL) should be treated. The decision to treat will depend on the patient's quality of life, the current symptoms, and the prospect for further cancer treatment. Treatment directed at hypercalcemia only extends life in patients for whom effective cancer treatment is available. Nonetheless, therapy may be indicated to reduce symptoms and improve the quality of life. Treatment of symptomatic hypercalcemia begins with intravenous saline to restore the depleted intravascular volume, which may be 4 to 8 L below normal at presentation. Rehydration usually has little effect on calcium levels, producing a median decrease of only 1 mg/dL. Antiresorptive agents are essential to decrease osteoclastic activity and control hypercalcemia. Bisphosphonates, which are potent inhibitors of bone resorption, are easy to administer, virtually free of side effects, and rapidly effective in lowering the serum calcium level. Pamidronate is the most effective of the commercially available bisphosphonates. The recommended dose of pamidronate is 60 mg for moderate hypercalcemia (corrected calcium 12 to 13.5 mg/dL) and 90 mg for severe hypercalcemia (corrected calcium >13.5 mg/dL). The dose is given as a single infusion over 4 or 24 h. SYNDROME OF INAPPROPRIATE SECRETION OF ANTIDIURETIC HORMONE (SIADH) SIADH is attributed to production of arginine vasopressin by the tumor cells and is characterized by hyponatremia, urine osmolarity inappropriately higher than plasma osmolarity, and high urinary sodium excretion in the absence of volume depletion. Renal, adrenal, and thyroid insufficiency must be excluded, because these disorders can also present with hyponatremia and impaired urinary dilution. Low serum levels of urea and uric acid are useful in distinguishing SIADH from conditions associated with renal hypoperfusion (Chaps. 100 and329). A broad spectrum of malignant tumors have been reported to causeSIADH. Ectopic vasopressin secretion may occur in some 38% of small cell carcinomas of the lung; often adrenocorticotropic hormone is also produced. The presence of hyponatremia in patients with small cell lung cancer confers a poor prognosis. SIADH may also be caused by various other conditions, such as CNS and pulmonary disorders and some surgical procedures. A variety of drugs have also been shown to produce SIADH, including antidepressants, angiotensin converting-enzyme inhibitors, and cytotoxic drugs such as vincristine, vinorelbine, ifosfamide, cyclophosphamide, cisplatin, levamisole, and melphalan. Most patients withSIADHare asymptomatic. The severity of symptoms and signs is related to the degree of hyponatremia and the rapidity with which it develops. Early changes include anorexia, depression, lethargy, irritability, confusion, muscle weakness, and marked personality changes. When the plasma sodium level falls below 110 mEq/L, extensor plantar responses, areflexia, and pseudobulbar palsy may be noted; and further reductions may cause coma, convulsions, and death.
TREATMENT The optimal therapy forSIADH is to treat the underlying malignancy. If that is not possible, other therapeutic approaches are available, such as water restriction or the administration of demeclocycline (900 to 1200 mg per os bid), urea, or lithium carbonate (300 mg per os tid). Demeclocycline is usually used first. Demeclocycline and lithium inhibit the effects of vasopressin on the distal renal tubule. Patients with seizure or coma from hyponatremia may require normal saline infusion plus furosemide to enhance free water clearance. The rate of sodium correction should be slow [0.5 to 1 (mEq/L)/h] to prevent rapid fluid shifts and central pontine myelinolysis. The serum calcium level should be monitored closely to avoid hypocalcemia. LACTIC ACIDOSIS Lactic acidosis is a rare and potentially fatal metabolic complication of cancer. Lactic acidosis associated with sepsis and circulatory failure is a common preterminal event in many malignancies. Lactic acidosis in the absence of hypoxemia may occur in patients with leukemia, lymphoma, or solid tumors. Extensive involvement of the liver by tumor is present in most cases. Alteration of liver function may be responsible for the lactate accumulation. Tachypnea, tachycardia, change of mental status, and hepatomegaly may be seen. The serum level of lactic acid may reach 10 to 20 meq/L (90 to 180 mg/dL). Treatment is aimed at the underlying disease. The danger from lactic acidosis is from the acidosis, not the lactate. Sodium bicarbonate should be added if acidosis is very severe or if hydrogen ion production is very rapid and uncontrolled. The prognosis is poor. HYPOGLYCEMIA Persistent hypoglycemia occasionally is associated with tumors other than pancreatic islet cell tumors. Usually these tumors are large, and often they are of mesenchymal origin or are hepatomas or adrenocortical tumors. Mesenchymal tumors are usually located in the retroperitoneum or thorax. In these patients, obtundation, confusion, and behavioral aberrations occur in the postabsorptive period and may precede the diagnosis of the tumor. Hypoglycemia is due to tumor overproduction of insulin-like growth factor, a peptide hormone with structural homology to proinsulin but having only about 1% of its biologic effects. Additionally, the development of hepatic dysfunction from liver metastases and increased glucose consumption by the tumor can contribute to hypoglycemia. If the tumor cannot be resected, treatment of the hypoglycemia has generally been relief of symptoms, with the administration of glucose, glucocorticoids, or glucagon. Hypoglycemia can be artifactual; hyperleukocytosis from leukemia, myeloproliferative diseases, leukemoid reactions, or colony stimulating factor treatment can increase glucose consumption in the test tube after blood is drawn, leading to pseudohypoglycemia. ADRENAL INSUFFICIENCY In patients with cancer, adrenal insufficiency may go unrecognized because the
symptoms, such as nausea, vomiting, anorexia, and orthostatic hypotension, are nonspecific and may be mistakenly attributed to progressive cancer or to cancer therapy. Primary adrenal insufficiency may develop owing to replacement of both glands by metastases (lung, breast, colon, or kidney cancer, lymphoma), to removal of both glands, or to hemorrhagic necrosis in association with sepsis or anticoagulation. Impaired adrenal steroid synthesis occurs in patients being treated for cancer with mitotane, ketoconazole, aminoglutethimide, or the investigational agent suramin or in those undergoing rapid reduction in glucocorticoid therapy. Rarely, metastatic replacement causes primary adrenal insufficiency as the first manifestation of an occult malignancy. Metastasis to the pituitary or hypothalamus is found at autopsy in up to 5% of patients with cancer, but associated secondary adrenal insufficiency is rare. Patients abruptly discontinuing megestrol acetate therapy (for cancer cachexia) may develop Addisonian crisis from central suppression of the pituitary-adrenal axis with decreased serum levels of cortisol and adrenocorticotropic hormone. Acute adrenal insufficiency is potentially lethal. Treatment of suspected adrenal crisis is initiated after the sampling of serum cortisol and ACTH levels (Chap. 331). TREATMENT-RELATED EMERGENCIES TUMOR LYSIS SYNDROME Tumor lysis syndrome is a well-recognized clinical entity that is characterized by various combinations of hyperuricemia, hyperkalemia, hyperphosphatemia, lactic acidosis, and hypocalcemia and is caused by the destruction of a large number of rapidly proliferating neoplastic cells. Frequently, acute renal failure develops as a result of the syndrome. Tumor lysis syndrome is most frequently associated with the treatment of Burkitt's lymphoma, acute lymphoblastic leukemia, and other high-grade lymphomas, but it also may be seen with chronic leukemias and, rarely, with solid tumors. This syndrome has been seen in patients with chronic lymphocytic leukemia after treatment with fludarabine and cladribine. Tumor lysis syndrome usually occurs during or shortly (1 to 5 days) after chemotherapy. Rarely, spontaneous necrosis of malignancies causes tumor lysis syndrome. Hyperuricemia may be present at the time of chemotherapy. Effective treatment accelerates the destruction of malignant cells and leads to increased serum uric acid levels from the turnover of nucleic acids. Owing to the acidic local environment, uric acid can precipitate in the tubules, medulla, and collecting ducts of the kidney, leading to renal failure. Lactic acidosis and dehydration may contribute to the precipitation of uric acid in the renal tubules. The finding of uric acid crystals in the urine is strong evidence for uric acid nephropathy. The ratio of urinary uric acid to urinary creatinine is >1 in patients with acute hyperuricemic nephropathy and100 mg) the iron preparation should be diluted in 5% dextrose in water or 0.9% NaCl solution. The iron solution can then be infused over a 60 to 90 min period (for larger doses) or at a rate convenient for the attending nurse or physician. While a test dose (25 mg) of parenteral iron is recommended, in reality a slow infusion of a larger dose of parenteral iron solution will afford the same kind of early warning as a separately injected test dose. Early in the infusion of iron, if chest pain, wheezing, a fall in blood pressure, or other systemic manifestations occur, the infusion of iron -- whether as a large solution or a test dose -- should be interrupted immediately. OTHER HYPOPROLIFERATIVE ANEMIAS In addition to mild to moderate iron deficiency anemia, the hypoproliferative anemias can be divided into four categories: (1) chronic inflammation/infection; (2) renal disease; (3) endocrine and nutritional deficiencies (hypometabolic states); and (4) marrow damage (Chap. 109). With chronic inflammation, renal disease, or hypometabolism, endogenous erythropoietin production is inadequate for the degree of anemia observed. For the anemia of chronic inflammation (anemia of chronic disease), the erythroid marrow also responds inadequately to stimulation in part due to defects in iron reutilization. As a result of the lack of adequate erythropoietin stimulation, an examination of the peripheral blood smear will disclose only an occasional polychromatophilic (shift) reticulocyte. In the cases of iron deficiency or marrow damage, appropriate elevations in endogenous erythropoietin levels are typically found, and "shift" reticulocytes will be present on the blood smear. ANEMIA OF ACUTE AND CHRONIC INFLAMMATION/INFECTION (THE ANEMIA OF CHRONIC DISEASE) The anemia of chronic disease -- which encompasses inflammation, infection, tissue injury, and conditions associated with the release of proinflammatory cytokines (such as cancer) -- is one of the most common forms of anemia seen clinically and is probably the most important in the differential diagnosis of iron deficiency, since many of the features of the anemia are brought about by inadequate iron delivery to the marrow, despite the presence of normal or increased iron stores. This is reflected by a low serum iron, increased red cell protoporphyrin, a hypoproliferative marrow, transferrin saturation in the range of 15 to 20%, and a normal or increased serum ferritin. The serum ferritin
values are often the most distinguishing feature between true iron deficiency anemia and the iron-deficient erythropoiesis associated with inflammation. Typically, serum ferritin values increase three-fold over basal levels in the face of inflammation. All of these changes are due to the effects of inflammatory cytokines at several levels of erythropoiesis (Fig. 105-4). IL-1 directly decreases erythropoietin production in response to anemia. IL-1, acting through accessory cell release of IFN-g, suppresses the response of the erythroid marrow to erythropoietin -- an effect that can be overcome by increased erythropoietin administration in vitro and in vivo. In addition, tumor necrosis factor (TNF), acting through the release of IFN-b by marrow stromal cells, also suppresses the response to erythropoietin; several of these same cytokines, acting in concert, block the release of iron fromREstorage sites. The overall result is a chronic hypoproliferative anemia with classic changes in iron metabolism. The anemia is further compounded by a mild to moderate shortening in red cell survival. With chronic inflammation/infection, the primary disease will determine the severity and characteristics of the anemia. For instance, many patients with cancer also have anemia that is typically normocytic and normochromic. In contrast, patients with long-standing active rheumatoid arthritis or chronic infections such as tuberculosis will have a microcytic, hypochromic anemia. In both cases, the bone marrow is hypoproliferative, but the differences in red cell indices reflect differences in the availability of iron for hemoglobin synthesis. Occasionally, conditions associated with chronic inflammation are also associated with chronic blood loss. Under these circumstances, a bone marrow aspirate stained for iron may be necessary to rule out absolute iron deficiency. However, the administration of iron in this case will correct the iron deficiency component of the anemia and leave the inflammatory component unaffected. The anemia associated with acute infection or inflammation is typically mild, but becomes more pronounced over time. Acute infection can produce a fall in hemoglobin levels of 2 to 3 g/dL within 1 or 2 days; this is largely related to the hemolysis of red cells near the end of their natural life span. The fever and cytokines released exert a selective pressure against cells with more limited capacity to maintain the red cell membrane. In most individuals the mild anemia is reasonably well tolerated, and symptoms, if present, are associated with the underlying disease. Occasionally, in patients with preexisting cardiac disease, moderate anemia (hemoglobin 10-11 g/dL) may be associated with angina, exercise intolerance, and shortness of breath. The red cell indices vary from normocytic, normochromic to microcytic, hypochromic. The serum iron values tend to correlate with the red cell indices. The erythropoietic profile that distinguishes the anemia of inflammation from the other causes of hypoproliferative anemias is shown inTable 105-6. ANEMIA OF RENAL DISEASE Chronic renal failure is usually associated with a moderate to severe hypoproliferative anemia; the level of the anemia correlates with the severity of the renal failure. Red cells are typically normocytic and normochromic. Reticulocytes are decreased. The anemia is due to a failure to produce adequate amounts of erythropoietin and a reduction in red cell survival. In certain forms of acute renal failure, the correlation between the anemia and renal function is weaker. Patients with the hemolytic-uremic syndrome increase erythropoiesis in response to the hemolysis, despite renal failure requiring dialysis.
Polycystic renal disease also shows a smaller degree of erythropoietin deficiency for a given level of renal failure. By contrast, patients with diabetes have more severe erythropoietin deficiency for a given level of renal failure. Assessment of iron status provides information to distinguish the anemia of renal disease from the other forms of hypoproliferative anemia (Table 105-6) and to guide management. Patients with the anemia of renal disease usually present with normal serum iron,TIBC, and ferritin levels. However, those maintained on chronic hemodialysis may develop iron deficiency from blood loss through the dialysis procedure. Iron must be replenished in these patients to ensure an adequate response to erythropoietin therapy (see below). ANEMIA IN HYPOMETABOLIC STATES Patients who are starving, particularly for protein, and those with a variety of endocrine disorders that produce lower metabolic rates may develop a mild to moderate hypoproliferative anemia. The release of erythropoietin from the kidney is sensitive to the need for O2, not just O2levels. Thus, erythropoietin production is triggered at lower levels of O2tension in disease states (such as hypothyroidism and starvation) where metabolic activity and thus O2demand is decreased. Endocrine Deficiency States The difference in the levels of hemoglobin between men and women is related to the effects of androgen and estrogen on erythropoiesis. Testosterone and anabolic steroids augment erythropoiesis; castration and estrogen administration to males decrease erythropoiesis. Patients who are hypothyroid or have deficits in pituitary hormones also may develop a mild anemia. Pathogenesis may be complicated by other nutritional deficiencies as iron and folic acid absorption can be affected by these disorders. Usually, correction of the hormone deficiency reverses the anemia. Anemia may be more severe in Addison's disease, depending on the level of thyroid and androgen hormone dysfunction; however, anemia may be masked by decreases in plasma volume. Once such patients are given cortisol and volume replacement, the hemoglobin level may fall rapidly. Mild anemia complicating hyperparathyroidism may be due to decreased erythropoietin production as a consequence of the renal effects of hypercalcemia or to impaired proliferation of erythroid progenitors. Protein Starvation Decreased dietary intake of protein may lead to mild to moderate hypoproliferative anemia; this form of anemia may be prevalent in the elderly. The anemia can be more severe in patients with a greater degree of starvation. In marasmus, where patients are both protein- and calorie-deficient, the release of erythropoietin is impaired in proportion to the reduction in metabolic rate; however, the degree of anemia may be masked by volume depletion and becomes apparent after refeeding. Deficiencies in other nutrients (iron, folate) may also complicate the clinical picture but may not be apparent at diagnosis. Changes in the erythrocyte indices on refeeding should prompt evaluation of iron, folate, and B12status. Anemia in Liver Disease A mild hypoproliferative anemia may develop in patients with chronic liver disease from nearly any cause. The peripheral blood smear may show burr
cells and stomatocytes from the accumulation of excess cholesterol in the membrane from a deficiency of lecithin cholesterol acyltransferase. Red cell survival is shortened, and the production of erythropoietin is inadequate to compensate. In alcoholic liver disease, nutritional deficiencies can add complexity to the management. Folate deficiency from inadequate intake and iron deficiency from blood loss and inadequate intake can alter the red cell indices. TREATMENT Many patients with hypoproliferative anemias experience recovery of normal hemoglobin levels when the underlying disease is appropriately treated. For those in whom such reversals are not possible -- such as patients with end-stage renal failure, cancer, and chronic inflammatory diseases -- symptomatic anemia requires treatment. The two major forms of treatment are transfusions and erythropoietin. Transfusions Thresholds for transfusion should be altered based on the patient's symptoms. In general, patients without serious underlying cardiovascular or pulmonary disease can tolerate hemoglobin levels above 8 g/dL and do not require intervention until the hemoglobin falls below that level. Patients with more physiologic compromise may need to have their hemoglobin levels kept above 11 g/dL. A typical unit of packed red cells increases the hemoglobin level by 1 g/dL. Transfusions are associated with certain infectious risks (Chap. 114) and chronic transfusions can produce iron overload. Erythropoietin Erythropoietin is particularly useful in anemias in which endogenous erythropoietin levels are inappropriately low, such as the hypoproliferative anemias. Iron status must be evaluated and iron repleted to obtain optimal effects from erythropoietin. In patients with chronic renal failure, the usual dose of erythropoietin is 50 to 150 U/kg three times a week subcutaneously. The dose needed to correct the anemia in patients with cancer is higher, up to 300 U/kg three times a week. Hemoglobin levels of 10 to 12 g/dL are usually reached within 4 to 6 weeks if iron levels are adequate. Once a target hemoglobin level is reached, the erythropoietin dose can be decreased to 75 U/kg three times a week. A fall in hemoglobin level occurring in the face of erythropoietin therapy usually signifies the development of an infection or iron depletion. Aluminum toxicity and hyperparathyroidism can also compromise the erythropoietin response. When an infection intervenes, it is best to interrupt the erythropoietin therapy and rely on transfusion to correct the anemia until the infection is adequately treated. ACKNOWLEDGEMENT Dr. Robert S. Hillman was the author of this chapter in the 14thedition, and material from his chapter has been retained. (Bibliography omitted in Palm version) Back to Table of Contents
106. HEMOGLOBINOPATHIES - Edward J. Benz, Jr. Hemoglobin is critical for normal oxygen delivery to tissues; it is also present in erythrocytes in such high concentrations that it can alter red cell shape, deformability, and viscosity. Hemoglobinopathies are disorders affecting the structure, function, or production of hemoglobin. These disorders are usually inherited and range in severity from asymptomatic laboratory abnormalities to death in utero. Different forms may present as hemolytic anemia, erythrocytosis, cyanosis, or vasoocclusive stigmata. PROPERTIES OF THE HUMAN HEMOGLOBINS HEMOGLOBIN STRUCTURE Different hemoglobins are produced during embryonic, fetal, and adult life (Fig. 106-1). Each consists of a tetramer of globin polypeptide chains: a pair of a-like chains 141 amino acids long and a pair of b-like chains 146 amino acids long. The major adult hemoglobin, HbA, has the structurea 2b2. HbF (a2g2) predominates during most of gestation, and HbA2(a2d2) is a minor adult hemoglobin. Each globin chain enfolds a single heme moiety, consisting of a protoporphyrin IX ring complexed with a single iron atom in the ferrous state (Fe2+), positioned in a manner optimal for reversible binding of oxygen. Each heme moiety can bind a single oxygen molecule; every molecule of hemoglobin can thus transport up to four oxygen molecules. The amino acid sequences of the various globins are highly homologous to one another. Each has a highly helical secondary structure. Their globular tertiary structures cause the exterior surfaces to be rich in polar (hydrophilic) amino acids that enhance solubility and the interior to be lined with nonpolar groups, forming a hydrophobic "pocket" into which heme is inserted. The tetrameric quaternary structure of HbA contains twoab dimers. Numerous tight interactions (i.e.,a 1b1contacts) hold the a and bchains together. The complete tetramer is held together by interfaces (i.e.,a1b2contacts) between the a-like chain of one dimer and the non-a chain of the other dimer. The hemoglobin tetramer is highly soluble, but individual globin chains are insoluble. Unpaired globin precipitates, forming inclusions (Heinz bodies) that damage the cell. Normal globin chain synthesis is balanced so that each newly synthesizeda or non-a globin chain will have an available partner with which to pair to form hemoglobin. Solubility and reversible oxygen binding are the key properties deranged in hemoglobinopathies. Both depend most on the hydrophilic surface amino acids, the hydrophobic amino acids lining the heme pocket, a key histidine in the F helix, and the amino acids forming thea1b1anda1b2contact points. Mutations in these strategic regions tend to be the ones that alter clinical behavior. FUNCTION OF HEMOGLOBIN To support oxygen transport, hemoglobin must bind O2efficiently at the partial pressure of oxygen (PO2) of the alveolus, retain it, and release it to tissues at the PO 2 of tissue
capillary beds. Oxygen acquisition and delivery over a relatively narrow range of oxygen tensions depend on a property inherent in the tetrameric arrangement of heme and globin subunits within the hemoglobin molecule called cooperativity or heme-heme interaction. At low oxygen tensions, the hemoglobin tetramer is fully deoxygenated (Fig. 106-2). Oxygen binding begins slowly as O2tension rises. However, as soon as some oxygen has been bound by the tetramer, an abrupt increase occurs in the slope of the curve. Thus, hemoglobin molecules that have bound some oxygen develop a higher oxygen affinity, greatly accelerating their ability to combine with more oxygen. This S-shaped oxygen equilibrium curve, along which substantial amounts of oxygen loading and unloading can occur over a narrow range of oxygen tensions, is physiologically more useful than the high-affinity hyperbolic curve of individual monomers. Oxygen affinity is modulated by several factors. The Bohr effect arises from the stabilizing action of protons on deoxyhemoglobin, which binds protons more readily than oxyhemoglobin because it is a weaker acid. Thus, hemoglobin has a lower oxygen affinity at low pH, facilitating delivery to tissues (Fig. 106-2). The major small molecule that alters oxygen affinity in humans is 2,3-bisphosphoglycerate (2,3-BPG, formerly 2,3-DPG), which lowers oxygen affinity when bound to hemoglobin. HbA has a reasonably high affinity for 2,3-BPG. HbF does not bind 2,3-BPG, so it tends to have a higher oxygen affinity in vivo. Hemoglobin may also bind nitric oxide reversibly, thereby contributing to vascular tone. To understand hemoglobinopathies, it is sufficient to understand that proper oxygen transport depends on the tetrameric structure of the proteins, the proper arrangement of the charged amino acids, and interaction with low-molecular-weight substances such as protons or2,3-BPG. DEVELOPMENTAL BIOLOGY Red cells first appearing at about 6 weeks after conception contain the embyronic hemoglobins Hb Portland (z2g2), Hb Gower I (z2e2), and Hb Gower II (a2e2). At 10 to 11 weeks, fetal hemoglobin (HbF;a 2g2) becomes predominant. The switch to nearly exclusive synthesis of adult hemoglobin (HbA;a 2b2) occurs at about 38 weeks (Fig. 106-1). Fetuses and newborns therefore requirea-globin but not b-globin for normal gestation. Small amounts of HbF are produced during postnatal life. A few red cell clones called F cells are progeny of a small pool of immature committed erythroid precursors (BFU-e) that retain the ability to produce HbF. Profound erythroid stress, such as that seen in severe hemolytic anemias, after bone marrow transplant, or during chemotherapy, cause more of the "F potent" BFU-e to be recruited. HbF levels thus tend to rise in some patients with sickle cell anemia or thalassemia. This phenomenon is also important because it probably explains the ability of hydroxyurea to increase levels of HbF in adults. Fetal globin genes can also be partially activated after birth by agents such as butyrate, which inhibit histone deacetylase and modify the structure of chromatin. GENETICS AND BIOSYNTHESIS OF HUMAN HEMOGLOBIN
The human hemoglobins are encoded in two tightly linked gene clusters; thea-like globin genes are clustered on chromosome 16, and the b-like genes on chromosome 11 (Fig. 106-1). The a-like cluster consists of two a-globin genes and a single copy of the z gene. The non-a gene cluster consists of a singlee gene, the Gg and Ag fetal globin genes, and the adultd and bgenes. Important regulatory sequences flank each gene. Immediately upstream are typical promoter elements needed for the assembly of the transcription initiation complex. Sequences in the 5¢ flanking region of theg and the b genes appear to be crucial for the correct developmental regulation of these genes, while elements that function like classic enhancers and silencers are in the 3¢ flanking regions. The locus control region (LCR) elements located far upstream appear to control the overall level of expression of each cluster. These elements achieve their regulatory effects by interacting with trans-acting transcription factors. Some of these factors are ubiquitous (e.g., Sp1 and YY1), while others are more or less limited to erythroid cells (e.g., GATA-1, NFE-2, and EKLF). The latter also appear to modulate genes specifically expressed during erythropoiesis, such as the genes that encode the enzymes of the heme biosynthetic pathway. This is relevant since normal red blood cell (RBC) differentiation also requires the coordinated expression of the globin genes with the genes responsible for heme and iron metabolism. CLASSIFICATION OF HEMOGLOBINOPATHIES There are five major classes of hemoglobinopathies (Table 106-1). Structural hemoglobinopathies occur when mutations alter the amino acid sequence of a globin chain, altering the physiologic properties of the variant hemoglobins and producing the characteristic clinical abnormalities. The variant hemoglobins relevant to this chapter polymerize abnormally, as in sickle cell anemia, or exhibit altered solubility or oxygen-binding affinity. Thalassemia syndromes arise from mutations that impair production or translation of globin mRNA, leading to deficient globin chain biosynthesis. Clinical abnormalities are attributable to the inadequate supply of hemoglobin and the imbalances in the production of individual globin chains, leading to premature destruction of erythroblasts and red cells. Thalassemic hemoglobin variants combine features of thalassemia (e.g., abnormal globin biosynthesis) and of structural hemoglobinopathies (e.g., an abnormal amino acid sequence). Hereditary persistence of fetal hemoglobin (HPFH) is characterized by synthesis of high levels of fetal hemoglobin in adult life. Acquired hemoglobinopathies include modifications of the hemoglobin molecule by toxins (e.g., acquired methemoglobinemia) and abnormal hemoglobin synthesis (e.g., high levels of HbF production in preleukemia and a-thalassemia in myeloproliferative disorders). EPIDEMIOLOGY Hemoglobinopathies are especially common in areas where malaria is endemic. This clustering of hemoglobinopathies is assumed to reflect a selective survival advantage for the abnormal red cells, which presumably provide a less hospitable environment during the obligate intraerythrocytic stages of the parasitic life cycle. Very young children with a-thalassemia are more susceptible to infection with the nonlethal Plasmodium vivax. Thalassemia might then favor a natural "vaccination" against
infection with the more lethal P. falciparum. Thalassemias are the most common genetic disorders in the world, affecting nearly 200 million people worldwide. About 15% of American blacks are silent carriers for a thalassemia;a-thalassemia trait (minor) occurs in 3% of American blacks and in 1 to 15% of persons of Mediterranean origin.b Thalassemia has a 10 to 15% incidence in individuals from the Mediterranean and Southeast Asia and 0.8% in American blacks. The number of severe cases of thalassemia in the United States is about 1000. Sickle cell disease is the most common structural hemoglobinopathy occurring in heterozygous form in about 8% of American blacks and in homozygous form in 1 in 400. Between 2 and 3% of American blacks carry a hemoglobin C allele. INHERITANCE AND ONTOGENY Hemoglobinopathies are autosomal "codominant" traits -- compound heterozygotes that inherit a different abnormal mutant allele from each parent exhibit composite features of each. For example, patients inheriting sickleb thalassemia exhibit features ofb thalassemia and sickle cell anemia. Thea-chain is present in HbA, HBA2, and HbF; a-chain mutations thus cause abnormalities in all three. The a-globin hemoglobinopathies are symptomatic in utero and after birth because normal function of thea-globin gene is required throughout gestation and adult life. In contrast, infants with b-globin hemoglobinopathies tend to be asymptomatic until 3 to 9 months of age, when HbA has largely replaced HbF. DETECTION AND CHARACTERIZATION OF HEMOGLOBINOPATHIES -- GENERAL METHODS Electrophoretic techniques are used for routine hemoglobin analysis. Electrophoresis at pH-8.6 on cellulose acetate membranes is simple, inexpensive, and reliable for initial screening. Hemoglobins S, G, and D have the same mobility at pH-8.6. Agar gel electrophoresis at pH-6.1 in citrate buffer is often used as a complementary method because it detects different variants (S migration differs from G and D). Comparison of results obtained in each system usually allows unambiguous diagnosis, but some important variants are electrophoretically silent. These mutant hemoglobins can usually be characterized by more specialized techniques such as isoelectric focusing and/or high-pressure liquid chromatography (HPLC). Quantitation of the hemoglobin profile is often desirable. HbA2 is frequently elevated in b-thalassemia trait and depressed in iron deficiency. HbF is elevated inHPFH, some b thalassemia syndromes, and occasional periods of erythroid stress or marrow dysplasia. For characterization of sickle cell trait, sickle thalassemia syndromes, or hemoglobin SC disease, and for monitoring the progress of exchange transfusion therapy to lower the percentage of circulating HbS, quantitation of individual hemoglobins is also required. In most laboratories, quantitation is performed only if the test is specifically ordered. Because some variants can comigrate with HbA or HbS (sickle hemoglobin), electrophoretic assessment should always be regarded as incomplete unless functional assays for hemoglobin sickling, solubility, or oxygen affinity are also performed, as dictated by the clinical presentation. The best sickling assays involve measurement of
the degree to which the hemoglobin becomes insoluble, or gelated, as it is deoxygenated (i.e., sickle solubility test). Unstable hemoglobins are detected by their precipitation in isopropanol or after heating to 50°C. High-O2affinity and low-O2affinity variants are detected by quantitating the partial pressure of oxygen at which the hemoglobin sample becomes 50% saturated with oxygen (P50test). Direct tests for the percentages of carboxyhemoglobin and methemoglobin, employing spectrophotometric techniques, can readily be obtained from most clinical laboratories on an urgent basis. Complete characterization, including amino acid sequencing or gene cloning and sequencing, is available from several investigational laboratories around the world. The advent of the polymerase chain reaction (PCR), allele-specific oligonucleotide hybridization, and automated DNA sequencing has made it possible to identify globin gene mutations in a few days. Diagnosis is best established by recognition of a characteristic history, physical findings, peripheral blood smear morphology, and abnormalities of the complete blood cell count (e.g., profound microcytosis with minimal anemia in thalassemia trait). Laboratory evaluation identifies the specific hemoglobinopathy suspected clinically. STRUCTURALLY ABNORMAL HEMOGLOBINS SICKLE CELL SYNDROMES The sickle cell syndromes are caused by a mutation in the b-globin gene that changes the sixth amino acid from glutamic acid to valine. HbS (a2b26Glu®Va1) polymerizes reversibly when deoxygenated to form a gelatinous network of fibrous polymers that stiffen the erythrocyte membrane, increase viscosity, and cause dehydration due to potassium leakage and calcium influx (Fig. 106-3). These changes also produce the characteristic sickle shape. Sickled cells lose the pliability needed to traverse small capillaries. They possess altered "sticky" membranes (especially reticulocytes) that are abnormally adherent to the endothelium of small venules. These abnormalities provoke unpredictable episodes of microvascular vasoocclusion and premature red cell destruction (hemolytic anemia). Hemolysis occurs because the abnormal erythrocytes are destroyed by the spleen. The rigid adherent cells also clog small capillaries and venules, causing tissue ischemia, acute pain, and gradual end-organ damage. This venoocclusive component usually dominates the clinical course. Prominent manifestations include episodes of ischemic pain (i.e., painful crises) and ischemic malfunction or frank infarction in the spleen, central nervous system, bones, liver, kidneys, and lungs. The prototype disease, sickle cell anemia, is the homozygous state for HbS (Table 106-2). Several sickle syndromes occur as the result of inheritance of HbS from one parent and another hemoglobinopathy, such as b thalassemia or HbC (a 2b26Glu®Lys) from the other parent. Clinical Manifestations Sickle Cell Anemia Most patients with sickling syndromes suffer from hemolytic anemia, with hematocrits of 15 to 30%, and significant reticulocytosis. Anemia was once thought
to exert protective effects against vasoocclusion by reducing blood viscosity. Natural history and drug therapy trials suggest that an increase in the hematocrit with feedback inhibition of reticulocytosis might be beneficial, even at the expense of increased blood viscosity. The role of adhesive reticulocytes in vasoocclusion might account for these paradoxical effects. Granulocytosis is common. The white cell count can fluctuate substantially and unpredictably during and between painful crises, infectious episodes, and other intercurrent illnesses. Vasoocclusion causes protean manifestations; intermittent episodes in connective and musculoskeletal structures produce painful ischemia manifested by acute pain and tenderness, fever, tachycardia, and anxiety. These recurrent episodes, called painful crises, are the most common clinical manifestation. Their frequency and severity vary greatly. Pain can develop almost anywhere in the body and may last from a few hours to 2 weeks. Repeated crises requiring hospitalization (more than three per year) correlate with reduced survival in adult life, suggesting that these episodes are associated with accumulation of chronic end-organ damage. Provocative factors include infection, fever, excessive exercise, anxiety, abrupt changes in temperature, hypoxia, or hypertonic dyes. Repeated microinfarction can destroy tissues having microvascular beds that promote sickling. Thus, the spleen is frequently infarcted within the first 18 to 36 months of life, causing susceptibility to infection, particularly from pneumococci. Acute venous obstruction of the spleen (splenic sequestration crisis), a rare occurrence in early childhood, may require emergency transfusion and/or splenectomy to prevent trapping of the entire arterial output in the obstructed spleen. Occlusion of retinal vessels can produce hemorrhage, neovascularization, and eventual detachments. Renal papillary necrosis invariably produces isosthenuria. More widespread renal necrosis leads to renal failure in adults, a common late cause of death. Bone and joint ischemia can lead to aseptic necrosis (especially of the femoral or humeral heads), chronic arthropathy, and unusual susceptibility to osteomyelitis, which may be caused by organisms such as Salmonella, rarely encountered in other settings. The hand-foot syndrome is caused by painful infarcts of the digits and dactylitis. Stroke is especially common in children, a small subset of whom tend to suffer repeated episodes; stroke is less common in adults and is often hemorrhagic. A particularly painful complication in males is priapism, due to infarction of the penile venous outflow tracts; permanent impotence is a frequent consequence. Chronic lower leg ulcers probably arise from ischemia and superinfection in the distal circulation. Acute chest syndrome is a distinctive manifestation characterized by chest pain, tachypnea, fever, cough, and arterial oxygen desaturation. It can mimic pneumonia, pulmonary emboli, bone marrow infarction and embolism, myocardial ischemia, or in situ lung infarction. Acute chest syndrome is thought to reflect in situ sickling within the lung, producing pain and temporary pulmonary dysfunction. Acute chest syndrome may be difficult or impossible to distinguish from other entities. Pulmonary infarction and pneumonia are the most frequent underlying or concomitant conditions in patients with this syndrome. Repeated episodes of acute chest pain correlate with reduced survival. Acutely, reduction in arterial oxygen saturation is especially ominous because it
promotes sickling on a massive scale. Repeated acute or subacute pulmonary crises lead to pulmonary hypertension and cor pulmonale, an increasingly common cause of death as patients survive further into adult life. Sickle cell syndromes are remarkable for their clinical heterogeneity. Some patients remain virtually asymptomatic into or even through adult life, while others suffer repeated crises requiring hospitalization from early childhood. At least five haplotypes of sickle cell disease are recognized based upon their origin: Senegal, Cameroon, Benin, Central African Republic, and India. Among these, patients of the Central African Republic have the worst disease and those of Senegal the least severe. Patients with sickle thalassemia and sickle-HbE tend to have similar, slightly milder, symptoms, perhaps because of the ameliorating effects of production of other hemoglobins within the red cell. Hemoglobin SC disease, one of the more common variants of sickle cell anemia, is frequently marked by lesser degrees of hemolytic anemia and a greater propensity for the development of retinopathy and aseptic necrosis of bones. In most respects, however, the clinical manifestations resemble sickle cell anemia. Some rare hemoglobin variants actually aggravate the sickling phenomenon. Sickle Cell Trait Sickle cell trait is usually asymptomatic. Anemia and painful crises are exceedingly rare. An uncommon, but highly distinctive, symptom is painless hematuria, often occurring in adolescent males, probably due to papillary necrosis. Sloughing of papillae with ureteral obstruction has been reported, as have isolated cases of massive sickling or sudden death due to exposure to high altitudes or extraordinary extremes of exercise and dehydration. Diagnosis Sickle cell syndromes are readily suspected on the basis of characteristic hemolytic anemia, red cell morphology (Plate V-39), and intermittent episodes of ischemic pain. Diagnosis is confirmed by hemoglobin electrophoresis and sickling tests. Thorough characterization of the exact hemoglobin profile of the patient is important, because sickle thalassemia and hemoglobin SC disease are correlated with alterations in prognosis or clinical features. The diagnosis is usually established in childhood, but occasional patients, often with compound heterozygous states, do not develop symptoms until the onset of puberty, pregnancy, or early adult life. Genotyping of family members and potential parental partners is critical for genetic counseling. Details of the childhood history help to establish prognosis and eligibility for aggressive or experimental therapies. Factors associated with increased morbidity and mortality are more than three crises requiring hospitalization per year, chronic neutrophilia, a history of splenic sequestration or hand-foot syndrome, and second episodes of acute chest syndrome. Patients with a history of cerebrovascular accidents are at higher risk for repeated episodes and require especially close monitoring. TREATMENT Patients with sickle cell syndromes require ongoing continuity of care. Familiarity with the pattern of symptoms provides the best safeguard against excessive use of the emergency room, hospitalization, and habituation to addictive narcotics. Additional preventive measures include regular slit-lamp examinations to monitor development of retinopathy; antibiotic prophylaxis appropriate for splenectomized patients during dental or other invasive procedures; vaccination against pneumococci and Haemophilus
influenzae; and vigorous oral hydration before or during periods of extreme exercise, exposure to heat or cold, emotional stress, or infection. The management of acute painful crisis includes vigorous hydration, thorough evaluation for underlying causes (such as infection), and aggressive narcotic analgesia administered by a standing order and/or PCA pump. Morphine (0.1 to 0.15 mg/kg every 3 to 4 h) or meperidine (0.75 to 1.5 mg/kg every 2 to 4 h) should control severe pain. Bone pain may respond as well to ketorolac (30 to 60 mg initial dose, then 15 to 30 mg every 6 to 8 h). Many crises can be managed at home with oral hydration and oral analgesia. Use of the emergency room should be reserved for especially severe symptoms or circumstances in which other processes (e.g., infection) are strongly suspected. Nasal oxygen should be employed as appropriate to protect arterial saturation. Most crises resolve in 1 to 7 days. Use of blood transfusion should be reserved for extreme cases; transfusion does not shorten the crisis. No tests are definitive to diagnose acute painful crisis. Critical to good management is an approach that recognizes that most patients reporting crisis symptoms do indeed have crisis or another significant medical problem. Diligent diagnostic evaluation for underlying causes is imperative, even though these are found infrequently. In adults, the possibility of aseptic necrosis or sickle arthropathy must be considered, especially if pain and immobility become repeated or chronic at a single site. Nonsteroidal anti-inflammatory agents are often effective for sickle cell arthropathy. Acute chest syndrome is a medical emergency that may require management in an intensive care unit. Hydration should be monitored carefully to avoid the development of pulmonary edema, and oxygen therapy should be especially vigorous for protection of arterial saturation. Diagnostic evaluation for pneumonia and pulmonary embolism should be thorough, since these may occur with atypical symptoms. Critical interventions are transfusion to maintain a hematocrit >30 and emergency exchange transfusion if arterial saturation drops below 90%. As patients with sickle cell syndromes increasingly survive into their fifth and sixth decades (median age at death is 42 years for men, 48 years for women), end-stage renal failure and pulmonary hypertension are becoming increasingly prominent causes of end-stage morbidity; anecdotal evidence suggests that a sickle cell cardiomyopathy and/or premature coronary artery disease may compromise cardiac function in later years. Sickle cell patients have received kidney transplants, but they often experience an increase in the frequency and severity of crises, possibly due to increased infection as a consequence of immunosuppression. The most significant advance in the therapy of sickle cell anemia has been the introduction of hydroxyurea as a mainstay of therapy for patients with severe symptoms. Hydroxyurea (10 to 30 mg/kg/per day) increases fetal hemoglobin and may also exert beneficial affects on red cell hydration, vascular wall adherence, and suppression of the granulocyte and reticulocyte counts; indeed, dosage is titrated to maintain a white cell count between 5,000 and 8,000. White cells and reticulocytes may play a major role in the pathogenesis of sickle cell crisis, and their suppression may be an important benefit of hydroxyurea therapy.
Hydroxyurea should be considered in patients experiencing repeated episodes of acute chest syndrome or more than three crises per year requiring hospitalization. The utility of this agent for reducing the incidence of other complications (e.g., priapism, retinopathy) is under evaluation, as are the long-term side effects. Therefore, when possible, treatment should be instituted as part of a clinical trial. Most patients respond within a few months with elevations of fetal hemoglobin. Bone marrow transplantation can provide definitive cures but is known to be effective and safe only in children. Prognostic features justifying bone marrow transplant are the presence of repeated crises early in life, a high neutrophil count, or the development of hand-foot syndrome. Children at risk for stroke can be identified through the use of Doppler ultrasound techniques. Prophylactic exchange transfusion appears to reduce the risk of stroke substantially in this population. Children who do suffer a cerebrovascular accident should be maintained for at least 3 to 5 years on a program of vigorous exchange transfusion, since the risk of second strokes is extremely high in this population. Gene therapy for sickle cell anemia is under investigation, but no safe therapy is currently available. Agents blocking red cell hydration or vascular adhesion, such as clotrimazole, may have value as an adjunct to hydroxyurea therapy; trials are ongoing. UNSTABLE HEMOGLOBINS Amino acid substitutions that reduce solubility or increase susceptibility to oxidation produce "unstable" hemoglobins that precipitate, forming inclusion bodies injurious to the red cell membrane. Representative mutations are those that interfere with contact points between the a and b subunits [e.g., Hb Philly (b 35Tyr®Phe)], alter the helical segments [e.g., Hb Genova (b28Leu®Pro)], or disrupt interactions of the hydrophobic pockets of the globin subunits with heme [e.g., Hb Koln (b 98Val®Met)] (Table 106-3). The inclusions, called Heinz bodies, are clinically detectable by staining with supravital dyes such as crystal violet (Heinz body test). Removal of these inclusions by the spleen generates pitted, rigid cells that have shortened life spans, producing hemolytic anemia of variable severity, sometimes requiring chronic transfusion support. Splenectomy may be needed to correct the anemia. Leg ulcers and premature gallbladder disease due to bilirubin turnover are frequent stigmata. Unstable hemoglobins occur sporadically, often by spontaneous new mutations. Heterozygotes are often symptomatic because a significant Heinz body burden can develop even when the unstable variant accounts for a portion of the total hemoglobin. Symptomatic unstable hemoglobins tend to be b-globin variants, because sporadic mutations affecting only one of the foura globins would generate only 20 to 30% abnormal hemoglobin. HEMOGLOBINS WITH ALTERED OXYGEN AFFINITY High-affinity hemoglobins [e.g., Hb Yakima (b99Asp®His)] bind oxygen more readily but deliver less O2 to tissues at normal capillary PO2levels (Fig. 106-2). Mild tissue hypoxia ensues, stimulatingRBCproduction and erythrocytosis (Table 106-3). In extreme cases, the hematocrit can rise to 60 to 65%, increasing blood viscosity and producing typical
symptoms (headache, somnolence, or dizziness). Phlebotomy may be required. Typical mutations alter interactions within the heme pocket or disrupt the Bohr effect or salt-bond site. Mutations that impair the interaction of HbA with2,3-BPGcan increase O2affinity, because 2,3-BPG binding lowers O2affinity. Low-affinity hemoglobins [e.g., Hb Kansas (b102Asn®Thr)] bind sufficient oxygen in the lungs, despite their lower oxygen affinity, to achieve nearly full saturation. At capillary oxygen tensions, they lose sufficient amounts of oxygen to maintain homeostasis at a low hematocrit (Fig. 106-2) (pseudoanemia). Capillary hemoglobin desaturation can also be sufficient to produce clinically apparent cyanosis. Despite these findings, patients usually require no specific treatment. METHEMOGLOBINEMIAS Methemoglobin is generated by oxidation of the heme iron moieties to the ferric state, causing a characteristic bluish-brown, muddy color resembling cyanosis. Methemoglobin has such high oxygen affinity that virtually no oxygen is delivered to tissues. Levels >50 to 60% are often fatal. Congenital methemoglobinemia arises from globin mutations that stabilize iron in the ferric state [e.g., HbM Iwata (a87His®Tyr),Table 106-3] or from mutations that impair the enzymes that reduce methemoglobin to hemoglobin (e.g., methemoglobin reductase, NADP diaphorase). Acquired methemoglobinemia is caused by toxins that oxidize heme iron, notably nitrate and nitrite-containing compounds. DIAGNOSIS AND MANAGEMENT OF PATIENTS WITH UNSTABLE HEMOGLOBINS, HIGH-AFFINITY HEMOGLOBINS, AND METHEMOGLOBINEMIA Unstable hemoglobin variants should be suspected in patients with nonimmune hemolytic anemia, jaundice, splenomegaly, or premature biliary tract disease. Severe hemolysis usually presents during infancy as neonatal jaundice or anemia. Milder cases may present in adult life with anemia or only as unexplained reticulocytosis, hepatosplenomegaly, premature biliary tract disease, or leg ulcers. Because spontaneous mutation is common, family history of anemia may be absent. The peripheral blood smear often shows anisocytosis, abundant cells with punctate inclusions, and irregular shapes (i.e., poikilocytosis). The two best tests for diagnosing unstable hemoglobins are the Heinz body preparation and the isopropanol or heat stability test. Many unstable Hb variants are electrophoretically silent. A normal electrophoresis does not rule out the diagnosis. Severely affected patients may require transfusion support for the first 3 years of life, because splenectomy before age 3 is associated with a significantly greater immune deficit. Splenectomy is usually effective thereafter, but occasional patients may require lifelong transfusion support. Even after splenectomy, patients can develop cholelithiasis and leg ulcers. Splenectomy can also be considered in patients exhibiting severe secondary complications of chronic hemolysis, even if anemia is absent. Precipitation of unstable hemoglobins is aggravated by oxidative stress, e.g., infection, antimalarial drugs.
High-O2-affinity hemoglobin variants should be suspected in patients with erythrocytosis. The best test for confirmation is measurement of the P50. A high-O2-affinity Hb causes a significant left shift (i.e., lower numeric value of the P 50); confounding conditions, e.g., tobacco smoking or carbon monoxide exposure, can also lower the P50. Patients with high-affinity hemoglobin are often asymptomatic; rubor or plethora may be telltale signs. When the hematocrit reaches 55 to 60%, symptoms of high blood viscosity and sluggish flow (headache, lethargy, dizziness, etc.) may be present. These symptoms respond to judicious phlebotomy. Erythrocytosis represents an appropriate attempt to compensate for the impaired oxygen delivery by the abnormal variant. Overzealous phlebotomy may stimulate increased erythropoiesis or aggravate symptoms by thwarting this compensatory mechanism. The guiding principle of phlebotomy should be to improve oxygen delivery by reducing blood viscosity and increasing blood flow rather than restoration of a normal hematocrit. Modest iron deficiency may aid in control. Low-affinity hemoglobins should be considered in patients with cyanosis or a low hematocrit with no other cause apparent after thorough evaluation. The P50 test confirms the diagnosis. Counseling and reassurance are the interventions of choice. Methemoglobin should be suspected in patients with hypoxic symptoms who appear cyanotic but have a PaO2sufficiently high that hemoglobin should be fully saturated with oxygen. A history of nitrite or other oxidant ingestions may not always be available; some exposures may be unapparent to the patient, and others may result from suicide attempts. The characteristic muddy appearance of freshly drawn blood can be a critical clue. The diagnostic test of choice is measurement of the methemoglobin content, which is usually available on an emergency basis. Methemoglobinemia often causes symptoms of cerebral ischemia at levels>15%; levels>60% are usually lethal. Intravenous injection of 1 mg/kg of methylene blue is effective emergency therapy. Milder cases and follow-up of severe cases can be treated orally with methylene blue (60 mg three to four times each day) or ascorbic acid (300 to 600 mg/d). THALASSEMIA SYNDROMES The thalassemia syndromes are inherited disorders of a- orb-globin biosynthesis. The reduced supply of globin diminishes production of hemoglobin tetramers, causing hypochromia and microcytosis. Unbalanced accumulation of a andb subunits occurs because the synthesis of the unaffected globins proceeds at normal rate. Unbalanced chain accumulation dominates the clinical phenotype. Clinical severity varies widely, depending on the degree to which the synthesis of the affected globin is impaired, altered synthesis of other globin chains, and coinheritance of other abnormal globin alleles. b-THALASSEMIA SYNDROMES Mutations causing thalassemia can affect any step in the pathway of globin gene
expression: transcription, processing of the mRNA precursor, translation, and posttranslational metabolism of the b-globin polypeptide chain. The most common forms arise from mutations that derange splicing of the mRNA precursor or prematurely terminate translation of the mRNA. Hypochromia and microcytosis due to reduced amounts of hemoglobin tetramers characterize all forms ofb thalassemia. In heterozygotes (b-thalassemia trait), this is the only abnormality seen; anemia is minimal. In homozygous states, unbalanceda- and b-globin accumulation causes accumulation of highly insoluble unpaired achains, which form toxic inclusion bodies that kill developing erythroblasts in the marrow. Few of the proerythroblasts beginning erythroid maturation survive. The few surviving red cells bear a burden of inclusion bodies, detected in the spleen, shortening the red cell life span and producing severe hemolytic anemia. The resulting profound anemia stimulates erythropoietin release and compensatory erythroid hyperplasia, but the marrow response is sabotaged by ineffective erythropoiesis. Anemia persists. Erythroid hyperplasia can become exuberant and produce extramedullary erythropoietic tissue in the liver and spleen. Massive bone marrow expansion deranges growth and development. Children develop characteristic "chipmunk" facies due to maxillary marrow hyperplasia and frontal bossing, thinning and pathologic fracture of long bones and vertebrae due to cortical invasion by erythroid elements, and profound growth retardation. Hemolytic anemia causes hepatosplenomegaly, leg ulcers, gallstones, and high-output congestive heart failure. The conscription of caloric resources to support erythropoiesis leads to inanition, susceptibility to infection, endocrine dysfunction, and, in the most severe cases, death during the first decade of life. Chronic transfusions with red cells improves oxygen delivery, suppresses the excessive ineffective erythropoiesis, and prolongs life, but the inevitable side effects, notably iron overload, usually prove fatal by age 30. Bone marrow transplantation in childhood is the only curative therapy. Severity is highly variable. Known modulating factors are those that ameliorate the burden of unpaireda-globin inclusions. Alleles associated with milder synthetic defects and coinheritance of a-thalassemia trait reduce clinical severity by reducing accumulation of excessa globin. HbF persists to various degrees inb thalassemias. g-Globin gene chains can substitute for b chains, simultaneously generating more hemoglobin and reducing the burden ofa-globin inclusions. The termsb-thalassemia major andb-thalassemia intermedia are used to reflect the clinical heterogeneity. Patients withb-thalassemia major require intensive transfusion support to survive. Patients with b-thalassemia intermedia have a somewhat milder phenotype and can survive without transfusion. The terms b-thalassemia minor andb-thalassemia trait describe asymptomatic heterozygotes for bthalassemia. a-THALASSEMIA SYNDROMES The four classica thalassemias, most common in Asians, area-thalassemia-2 trait, in which one of the foura-globin loci is deleted;a-thalassemia-1 trait, with two deleted loci; HbH disease, with three loci deleted; and hydrops fetalis with Hb Bart's, with all four loci deleted (Table 106-4). Nondeletion forms ofa thalassemia also exist.
a-Thalassemia-2 trait is an asymptomatic, silent carrier state.a-Thalassemia-1 trait resemblesb-thalassemia minor. Offspring doubly heterozygous fora-thalassemia-2 anda-thalassemia-1 exhibit a more severe phenotype, called HbH disease. Heterozygosity for a deletion that removes both genes from the same chromosome (cis deletion) is common in Asians and Mediterranean individuals, as is homozygosity for a-thalassemia-2 (trans deletion). Both produce asymptomatic hypochromia and microcytosis. In HbH disease, HbA production is only 25 to 30% of normal. Fetuses accumulate some unpaired b chains. In adults, unpairedb chains accumulate and are soluble enough to formb4tetramers called HbH. HbH forms few inclusions in erythroblasts but does precipitate in circulating red cells. Patients with HbH disease have thalassemia intermedia characterized by moderately severe hemolytic anemia but milder ineffective erythropoiesis. Survival into midadult life without transfusions is common. The homozygous state for thea-thalassemia-1 cis deletion (hydrops fetalis) causes total absence of a-globin synthesis. No physiologically useful hemoglobin is produced beyond the embryonic stage. Excessg globin forms tetramers called Hb Bart's (g4), which has an extraordinarily high oxygen affinity. It delivers almost no O2 to fetal tissues, causing tissue asphyxia, edema (hydrops fetalis), congestive heart failure, and death in utero. a-Thalassemia-2 trait is common (15 to 20%) among people of African descent. The cis a-thalassemia-1 deletion is almost never seen, however. Thus, a-thalassemia-2 and the trans form of a-thalassemia-1 are very common, but HbH disease and hydrops fetalis are almost never encountered. DIAGNOSIS AND MANAGEMENT The diagnosis ofb-thalassemia major is readily made during childhood on the basis of severe anemia accompanied by hepatosplenomegaly; profound microcytosis; a characteristic blood smear (Plate V-2); and elevated levels of HbF, HbA2, or both. Many patients require chronic hypertransfusion therapy designed to maintain a hematocrit of at least 27 to 30% so that erythropoiesis is suppressed. Splenectomy is required if the annual transfusion requirement (volume ofRBCs per kilogram body weight per year) increases by>50%. Folic acid supplements may be useful. Vaccination with pneumococcal vaccine in anticipation of eventual splenectomy is advised, as is close monitoring for infection, leg ulcers, and biliary tract disease. Early endocrine evaluation is required for glucose intolerance, thyroid dysfunction, and delayed onset of puberty or secondary sexual characteristics. Many patients develop endocrine deficiencies as a result of iron overload. Patients withb-thalassemia intermedia exhibit similar stigmata but can survive without chronic hypertransfusion. Management is particularly challenging because a number of factors can aggravate the anemia, including infection, onset of puberty, and development of splenomegaly and hypersplenism. Some patients may eventually benefit from splenectomy. The expanded erythron can cause excess absorption of dietary iron and hemosiderosis, even without transfusion. b-Thalassemia minor (i.e., thalassemia trait) usually presents as profound microcytosis and hypochromia with target cells but only minimal or mild anemia. The mean
corpuscular volume is rarely >75 fL; the hematocrit is rarely 100 fL] suggests the presence of a megaloblastic anemia. Other causes of macrocytosis include hemolysis, liver disease, alcoholism, hypothyroidism, and aplastic anemia. If the macrocytosis is marked (MCV > 110 fL), the patient is much more likely to have a megaloblastic anemia. Macrocytosis is less marked with concurrent iron deficiency or thalassemia. The reticulocyte count is low, and the leukocyte and platelet count may also be decreased, particularly in severely anemic patients. The blood smear (see Plate V-24) demonstrates marked anisocytosis and poikilocytosis, together with macroovalocytes, which are large, oval, fully hemoglobinized erythrocytes typical of megaloblastic anemias. There is some basophilic stippling, and an occasional
nucleatedRBC may be seen. In the white blood cell series, the neutrophils show hypersegmentation of the nucleus (see Plate V-38). This is such a characteristic finding that a single cell with a nucleus of six lobes or more should raise the immediate suspicion of a megaloblastic anemia. A rare myelocyte may also be seen. Bizarre, misshapen platelets are also observed. The reticulocyte index is low. The bone marrow is hypercellular with a decreased myeloid/erythroid ratio and abundant stainable iron. RBC precursors are abnormally large and have nuclei that appear much less mature than would be expected from the development of the cytoplasm (nuclear-cytoplasmic asynchrony). The nuclear chromatin is more dispersed than expected, and it condenses in a peculiar fenestrated pattern that is very characteristic of megaloblastic erythropoiesis. Abnormal mitoses may be seen. Granulocyte precursors are also affected, many being larger than normal, including giant bands and metamyelocytes. Megakaryocytes are decreased and show abnormal morphology. Megaloblastic anemias are characterized by ineffective erythropoiesis (Chap. 61). In a severely megaloblastic patient, as many as 90% of theRBCprecursors may be destroyed before they are released into the bloodstream, compared with 10 to 15% in normal individuals. Enhanced intramedullary destruction of erythroblasts results in an increase in unconjugated bilirubin and lactic acid dehydrogenase (isoenzyme 1) in plasma. Abnormalities in iron kinetics also attest to the presence of ineffective erythropoiesis, with increased iron turnover but low incorporation of labeled iron into circulating RBCs. In evaluating a patient with megaloblastic anemia, it is important to determine whether there is a specific vitamin deficiency by measuring serum cobalamin and folate levels. The normal range of cobalamin in serum is 200 to 900 pg/mL; values34 umol/L or 2 mg/dL). The unconjugated (indirect) bilirubin level can be further elevated by a commonly encountered defect in conjugation of bilirubin (Gilbert's syndrome) (Chap. 294). In patients with hemolysis, the level of unconjugated bilirubin never exceeds 70 to 85 umol/L (4 to 5 mg/dL), unless liver function is impaired. In the absence of tissue damage in other organs, serum enzyme levels can be useful in
the diagnosis and monitoring of patients with hemolysis. Lactate dehydrogenase (LDH), particularly LDH-2, is elevated by acceleratedRBCdestruction. Serum AST (SGOT) may be somewhat elevated, whereas ALT (SGPT) is not. Haptoglobin is ana globulin that is present in high concentration (~1.0 g/L) in the plasma (and serum). It binds specifically and tightly to the globin in hemoglobin. The hemoglobin-haptoglobin complex is cleared within minutes by the mononuclear phagocyte system. Thus patients with significant hemolysis, either intravascular or extravascular, have low or absent levels of serum haptoglobin. The fact that haptoglobin synthesis is decreased in patients with hepatocellular disease and increased in inflammatory states must be considered in the interpretation of serum haptoglobin. Intravascular hemolysis (which is uncommon) results in the release of hemoglobin into the plasma. In these cases, plasma hemoglobin is increased in proportion to the degree of hemolysis. Plasma hemoglobin may be falsely elevated due to lysis ofRBC in vitro. If the haptoglobin-binding capacity of the plasma is exceeded, free hemoglobin passes through renal glomeruli. This filtered hemoglobin is reabsorbed by the proximal tubule, where it is catabolized in situ, and the heme iron is incorporated into storage proteins (ferritin and hemosiderin). The presence of hemosiderin in the urine, detected by staining the sediment with Prussian blue, indicates that a significant amount of circulating free hemoglobin has been filtered by the kidneys. Hemosiderin appears 3 to 4 days after the onset of hemoglobinuria and may persist for weeks after its cessation. When the absorptive capacity of the tubular cells is exceeded, hemoglobinuria ensues. Hemoglobinuria indicates severe intravascular hemolysis. Hemoglobinuria must be distinguished from hematuria (in which case RBC are seen on urine examination) and from myoglobin due to rhabdomyolysis; in all three cases, the urine is positive with the benzidine reaction, commonly used in analysis of urine. The distinction between hemoglobinuria and myoglobinuria can best be made by specific tests that exploit immunologic differences or differences in solubility. After centrifugation of an anticoagulated blood specimen, the plasma of patients with hemoglobinuria has a reddish-brown color, whereas that of patients with myoglobinuria is normal in color. Because of its higher molecular weight, hemoglobin has lower glomerular permeability than myoglobin and is less rapidly cleared by the kidneys. CLASSIFICATION The hemolytic anemias can be grouped in three different ways, shown inTable 108-3. The cause of acceleratedRBCdestruction can be regarded as (1) a molecular defect (hemoglobinopathy or enzymopathy) inside the red cell, (2) an abnormality in membrane structure and function, or (3) an environmental factor such as mechanical trauma or an autoantibody. In intracorpuscular types of hemolysis, the patient's RBC have an abnormally short life span in a normal recipient (with a compatible blood type), while compatible normal RBC survive normally in the patient. The opposite is true in extracorpuscular types of hemolysis. Finally, hemolytic disorders can be classified as either inherited or acquired. INHERITED HEMOLYTIC ANEMIAS The inherited hemolytic anemias are due to inborn defects in one of three main
components of red cells: the membrane, the enzymes, or hemoglobin. These defects are often known at the genomic level, but their identification still largely depends on their clinical and laboratory manifestations. Red Cell Membrane Disorders These are usually readily detected by morphologic abnormalities of the RBC on the blood film. There are three types of inherited RBC membrane abnormalities: hereditary spherocytosis, hereditary elliptocytosis (including hereditary pyropoikilocytosis), and hereditary stomatocytosis. Hereditary spherocytosis This condition is characterized by spherical RBC due to a molecular defect in one of the proteins in the cytoskeleton of the RBC membrane, leading to a loss of membrane and hence decreased ratio of surface area to volume and consequently spherocytosis. This disorder usually has an autosomal dominant inheritance pattern and an incidence of approximately 1:1000 to 1:4500. In ~20% of patients, the absence of hematologic abnormalities in family members suggests either autosomal recessive inheritance or a spontaneous mutation. The disorder is sometimes clinically apparent in early infancy but often escapes detection until adult life. CLINICAL MANIFESTATIONS The major clinical features of hereditary spherocytosis are anemia, splenomegaly, and jaundice. The prominence of jaundice accounts for the disorder's prior designation as "congenital hemolytic jaundice" and is due to an increased concentration of unconjugated (indirect-reacting) bilirubin in plasma. Jaundice may be intermittent and tends to be less pronounced in early childhood. Because of the increased bile pigment production, pigmented gallstones are common, even in childhood. Compensatory erythroid hyperplasia of the bone marrow occurs, with the extension of red marrow into the midshafts of long bones and occasionally with extramedullary erythropoiesis, at times leading to the formation of paravertebral masses visible on chest x-ray. Because the bone marrow's capacity to increase erythropoiesis by six- to eightfold exceeds the usual rate of hemolysis, anemia is usually mild or moderate and may even be absent in an otherwise healthy individual. Compensation may be temporarily interrupted by episodes of relative erythroid hypoplasia precipitated by infections, particularly parvovirus, trauma, surgery, and pregnancy. Splenomegaly is very common. The hemolytic rate may increase transiently during systemic infections, which induce further splenic enlargement. Chronic leg ulcers, similar to those observed in sickle cell anemia, occur occasionally. The characteristic erythrocyte abnormality is the spherocyte (Plate V-26). The mean corpuscular volume (MCV) is usually normal or slightly decreased, and the mean corpuscular hemoglobin concentration (MCHC) is increased to 350 to 400 g/L. Spheroidicity may be quantitatively assessed by measurement of the osmotic fragility of theRBC on exposure to hypoosmotic solutions causing a net influx of water (Fig. 108-1). Because spherocytes have a decreased surface area per unit volume, they are able to take in less water and hence lyse at a higher concentration of saline than normal cells. On microscopic examination, spherocytes are usually detected as small cells without central pallor. They will ordinarily not influence the osmotic fragility test unless they constitute more than 1 or 2% of the total cell population. The autohemolysis test, which measures the amount of spontaneous hemolysis occurring after 48 h of sterile incubation, is also useful.
PATHOGENESIS The molecular abnormality in hereditary spherocytosis primarily involves the proteins responsible for tethering the lipid bilayer to the underlying cytoskeletal network. Nearly all patients have a significant deficiency of spectrin, which is sometimes secondary to an inherited molecular defect in that protein. About 50% of patients have a defect in ankyrin, the protein that forms a bridge between protein 3 and spectrin (Fig. 108-2). Homozygotes who have a recessive inheritance pattern for ankyrin deficiency have more severe anemia than heterozygotes with the more common dominant form. About 25% of patients have a mutation of protein 3, resulting in a deficiency of that protein and mild anemia with dominant inheritance. Most of the remaining 25% have mutations of spectrin, leading to impaired synthesis or self-association;b-spectrin deficiency is generally mild, with dominant inheritance, while a-spectrin deficiency is severe, with a recessive inheritance pattern. Because the lipid bilayer is not well anchored when these proteins are defective, part of it is lost by vesiculation, resulting in a more spherical and less deformable cell. Because of their shape and rigidity, spherocytes cannot traverse the interstices of the spleen where their increased metabolic rate cannot be sustained, causing a further loss of surface membrane. This "conditioning" produces a subpopulation of hyperspheroidalRBC in the peripheral blood. DIAGNOSIS Hereditary spherocytosis must be distinguished primarily from the spherocytic hemolytic anemias associated withRBCantibodies. The family history of anemia and/or splenectomy is helpful, when present. The diagnosis of immune spherocytosis is usually readily established by a positive direct Coombs test (see below). Spherocytes are also seen in association with hemolysis induced by splenomegaly in patients with cirrhosis, in clostridial infections, and in certain snake envenomations (due to the action of phospholipases on the membrane). A few spherocytes are seen in the course of a wide variety of hemolytic disorders, particularly glucose-6-phosphate dehydrogenase (G6PD) deficiency. TREATMENT Splenectomy reliably corrects the anemia, although theRBCdefect and its consequent morphology persist. The operative risk is low. RBC survival after splenectomy is normal or nearly so; if it is not, an accessory spleen or another diagnosis should be sought. Because of the potential for gallstones and for episodes of bone marrow hypoplasia or hemolytic crises, splenectomy should be performed in symptomatic individuals; cholecystectomy should not be performed without splenectomy, as intrahepatic gallstones may result. Splenectomy in children should be postponed until age 4, if possible, to minimize the risk of severe infections with gram-positive encapsulated organisms. Polyvalent pneumococcal vaccine should be administered at least 2 weeks before splenectomy. In patients with severe hemolysis, folic acid (1 mg/d) should be administered prophylactically. Hereditary elliptocytosis and hereditary pyropoikilocytosis Oval or ellipticRBC are normally found in birds, reptiles, camels, and llamas; however, they occur in appreciable numbers in humans only in hereditary elliptocytosis, a disorder that is transmitted as an autosomal dominant trait and affects 1 per 4000 to 5000 people, a frequency similar to that of hereditary spherocytosis (rarely, patients with myelodysplastic disorders of the bone marrow may have acquired elliptocytosis). The elliptic shape is acquired as the
cell deforms to traverse the microcirculation but does not spring back to its initial biconcave shape. In most affected individuals, a structural abnormality of erythrocyte spectrin that leads to impaired assembly of the cytoskeleton. In some families, affected individuals have a deficiency of erythrocyte membrane protein 4.1, which stabilizes the interaction of spectrin and actin in the cytoskeleton (Fig. 108-2); homozygotes with absence of this protein have more marked hemolysis. In Southeast Asia, there is a high incidence of hereditary ovalocytosis, in which a small internal deletion of protein 3 makes the membrane rigid and confers resistance against malaria. The great majority of patients manifest only mild hemolysis, with hemoglobin levels>120 g/L, reticulocytes75% ofRBC are elliptic, with an axial ratio (width/length) of90% of patients succumb to their underlying liver disease within 1 year of the diagnosis of spur cell anemia. Paroxysmal Nocturnal Hemoglobinuria (PNH) This hemolytic disorder is distinctive because it is an intracorpuscular defect acquired at the stem cell level. Clinical manifestations The three common manifestations ofPNH are: hemolytic anemia, venous thrombosis, and deficient hematopoiesis. Anemia is highly variable with hematocrit values ranging from£20% to normal.RBC are normochromic and normocytic unless iron deficiency has occurred from chronic iron loss in the urine. Granulocytopenia and thrombocytopenia are common and reflect deficient hematopoiesis. Clinical hemoglobinuria is intermittent in most patients and never occurs in some, but hemosiderinuria is usually present. The lack of two proteins, decay-accelerating factor (DAF, CD55) and a membrane inhibitor of reactive lysis (MIRL, CD59) (see below) make theRBC more sensitive to the lytic effect of
complement. DAF normally disrupts the enzyme complexes from either the classical (antibody-driven) pathway or the alternative pathway that activate C3 and C5; CD59 inhibits the conversion of C9 by the membrane attack complex C5b-8 to a polymeric complex capable of penetrating the membrane. The platelets also lack these proteins, but the life span of the platelet is normal. However, the activation of complement indirectly stimulates platelet aggregation and hypercoagulability; this probably accounts for the tendency to thrombosis seen inPNH. Venous thrombosis is a common complication of patients of European origin, affecting ~40% at one time or another; it is less common in Asian patients. It occurs primarily in intraabdominal veins (hepatic, portal, mesenteric) and results in the Budd-Chiari syndrome, congestive splenomegaly, and abdominal pain. It may occur in cerebral venous sinuses and is a common cause of death in patients withPNH. The bone marrow may appear normocellular, but in vitro marrow progenitor assays are abnormal. In about 15 to 30% of long-term survivors of aplastic anemia, PNH cells appear in the circulation; in some patients, the manifestations of PNH become dominant. Patients with PNH may have aplastic periods lasting from weeks to years. PNH may be seen in association with other stem cell disorders, including myelofibrosis, and (rarely) other myelodysplastic or myeloproliferative disorders. Pathogenesis PNH is an acquired clonal disease, arising from an inactivating somatic mutation in a single abnormal stem cell of a gene on the X-chromosome (pig-A) important for the biosynthesis of the glycosylphosphatidylinositol (GPI) anchor. This anchor is necessary for the attachment of a number of proteins to the external membrane surface, and its partial or complete absence results in the absence of those proteins; to date, about 20 proteins have been found to be missing on the blood cells of patients with PNH. The normal clone of stem cells does not completely disappear, and the proportion of cells that are abnormal varies among patients and over time in a single patient. Diagnosis PNHshould be suspected in anyone with otherwise unexplained hemolytic anemia, especially with leukopenia and/or thrombocytopenia and with evidence of intravascular hemolysis (hemoglobinemia, hemoglobinuria, hemosiderinuria, elevatedLDH). Anyone recovering from aplastic anemia should be examined at intervals for the appearance of the diagnostic cells. The diagnosis is often delayed because (1) it is not considered, (2) hemoglobinuria is confused with hematuria, (3) elevation of the LDH is attributed to liver disease, and (4) the diagnostic tests (Ham's test and the sucrose lysis test) are not reliable. For many years, the diagnosis ofPNHdepended on the demonstration of the lysis ofRBCafter complement activation either by acid (Ham or acidified serum lysis test) or by reduction in ionic strength (sucrose lysis test). These tests are inferior to the analysis of GPI-linked proteins (e.g., CD59, DAF) on RBC and granulocytes by flow cytometry. TREATMENT
Transfusion therapy is useful inPNH not only for raising the hemoglobin level but also for suppressing the marrow production ofRBCduring episodes of sustained hemoglobinuria. Washed RBC are the preferred source to prevent exacerbation of hemolysis. Therapy with androgens sometimes results in a rise in hemoglobin level. Glucocorticoids reduce the rate of hemolysis in moderate doses (15 to 30 mg prednisone) on alternate days. Iron deficiency is common. Iron replacement may exacerbate hemolysis because of the formation of many newRBC, which may be sensitive to complement. This occurrence may be minimized by giving prednisone (60 mg/d) or by suppressing the bone marrow with transfusions. Acute thrombosis inPNH, particularly the Budd-Chiari syndrome and cerebral thrombosis, should be treated with thrombolytic agents. Heparin therapy should be instituted rapidly and maintained for several days before changing to coumadin therapy. Antithymoctye globulin (total dose of 150 mg/kg over 4 to 10 days) is often of use in treating marrow hypoplasia; prednisone counteracts the immune-complex disease that results from the administration of this foreign protein. In patients with either hypoplasia or thrombosis who have an appropriate sibling donor, marrow transplantation should be considered early in the course of the disease. The usual conditioning programs are sufficient to eradicate the aberrant clone. ANEMIA OF ACUTE BLOOD LOSS The normal capacity to compensate for acute blood loss involves cardiovascular mechanisms, an adjustment in the oxygen affinity of hemoglobin, and an increase in erythropoiesis in the marrow. The signs and symptoms of blood loss relate to the volume of the blood loss and the time frame over which the hemorrhage occurs (Table 108-9). Losses of up to 20% of the blood volume are normally tolerated by redistribution of blood flow mediated by reflex venospasm, but the presence of fever or pain may interfere with this compensation. With larger losses, blood volume redistribution is not adequate to maintain normal blood pressure: initially, hypotension is only seen on standing, but with greater losses progressively greater problems are encountered in maintaining blood pressure in sitting or supine positions. If the blood loss is more gradual, plasma volume increases, but albumin production usually lags behind the fluid shifts. It may take 2 to 3 days for the liver to generate the albumin lost in a 1500-mL bleed. The most rapid hematologic adjustment to acute blood loss is an increase in oxygen delivery to the tissues. This is first mediated by the Bohr effect, where the more acidic milieu of the hypoperfused hypoxic tissues shifts the hemoglobin oxygen dissociation curve to the right. Over several hours theRBCincrease their production of 2,3-bisphosphoglycerate, which also enhances the unloading of oxygen to tissues. These two mechanisms can substantially increase the capacity of RBC to deliver oxygen to the tissues. The marrow response to hemorrhage is related to the generation of erythropoietin in the kidney in response to decreased oxygen tensions. A normal response depends on the production of erythropoietin, the presence of normal erythroid progenitors in the marrow,
and an adequate supply of iron. If these three elements are normal, reticulocytes begin to increase in number in the first 2 days based on early release of reticulocytes from the marrow. However, it takes 3 to 6 days for erythroid hyperplasia to appear and 7 to 10 days before the erythropoietic response is maximal, producing reticulocyte counts up to 20 to 30%, a reticulocyte index of³3, and a marked increase in the marrow erythoid/granulocytic ratio. DIAGNOSIS Usually it is clear that a patient is bleeding; however, in some cases, large volumes of blood loss can occur internally from the gastrointestinal tract (esophageal varicies, cancer in the stomach or colon), a ruptured spleen, fractures and other trauma, or other lesions that can cause massive hemorrhage into the peritoneal cavity, pleural cavity, or the retroperitoneal space. Patients who have bled sufficiently to develop hypotension generally develop anemia, which is apparent only after volume replacement. The granulocyte count may increase to³20,000 cells/uL and include immature cell types such as metamyelocytes and myelocytes. Epinephrine-induced demargination of peripheral granulocytes and release of cells from the marrow may account for this change. NucleatedRBC may appear in the circulation, and platelet counts may exceed 1´106/uL. The basis for the increased platelet count is unclear. Hemorrhage in an internal cavity is accompanied by a rise in unconjugated bilirubin and a fall in serum haptoglobin. TREATMENT Treatment of the underlying cause of the hemorrhage is of paramount importance. If the patient is severely anemic or sufficiently hypovolemic, packedRBCshould be transfused. In less severe cases, if the patient has normal kidneys (and presumably a normal erythropoietin response to anemia), normal bone marrow function, and an adequate supply of iron, no specific therapy for the anemia is required. (Bibliography omitted in Palm version) Back to Table of Contents
109. APLASTIC ANEMIA, MYELODYSPLASIA, AND RELATED BONE MARROW FAILURE SYNDROMES - Neal S. Young The hypoproliferative anemias associated with marrow damage include aplastic anemia, myelodysplasia (MDS), pure red cell aplasia (PRCA), and myelopthisis. Anemia in these disorders, which is normochromic, normocytic, or macrocytic and characterized by low reticulocyte count, is not a solitary or even the major finding in these diseases, which are better described as marrow failure states. In bone marrow failure, pancytopenia -anemia, leukopenia, and thrombocytopenia (sometimes in various combinations) -results from deficient hematopoiesis, as distinguished from blood count depression due to peripheral destruction of red cells (hemolytic anemias), platelets (idiopathic thrombocytopenic purpura or due to splenomegaly), and granulocytes (as in the immune leukopenias). Hematopoietic failure syndromes are classified by dominant morphologic features of the bone marrow (Table 109-1). While practical distinction among these syndromes is clear in stereotypical cases, they can, of course, occur secondary to other diseases, and are so closely related that the differential diagnosis may be arbitrary, patients may seem to suffer from two or three related diseases simultaneously, or one diagnosis may appear to evolve into another. Finally, there is an important pathophysiologic relationship among these syndromes in their sharing of immune-mediated mechanisms of marrow destruction and some element of genomic instability resulting in a higher rate of malignant transformation. APLASTIC ANEMIA DEFINITION Aplastic anemia is pancytopenia with bone marrow hypocellularity. Acquired aplastic anemia is distinguished from iatrogenic marrow aplasia, the common occurrence of marrow hypocellularity after intensive cytotoxic chemotherapy for cancer. Aplastic anemia can also be constitutional: the genetic disease Fanconi's anemia, while frequently associated with typical physical anomalies and the development of pancytopenia early in life, can also present as marrow failure in normal-appearing adults. Acquired aplastic anemia is often stereotypical in its manifestations, with the abrupt onset of low blood counts in a previously well young adult; seronegative hepatitis or a course of an incriminated medical drug may precede the onset. The diagnosis in these instances is uncomplicated. Sometimes blood count depression is moderate or incomplete, resulting in anemia, leukopenia, and thrombocytopenia in some combination. Aplastic anemia is related to both paroxysmal nocturnal hemoglobinuria (PNH;Chap. 108) and toMDS, and in some cases a clear distinction among these disorders may not be possible. EPIDEMIOLOGY The incidence of acquired aplastic anemia in Europe and Israel is 2 cases per million persons annually. In Thailand and China, rates of 5 to 7 per million have been established. In general, men and women are affected with equal frequency, but there is a biphasic age distribution, with the major peak among older children and young adults
and a second rise in the elderly. ETIOLOGY The origins of aplastic anemia have been inferred from several recurring clinical associations (Table 109-2); unfortunately, these relationships are neither a reliable guide in an individual patient nor necessarily etiologic. In addition, while most cases of aplastic anemia are idiopathic, little other than history separates these cases from those with a presumed etiology such as a drug exposure. Radiation Marrow aplasia is a major acute sequela of radiation. Radiation damages DNA; tissues dependent on active mitosis are particularly susceptible. Nuclear accidents can involve not only power plant workers but also employees of hospitals, laboratories, and industry (food sterilization, metal radiography, etc.), as well as innocents exposed to stolen, misplaced, or misused sources. While the radiation dose can be approximated from the rate and degree of decline in blood counts, dosimetry by reconstruction of the exposure can help to estimate the patient's prognosis and also to protect medical personnel from contact with radioactive tissue and excreta.MDS and leukemia, but probably not aplastic anemia, are late effects of irradiation. Chemicals Benzene is a notorious cause of bone marrow failure. Vast quantities of epidemiologic, clinical, and laboratory data link benzene to aplastic anemia, acute leukemia, and blood and marrow abnormalities. The occurrence of leukemia is roughly correlated with cumulative exposure, but susceptibility must also be important, as only a minority of even heavily exposed workers develop benzene myelotoxicity. The employment history is important, especially in industries where benzene is used for a secondary purpose, usually as a solvent. Benzene-related blood diseases have declined with regulation of industrial exposure. Although benzene is no longer generally available as a household solvent, exposure to its metabolites occurs in the normal diet and in the use of lead-free gasoline. The association between marrow failure and other chemicals that contain a benzene ring is much less well substantiated; these chemicals may have been contaminated with benzene in manufacture, or petroleum distillates may have been used to dissolve the product. Drugs (SeeTable 109-3) Many chemotherapeutic drugs have marrow suppression as a major toxicity; effects are dose-dependent and will occur in all recipients. In contrast, idiosyncratic reactions to a large and diverse group of drugs may lead to aplastic anemia without a clear dose-response relationship. These associations rest largely on accumulated case reports, but a massive international study in Europe in the 1980s quantitated drug relationships, especially for nonsteroidal analgesics, sulfonamides, thyrostatic drugs, some psychotropics, penicillamine, allopurinol, and gold. Not all associations necessarily reflect causation: a drug may have been used to treat the first symptoms of bone marrow failure (antibiotics for fever or the preceding viral illness) or provoked the first symptom of a preexisting disease (petechiae by nonsteroidal anti-inflammatory agents administered to the thrombocytopenic patient). In the context of total drug use, idiosyncratic reactions, while individually devastating, are exceedingly rare events. Chloramphenicol, the most infamous culprit, reportedly produced aplasia in only about 1/60,000 therapy courses, and even this number is almost certainly an overestimate (risks are almost invariably exaggerated when based on collections of
cases; although the introduction of chloramphenicol was perceived to have created an epidemic of aplastic anemia, its diminished use was not followed by a changed frequency of marrow failure). Risk estimates are usually lower when determined in population-based studies; furthermore, the low absolute risk is also made more obvious: even a 10- or 20-fold increase in risk translates, in a rare disease, to but a handful of drug-induced aplastic anemia cases among hundreds of thousands of exposed patients. Infections Hepatitis is the most common preceding infection, and posthepatitis marrow failure accounts for about 5% of etiologic associations in most series. Patients are usually young men who have recovered from a mild bout of liver inflammation 1 to 2 months earlier; the subsequent pancytopenia is very severe. The hepatitis is almost invariably seronegative (non-A, non-B, non-C, non-G) and presumably due to a novel, as yet undiscovered, virus. Fulminant liver failure in childhood can follow seronegative hepatitis, and marrow failure occurs at a high rate in these patients as well. Aplastic anemia can rarely follow infectious mononucleosis, and Epstein-Barr virus has been found in the marrow of a few aplastic anemia patients, some without a suggestive preceding history. Parvovirus B19, the cause of transient aplastic crisis in hemolytic anemias and of some pure red cell aplasia (see below), does not usually cause generalized bone marrow failure. Blood count depression is frequent in the course of many viral and bacterial infections but is comparatively moderate and resolves with the infection. Immunologic Diseases Aplasia is a major consequence and the cause of death in transfusion-associated graft-versus-host disease, which can occur after infusion of unirradiated blood products to an immunodeficient recipient. Aplastic anemia is strongly associated with the rare collagen vascular syndrome called eosinophilic fasciitis, which is characterized by painful induration of subcutaneous tissues (Chap. 313). Pancytopenia with marrow hypoplasia can also occur in systemic lupus erythematosus. Pregnancy Aplastic anemia very rarely may occur and recur during pregnancy and resolve with delivery or with spontaneous or induced abortion. Paroxysmal Nocturnal Hemoglobinuria An acquired mutation in the PIG-A gene in a hematopoietic stem cell is required for the development ofPNH, but PIG-A mutations probably occur commonly in normal individuals. If the PIG-A mutant stem cell proliferates, the result is a clone of progeny deficient in glycosylphosphatidylinositol-linked cell surface membrane proteins (Chap. 108). Such PNH cells are now most accurately enumerated using fluorescence-activated flow cytometry of CD55 or CD59 expression on granulocytes rather than Ham or sucrose lysis tests on red cells. Deficient cells can be detected in about a quarter of patients with aplastic anemia at the time of presentation [and PNH cells are also seen in cases of hypocellularMDS (see below)]. In addition, functional studies of bone marrow from PNH patients, even those with mainly hemolytic manifestations, show evidence of defective hematopoiesis. Patients with an initial clinical diagnosis of PNH, especially younger individuals, may later develop frank marrow aplasia and pancytopenia; patients with an initial diagnosis of aplastic anemia may suffer from hemolytic PNH years after recovery of blood counts. One explanation for the aplastic anemia/PNH syndrome is selection of the deficient clones, perhaps because they are favored for proliferation in the peculiar environment of immune-mediated marrow destruction.
Congenital Disorders Fanconi's anemia, an autosomal recessive disorder, manifests as progressive pancytopenia, increased chromosome fragility, congenital developmental anomalies, and an increased risk of malignancy. Patients with Fanconi's anemia typically have short stature; cafe au lait spots; and anomalies involving the thumb, radius, and genitourinary tract. At least seven different genetic defects have been defined by complementation analysis. The most common, type A Fanconi's anemia, is due to a mutation in FANCA. The function of the four cloned genes so far identified in Fanconi's anemia remains unknown. Patients with Shwachman-Diamond syndrome may develop pancreatic insufficiency, malabsorption, and neutropenia and are at risk of aplastic anemia. Dyskeratosis congenita is an X-linked disorder characterized by mucous membrane leukoplasia, dystrophic nails, reticular hyperpigmentation, and the later development of aplastic anemia in about half of patients. Mutation in the DKC1 (dyskerin) gene has been found in some cases. PATHOPHYSIOLOGY Bone marrow failure results from severe damage to the hematopoietic cell compartment. In aplastic anemia, replacement of the bone marrow by fat is apparent in the morphology of the biopsy specimen (Fig. 109-1;Plate V-13) and magnetic resonance imaging of the spine; cells bearing the CD34 antigen, a marker of early hematopoietic cells, are greatly diminished; and in functional studies, committed and primitive progenitor cells are virtually absent -- in vitro assays have suggested that the stem cell pool is reduced to 1% of normal in severe disease at the time of presentation. Qualitative abnormalities, such as limited number of operating stem cell clones or shortened telomere length, may follow from the quantitative deficiency, reflecting the shrunken and stressed state of hematopoiesis. An intrinsic stem cell defect exists for constitutional aplastic anemia, as cells from patients with Fanconi's anemia exhibit chromosome damage and death on exposure to certain chemical agents, but there is no convenient mechanism for the propagation of an acquired genetic abnormality that would produce a hypoproliferative (as opposed to neoplastic) disease. Aplastic anemia does not appear to result from defective stroma or growth factor production. Drug Injury Extrinsic damage to the marrow follows massive physical or chemical insults such as high doses of radiation and toxic chemicals. For the more common idiosyncratic reaction to modest doses of medical drugs, altered drug metabolism has been invoked as a likely mechanism. The metabolic pathways of many drugs and chemicals, especially if they are polar and have limited water solubility, involve enzymatic degradation to highly reactive electrophilic compounds; these intermediates are toxic because of their propensity to bind to cellular macromolecules. For example, derivative hydroquinones and quinolones are responsible for benzene-induced tissue injury. Excessive generation of toxic intermediates or failure to detoxify the intermediates may be genetically determined and apparent only on specific drug challenge; the complexity and specificity of the pathways imply multiple susceptible loci and would provide an explanation for the rarity of idiosyncratic drug reactions. Immune-Mediated Injury The recovery of marrow function in some patients prepared
for bone marrow transplantation with antilymphocyte globulin (ALG) first suggested that aplastic anemia might be immune-mediated. Consistent with this hypothesis was the frequent failure of simple bone marrow transplantation from a syngeneic twin, without conditioning cytotoxic chemotherapy, which also argued both against simple stem cell absence as the cause and for the presence of a host factor producing marrow failure. Laboratory data support an important role for the immune system in aplastic anemia. Blood and bone marrow cells of patients can suppress normal hematopoietic progenitor cell growth, and removal of T cells from aplastic anemia bone marrow improves colony formation in vitro. Increased numbers of activated cytotoxic T cells are observed in aplastic anemia patients and usually decline with successful immunosuppressive therapy; cytokine measurements suggest a predominant T H1 immune response (interferon g, interleukin 2, and tumor necrosis factor). Interferon and tumor necrosis factor induce Fas expression on CD34 cells, leading to apoptotic cell death; localization of activated T cells to bone marrow and local production of their soluble factors are probably important in stem cell destruction. Early immune system events in aplastic anemia are not well understood. Many different exogeneous antigens appear capable of initiating a pathologic immune response, but at least some of the active T cells recognize true self-antigens. The rarity of occurrence of aplastic anemia despite common exposures (medical drugs, hepatitis virus) suggests that genetically determined features of the immune response can convert a normal physiologic response into a sustained abnormal autoimmune process. CLINICAL FEATURES History Aplastic anemia can appear with seeming abruptness or have a more insidious onset. Bleeding is the most common early symptom; a complaint of days to weeks of easy bruising, oozing from the gums, nose bleeds, heavy menstrual flow, and sometimes petechiae will have been noticed. With thrombocytopenia, massive hemorrhage is unusual, but small amounts of bleeding in the central nervous system can result in catastrophic intracranial or retinal hemorrhage. Symptoms of anemia are also frequent, including lassitude, weakness, shortness of breath, and a pounding sensation in the ears. Infection is an unusual first symptom in aplastic anemia (unlike in agranulocytosis, where pharyngitis, anorectal infection, or frank sepsis occur early). A striking feature of aplastic anemia is the restriction of symptoms to the hematologic system, and patients often feel and look remarkably well despite drastically reduced blood counts. Systemic complaints and weight loss should point to other etiologies of pancytopenia. History of drug use, chemical exposure, and preceding viral illnesses must often be elicited with repeated questioning. Physical Examination Petechiae and ecchymoses are often present, and retinal hemorrhages may be present. Pelvic and rectal examinations should be performed with great gentleness to avoid trauma; these will often show bleeding from the cervical os and blood in the stool. Pallor of the skin and mucous membranes is common except in the most acute cases or those already transfused. Infection on presentation is unusual but may be present if the patient has been symptomatic for a few weeks. Lymphadenopathy and splenomegaly are highly atypical of aplastic anemia. Cafe au lait spots and short stature suggest Fanconi's anemia; peculiar nails, dyskeratosis congenita.
LABORATORY STUDIES Blood The smear shows large erythrocytes and a paucity of platelets and granulocytes. Mean corpuscular volume (MCV) is commonly increased. Reticulocytes are absent or few, and lymphocyte numbers may be normal or reduced. The presence of immature myeloid forms suggests leukemia orMDS; nucleated red blood cells suggest marrow fibrosis or tumor invasion; abnormal platelets suggest either peripheral destruction or MDS. Bone Marrow The bone marrow is usually readily aspirated but appears dilute on smear, and the fatty biopsy specimen may be grossly pale on withdrawal; a "dry tap" suggests fibrosis or myelophthisis. In severe aplasia the smear of the aspirated specimen shows only red cells, residual lymphocytes, and stromal cells; the biopsy, which should be>1 cm in length, is superior for determination of cellularity and shows mainly fat under the microscope, with hematopoietic cells occupying, by definition, 10,000/uL (oozing from the gut, and presumably also from other vascular beds, increases precipitously at counts100 per million persons in the general population and 120 to >500 per million in the aged. MDS is rare in children, but monocytic leukemia can be seen. Therapy-related MDS is not age-related and may occur in as many as 15% of patients within a decade following intensive combined modality treatment for cancer. Rates of MDS have increased over time, due to the recognition of the syndrome by physicians and the aging of the population. ETIOLOGY AND PATHOPHYSIOLOGY The myelodysplastic syndromes have been convincingly linked to environmental exposures such as radiation and benzene; other risk factors have been reported inconsistently. SecondaryMDSoccurs as a stereotypical late toxicity of cancer treatment, usually with a combination of radiation and the radiomimetic alkylating agents such as busulfan, nitrosourea, or procarbazine (with a latent period of 5 to 7 years) or the DNA topoisomerase inhibitors (2 years). Both acquired aplastic anemia following immunosuppressive treatment and Fanconi's anemia can evolve into MDS. MDSis a clonal hematopoietic stem cell disorder leading to impaired cell proliferation and differentiation. Cytogenetic abnormalities are found in about half of patients, and
some of the same specific lesions are also seen in frank leukemia; deletions are more frequent than translocations. Both presenting and evolving hematologic manifestations result from the accumulation of multiple genetic lesions, loss of tumor suppressor genes, activating oncogene mutations, or other harmful alterations. Cytogenetic abnormalities are not random (loss of all or part of 5, 7, and 20, trisomy of 8) and may be related to etiology (11q23 following topoisomerase II inhibitors); chronic myelomonocytic leukemia is often associated with t(5;12) that creates a chimeric tel-PDGFb gene. The type and number of cytogenetic abnormalities strongly correlate with the probability of leukemic transformation and survival. Mutations of N-ras (an oncogene), p53 and IRF-1 (tumor suppressor genes), Bcl-2 (an antiapoptotic gene), and others have been reported in some patients but may occur relatively late in the sequence leading to leukemic transformation. Apoptosis of marrow cells is increased in MDS, presumably due to these acquired genetic alterations or possibly to an overlaid immune response. Sideroblastic anemia may be related to mutations in mitochondrial genes. Ineffective erythropoiesis and disordered iron metabolism are the functional consequences of the genetic alterations. CLINICAL FEATURES Anemia dominates the early course. Most symptomatic patients complain of the gradual onset of fatigue and weakness, dyspnea, and pallor, but at least half the patients are asymptomatic andMDS is discovered only incidentally on routine blood counts. Previous chemotherapy or radiation exposure is an important historic fact. Fever and weight loss should point to a myeloproliferative rather than myelodysplastic process. Children with Down's syndrome are susceptible to MDS, and a family history may indicate a hereditary form of sideroblastic anemia or Fanconi's anemia. The physical examination is remarkable for signs of anemia; about 20% of patients have splenomegaly. Some unusual skin lesions, including Sweet's syndrome (febrile neutrophilic dermatosis), have been associated withMDS. LABORATORY STUDIES Blood Anemia is present in the majority of cases, either alone or as part of bi- or pancytopenia; isolated neutropenia or thrombocytopenia is more unusual. Macrocytosis is common, and the smear may be dimorphic with a distinctive population of large red blood cells. Platelets are also large and lack granules. In functional studies, they may show marked abnormalities, and patients may have bleeding symptoms despite seemingly adequate numbers. Neutrophils are hypogranulated; have hyposegmented, ringed, or abnormally segmented nuclei; and contain Dohle bodies and may be functionally deficient. Circulating myeloblasts usually correlate with marrow blast numbers, and their quantitation is important for classification and prognosis. The total white blood cell count is usually normal or low, except in chronic myelomonocytic leukemia. As in aplastic anemia,MDS also can be associated with a clonal population ofPNHcells. Bone Marrow The bone marrow is usually normal or hypercellular but in 20% of cases is sufficiently hypocellular to be confused with aplasia. No single characteristic feature of marrow morphology distinguishesMDS, but the following are commonly observed:
dyserythropoietic changes (especially nuclear abnormalities) and ringed sideroblasts in the erythroid lineage; hypogranulation and hyposegmentation in granulocytic precursors, with an increase in myeloblasts; and megakaryocytes showing reduced numbers of disorganized nuclei. Prognosis strongly correlates with the proportion of marrow blasts. Cytogenetic analysis also is important. A much more sensitive method to detect infrequent chromosome aberrations is fluorescent in situ hybridization, and gene amplification by polymerase chain reaction can detect known chromosomal translocations. DIFFERENTIAL DIAGNOSIS Deficiencies of vitamin B12 or folate should be suggested by history and excluded by appropriate blood tests; vitamin B6deficiency can be assessed by a therapeutic trial of pyridoxine if the bone marrow shows ringed sideroblasts. Marrow dysplasia can be observed in acute viral infections, drug reactions, or chemical toxicity but should be transient. More difficult (arbitrary) are the distinctions between hypocellularMDS and aplasia or between refractory anemia with excess blasts in transformation and early acute leukemia. PROGNOSIS The median survival varies greatly withFAB type and, according toIPSScalculations, ranges from years for patients with 5q- or sideroblastic anemia to a few months in refractory anemia with excess blasts or severe pancytopenia associated with monosomy 7. Most patients die as a result of complications of pancytopenia and not due to leukemic transformation; perhaps one-third will succumb to other diseases unrelated to theirMDS. Precipitous worsening of pancytopenia, acquisition of new chromosomal abnormalities on serial cytogenetic determination, and increase in the number of blasts are all poor prognostic indicators. The outlook in therapy-related MDS, regardless of FAB type, is very poor, and most patients will progress within a few months to refractory acute myeloid leukemia. TREATMENT The therapy ofMDS is generally unsatisfactory. Only stem cell transplantation offers cure: survival rates of 40% have been reported, but older patients are particularly prone to develop treatment-related mortality and morbidity. Those with better prognostic features (and a more favorable natural history) have much better outcomes than patients with more malignant subtypes. Surprisingly, results of transplant using matched unrelated donor are comparable, although most series contain younger and more highly selected cases. MDShas been regarded as particularly refractory to cytotoxic chemotherapy regimens but is probably no more resistant to effective treatment than acute myeloid leukemia in the elderly, in whom drug toxicity is often fatal and remissions, if achieved, are brief. Low doses of cytotoxic drugs have been administered for their "differentiating" potential: responses to cytosine arabinoside did not translate into a survival advantage; etoposide and 5-azacytidine are under active study. Amifostine, an organic thiophosphonate that blocks apoptosis, can improve blood counts but has significant toxicities.
Immunosuppressive therapies, includingATG and cyclosporine, that are effective in aplastic anemia may induce sustained remissions in a high proportion of patients with refractory anemia, especially in those with hypocellular marrows or without cytogenetic abnormalities. Hematopoietic growth factors can improve blood counts but, as in most other marrow failure states, have been most beneficial in patients with the least severe pancytopenia. G-CSF treatment alone failed to improve survival in a controlled trial. The combination of G-CSF and erythropoietin increased blood counts in one-third to one-half of patients, but survival advantage is not yet proven. The same principles of supportive care described for aplastic anemia apply toMDS. Because many patients will be anemic for years, erythrocyte transfusion support should be accompanied by iron chelation in order to prevent secondary hemochromatosis. MYELOPHTHISIC ANEMIAS Fibrosis of the bone marrow (seePlate V-19), usually accompanied by a characteristic blood smear picture called leukoerythroblastosis, can occur as a primary hematologic disease, called myelofibrosis or myeloid metaplasia (Chap. 110), and as a secondary process, called myelophthisis. Myelophthisis, or secondary myelofibrosis, is reactive. Fibrosis can be a response to invading tumor cells, usually of an epithelial cancer of breast, lung, and prostate or neuroblastoma. Marrow fibrosis may occur with infection of mycobacteria (both Mycobacterium tuberculosis and M. avium) fungi, or HIV, and in sarcoidosis. Intracellular lipid deposition in Gaucher's disease and obliteration of the marrow space related to absence of osteoclast remodeling in congenital osteopetrosis also can produce fibrosis. Secondary myelofibrosis is a late consequence of radiation therapy or treatment with radiomimetic drugs. Usually, the infectious or malignant underlying processes are obvious. Marrow fibrosis can also be a feature of a variety of hematologic syndromes, especially chronic myeloid leukemia, multiple myeloma, lymphomas, myeloma, and hairy cell leukemia. The pathophysiology has three distinct features: proliferation of fibroblasts in the marrow space (myelofibrosis); the extension of hematopoiesis into the long bones and most particularly into extramedullary sites, usually the spleen, liver, and lymph nodes (myeloid metaplasia); and ineffective erythropoiesis. The etiology of fibrosis is unknown but most likely involves dysregulated production of growth factors: platelet-derived growth factor and transforming growth factorb have been implicated. Abnormal regulation of other hematopoietins would lead to localization of blood-producing cells in nonhematopoietic tissues and uncoupling of the usually balanced processes of stem cell proliferation and differentiation. Myelofibrosis is remarkable for pancytopenia despite extraordinarily large numbers of circulating hematopoietic progenitor cells. Anemia is dominant in secondary myelofibrosis, usually normocytic and normochromic. The diagnosis is suggested by the characteristic leukoerythroblastic smear (seePlate V-9). Erythrocyte morphology is very abnormal, with circulating nucleated red blood cells, teardrops, and shape distortions. White blood cell numbers are often elevated, sometimes mimicking a leukemoid reaction, with circulating myelocytes, promyelocytes, and myeloblasts. Platelets may be abundant and are often giant size. Inability to
aspirate the bone marrow, the characteristic "dry tap," can allow a presumptive diagnosis before the biopsy is decalcified. The course of secondary myelofibrosis is determined by its cause, usually a metastatic tumor or an advanced hematologic malignancy. Treatable causes must be excluded, especially tuberculosis and fungus. Transfusion support can relieve symptoms. (Bibliography omitted in Palm version) Back to Table of Contents
110. POLYCYTHEMIA VERA AND OTHER MYELOPROLIFERATIVE DISEASES Jerry L. Spivak Polycythemia vera, idiopathic myelofibrosis, essential thrombocytosis, and chronic myeloid leukemia (CML) are commonly classified together under the rubric the chronic myeloproliferative disorders, because their pathophysiology involves the clonal expansion of a multipotent hematopoietic progenitor cell with the overproduction of one or more of the formed elements of the blood. These entities may transform into acute leukemia naturally or as a consequence of mutagenic treatment. However, while polycythemia vera, idiopathic myelofibrosis, essential thrombocytosis, and CML share similar phenotypic characteristics, CML is genotypically distinct from the other three disorders because it alone is associated with translocation of genetic material between the long arms of chromosomes 9 and 22, resulting in the production of the unique fusion protein, bcr-abl. Furthermore, based on its natural history, CML is more appropriately considered as a form of leukemia.*CML is discussed with the acute myeloid leukemias in Chap. 111. POLYCYTHEMIA VERA Polycythemia vera is a clonal disorder involving a multipotent hematopoietic progenitor cell in which there is accumulation of phenotypically normal red cells, granulocytes, and platelets in the absence of a recognizable physiologic stimulus. Polycythemia vera, the most common of the chronic myeloproliferative disorders, occurs in about 2 per 100,000 people. It spares no adult age group. Vertical transmission has been documented, establishing a genetic basis for the disorder. A slight overall male predominance has been observed, but females predominate within the reproductive age range. ETIOLOGY The etiology of polycythemia vera is unknown. Although nonrandom chromosome abnormalities such as 20q-, trisomy 8 or 9 have been documented in a small percentage of untreated polycythemia vera patients, no consistent cytogenetic abnormality has been associated with the disorder and no specific genetic defect has yet been identified. Impaired posttranslational processing of the thrombopoietin receptor, Mpl, has been noted in polycythemia vera patients; the extent of the defect correlated with disease duration and splenomegaly. While this defect is specific for polycythemia vera and is not found in secondary polycythemias, its role in the pathophysiology of the disorder is still undefined. In contrast to normal erythroid progenitor cells, polycythemia vera erythroid progenitor cells can grow in vitro in the absence of erythropoietin due to hypersensitivity to insulin-like growth factor I. However, this phenotypic abnormality is not specific for polycythemia vera and has been documented in essential thrombocytosis and secondary polycythemias. Polycythemia vera erythroid progenitor cells are more resistant to apoptosis induced by erythropoietin deprivation, due to upregulation of bcl-XL, an antiapoptotic protein. The polycythemia vera erythroid progenitors do not divide more rapidly than their normal counterparts, but they accumulate because they do not die normally. Additionally, the transformed hematopoietic progenitor cells in polycythemia vera, as in other neoplastic disorders, exhibit clonal dominance and suppress the proliferation of normal hematopoietic progenitor cells by an unknown mechanism. Consequently, the circulating formed
elements of the blood represent only progeny of the transformed clone. CLINICAL FEATURES Although massive splenomegaly may be the initial presenting sign in polycythemia vera, most often the disorder is first recognized by the discovery of a high hemoglobin or hematocrit, and with the exception of aquagenic pruritus, no symptoms distinguish polycythemia vera from other causes of erythrocytosis. Uncontrolled erythrocytosis can lead to neurologic symptoms such as vertigo, tinnitus, headache, and visual disturbances. Systolic hypertension also accompanies an elevated red cell mass. In some patients, venous or arterial thrombosis may be the presenting manifestation of polycythemia vera. Intraabdominal venous thrombosis is particularly common and may be catastrophic when there is sudden compromise of the hepatic vein. Polycythemia vera should be suspected in any patient who develops the Budd-Chiari syndrome. Digital ischemia may also occur. Easy bruising, epistaxis, or gastrointestinal hemorrhage may be observed, and polycythemia vera patients are frequently hypermetabolic. Hypercuricemia with secondary gout and uric acid stones and acid-peptic disease also complicate the disorder. Because isolated erythrocytosis is a common initial presentation for polycythemia vera but no clonal marker is available for the disease, the first task of the physician is to distinguish this autonomous clonal form of erythrocytosis from the many other types of erythrocytosis, most of which are correctable (Table 110-1). Erythropoiesis is normally regulated by the glycoprotein hormone erythropoietin. Erythropoietin, which in adults is produced primarily in the kidneys and to a small extent in the liver, promotes the proliferation of erythroid progenitor cells, maintains their survival, and facilitates their differentiation. Because erythropoietin acts as a survival factor, it is constitutively produced and, like the red cell mass, its level is constant as long as tissue oxygenation is adequate. The plasma erythropoietin level, like the red cell mass, differs among individuals but in adults is not affected by either age or gender. Erythropoietin production is regulated at the level of gene transcription. Hypoxia is the only physiologic stimulus that increases the number of cells producing erythropoietin, and thus the production and metabolism of erythropoietin are independent of its plasma level. In the absence of renal or hepatic disease, plasma erythropoietin levels reflect erythropoietin production, and therefore the assay for plasma erythropoietin is a surrogate assay for tissue hypoxia. Erythropoietin is active at the picomolar level, and its production is tightly regulated. Thus, the plasma erythropoietin level does not rise outside the normal range until the hemoglobin level falls below 105 g/L. This is not meant to imply that an increase in erythropoietin production does not occur as the hemoglobin level falls below normal, but because the normal range for plasma erythropoietin is wide (4 to 26 mU/mL), unless the patient's baseline level is known, any increase will not be recognized until the hemoglobin falls below 105 g/L. Thereafter, there is a log-linear inverse correlation between the levels of plasma erythropoietin and hemoglobin. With erythrocytosis, erythropoietin production is suppressed; this suppression reflects not only the increase in tissue oxygen transport associated with the increase in red cell number but also additional negative-feedback mechanism unrelated to oxygen transport but related to the increase in blood viscosity and an increase in red cell precursors capable of taking up erythropoietin. The summation of these
mechanisms accounts for the paradoxical observation that many patients with hypoxic erythrocytosis due to cyanotic congenital heart disease or obstructive lung disease have a "normal" plasma erythropoietin level. The plasma erythropoietin level is a useful diagnostic test in patients with isolated erythrocytosis, because an elevated level essentially excludes polycythemia vera as the cause for the erythrocytosis. DIAGNOSIS When confronted with an elevated hemoglobin or hematocrit level, it is important to obtain previous values to determine the duration of this laboratory abnormality. Because the hemoglobin or hematocrit level is affected by the plasma volume, and hematocrit and red cell mass are not linearly related, a red cell mass determination must also be performed to distinguish absolute erythrocytosis from relative erythrocytosis due to a reduction in plasma volume alone (also known as stress or spurious erythrocytosis or Geisbock's syndrome). Red cell mass determination is important because in polycythemia vera, in contrast to erythropoietin-driven erythrocytosis, the plasma volume is frequently elevated, not only masking the true extent of red cell mass expansion but often its presence. Indeed, a significant proportion of patients with polycythemia vera have a hematocrit within the normal range, particularly in patients with a substantial splenomegaly. Failure to recognize this phenomenon is undoubtedly the basis for many of the reported instances of hepatic or portal vein thrombosis in patients with a so-called undefined myeloproliferative disorder. Red cell mass is reliably determined by isotope dilution using the patient's51Cr-tagged red cells; extrapolations made by determining directly only the plasma volume are unacceptable. Furthermore, to allow ample time for equilibration of the labeled red cells, measurements should be made over a period of³90 min. Once the presence of absolute erythrocytosis has been established, its cause must be determined. An elevated plasma erythropoietin level suggests either an hypoxic cause for erythrocytosis or autonomous erythropoietin production, in which case assessment of pulmonary function and an abdominal computed tomography scan to evaluate renal and hepatic anatomy are appropriate. A normal erythropoietin level does not exclude an hypoxic cause for erythrocytosis. In polycythemia vera, in contrast to hypoxic erythrocytosis, the arterial oxygen saturation is normal. However, a normal oxygen saturation does not exclude a high-affinity hemoglobin as a cause for erythrocytosis, and it is here that documentation of previous hemoglobin levels and a family study become important. Because there is no clonal marker for polycythemia vera, clinical guidelines have been proposed to define the disease. A modified version is provided inTable 110-2. However, these guidelines do not establish clonality, and in some patients only with time will the underlying disorder become apparent. Diagnostic ambiguity does not preclude the initiation of therapy. Other laboratory studies that may aid in diagnosis include the red cell count, mean corpuscular volume, and red cell distribution width (RDW). Only three situations cause microcytic erythrocytosis:b-thalassemia trait, hypoxic erythrocytosis, and polycythemia vera. However, with b-thalassemia trait the RDW is normal, whereas with hypoxic erythrocytosis and polycythemia vera, the RDW is usually elevated. A properly made blood smear from a patient with erythrocytosis will be virtually unreadable due to the
marked elevation in red cell count, but no specific morphologic abnormalities are seen in the leukocytes or platelets in polycythemia vera. However, when these are also elevated the diagnosis is assured. In many patients, the leukocyte alkaline phosphatase level is also increased, as is the uric acid level. Elevated serum vitamin B12 or B12-binding capacity may be present. In patients with associated acid-peptic disease, occult gastrointestinal bleeding may lead to presentation with hypochromic, microcytic anemia. A bone marrow aspirate and biopsy will provide no specific diagnostic information, and unless there is a need to establish the presence of myelofibrosis or exclude some other disorder, these procedures need not be done. Although the presence of a cytogenetic abnormality such as trisomy 8 or 9 or 20q- in the setting of an expansion of the red cell mass supports the clonal etiology, no specific cytogenetic abnormality is associated with polycythemia vera, and the absence of a cytogenetic marker does not exclude the diagnosis. COMPLICATIONS The major clinical complications of polycythemia vera relate directly to the increase in blood viscosity associated with elevation of the red cell mass and indirectly to the increased turnover of red cells, leukocytes, and platelets and the attendant increase in uric acid and histamine production. The latter appears to be responsible for the increase in peptic ulcer disease and for the pruritus associated with this disorder, although little formal proof for this has been obtained. A sudden massive increase in spleen size is another problem and can be associated with splenic infarction or progressive cachexia. Myelofibrosis and myeloid metaplasia can also develop with transfusion-dependent anemia, but the frequency is low in those not receiving chemotherapy or irradiation. Although acute nonlymphocytic leukemia is reported to be increased in polycythemia vera, the incidence of acute leukemia in patients not exposed to chemotherapy or radiation is low and the development of leukemia is not related to disease duration, suggesting that the treatment exposure may be a more important risk factor than the disease itself. Erythromelalgia is a curious syndrome of unknown etiology involving primarily the lower extremities and manifested usually by erythema, warmth, and pain of the affected appendage and occasionally digital infarction. It occurs with a variable frequency in patients with a myeloproliferative disorder and is usually responsive to salicylates. Some of the central nervous system symptoms observed in patients with polycythemia vera may represent a variant of erythromelalgia. If left uncontrolled, erythrocytosis can lead to intravascular thrombosis involving vital organs such as the liver, heart, brain, or lungs. Patients with massive splenomegaly are particularly prone to thrombotic events because the associated increase in plasma volume masks the true extent of the red cell mass elevation as measured by the hematocrit or hemoglobin level. A "normal" hematocrit or hemoglobin level in a polycythemia vera patient with massive splenomegaly should be considered as indicative of an elevated red cell mass until proven otherwise. TREATMENT
Polycythemia vera is generally an indolent disorder whose clinical course can run many decades, and its medical management should reflect the tempo of the disorder. Maintenance of the hemoglobin level at £140 g/L in men and £120 g/L in women is mandatory to avoid the thrombotic complications. Thrombosis due to erythrocytosis is the most significant complication of this disorder. Phlebotomy serves initially to reduce hyperviscosity by bringing the red cell mass into the normal range. Periodic phlebotomies thereafter serve to maintain the red cell mass within the range of normal and to induce a state of iron deficiency, which prevents an accelerated reexpansion of the red cell mass. In most polycythemia vera patients, once an iron-deficient state is achieved, phlebotomy is usually required only at 3-month intervals. Although both phlebotomy and iron deficiency, in addition to the disease itself, tend to increase the platelet count, thrombocytosis is not correlated with thrombosis in polycythemia vera, in contrast to the strong correlation between erythrocytosis and thrombosis in this disease. The use of salicylates as a tonic against thrombosis in polycythemia vera patients is potentially harmful, and salicylates should be employed only to treat erythromelalgia. Oral anticoagulants are not routinely indicated and are difficult to assess owing to the artifactual imbalance between the test tube anticoagulant and plasma that occurs when blood from these patients is assayed for prothrombin or partial thromboplastin activity. Asymptomatic hyperuricemia requires no therapy, but allopurinol should be administered to avoid further elevation of the uric acid when chemotherapy is employed to reduce splenomegaly or leukocytosis-associated pruritus. Generalized pruritus intractable to antihistamines can be a major problem in polycythemia vera, and hydroxyurea, interferon (IFN)-a, and psoralens with ultraviolet light in the A range (PUVA) therapy may have some palliative effects. Asymptomatic thrombocytosis requires no therapy. Symptomatic thrombocytosis or splenomegaly can be treated with hydroxyurea or IFN-a, although each can be associated with significant side effects. Anagrelide, a quinazolin derivative and platelet antiaggregant that also lowers the platelet count, can control thrombocytosis. A reduction in platelet number may be necessary in the treatment of erythromelalgia if salicylates are not effective or if the thrombocytosis is associated with migraine-like symptoms. However, the highest priority for treatment is reduction of the red cell mass to normal. Alkylating agents and32P are leukemogenic in polycythemia vera, and their use should be avoided. If a cytotoxic agent must be used, hydroxyurea is preferred, but it also may be leukemogenic with chronic use. Chemotherapy should be used for as short a time as possible. In some patients, massive splenomegaly unresponsive to reduction by hydroxyurea or IFN-a therapy and associated with intractable weight loss will require splenectomy. Allogeneic bone marrow transplantation may be effective in young patients. Patients with polycythemia vera can be expected to live long and useful lives when their red cell mass is effectively managed with phlebotomy. Chemotherapy is never indicated to control the red cell mass unless venous access is impossible. IDIOPATHIC MYELOFIBROSIS Idiopathic myelofibrosis (other designations include agnogenic myeloid metaplasia or myelofibrosis with myeloid metaplasia) is a clonal disorder of a multipotent hematopoietic progenitor cell of unknown etiology characterized by marrow fibrosis, myeloid metaplasia with extramedullary hematopoiesis, and splenomegaly. Idiopathic myelofibrosis is uncommon; in the absence of a specific clonal marker, establishing this
diagnosis is difficult because myelofibrosis and myeloid metaplasia with splenomegaly are also features of both polycythemia vera andCML. Furthermore, myelofibrosis and splenomegaly occur in a variety of benign and malignant disorders (Table 110-3), many of which are amenable to specific therapies not effective in idiopathic myelofibrosis. In contrast to the other chronic myeloproliferative disorders and so-called acute or malignant myelofibrosis, which can occur at any age, idiopathic myelofibrosis primarily afflicts individuals in their sixth decade or later. ETIOLOGY The etiology of idiopathic myelofibrosis is unknown. Although nonrandom chromosome abnormalities such as 20q-, 13q-, and trisomy 1q are not uncommon, no specific cytogenetic abnormality has been identified. The degree of myelofibrosis and the extent of extramedullary hematopoiesis are not related. This disorder is associated with overproduction of type III collagen, a finding that has been attributed to platelet-derived growth factor or transforming growth factorb, but no proof has been forthcoming. Importantly, fibroblasts in idiopathic myelofibrosis are not part of the neoplastic clone. CLINICAL FEATURES No specific signs or symptoms are associated with idiopathic myelofibrosis. Most patients are asymptomatic at presentation and are usually detected by the discovery of splenic enlargement and/or abnormal blood counts during a routine examination. A blood smear reveals the characteristic features of extramedullary hematopoiesis: teardrop-shaped red cells, nucleated red cells, myelocytes, and promyelocytes; myeloblasts may also be present but have no prognostic significance. Anemia, usually mild initially, is the rule, while the leukocyte and platelet counts are either normal or increased but either can be depressed. Mild hepatomegaly may accompany the splenomegaly, and both the lactate dehydrogenase and serum alkaline phosphatase levels can be elevated. The level of leukocyte alkaline phosphatase can be low, normal, or elevated. Marrow may be unaspirable due to the myelofibrosis, and bone x-rays may reveal osteosclerosis. Exuberant extramedullary hematopoiesis can cause ascites, pulmonary hypertension, intestinal or ureteral obstruction, intracranial hypertension, pericardial tamponade, spinal cord compression, or skin nodules. Splenic enlargement can be sufficiently rapid to cause splenic infarctions with fever and pleuritic chest pain. Hyperuricemia and secondary gout may ensue. DIAGNOSIS While the clinical picture described above is characteristic of idiopathic myelofibrosis, all of the clinical features described can be observed in polycythemia vera orCML. Massive splenomegaly commonly masks erythrocytosis in polycythemia vera, and reports of intraabdominal thromboses in idiopathic myelofibrosis likely represent instances of unrecognized polycythemia vera. Furthermore, many other disorders have features that overlap with idiopathic myelofibrosis but respond to distinctly different therapies. Therefore, the diagnosis of idiopathic myelofibrosis is one of exclusion, which requires that the disorders listed inTable 110-3 be ruled out. The presence of teardrop-shaped red cells, nucleated red cells, myelocytes, and
promyelocytes establishes the presence of extramedullary hematopoiesis; the presence of leukocytosis, thrombocytosis with large and bizarre platelets, as well as circulating myeloblasts suggests the presence of a myeloproliferative disorder as opposed to a secondary form of myelofibrosis (Table 110-3). Marrow is usually not aspirable due to increased marrow reticulin, but marrow biopsy will reveal a hypercellular marrow with trilineage hyperplasia and, in particular, increased megakaryocytes, but there are no characteristic morphologic abnormalities that distinguish idiopathic myelofibrosis from the other chronic myeloproliferative disorders. Splenomegaly due to extramedullary hematopoiesis may be sufficiently massive to cause portal hypertension and variceal formation. In some patients, exuberant extramedullary hematopoiesis can dominate the clinical picture. An intriguing feature of idiopathic myelofibrosis is the occurrence of autoimmune abnormalities such as immune complexes, antinuclear antibodies, rheumatoid factor, or a positive Coombs' test. Whether these represent a host reaction to the disorder or are involved in its pathogenesis is unknown. Cytogenetic analysis of blood or marrow is useful both to excludeCML and for prognostic purposes, because complex karyotype abnormalities portend a poor prognosis in idiopathic myelofibrosis. COMPLICATIONS Idiopathic myelofibrosis is a chronic disorder but with a median survival of only 5 years (range 1 to 15 years), a duration much shorter than for polycythemia vera or essential thrombocytosis. The natural history of idiopathic myelofibrosis is one of inexorable marrow failure with transfusion-dependent anemia and increasing organomegaly. Patients are prone to deep-seated tissue infections, particularly of the lungs. As withCML, idiopathic myelofibrosis can evolve from a chronic phase to an accelerated phase with constitutional symptoms and increasing marrow failure. About 10% of patients develop an aggressive form of acute leukemia for which therapy is usually ineffective. Important prognostic factors for disease acceleration include anemia; thrombocytopenia; age; the presence of complex cytogenetic abnormalities; and constitutional symptoms such as unexplained fever, night sweats, or weight loss. Any nonrandom cytogenetic abnormality is associated with a shortened life span, and the presence or development of multiple cytogenetic abnormalities is highly indicative of disease acceleration. TREATMENT There is no specific therapy for idiopathic myelofibrosis. Anemia may be exacerbated by deficiency of folic acid or iron, and in rare instances, pyridoxine therapy has been effective. However, anemia is more often due to ineffective erythropoiesis not compensated for by the extramedullary hematopoiesis in the spleen and liver; neither androgens nor erythropoietin has been consistently effective therapy. Erythropoietin may worsen splenomegaly. A red cell splenic sequestration study can establish the presence of hypersplenism, for which splenectomy is indicated. Splenectomy may also be necessary if splenomegaly impairs alimentation and should be performed before cachexia sets in. In this situation, splenectomy should not be avoided because of concern over rebound thrombocytosis, loss of hematopoietic capacity, or compensatory hepatomegaly. However, for unexplained reasons, splenectomy increases the risk of blastic transformation. Allopurinol can control significant hyperuricemia and hydroxyurea has proved useful for controlling organomegaly. The role of interferon-a is undefined,
and its side effects are more pronounced in the older individuals who are affected with this disorder, but reversal of myelofibrosis has been observed. Glucocorticoids are used to control autoimmune complications. Allogeneic bone marrow transplantation should be considered in younger patients. ESSENTIAL THROMBOCYTOSIS Essential thrombocytosis (other designations include essential thrombocythemia, idiopathic thrombocytosis, primary thrombocytosis, hemorrhagic thrombocythemia) is a clonal disorder of unknown etiology involving a multipotent hematopoietic progenitor cell and is manifested clinically by the overproduction of platelets without a definable cause. Essential thrombocytosis is an uncommon disorder, but its exact frequency is unknown. No clonal marker distinguishes it from the more common nonclonal, reactive forms of thrombocytosis (Table 110-4). Clinical recognition of thrombocytosis is unlikely in the largely asymptomatic persons affected by this disorder. As a consequence, essential thrombocytosis was formerly considered to be a disease of the elderly and to be responsible for significant morbidity due to hemorrhage or thrombosis. However, with the widespread application of platelet counting, it is now clear that essential thrombocytosis can occur at any age in adults and often occurs without symptoms or disturbances of hemostasis. There is an unexplained female predominance, in contrast to the reactive forms of thrombocytosis where no sex bias exists. Because no clonal marker is available for the disorder, clinical criteria have been proposed to distinguish it from the other chronic myeloproliferative disorders, which may also present with thrombocytosis but have distinct prognosis and treatment (Table 110-5). These criteria do not establish clonality; therefore, they are truly useful only in identifying disorders such asCML, polycythemia vera, or myelodysplasia, which can masquerade as essential thrombocytosis, as opposed to establishing the presence of essential thrombocytosis. Furthermore, as with "primary" erythrocytosis, nonclonal, benign forms of thrombocytosis exist (such as hereditary overproduction of thrombopoietin) that are not widely recognized because we currently lack the diagnostic tools to do so. ETIOLOGY Megakaryocytopoiesis and platelet production depend upon thrombopoietin and its receptor, Mpl. As in the case of early erythroid and myeloid progenitor cells, early megakaryocytic progenitors require the presence of interleukin (IL) 3 and stem cell factor for optimal proliferation, and their subsequent development is enhanced by IL-6 and -11. However, megakaryocyte maturation and differentiation require thrombopoietin. Megakaryocytes are unique amongst hematopoietic progenitor cells because they undergo endomitotic as opposed to mitotic reduplication of their genome. In the absence of thrombopoietin, endomitotic megakaryocytic reduplication and, by extension, the cytoplasmic development necessary for platelet production are impaired. Like erythropoietin, thrombopoietin is produced in both the liver and the kidneys, and an inverse correlation between the platelet count and plasma thrombopoietic activity exists. Like erythropoietin, plasma levels of thrombopoietin are controlled in part by the size of its progenitor cell pool. In contrast to erythropoietin, but like its myeloid counterparts granulocyte and granulocyte-macrophage colony stimulating factors, thrombopoietin not only enhances the proliferation of its target cells but also enhances the reactivity of their
end-stage product, the platelet. In addition to its role in thrombopoiesis, thrombopoietin enhances the survival of multipotent hematopoietic stem cells. The clonality of essential thrombocytosis has been established by the use of the isoenzymes of glucose-6-phosphate dehydrogenase in patients who are hemizygous for this gene, by the use of X-linked DNA polymorphisms, and by the identification of nonrandom, although variable cytogenetic abnormalities. The multipotent hematopoietic progenitor cell involved in this disorder can vary; in some patients lymphocytes contained the same clonal marker as the megakaryocytes, erythrocytes, and myeloid cells, whereas in others the lymphocytes were not involved. Similar observations have been made in polycythemia vera. Furthermore, a number of families have been described in which essential thrombocytosis was inherited, in one instance as an autosomal dominant trait. In one kindred, in addition to essential thrombocytosis, idiopathic myelofibrosis and polycythemia vera were also individually documented. CLINICAL FEATURES Clinically, essential thrombocytosis is most often identified incidentally when a platelet count is obtained during the course of a routine evaluation. Occasionally, review of previous platelet counts will reveal that an elevation was present but overlooked. No symptoms or signs are specific for essential thrombocytosis, but patients do have hemorrhagic and thrombotic tendencies expressed as easy bruising for the former or microvascular occlusions for the latter, which may be manifested by erythromelalgia, migraine, or transient ischemic attacks. Physical examination is generally unremarkable except for the presence of mild splenomegaly. Massive splenomegaly is more characteristic of the other myeloproliferative disorders, particularly polycythemia vera or idiopathic myelofibrosis. Anemia is unusual, but a mild neutrophilic leukocytosis is not. The blood smear, however, is most remarkable for the number of platelets present, some of which may be very large. The leukocyte alkaline phosphatase score is either normal or elevated. The large mass of circulating platelets may prevent the accurate measurement of serum potassium due to the release of platelet potassium upon blood clotting. This hyperkalemia is a laboratory artifact and is not associated with any electrocardiographic abnormalities. Similarly, arterial oxygen measurements can be inaccurate unless the blood is collected on ice. The prothrombin and partial thromboplastin times are normal, while abnormalities of platelet function such as a prolonged bleeding time and impaired platelet aggregation can be present. However, in spite of much study, characteristic platelet function abnormalities associated are not defined, and no platelet function test predicts the presence of clinically significant bleeding or thrombosis. The elevated platelet count may hinder the collection of a marrow aspirate, but marrow biopsy usually reveals both megakaryocyte hyperplasia and hypertrophy, as well as an overall increase in marrow cellularity. An increase in marrow reticulin may be present, but if extensive, another diagnosis should be considered. The absence of stainable iron demands an explanation, because iron deficiency alone can cause thrombocytosis and absent marrow iron is a feature of polycythemia vera. While nonrandom cytogenetic abnormalities have been identified in essential
thrombocytosis, no consistently identifiable abnormality is noted, even involving chromosomes 3 and 1 where the genes for thrombopoietin and its receptor Mpl, respectively, are located. DIAGNOSIS Thrombocytosis is encountered in a variety of clinical disorders (Table 110-4) in which production of cytokines is increased. Thus, the first obligation when confronted with a high platelet count is to determine if it is a consequence of another disorder. Cytogenetic evaluation is mandatory to determine if the thrombocytosis is due toCML or a myelodysplastic disorder such as the 5q-syndrome. Because the bcr-abl translocation can be present in the absence of the Ph chromosome, polymerase chain reaction analysis for bcr-abl expression should be performed in all patients with thrombocytosis in whom a cytogenetic study is normal. Anemia and ringed sideroblasts are not features of essential thrombocytosis, but they are features of idiopathic refractory sideroblastic anemia, in which thrombocytosis can also occur. The presence of massive splenomegaly should suggest the possibility of another myeloproliferative disorder, and in this setting a red cell mass determination is mandatory because substantial splenomegaly can mask the presence of erythrocytosis. What appears to be essential thrombocytosis can evolve into polycythemia vera, revealing the true nature of the underlying myeloproliferative disorder. COMPLICATIONS Perhaps no other condition in clinical medicine has caused otherwise astute physicians to intervene inappropriately more often than thrombocytosis, particularly if the platelet count is greater than 1 ´106/uL. It is commonly believed that a high platelet count must cause intravascular stasis and thrombosis; however, no controlled clinical study has ever established either association. To the contrary, very high platelet counts are associated primarily with hemorrhage, while platelet counts of 1 month before the diagnosis of AML when compared to those without such a history. Responsiveness to chemotherapy declines as the duration of the antecedent disorder(s)
increases. SecondaryAMLdeveloping after treatment with cytotoxic agents and/or irradiation for other malignancies is extremely difficult to treat successfully. A high presenting leukocyte count is an independent prognostic factor; duration ofCR is inversely related to the presenting leukocyte count or absolute circulating myeloblast count. Among patients with hyperleukocytosis (>100,000/uL), early central nervous system bleeding and pulmonary leukostasis and late relapse contribute to poor outcome. TheFABclassification diagnosis has been found to be an independent prognostic factor in some series. Other characteristics of the leukemic cell have been reported to have prognostic significance, including Auer rods, ultrastructural features, in vitro and in vivo growth characteristics and chemotherapeutic sensitivity, and immunophenotype. Expression of the MDR1 gene adversely influences outcome. This gene encodes a protein that actively pumps out a variety of lipophilic compounds (e.g., anthracyclines) from the cell. In addition to pretreatment variables, several treatment factors have been reported to correlate with prognosis inAML, in particular withCRduration. One is the rapidity with which the blast cells disappear from the blood after the institution of therapy. In addition, patients who achieve CR after one induction cycle have longer CR than those requiring multiple cycles. TREATMENT Treatment of the newly diagnosed patient withAML is usually divided into two phases, induction and postremission management (Fig. 111-1). The initial goal is to quickly induceCR. Once CR is obtained, further therapy must be used to prolong survival. Induction Chemotherapy The most commonly usedCRinduction regimens (for patients with allFABsubtypes except M3) consist of combination chemotherapy with cytarabine (cytosine arabinoside) and an anthracycline. Cytarabine is a cell cycle S-phase-specific antimetabolite that becomes phosphorylated to an active triphosphate form that interferes with DNA synthesis. Anthracyclines are DNA intercalaters. Their primary mode of action is thought to be inhibition of topoisomerase II, leading to DNA breaks. Cytarabine is usually administered as a continuous intravenous infusion at 100 to 200 mg/m2per day for 7 days. Anthracycline therapy generally consists of daunorubicin, 45 mg/m2intravenously on days 1, 2, and 3 (the 7 and 3 regimen). Treatment with idarubicin at 12 or 13 mg/m2per day for 3 days in conjunction with cytarabine by 7-day continuous infusion is at least as effective and may be superior to daunorubicin in younger patients. The addition of etoposide or other agents does not increase the CR rate but may improve the CR duration. After induction chemotherapy, the bone marrow is examined to determine if the leukemia has been eliminated. If >5% blasts exist with³20% cellularity, the patient has traditionally been retreated with cytarabine and an anthracycline in doses similar to those given initially, but for 5 and 2 days, respectively. Our recommendation, however, is to consider changing therapy in this setting. Patients who fail to attainCR after two induction courses should immediately proceed to an allogeneic stem cell transplant
(SCT) if an appropriate donor exists. With the 7 and 3 cytarabine/daunorubicin regimen outlined above, 65 to 75% of adults with de novoAMLachieveCR. Two-thirds achieve CR after a single course of therapy, and one-third require two courses. About 50% of patients who do not achieve CR have a drug-resistant leukemia, and 50% do not achieve CR because of fatal complications of bone marrow aplasia or impaired recovery of normal stem cells. High-dose cytarabine-based regimens have very highCR rates after a single cycle of therapy. When given in high doses, more cytarabine may enter the cells, saturate the cytarabine-inactivating enzymes, and increase the intracellular levels of 1-b-D-arabinofuranylcytosine-triphosphate, the active metabolite incorporated into DNA. Thus, higher doses of cytarabine may increase the inhibition of DNA synthesis and thereby overcome resistance to standard-dose cytarabine. In two randomized studies, one by the Southwest Oncology Group (SWOG) and one by the Australian Leukemia Study Group (ALSG), high-dose cytarabine with an anthracycline produced CR rates similar to those achieved with standard 7 and 3 regimens. However, the ALSG demonstrated that the CR duration was much longer after high-dose cytarabine than after standard-dose cytarabine. The hematologic toxicity of high-dose cytarabine-based induction regimens has typically been greater than that associated with 7 and 3 regimens. Toxicity with high-dose cytarabine includes myelosuppression, pulmonary toxicity, and significant and occasionally irreversible cerebellar toxicity. All patients treated with high-dose cytarabine must be closely monitored for cerebellar toxicity. Full cerebellar testing should be performed before each dose, and further high-dose cytarabine should be withheld if evidence of cerebellar toxicity develops. Supportive Care Measures geared to supporting patients through several weeks of granulocytopenia and thrombocytopenia are critical to the success ofAMLtherapy. Patients with AML should be treated in centers expert in providing supportive measures for their management. Recombinant hematopoietic growth factors have been incorporated into clinical trials inAML. These trials have been designed to lower the infection rate after chemotherapy or to sensitize (prime) the leukemic blasts to chemotherapy, or both. Both granulocyte colony stimulating factor (G-CSF) and granulocyte-macrophage colony stimulating factor (GM-CSF) have reduced the median time to neutrophil recovery by an average of 5 to 7 days. This accelerated rate of neutrophil recovery, however, has not always translated into significant reductions in infection rates. In most randomized studies, both G-CSF and GM-CSF have failed to improve the CR rate, disease-free survival, or overall survival. Although receptors for both G-CSF and GM-CSF are present on AML blasts, therapeutic efficacy is neither enhanced nor inhibited by these agents. The use of growth factors as supportive care for AML patients is controversial. We favor their use in elderly patients, those receiving intensive regimens, patients with uncontrolled infections, or those participating in clinical trials. Multilumen right atrial catheters should be inserted through a subcutaneous tunnel as soon as patients with newly diagnosedAML have been stabilized. They should be used
thereafter for administration of intravenous medications and transfusions, as well as for blood drawing. The separation between the vascular access site and the exit site and the presence of a Dacron cuff in the subcutaneous channel reduce the risk of infection. With meticulous attention to sterile technique in catheter placement and maintenance, catheters may often be left in place for months. Adequate and prompt blood bank support is critical to therapy ofAML. Platelet transfusions should be given as needed to maintain a platelet count>10,000 to 20,000/uL. We believe that the platelet count should be kept at higher levels in febrile patients and during episodes of active bleeding orDIC. Patients with poor posttransfusion platelet count increments may benefit from administration of platelets from human leukocyte antigen (HLA)-matched donors. Red blood cell transfusions should be administered to keep the hemoglobin level>80 g/L (8 g/dL) in the absence of active bleeding or DIC. Blood products leukodepleted by filtration should be used to avert or delay alloimmunization as well as febrile reactions. Blood products should also be irradiated to prevent graft-versus-host disease (GVHD). Cytomegalovirus (CMV)-negative blood products should be used for CMV-seronegative patients who are potential candidates for allogeneicSCT. Leukodepleted products are also effective for these patients if CMV-negative products are not available. Infectious complications remain the major cause of morbidity and death during induction and postremission chemotherapy forAML. Prophylactic administration of antibiotics in the absence of fever is controversial. Oral nystatin or clotrimazole are recommended to prevent localized candidiasis. For patients who are herpes simplex virus antibody titer-positive, acyclovir prophylaxis is effective in preventing reactivation of latent oral herpes infections. Fever develops in most patients withAML, but infections are documented in only half of febrile patients. Early initiation of empiric broad-spectrum antibacterial and antifungal antibiotics has significantly reduced the number of patients dying of infectious complications (Chap. 85). An antibiotic regimen adequate to treat gram-negative and gram-positive organisms should be instituted at the onset of fever in a granulocytopenic patient after clinical evaluation, including a detailed physical examination with inspection of the indwelling catheter exit site and a perirectal examination, as well as procurement of cultures and radiographs aimed at documenting the source of fever. Specific antibiotic regimens should be based on antibiotic sensitivity data obtained from the institution at which the patient is being treated. Acceptable regimens include imipenem-cilastin, an antipseudomonal semisynthetic penicillin (e.g., piperacillin) combined with an aminoglycoside, a third-generation cephalosporin with antipseudomonal activity (i.e., ceftazidine or cefapime) or doubleb-lactam combinations (ceftazidine and piperacillin). Empiric vancomycin is not given initially in the absence of suspected gram-positive infection or mucositis. Aminoglycosides should be avoided if possible in patients with renal insufficiency. For patients with known immediate-type hypersensitivity reactions to penicillin, aztreonam may be substituted for b-lactams. Aztreonam should be combined with an aminoglycoside or a quinolone antibiotic rather than used alone. Empiric vancomycin should be initiated in neutropenic patients who remain febrile for 3 days, and amphotericin B is added at 7 days if fever persists. Liposomal amphotericin is at least equivalent to regular amphotericin for empiric antifungal treatment and has less renal toxicity. Antibacterial and antifungal antibiotics
should be continued until patients are no longer neutropenic, regardless of whether a specific source has been found for the fever. Treatment of Promyelocytic Leukemia ATRA is an oral drug that induces the differentiation of leukemic cells bearing the t(15;17); it is not effective in other forms ofAML. Acute promyelocytic leukemia is responsive to cytarabine and daunarubicin, but about 10% of patients treated with these drugs die fromDICinduced by the release of granule components by dying tumor cells. ATRA does not produce DIC but produces another complication called the retinoic acid syndrome. Occurring within the first 3 weeks of treatment, it is characterized by fever, dyspnea, chest pain, pulmonary infiltrates, pleural and pericardial effusions, and hypoxia. The syndrome is related to the adhesion of differentiated neoplastic cells in the pulmonary vasculature. Glucocorticoids, chemotherapy, and/or supportive measures can be effective. About 10% of patients die from this syndrome. ATRA (45 mg/m2per day orally until remission is documented) plus concurrent chemotherapy (7 and 3) appears to be the safest and most effective treatment for acute promyelocytic leukemia. Unlike patients with other types ofAML, patients with this subtype may benefit from maintenance therapy with either ATRA or chemotherapy. The optimal regimen is being sought in clinical studies. Arsenic trioxide produces meaningful responses in patients refractory to ATRA. The detection of minimal residual disease byRT-PCRamplification of the t(15;17) chimeric gene product appears to predict relapse. Disappearance of the signal is associated with long-term disease-free survival; its persistence predicts relapse. With increases in the sensitivity of the assay, some patients with persistent abnormal gene product have been found who do not suffer a relapse. It is not known whether there is a critical threshold level of transcripts that predicts for leukemia relapse. Postremission Therapy Induction of a durable firstCR is critical to long-term disease-free survival inAML. Without further therapy virtually all patients experience relapse. Once relapse has occurred, AML is generally curable only bySCT. Postremission therapy is designed to eradicate any residual leukemic cells. Therefore, it should prevent relapse and prolong survival. Approaches to postremission therapy in AMLinclude intensive chemotherapy and allogeneic or autologousSCT. In patients older than 65 years, such therapy is of uncertain benefit. High-dose cytarabine is more effective than standard-dose cytarabine. The Cancer and Leukemia Group B, for example, compared the duration ofCR in patients randomly assigned postremission to four cycles of high (3 g/m2every 12 h on days 1, 3, and 5), intermediate (400 mg/m2for 5 days by continuous infusion), or standard (100 mg/m 2per day for 5 days by continuous infusion) doses of cytarabine. A dose-response effect for cytarabine in patients with AML who were age 60 years or younger was demonstrated. High-dose cytarabine significantly prolonged CR and increased the fraction cured in patients with favorable [t(8;21) and inv(16)] and normal cytogenetics, but it had no significant effect on patients with other abnormal karyotypes. Allogeneic and autologousSCT in first CRhas been studied extensively in younger
patients with no major organ dysfunction. Allogeneic SCT is used in patients 80%. Classical Hodgkin's Disease Hodgkin's disease occurs in ~8000 patients in the United States each year, and the disease does not appear to be increasing in frequency. Most patients present with palpable lymphadenopathy that is nontender; in most patients, these lymph nodes are in the neck, supraclavicular area, and axilla. More than half the patients will have mediastinal adenopathy at diagnosis, and this is sometimes the initial manifestation. Subdiaphragmatic presentation of Hodgkin's disease is unusual and more common in older males. Approximately one-third of patients present with fevers, night sweats, and/or weight loss -- B symptoms in the Ann Arbor staging classification (Table 112-8). Occasionally, Hodgkin's disease can present as a fever of unknown origin. This is more common in older patients who are found to have mixed-cellularity Hodgkin's disease in an abdominal site. Rarely, the fevers persist for days to weeks, followed by afebrile intervals and then recurrence of the fever. This pattern is known as Pel-Epstein fever. Hodgkin's disease can occasionally present with unusual manifestations. These include severe and unexplained itching, cutaneous disorders such as erythema nodosum and ichthyosiform atrophy, paraneoplastic cerebellar degeneration and other distant effects on theCNS, nephrotic syndrome, immune hemolytic anemia and thrombocytopenia, hypercalcemia, and pain in lymph nodes on alcohol ingestion. The diagnosis of Hodgkin's disease is established by review of an adequate biopsy specimen by an expert hematopathologist. In the United States, most patients would be classified as having nodular sclerosing Hodgkin's disease, with a minority of patients having mixed-cellularity Hodgkin's disease. Lymphocyte-predominant and lymphocyte-depleted Hodgkin's disease are rare. Mixed-cellularity Hodgkin's disease or lymphocyte-depletion Hodgkin's disease are seen more frequently in patients infected by HIV (see Plate V-18). The differential diagnosis of a lymph node biopsy suspicious for Hodgkin's disease includes inflammatory processes, mononucleosis, non-Hodgkin's lymphoma, diphenylhydantoin-induced lymphadenopathy, and nonlymphomatous malignancies. The staging evaluation for a patient with Hodgkin's disease would typically include a careful history and physical examination; complete blood count; erythrocyte sedimentation rate; serum chemistry studies includingLDH; chest radiograph;CT scan of the chest, abdomen, and pelvis; and bone marrow biopsy. Many patients would also have a gallium scan. If radiologic expertise is available, a bipedal lymphangiogram can be helpful. Gallium scans are most useful at the completion of therapy to document remission. Staging laparotomies were once popular for most patients with Hodgkin's disease but are now done rarely because of an increased reliance on systemic rather than local therapy. TREATMENT
Patients with localized Hodgkin's disease are cured >90% of the time. In patients with good prognostic factors, extended field radiotherapy has a high cure rate. Increasingly, patients with all stages of Hodgkin's disease are treated initially with chemotherapy. Patients with localized or good-prognosis disease receive a brief course of chemotherapy followed by radiotherapy to sites of node involvement. Patients with more extensive disease or those with B symptoms receive a complete course of chemotherapy. The most popular chemotherapy regimens used in the treatment of Hodgkin's disease include doxorubicin, bleomycin, vinblastine, and dacarbazine (ABVD) and mechlorethamine, vincristine, procarbazine, and prednisone (MOPP), or combinations of the drugs in these two regimens. Today, most patients in the United States receive ABVD. Long-term disease-free survival in patients with advanced disease can be achieved in >75% of patients who lack systemic symptoms and in 50 to 70% of patients with systemic symptoms. Patients who relapse after primary therapy of Hodgkin's disease can frequently still be cured. Patients who relapse after initial treatment only with radiotherapy have an excellent outcome when treated with chemotherapy. Patients who relapse after an effective chemotherapy regimen are usually not curable with subsequent chemotherapy administered at standard doses. However, patients with a long initial remission can be an exception to this rule. Autologous bone marrow transplantation can cure half of patients who fail effective chemotherapy regimens. Because of the very high cure rate in patients with Hodgkin's disease, long-term complications have become a major focus for clinical research. In fact, in some series of patients with early-stage disease, more patients died from late complications of therapy than from Hodgkin's disease itself. This is particularly true in patients with localized disease. The most serious late side effects include second malignancies and cardiac injury. Patients are at risk for the development of acute leukemia in the first 10 years after treatment with combination chemotherapy regimens that contain alkylating agents. The risk for development of acute leukemia appears to be greater afterMOPP-like regimens than withABVD. The development of carcinomas as a complication of treatment for Hodgkin's disease has become a major problem. These tumors usually occur³10 years after treatment and are associated more with radiotherapy than with chemotherapy. For this reason, young women treated with thoracic radiotherapy for Hodgkin's disease should institute screening mammograms 5 to 10 years after treatment, and all patients who receive thoracic radiotherapy for Hodgkin's disease should be discouraged from smoking. Thoracic radiation also accelerates coronary artery disease, and patients should be encouraged to minimize risk factors for coronary artery disease such as smoking and elevated cholesterol levels. A number of other late side effects from the treatment of Hodgkin's disease are well known. Patients who receive thoracic radiotherapy are at very high risk for the eventual development of hypothyroidism and should be observed for this complication; intermittent measurement of thyrotropin should be made to identify the condition before it becomes symptomatic. Lhermitte's syndrome occurs in ~15% of patients who receive thoracic radiotherapy. This syndrome is manifested by an "electric shock" sensation into the lower extremities on flexion of the neck. Infertility is a concern for all patients undergoing treatment for Hodgkin's disease. In both women and men, the risk of permanent infertility is age-related, with younger patients more likely to recover fertility.
In addition, treatment withABVDrather thanMOPPincreases the chances to retain fertility. LYMPHOMA-LIKE DISORDERS The most common condition that pathologists and clinicians might confuse with lymphoma is reactive, atypical lymphoid hyperplasia. Patients might have localized or disseminated lymphadenopathy and might have the systemic symptoms characteristic of lymphoma. Underlying causes include a drug reaction to diphenylhydantoin or carbamezepine. Immune disorders such as rheumatoid arthritis and lupus erythematosus, viral infections such as cytomegalovirus andEBV, and bacterial infections such as cat-scratch disease may cause adenopathy (Chap. 63). In the absence of a definitive diagnosis after initial biopsy, continued follow-up, further testing, and repeated biopsies, if necessary, are the appropriate approach rather than instituting therapy. Specific conditions that can be confused with lymphoma include Castleman's disease, which can present with localized or disseminated lymphadenopathy; some patients have systemic symptoms. The disseminated form is often accompanied by anemia and polyclonal hypergammaglobulinemia, and the condition seems to be related to an overproduction of interleukin 6, possibly produced by human herpesvirus 8. Patients with localized disease can be treated effectively with local therapy, while the initial treatment for patients with disseminated disease is usually with systemic glucocorticoids. Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman's disease) usually presents with bulky lymphadenopathy in children or young adults. The disease is usually nonprogressive and self-limited, but patients can manifest autoimmune hemolytic anemia. Lymphomatoid papulosis is a cutaneous lymphoproliferative disorder that is often confused with anaplastic large-cell lymphoma involving the skin. The cells of lymphomatoid papulosis are similar to those seen in lymphoma and stain for CD30, and T cell receptor gene rearrangements are sometimes seen. However, the condition is characterized by waxing and waning skin lesions that usually heal, leaving small scars. In the absence of effective communication between the clinician and the pathologist regarding the clinical course in the patient, this disease will be misdiagnosed. Since the clinical picture is usually benign, misdiagnosis is a serious mistake. ACKNOWLEDGEMENT Dr. Arnold Freedman and Dr. Lee Nadler contributed this chapter to the 14th edition, and some elements of that chapter were retained here. (Bibliography omitted in Palm version) Back to Table of Contents
113. PLASMA CELL DISORDERS - Dan L. Longo GENERAL PRINCIPLES The plasma cell disorders are monoclonal neoplasms related to each other by virtue of their development from common progenitors in the B lymphocyte lineage. Multiple myeloma, Waldenstrom's macroglobulinemia, primary amyloidosis (Chap. 319), and the heavy chain diseases comprise this group and may be designated by a variety of synonyms such as monoclonal gammopathies, paraproteinemias, plasma cell dyscrasias, and dysproteinemias. Mature B lymphocytes destined to produce IgG bear surface immunoglobulin molecules of both M and G heavy chain isotypes with both isotypes having identical idiotypes (variable regions). Under normal circumstances, maturation to antibody-secreting plasma cells is stimulated by exposure to the antigen for which the surface immunoglobulin is specific; however, in the plasma cell disorders the control over this process is lost. The clinical manifestations of all the plasma cell disorders relate to the expansion of the neoplastic cells, to the secretion of cell products (immunoglobulin molecules or subunits, lymphokines), and to some extent to the host's response to the tumor.*Normal development of B lymphocytes is discussed in Chap. 305. There are three categories of structural variation among immunoglobulin molecules that form antigenic determinants, and these are used to classify immunoglobulins (Chap. 305). Isotypes are those determinants that distinguish among the main classes of antibodies of a given species and are the same in all normal individuals of that species. Therefore, isotypic determinants are, by definition, recognized by antibodies from a distinct species (heterologous sera) but not by antibodies from the same species (homologous sera). There are five heavy chain isotypes (M, G, A, D, E) and two light chain isotypes (k, l). Allotypes are distinct determinants that reflect regular small differences between individuals of the same species in the amino acid sequences of otherwise similar immunoglobulins. These differences are determined by allelic genes; by definition, they are detected by antibodies made in the same species. Idiotypes are the third category of antigenic determinants. They are unique to the molecules produced by a given clone of antibody-producing cells. Idiotypes are formed by the unique structure of the antigen-binding portion of the molecule. Antibody molecules (Fig. 305-8) are composed of two heavy chains (mol wt ~50,000) and two light chains (mol wt ~25,000). Each chain has a constant portion (limited amino acid sequence variability) and a variable region (extensive sequence variability). The light and heavy chains are linked by disulfide bonds and are aligned so that their variable regions are adjacent to one another. This variable region forms the antigen recognition site of the antibody molecule; its unique structural features form a particular set of determinants, or idiotypes, that are reliable markers for a particular clone of cells because each antibody is formed and secreted by a single clone. Each chain is specified by distinct genes, synthesized separately, and assembled into an intact antibody molecule after translation (Fig. 113-1). Because of the mechanics of the gene rearrangements necessary to specify the immunoglobulin variable regions (VDJ joining for the heavy chain, VJ joining for the light chain), a particular clone rearranges only one of the two chromosomes to produce an immunoglobulin molecule of only one light chain isotype and only one allotype (allelic exclusion). After exposure to antigen, the variable
region may become associated with a new heavy chain isotype (class switch). Each clone of cells performs these sequential gene arrangements in a unique way. This results in each clone producing a unique immunoglobulin molecule. In most cells, light chains are synthesized in slight excess, are secreted as free light chains by plasma cells, and are cleared by the kidney, but 10%), lytic bone lesions, and a serum and/or urine M component. The diagnosis may be made in the absence of bone lesions if the plasmacytosis is associated with a progressive increase in the M component over time or if extramedullary mass lesions develop. There are two important variants of myeloma, solitary bone plasmacytoma and extramedullary plasmacytoma. These lesions are associated with an M component in fewer than 30% of the cases, they may affect younger individuals, and both are associated with median survivals of 10 or more years. Solitary bone plasmacytoma is a single lytic bone lesion without marrow plasmacytosis. Extramedullary plasmacytomas usually involve the submucosal lymphoid tissue of the nasopharynx or paranasal sinuses without marrow plasmacytosis. Both tumors are highly responsive to local radiation therapy. If an M component is present, it should disappear after treatment. Solitary bone plasmacytomas may recur in other bony sites or evolve into myeloma. Extramedullary plasmacytomas rarely recur or progress. The most difficult differential diagnosis in patients with myeloma involves their separation from individuals with benign monoclonal gammopathies or monoclonal gammopathies of uncertain significance (MGUS). MGUS are vastly more common than
myeloma, occurring in 1% of the population over age 50 and in up to 10% over age 75. Patients with MGUS usually have 50% of patients. INFECTION The general problem of infection in the immunocompromised host is discussed inChap. 136. Posttransplant patients, particularly recipients of allogeneic transplantation, require unique approaches. Early after transplantation, patients are profoundly neutropenic, and because the risk of bacterial infection is so great, most centers initiate antibiotic treatment once the granulocyte count falls to 20 million. A single large vWF precursor subunit is synthesized in endothelial cells and megakaryocytes, where it is cleaved and assembled into the disulfide-linked multimers present in plasma, platelets, and vascular subendothelium. A modest reduction in plasma vWF concentration or a selective loss in the high-molecular-weight multimers decreases platelet adhesion and causes clinical bleeding. AlthoughvWD is heterogeneous, certain clinical features are common to all the syndromes. With one exception (type III disease), all forms are inherited as autosomal dominant traits, and affected patients are heterozygous with one normal and one abnormalvWFallele. In mild cases, bleeding occurs only after surgery or trauma. More severely affected patients have spontaneous epistaxis or oral mucosal, gastrointestinal, or genitourinary bleeding. The laboratory findings are variable. The most diagnostic pattern is the combination of (1) a prolonged bleeding time, (2) a reduction in plasma vWF concentration, (3) a parallel reduction in biologic activity as measured with the ristocetin cofactor assay, and (4) reduced factor VIII activity. The variability in laboratory tests is related to both the heterogeneous nature of the defects in vWD and the fact that plasma levels are influenced by ABO blood group type, central nervous system disorders, systemic inflammation, and pregnancy. Since vWD is an autosomal dominant disorder, some vWF is produced by the remaining normal allele. Thus patients with mild defects may have laboratory values that fluctuate over time and may occasionally be within the normal range. There are three major types ofvWD. Their mode of inheritance and laboratory findings are shown inFig. 116-2. Patients with type I disease, the most common abnormality, have a mild to moderate decrease in plasmavWF. In the milder cases, although hemostasis is impaired, the vWF level is just below normal (50% activity, or 5 mg/L). In type I disease, vWF antigen, factor VIII activity, and ristocetin cofactor activity are decreased with a normal spectrum of multimers detected by sodium dodecyl sulfate (SDS)-agarose gel electrophoresis. The variant forms ofvWD(type II disease) are much less common and characterized by normal or near-normal levels of a dysfunctional protein. Patients with the type IIa variant of vWD have a deficiency in the high- and medium-molecular-weight forms ofvWFmultimer detected bySDS-agarose electrophoresis. This is due either to an inability to secrete the high-molecular-weight vWF multimers or to proteolysis of the multimers soon after they leave the endothelial cell and enter the circulation. Mutations in a localized region of the vWF A-2 domain have been identified in families with type IIa vWD (Fig. 116-3). The quantity of vWF antigen and the amount of associated factor VIII are usually normal. In the type IIb variant, high-molecular-weight multimers are also decreased; however, the decrease is due to the inappropriate binding of vWF to platelets. Intravascular platelet aggregates form that are rapidly cleared from the circulation, causing mild, variable thrombocytopenia. Mutations in a disulfide-bonded loop in the A-1 domain that binds to glycoprotein Ib-IX are the cause of the type IIb defect (Fig. 116-3). A few patients have a platelet membrane disorder that mimics type IIb vWD -- platelet-type vWD. It is due to mutations in the portion of glycoprotein Ib-IX
that interacts with vWF. Levels of total vWF antigen and factor VIII are normal. Approximately 1 in 1 million individuals has a very severe form ofvWD that is phenotypically recessive (type III disease). Type III patients are usually the offspring of two parents (usually asymptomatic) with mild type I disease. Type III patients may inherit a different abnormality from each parent (a doubly heterozygous or compound heterozygous state) or be homozygous for a single defect. Type III patients have severe mucosal bleeding and no detectablevWFantigen or activity and, like patients with mild hemophilia, may have sufficiently low factor VIII that they have occasional hemarthroses. Major deletions in the vWF gene have been found in some type III families. Families with nonsense mutations and the combination of a deleted and nonsense mutant allele have also been described. Type IIN disease is due to a defect in the factor VIII binding site ofvWF. Patients resemble those with mild hemophilia and have low levels of factor VIII. The presence of disease in both males and females in a family is a clue to the role of vWF in this disease. TREATMENT There are two therapeutic options. Factor VIII concentrates retain high-molecular-weightvWFmultimers (Humate-P, Alfanate), are highly purified and heat-treated to destroy HIV, and are appropriate treatments for all the inherited forms ofvWD. During surgery or after major trauma, patients should receive factor VIII concentrates twice daily for 2 or 3 days to assure optimal hemostasis. Minor bleeding episodes such as prolonged epistaxis or severe menorrhagia may respond to a single infusion. Recurrent menorrhagia, a major problem for women with severe vWD, can be treated effectively with oral contraceptive agents that suppress menses. A second therapeutic option, which avoids the use of plasma, is the use ofDDAVP or desmopressin, a vasopressin analogue that has minimal blood pressure-elevating and fluid-retaining properties and raises the plasmavWFlevel in both normal individuals and patients with mildvWD. Patients with type I disease are the best candidates for DDAVP therapy. However, they must be tested for an adequate response before anticipated surgery, and vWF levels must be monitored closely during therapy, since the patient may develop tachyphylaxis when therapy is continued for more than 48 h. DDAVP should not be given to patients with variant forms of vWD without prior testing, since it may not improve multimer pattern or hemostasis in type IIa patients and may actually worsen the defect by depleting high-molecular-weight multimers, inducing intravascular platelet aggregation, and lowering the platelet count in type IIb patients. It is ineffective therapy for the severe (type III) form of vWD. Acquired vWD Although most cases of vWDare inherited, acquired vWD may be caused by antibodies that inhibitvWFfunction or by lymphoid or other tumors that selectively adsorb vWF multimers onto their surfaces. Anti-vWF antibodies have developed in patients with severe vWD following multiple transfusions, as well as in patients with autoimmune and lymphoproliferative disorders. Adsorption of vWF to tumor surfaces has been documented in patients with Waldenstrom's macroglobulinemia and Wilms' tumor and inferred in other patients with lymphoma. Treatment of acquired vWD should
focus on the underlying disease, since plasma derivatives andDDAVPare often not effective and the disorder can be fatal. Platelet Membrane Defects Receptors that modulate platelet adhesion and aggregation are located on the two major platelet surface glycoproteins.vWFfacilitates platelet adhesion by binding to glycoprotein Ib-IX, while fibrinogen links platelets into aggregates via sites on the glycoprotein IIb-IIIa complex. Two rare platelet defects are characterized by a loss of or a defect in these glycoprotein receptors. Patients with the Bernard-Soulier syndrome have markedly reduced platelet adhesion and cannot bind vWF to their platelets due to deficiency or dysfunction of the glycoprotein Ib-IX complex. They also have reduced levels of several other membrane proteins, mild thrombocytopenia, and extremely large, lymphocytoid platelets. Platelets from patients with Glanzmann's disease or thrombasthenia are deficient or defective in the glycoprotein IIb-IIIa complex. Their platelets do not bind fibrinogen and cannot form aggregates, although the platelets undergo shape change and secretion and are of normal size. Both these disorders are autosomal recessive traits and are characterized by markedly impaired hemostasis and recurrent episodes of severe mucosal hemorrhage. Bernard-Soulier platelets react normally to all stimuli except ristocetin. In contrast, thrombasthenic platelets adhere normally and will agglutinate with ristocetin but will not aggregate with any of the agonists that require fibrinogen binding, such as adenosine diphosphate (ADP), thrombin, or epinephrine. The only effective therapy for hemorrhagic episodes in these two disorders is transfusion with normal platelets. Alloimmunization will eventually limit the life span of infused platelets. In addition, a few patients have developed inhibitor antibodies with specificity for the missing protein. These antibodies bind to the protein that is expressed on the transfused normal platelets and impair their function. Platelet Release Defects The most common mild bleeding disorders arise from the ingestion of aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) that inhibit platelet production of thromboxane A2, an important mediator of platelet secretion and aggregation (Figs. 62-3 and62-4). These drugs inhibit cyclooxygenase, which converts arachidonic acid to a labile endoperoxide intermediate that is critical for thromboxane formation. Aspirin is the most potent agent, since it irreversibly acetylates the platelet enzyme so that a single dose impairs hemostasis for 5 to 7 days. The other agents are competitive and reversible inhibitors with more transient effects. Blocking thromboxane A2synthesis partially inhibits platelet release and aggregation with weak agonists, such asADP and epinephrine, and produces a mild hemostatic defect. The administration of high doses of certain antibiotics, particularly penicillin, can coat the platelet surface, block platelet release, and impair hemostasis. Patients generally have minimal symptoms such as easy bruising, and bleeding is usually confined to the skin. Occasional patients will have prolonged oozing after surgery, particularly with procedures involving mucous membranes such as periodontal, oral, or reconstructive plastic surgery. The antiplatelet effect of drugs such as aspirin is more dramatic when they are administered to patients with underlying defects such asvWD or hemophilia. Patients with drug-induced cyclooxygenase deficiency often have
a mildly prolonged bleeding time, and their platelets fail to aggregate when incubated with arachidonic acid, epinephrine, or low doses ofADP. Patients who have taken aspirin should be treated as if they have a mild hemostatic defect for the next 5 to 7 days. Platelet responses to collagen and thrombin are impaired at low doses but normal at higher doses. Symptomatic patients should be encouraged to use drugs such as acetaminophen that do not impair platelet function. Although most cases of cyclooxygenase deficiency are drug-induced, occasional patients have inherited disorders in platelet cyclooxygenase activity that impair thromboxane production or receptor level defects that prevent platelets from responding to thromboxane A2. Of the metabolic disorders that can perturb hemostasis, uremic platelet dysfunction is clinically the most important. The mechanism by which uremia impairs platelet function is not well understood, and retention of phenolic and guanidinosuccinic acids, excess prostacyclin production, or impairedvWF-platelet interactions have all been implicated. The degree of uremia correlates with bleeding symptoms and anemia. Bleeding can usually be reversed by dialysis and often improves after red cell transfusion or treatment with erythropoietin. In addition, factor VIII concentrate orDDAVP, both of which raise plasma vWF levels, can also improve hemostasis. Conjugated estrogens improve hemostasis and can be used as long-term therapy. Storage Pool Defects Platelet granules have considerable amounts of adenine nucleotides, calcium, and adhesive glycoproteins such as thrombospondin, fibronectin, andvWF, all of which promote platelet adhesion and aggregation. Patients with defective platelet granules have a mild bleeding disorder. Platelet storage pool defects may be inherited as an isolated disorder or be part of systemic granule packaging defects such as oculocutaneous albinism or the Hermansky-Pudlak or Chediak-Higashi syndromes. Clinically, these patients cannot be distinguished from those with other functional platelet disorders, since they all have easy bruising, mucosal bleeding, and a prolonged bleeding time. They can be differentiated from patients with the cyclooxygenase defects because their platelets will usually aggregate in response to arachidonic acid. In addition, their platelets have decreased levels of specific granule constituents such asADP and serotonin and abnormalities in granule morphology that are best visualized by electron microscopy. Occasionally, patients with acute or chronic leukemia or one of the myeloproliferative disorders develop an acquired storage pool disorder due to dysplastic megakaryocyte development. In addition, patients with liver disease and some patients withSLE or other immune complex-mediated disorders may have circulating platelets that have degranulated prematurely. Platelet degranulation and a transient storage pool disorder may occur after prolonged cardiopulmonary bypass. Fortunately, most patients with storage pool defects have only mildly impaired hemostasis. They can be treated with platelet transfusions. Occasional patients have responded toDDAVP. VESSEL WALL DISORDERS Bleeding from vascular disorders (nonthrombocytopenic purpura) is usually mild and confined to the skin and mucous membranes. The pathogenesis of bleeding is poorly defined in many of the syndromes, and classic tests of hemostasis, including the bleeding time and tests of platelet function, are usually normal. Vascular purpura arises
from damage to capillary endothelium, abnormalities in the vascular subendothelial matrix or extravascular connective tissues that support blood vessels, or from the formation of abnormal blood vessels. Several idiopathic disorders involve the vessel wall and can cause more severe bleeding and organ dysfunction. THROMBOTIC THROMBOCYTOPENIC PURPURA TTPis a fulminant, often lethal disorder that may be initiated by endothelial injury and subsequent release ofvWF and other procoagulant materials from the endothelial cell. Causes include pregnancy, metastatic cancer, mitomycin C, high-dose chemotherapy, HIV infection, and certain drugs, such as the antiplatelet angent ticlopidine. Characteristic findings include the microvascular deposition of hyaline fibrin thrombi, thrombocytopenia, microangiopathic hemolytic anemia, fever, renal failure, fluctuating levels of consciousness, and evanescent focal neurologic deficits. The presence of hyaline thrombi in arterioles, capillaries, and venules without any inflammatory changes in the vessel wall is diagnostic. The presence of a severe Coombs-negative hemolytic anemia with schistocytes or fragmented red blood cells in the peripheral blood smear, coupled with thrombocytopenia, and minimal activation of the coagulation system help to confirm the clinical suspicion of TTP. This disorder should be distinguished from vasculitis andSLE, which can predispose patients to TTP. Platelet-associated IgG and complement levels are usually normal in TTP. The treatment of acuteTTP has focused on the use of exchange transfusion or intensive plasmapheresis coupled with infusion of fresh frozen plasma. Patients with TTP become transiently deficient in a plasma enzyme that depolymerizes ultra-high-molecular-weightvWFreleased from endothelial cells. Therapy may remove abnormal forms of vWF and replenish the deficient enzyme. Overall mortality has been markedly reduced, and the majority of patients with TTP recover from this formerly fatal disorder. Most patients surviving the acute illness recover completely, with no residual renal or neurologic disease. Occasional patients with a chronic, relapsing form of TTP require maintenance plasmapheresis and plasma infusion, and a few patients are controlled only with glucocorticoids. HEMOLYTIC-UREMIC SYNDROME HUSis a disease of infancy and early childhood that closely resemblesTTP. Patients present with fever, thrombocytopenia, microangiopathic hemolytic anemia, hypertension, and varying degrees of acute renal failure. In many cases, onset is preceded by a minor febrile or viral illness, and an infectious or immune complex-mediated cause has been proposed. Epidemics related to infection with a specific strain of Escherichia coli (O157:H7) have been documented. As in TTP, disseminated intravascular coagulation is not found. In contrast to TTP, the disorder remains localized to the kidney, where hyaline thrombi are seen in the afferent arterioles and glomerular capillaries. Such thrombi are not present in other vessels, and neurologic symptoms, other than those associated with uremia, are uncommon. No therapy is proven effective; however, with dialysis for acute renal failure, the initial mortality is only 5%. Between 10 and 50% of patients have some chronic renal impairment.
HENOCH-SCHONLEIN PURPURA Henoch-Schonlein, or anaphylactoid, purpura is a distinct, self-limited type of vasculitis that occurs in children and young adults. Patients have an acute inflammatory reaction in capillaries, mesangial tissues, and small arterioles that leads to increased vascular permeability, exudation, and hemorrhage. Vessel lesions contain IgA and complement components. The syndrome may be preceded by an upper respiratory infection or streptococcal pharyngitis or be associated with food or drug allergies. Patients develop a purpuric or urticarial rash on the extensor surfaces of the arms and legs and on the buttocks as well as polyarthralgias or arthritis, colicky abdominal pain, and hematuria from focal glomerulonephritis. Despite the hemorrhagic features, all coagulation tests are normal. A small number of patients may develop fatal acute renal failure, and 5 to 10% develop chronic nephritis. Glucocorticoids provide symptomatic relief of the joint and abdominal pains but do not alter the course of the illness. METABOLIC AND INFLAMMATORY DISORDERS Acute febrile illnesses may cause capillary fragility and skin bleeding. Immune complexes containing viral antigens or the viruses themselves may damage endothelial cells. In addition, certain pathogens such as the rickettsiae that cause Rocky Mountain spotted fever replicate in endothelial cells and damage them. Thrombocytopenia is also a frequent finding in acute infectious disorders and may contribute to skin bleeding. In addition, whenever the platelet count is95% of those immunized at this age and there is little measles morbidity/mortality among very young infants. In contrast, in the developing world, measles accounts for a significant proportion of deaths of young infants. Thus it is desirable to immunize children during the first few months of life in order to narrow the window of vulnerability between the rapid decline of maternal antibody after 4 to 6 months and the development of vaccine-induced active immunity. Hibcauses meningitis, epiglottitis, and pneumonia in early childhood, with rates rising sharply after the disappearance of maternally derived antibody. The first Hib vaccines often failed when administered during infancy; this failure was due mainly to an age-related inability to respond to polysaccharide antigens. To overcome this problem, the protective polysaccharide was coupled to protein and converted to a T
cell-dependent antigen to which young infants could respond. In contrast to measles andHibinfection, rubella is primarily a threat to the fetus; young infants and children are not at risk of serious illness. Given the susceptibility of the fetus, immunization of all women of reproductive age before pregnancy would be an ideal strategy. However, it is difficult to systematically vaccinate adolescent and young-adult females. Thus, to assure the protection of as many women as possible, the rubella component is included in a combination vaccine with mumps and measles (MMR) that is administered during infancy. Some vaccines are now used primarily for adults. For example, influenza virus and polyvalent pneumococcal polysaccharide vaccines are used to prevent pneumonia deaths in the elderly. Unfortunately, these vaccines are underutilized, in part because physicians and otherwise healthy individuals in the target group ignore the indications and in part because there is still a tendency to think about disease prevention with vaccines as a strategy for children. Pneumococcal polysaccharide vaccine is also recommended for children>2 years old who are at risk of severe or even life-threatening pneumococcal infection, such as those with sickle cell disease, asplenia (whether functional or anatomic), renal failure with nephrotic syndrome, cerebrospinal fluid leak, and HIV infection or other immunosuppressive disease states. THE DEVELOPMENT OF VACCINES BIOLOGIC IMPEDIMENTS There are often major technical problems to overcome in vaccine development. Although just one major antigenic type of influenza virus is typically in circulation at any one time, the virus is characterized biologically by its antigenic drift. Thus, a new antigenic version capable of causing a global pandemic emerges periodically, and a new vaccine must be rapidly devised, produced, and distributed. In contrast, many prevalent pneumococcal polysaccharide serotypes circulate at all times. Because immunity to the pneumococcus is serotype specific, an individual is susceptible to all serotypes against which he or she lacks antibody. Serotype-specific protection has made it more difficult to develop an effective pneumococcal vaccine than it was to develop a vaccine against H. influenzae, of which one capsular serotype (type b) is associated with nearly all cases of severe disease. To overcome this problem, pneumococcal vaccine currently includes 23 polysaccharides that represent ~80% of the virulent serotypes commonly encountered in the United States. Unfortunately, some serotypes are poorly immunogenic, and immunized individuals remain susceptible to the serotypes not included in the vaccine. STRATEGY FOR VACCINE DEVELOPMENT Vaccine development depends on the systematic application of a four-phase strategy: (1) studies in animals to identify protective antigen, (2) determination of how to present this antigen effectively to the immune system, (3) assessment of the safety and immunogenicity of the preparation in small and then in large human populations at various ages, and (4) evaluation of safety and efficacy in the target population. Each of these steps is simple in concept but difficult in execution, not least because of the
clinical trials necessary to assess safety and efficacy; failure at any level stops the process. Thus, in 1995,>190 candidate vaccines were under investigation, but just 5 new products were licensed in the United States. Progress in immunology has taught us much about the organization and function of the immune system (Chap. 305); it has also taught us that the immune system is complicated and that details of antigen composition and presentation are critical for stimulating desired immune responses. Ultimately, vaccines for humans must be tested in humans. After initial animal studies and small phase 1 and 2 human studies to assess immune responses, optimal dosage, and safety, clinical trials of vaccine efficacy are performed, sometimes with informed volunteers who are challenged with a virulent strain. Larger clinical effectiveness trials in the community, typically involving 1000 to 10,000 vaccinees, may lead to application for licensure. Because of their limited size, however, these trials cannot be expected to detect rare adverse effects. Thus, licensing does not guarantee that a new vaccine is completely safe, and postlicensing monitoring is needed to ensure effectiveness and to document the occurrence of adverse events of low frequency. In 1999, the recently licensed rhesus rotavirus vaccine was withdrawn because postmarketing surveillance uncovered an association with a rare event in infants, intussusception of the bowel. The development of vaccines goes beyond technology and proof of principle to issues such as development costs, manufacturers' liability and indemnity, perceived public health needs, and the likelihood that a product will be used or sold. Given the complex science required, the costs of vaccine development are high and success is uncertain, adding risk to the development decision. It is unfortunate that the one sure implication of uncertainty in vaccine development is increased cost. In addition, a rational assignment of costs for development between the public and private sectors in the United States has never been achieved. VACCINE FORMULATIONS Studies of clinical immunology have shown that living and dead antigens do not necessarily induce the same immune responses and that the requirements for the development of protective immunity differ with the organism. These insights, together with the refinement of epidemiologic concepts surrounding immunization, have changed the strategy of vaccine development. The goal is not only to select the correct antigens but also to ensure that the vaccines will result in the type of immune response needed for protection, whether the T cell-mediated activation of macrophages or the generation of cytotoxic T cells, B cell-mediated secretory IgA, or a particular IgG subtype response to a specific polysaccharide epitope. Live vaccines consist of selected or genetically altered organisms that are avirulent or dramatically attenuated yet remain immunogenic. These agents are expected to cause a subclinical illness that mimics natural infection except for the lack of clinically significant disease. They offer the advantage of replication in vivo, which increases the antigenic load presented to the host's immune system; they may confer lifelong protection with one dose; they present all expressed antigens, thus overcoming immunogenetic restrictions in some hosts; they may reach the local sites most relevant to the induction of protective immunity; and they may produce important protective antigens in vivo that are not efficiently expressed in vitro.
Nonviable vaccines may fail to elicit mucosal IgA-mediated immunity, as they lack a delivery system that will effectively transport them to local antigen-processing cells. Moreover, except for pure polysaccharide antigens, these preparations must almost always be given in multiple doses to induce effective responses. However, killed vaccines can be extremely effective. For example, the nonviable hepatitis A vaccine formulation appears to be close to 100% effective in inducing protective immunity. Methods are under development to incorporate vaccine antigens into degradable polymers that may release antigen at predictable times after a single inoculation and simulate multiple injections over time of the same vaccine. In spite of their advantages, live vaccines are not always to be preferred. For example, liveOPV is contraindicated for use in children with immune-deficiency diseases and in their adult contacts. In addition, even though killed poliovirus vaccine does not completely immunize the gut and can neither reduce the circulation of wild-type poliovirus nor immunize contacts of vaccine recipients, the United States has now switched to a four-dose schedule of this vaccine because of the risk of vaccine-associated polio posed by live OPV. To create a deliverable vaccine, constituents other than the antigens are required (Table 122-3). These constituents can affect the immunogenicity, efficacy, and safety of a vaccine and can render one formulation superior to another. NEW VACCINE APPROACHES The first generation of vaccines included whole killed -- or, more recently, live -attenuated microorganisms or partially purified microbial products, such as tetanus toxoid, that induced protective antibodies. The second generation of vaccines has taken advantage of molecular genetics and protein chemistry to isolate and manipulate purified proteins or components or subunits of organisms or to generate genetically engineered and attenuated live native organisms or cloned antigens expressed by harmless vector organisms. One conceptual leap is the production of transgenic plants expressing protective vaccine antigens (cloned, for example, in potatoes or bananas) that, when ingested orally, induce mucosal and systemic immune responses to homologous infectious challenges. While the practical use of this technique awaits further refinement, the concept that protective immunity can be induced in this manner has been proven in both animals and humans. Ease of production, stability, ease of administration without equipment, and low cost are the obvious advantages. Another conceptual leap has led to a third generation of vaccines, in which nucleic acids (either DNA or RNA) are used to induce immunity. Development of DNA vaccines is at a more advanced stage. The principle is simple. First, a DNA plasmid containing the gene sequence for the immunogenic protein or fragment of interest is assembled, and the gene is placed under the control of a strong promoter and an appropriate transcription termination sequence. A single immunization with the plasmid (via intramuscular or intradermal injection, helium-accelerated gene gun injection of DNA-coated gold particles, compressed-air pneumatic jet injection of soluble DNA, direct skin application after suitable preparation, or even insertion of biodegradable stents loaded with the DNA of interest) results in DNA uptake into cells where the gene is expressed and
processed normally; thus the product stimulates an immune response. Alteration of the DNA construct or of the mode of administration or the coadministration of cytokine genes can determine whether the immune response is humoral or cellular or whether it involves primarily Th1, Th2, or cytotoxic T cells. Such decisions can be used to optimize the protective immunity induced. This form of immunization offers real advantages and only theoretical and remote disadvantages (Table 122-4). Moreover, DNA vaccines may be useful in inducing tumor immunity, treating allergy (by suppressing IgE production), or even administering genes for gene therapy. RNA vaccines would avoid some of the potential concerns raised by DNA vaccines because RNA is less stable and does not persist or integrate into the chromosome or cause insertional mutagenesis. However, this lack of stability and the likely need for multiple doses, along with the increased cost of producing, storing, and transporting RNA, are significant disadvantages that remain to be overcome. The concept of nucleic acid vaccines -- whether based on DNA or RNA -- has been validated experimentally, and early human trials have begun. There is great optimism for the future but much to be learned if we are to apply this powerful new immunization technique successfully. PRODUCTION OF VACCINES As products to be given to healthy individuals to prevent disease, vaccines must not only be efficacious but also cause no harm. In the United States, quality assurance is the responsibility of vaccine manufacturers. Standards of manufacture of biologics [known as good manufacturing practices (GMPs)] are regulated and supervised by the U.S. Food and Drug Administration (FDA). Proof of the safety, efficacy, sterility, and purity of products is required before licensure, and sterility and purity are continually monitored for all lots of vaccine after licensure. Postmarketing studies of safety (phase IV studies) are part of routine regulatory control. On rare occasions, either GMP or quality assurance is inadequate; for example, the release of incompletely killed Salk polio vaccine in 1955 caused an outbreak of poliomyelitis in nearly 200 vaccine recipients and their contacts. Unregulated and uncontrolled manufacture of vaccines in developing countries has sometimes led to immunization with inactive products that fail to provide the expected protective immunity. Another problem in the production of vaccines has unexpectedly cropped up in the past decade. For various reasons, including the high costs of vaccine development and the prospect of much higher profitability from investments in other products, the number of vaccine manufacturers in the United States has declined and the cost of some basic childhood vaccines has increased. Concern therefore exists about the future availability of these essential biologics for national use. Furthermore, pricing decisions made within the private-sector pharmaceutical industry can have a major impact on vaccine use. This situation has stimulated an initiative toward increased public involvement in supplying vaccine to individuals for whom price is an issue as well as in oversight of the vaccine supply and of price negotiations with the industry. ADMINISTRATION OF VACCINES Health care workers administering vaccines must take the precautions necessary to
minimize the risk of spreading disease -- for example, hand washing between immunizations. They should be immunized against hepatitis B, measles, rubella, influenza, and varicella. Different vaccines should not be mixed in the same syringe unless such a practice is specifically endorsed by licensure. Disposable needles and syringes should be discarded in labeled, puncture-proof containers to prevent inadvertent needlestick injury or reuse. The addition of new, individually injectable vaccines to the immunization schedule has heightened parental concerns about the administration of up to four injections at a single clinic visit. The development and use of combinations of vaccines are intended to mitigate these concerns. Even when multiple injections are required, providers must make every effort to administer all indicated vaccines at each visit. Wherever effective primary health care systems ensure access to medical services for the majority and the population is educated about the need for and efficacy of vaccines, coverage rates for basic immunization are usually high, regardless of the route of vaccine administration or the number of doses necessary. However, without systematic attention to the completion of multiple-dose vaccine schedules, coverage rates for second, third, and booster doses may drop off significantly. USE OF VACCINES Recommendations for vaccine use in the United States are developed by several different groups. These recommendations are the result of a collaborative process among the recommending groups, the pharmaceutical industry, and theFDA. Vaccines recommended in 1999 for routine administration to infants, children, and adults are shown in Table 122-5; vaccines recommended for special use are shown in Table 122-6; and schedules for immunization of children and adults are shown in Fig. 122-1 andTable 122-7, respectively. The recommendations on route, site, and dosages for vaccination are derived from theoretical considerations, experimental trials, and clinical experience; deviation from these recommendations can result in inadequate protection. The administration of doses at intervals longer than those recommended does not diminish the ultimate protective response but merely delays it. It is not necessary to restart an interrupted series from the beginning or to add an extra dose. In contrast, giving vaccines at shorter-than-recommended intervals may result in poor responses. RECORDING AND REPORTING REQUIREMENTS Certain aspects of vaccine use are regulated by the National Childhood Vaccine Injury Act (NCVIA) of 1986 (modified in 1995). The act requires that all mandated childhood vaccinations be recorded by health care providers in the child's permanent medical record, including date of administration, manufacturer and lot number, and name of the provider administering the vaccine. State-based immunization information systems and registries are being developed to help public and private providers manage their immunization activities and particularly to address the problem of assessing immunization coverage when an individual's records are divided among multiple medical facilities.
Parents must be informed about the benefits and risks of immunization and should maintain an up-to-date immunization record on their children. Educational materials providing the required information (Vaccine Information Statements, VISs) are available from the American Academy of Pediatrics (AAP) or the Centers for Disease Control and Prevention (CDC). VACCINES FOR ROUTINE USE Infants and Children Recommended routine-use vaccines and schedules for their administration to infants and children are shown inTable 122-5 andFig. 122-1, respectively. It is current practice for all children in the United States to receiveDTaP, poliovirus,MMR,Hib,HBV, and varicella vaccines unless there are specific contraindications. Hepatitis A vaccine is currently recommended when there is a special risk of exposure to infection due to residence in communities with elevated rates of hepatitis A or travel to highly endemic countries. Adults (See Table 122-7) All adults should be immune to diphtheria and tetanus. If not previously immunized, adults require a primary immunizing course ofTd. Many individuals remain immune to tetanus into adulthood because they have received tetanus toxoid rather than Td after injuries, but they are commonly at risk of diphtheria because of the decline in titer of diphtheria antitoxin and the lack of boosting against diphtheria. The development of acellular pertussis vaccines that appear to be safe in adults may lead to a recommendation for booster immunization of adults if clinical trials confirm safety and efficacy. Routine immunization against polio is not recommended for adults unless they are at particular risk of exposure (e.g., through travel to endemic regions, as discussed below) or are the parents or guardians of a child with an immunodeficiency disorder. Adults should be protected from measles, mumps, and rubella; they should be vaccinated unless they are known to have received vaccine on or after their first birthday or to have had physician-diagnosed disease. Rubella vaccine should be given to all women of childbearing age unless they have documentary proof of immunization after their first birthday or laboratory evidence of immunity. An unsupported history of rubella disease is unreliable and should not be accepted. Adults without a clear history of chickenpox should receive varicella vaccine. College students, particularly freshmen living in a dormitory, are at increased risk of meningococcal meningitis. They should be made aware of the polysaccharide vaccine for serogroups A, C, Y, and W-135 and should be offered the option of immunization. Current recommendations also include influenza vaccine for routine annual administration to adults³55 years of age and to individuals with chronic illness at any age. Polyvalent pneumococcal polysaccharide vaccine is similarly recommended for the elderly or chronically ill.HBVvaccine is recommended for individuals at high risk of exposure, including health care workers exposed to potentially infected blood or blood products, homosexuals, injection drug users, individuals living and working in institutions for the mentally retarded, and household contacts of known carriers of hepatitis B surface antigen (HBsAg). A new recombinant outer-surface protein A (rOspA) is licensed for persons 15 to 70 years of age for Lyme disease (LYMErix, SmithKline Beecham Pharmaceuticals), with use based on individual risk (geography and risk of exposure to ticks).
Adverse Events Modern vaccines, while safe and effective, are associated with adverse effects that range from infrequent and very mild to rare and life-threatening. The decision to use a vaccine involves an assessment of the risks of disease, the benefits of vaccination, and the risks associated with vaccination. Because these factors may change over time, continued assessment is essential.Table 122-8 lists valid and invalid contraindications to immunization and describes appropriate precautions in the use of specific vaccines. Antivaccine advocacy groups actively encourage avoidance of immunization because of their unproven belief that vaccines may cause certain disorders (for example, autism). This situation presents a challenge to the physician in educating parents about vaccine benefits and risks. Vaccine components, including protective antigens, animal proteins introduced during vaccine production, and antibiotics or other preservatives or stabilizers, can cause allergic reactions in some recipients. These reactions may be local or systemic and may include urticaria and serious anaphylaxis. The most common extraneous allergen is egg protein introduced when vaccines such as those for measles, mumps, influenza, and yellow fever are prepared in embryonated eggs. Local or systemic reactions can result from too-frequent administration of vaccines such asTd, diphtheria/tetanus (DT), or rabies; these reactions are probably due to antigen-antibody complexes. In addition, live-virus vaccines can interfere with tuberculin test responses. When a tuberculin skin test is indicated, it should be done either on the day of immunization or 6 weeks later. When influenza vaccine is given to children 95%. The monthly incidence of hepatitis B infection, both symptomatic and asymptomatic, is 80 to 240 cases per 100,000. For reasons that are not entirely clear, long-stay overseas workers are at considerable risk for hepatitis B infection. In the near future, a combined hepatitis A and B vaccine should become available in the United States. Typhoid Fever The attack rate for typhoid fever is 1 case per 30,000 per month of travel to the developing world (Chap. 156). However, the rates in India, Senegal, and North
Africa are tenfold higher, and, within these areas, rates are especially high among travelers to relatively remote destinations and among persons who are returning to their homelands to stay with relatives or friends. Each of the three available vaccines -- one oral and two injectable -- has an efficacy rate of approximately 70%. Meningococcal Meningitis Although the risk of meningococcal disease among travelers has not been quantified, it is likely to be higher among those who live with poor indigenous populations in overcrowded conditions. The vaccine is recommended for persons traveling to sub-Saharan Africa during the dry season or to areas of the world where there are epidemics. Meningococcal vaccine, which protects against serogroups A/C/Y/W-135, has an efficacy rate of>90%. Japanese Encephalitis The risk of Japanese encephalitis, an infection transmitted by mosquitoes in rural Asia and Southeast Asia, is approximately 1 case per 5000 per month of stay in an endemic area. Most symptomatic infections in U.S. residents have involved military personnel or their families. The vaccine efficacy rate is >80%. Serious allergic reactions sometimes occur; these reactions may be delayed in onset, developing up to 10 days after immunization. The vaccine is recommended for persons staying>1 month in endemic areas. Cholera The risk of cholera is extremely low, with approximately 1 case per 500,000 journeys to endemic areas. Cholera vaccine is rarely recommended but should be considered for aid workers in refugee camps or in disaster/war-torn areas. The injectable vaccine available in the United States is only 30 to 50% effective, and its protective effect persists for only a short period. A more effective oral cholera vaccine is available in other countries. Rabies Many cases of rabies have been reported in travelers, but there are no data on the risk of infection. Domestic animals are the major transmitters of rabies in developing countries (Chap. 197). Countries where canine rabies is highly endemic include Mexico, the Philippines, Sri Lanka, India, Thailand, and Vietnam. Each of the three vaccines available in the United States provides>90% protection. Rabies vaccine is recommended for long-stay travelers, particularly children, and persons who may be occupationally exposed in endemic areas. PREVENTION OF MALARIA AND OTHER INSECT-BORNE DISEASES It is estimated that more than 30,000 American and European travelers develop malaria each year (Chap. 214). Nevertheless, several studies indicate that fewer than 50% of U.S. travelers to malaria-endemic regions adhere to basic recommendations for malaria prevention. The risk of malaria is highest in sub-Saharan Africa and Oceania (1:50 to 1:1000) and during the past decade has increased by more than fivefold for travelers to Kenya. The risk is intermediate (1:1000 to 1:12,000) for travelers to Haiti and the Indian subcontinent and is low (4 weeks for outpatients or>3 days for hospitalized patients with HIV infection. This diagnosis is invoked if appropriate investigation over 3 days, including 2 days' incubation of cultures, reveals no source. Adoption of these categories ofFUO on a wide scale in the literature would allow a more rational compilation of data regarding these disparate groups. In the remainder of this chapter, the discussion will focus on classic FUO unless otherwise specified. CAUSES OF CLASSIC FUO Table 125-2 summarizes the findings of several large studies ofFUOcarried out since the advent of the antibiotic era, including a prospective study of 167 adult patients with FUO encompassing all 8 university hospitals in the Netherlands and using a standardized protocol in which the first author reviewed every patient. Coincident with the widespread use of antibiotics, increasingly useful diagnostic technologies -- both noninvasive and invasive -- have been developed. Newer studies reflect not only changing patterns of disease but also the impact of diagnostic techniques that make it possible to eliminate many patients with specific illness from the FUO category. The ubiquitous use of microbiologic cultures and the widespread use of potent broad-spectrum antibiotics may have decreased the number of infections causing FUO.
The wide availability of ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI) has enhanced the detection of occult neoplasms and lymphomas in patients previously thought to have FUO. Likewise, the widespread availability of highly specific and sensitive immunologic testing has reduced the number of undetected cases of systemic lupus erythematosus and other autoimmune diseases. Several generalizations can be made. Infections, especially extrapulmonary tuberculosis, remain the leading diagnosable cause ofFUO. Prolonged mononucleosis syndromes caused by Epstein-Barr virus, cytomegalovirus (CMV), or HIV are conditions whose consideration as a cause of FUO is sometimes confounded by delayed antibody responses. Intraabdominal abscesses (sometimes poorly localized) and renal, retroperitoneal, and paraspinal abscesses continue to be difficult to diagnose. Renal malacoplakia, with submucosal plaques or nodules involving the urinary tract, may cause FUO and is often fatal if untreated. It is associated with coliform infection, is seen most often in patients with defects of intracellular bacterial killing, and is treated with fluoroquinolones or trimethoprim-sulfamethoxazole. Occasionally, other organs may be involved. Osteomyelitis, especially where prosthetic devices have been implanted, and infective endocarditis must be considered. Although true culture-negative infective endocarditis is rare, one may be misled by slow-growing organisms of the HACEK group (Haemophilus aphrophilus, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, and Kingella kingae;Chap. 150), Bartonella spp. (previously Rochalimaea), Legionella spp., Coxiella burnetii, Chlamydia psittaci, and fungi. Prostatitis, dental abscesses, sinusitis, and cholangitis continue to be sources of occult fever. Fungal disease, most notably histoplasmosis involving the reticuloendothelial system, may causeFUO. FUO with headache should prompt examination of spinal fluid for Cryptococcus neoformans. Malaria (which may result from transfusion, the failure to take a prescribed prophylactic agent, or infection with a drug-resistant strain) continues to be a cause, particularly of nonsynchronized FUO. A related protozoan species, Babesia, may cause FUO and is increasing in incidence. In most earlier series, neoplasms were the next most common cause ofFUOafter infections (Table 125-3). In the two most recent series, a decrease in the percentage of FUO cases due to malignancy was attributed to improvement in diagnostic technologies. This observation does not diminish the importance of considering neoplasia in the initial diagnostic evaluation of a patient with fever. A number of patients in these series had temporal arteritis, adult Still's disease, drug-related fever, and factitious fever. In recent series, approximately 25 to 30% of cases of FUO have remained undiagnosed. The general term noninfectious inflammatory diseases applies to systemic rheumatologic or vasculitic diseases such as polymyalgia rheumatica, lupus, and adult Still's disease as well as to granulomatous diseases such as sarcoidosis and Crohn's and granulomatous hepatitis. In the elderly, multisystem disease is the most frequent cause ofFUO, giant cell arteritis being the leading etiologic entity in this category. Tuberculosis is the most common infection causing FUO in the elderly, and colon cancer is an important cause of FUO with malignancy.
Many diseases have been grouped in the various studies as "miscellaneous." On this list are drug fever, pulmonary embolism, factitious fever, familial Mediterranean fever, and Fabry's disease. A drug-related etiology must be considered in any case of prolonged fever. Any febrile pattern may be elicited by a drug, and both relative bradycardia and hypotension are uncommon. Eosinophilia and/or rash is found in only one-fifth of patients with drug fever, which usually begins 1 to 3 weeks after the start of therapy and remits 2 to 3 days after therapy is stopped. Virtually all classes of drugs cause fever, but antimicrobials (especially b-lactam antibiotics), cardiovascular drugs (e.g., quinidine), antineoplastic drugs, and drugs acting on the central nervous system (e.g., phenytoin) are particularly common causes. It is axiomatic that, as the duration of fever increases, the likelihood of an infectious cause decreases (Table 125-4). In a series of 347 patients referred to the National Institutes of Health from 1961 to 1977, only 6% had an infection. A significant proportion (9%) had factitious fevers -- i.e., fevers due either to false elevations of temperature or to self-induced disease. A substantial number of these factitious cases were in young women in the health professions. It is worth noting that 8% of the patients with prolonged fevers (some of whom had completely normal liver function studies) had granulomatous hepatitis, and 6% had adult Still's disease. After prolonged investigation, 19% of cases still had no specific diagnosis. A total of 27% of patients either had no actual fever during the weeks of inpatient observation or had an exaggerated circadian temperature rhythm without chills, elevated pulse, or other abnormalities. The conditions that may be considered in a differential diagnosis of classicFUO in adults are listed in Table 125-5. This list applies strictly to the United States; the frequency of global travel underscores the need for a detailed travel history, and the continuing emergence of new infectious diseases makes this listing potentially incomplete. SPECIALIZED DIAGNOSTIC STUDIES Classic FUO Certain specific diagnostic maneuvers become critical in dealing with prolonged fevers. If factitious fever is suspected, electronic thermometers should be used, temperature-taking should be supervised, and simultaneous urine and body temperatures should be measured. Any tissue removed during prior relevant surgery should be reexamined; slides should be requested, and, if need be, paraffin blocks of fixed pathologic material should be reexamined and additional special studies performed. Relevant x-rays should be reexamined; reviewing of prior radiologic reports may be insufficient. Serum should be set aside in the laboratory as soon as possible and retained for future examination for rising antibody titers. Febrile agglutinins is a vague term that in most laboratories refers to serologic studies for salmonellosis, brucellosis, and rickettsial diseases. These studies are seldom useful, having low sensitivity and variable specificity. Rising titers of antibody to Brucella (Chap. 160) are usually diagnostic, but false-positive results may be obtained in typhoid fever, tularemia, and yersinial infections. Infection with Brucella canis may be missed with standard antibody tests for Brucella. Salmonella infection (Chap. 156) elevates antibody titers to the H and O antigens. High titers of antibody to the H antigen persist for years and may reflect previous infection or immunization. Serology for Yersinia enterocolitica may be
useful. The measurement of specific antirickettsial titers should be requested for the diagnosis of Rocky Mountain spotted fever and Q fever. Multiple blood samples (no fewer than three, rarely more than six), including samples for anaerobic culture, should be cultured in the laboratory for at least 2 weeks to ensure that any HACEK-group organisms that may be present have ample time to grow (Chap. 150). Lysis-centrifugation blood culture techniques should be employed in cases where prior antimicrobial therapy or fungal or atypical mycobacterial infection is suspected. Blood culture media should be supplemented with L-cysteine or pyridoxal to assist in the isolation of nutritionally variant streptococci. It should be noted that sequential cultures positive for multiple organisms may reflect self-injection of contaminated substances. Urine cultures, including cultures for mycobacteria, fungi, andCMV, are indicated. Blood, urine, or cerebrospinal fluid (CSF) can now be tested for a variety of pathogens such as CMV or hepatitis C virus by using the polymerase chain reaction (PCR) to amplify and hence detect viral nucleic acid (Chap. 121). Liver biopsy, even when the results of liver function studies are normal, should be considered and pursued if the diagnosis remains elusive. Specimens should be cultured for mycobacteria and fungi. Likewise, bone marrow biopsy (not simple aspiration) should be used to obtain specimens for histology and culture. The blood smear should be examined for Plasmodium, Babesia, Trypanosoma, Leishmania, and Borrelia. In anFUOworkup, the erythrocyte sedimentation rate (ESR) should be determined. Striking elevation of the ESR and anemia of chronic disease are frequently seen in association with giant cell arteritis or polymyalgia rheumatica, common causes of FUO in patients over 50 years of age. Still's disease is also suggested by elevations of ESR, leukocytosis, and anemia and is often accompanied by arthralgias, polyserositis (pleuritis, pericarditis), lymphadenopathy, splenomegaly, and rash. Antinuclear antibody, antineutrophil cytoplasmic antibody, rheumatoid factor, and serum cryoglobulins should be measured to rule out other collagen vascular diseases and vasculitis. Another cause of an extremely high ESR may be a false-positive value attributable to a cold agglutinin with a broad thermal amplitude. The ESR test is nonspecific, yielding values that depend on certain serum proteins (most notably fibrinogen) known to interfere with the zeta-potential that keeps erythrocytes from clumping. When fibrinogen levels go up, the zeta-potential is inhibited, erythrocytes clump, and the ESR is high. A cold agglutinin, by binding to erythrocytes, can produce a false-positive agglutinin that mimics an acute-phase response; cold agglutinins may be seen in Mycoplasma and Epstein-Barr virus infections and in lymphomas. With rare exceptions, the intermediate-strength purified protein derivative (PPD) skin test should be used to screen for tuberculosis in patients with classicFUO. Concurrent control tests, such as the CMI test (Connaught Labs, Swiftwater, PA), which is especially effective, should be employed. It should be kept in mind that both the PPD skin test and control tests may yield negative results in miliary tuberculosis, sarcoidosis, Hodgkin's disease, malnutrition, or AIDS. Noninvasive procedures should include an upper gastrointestinal contrast study with small-bowel follow-through and barium enema to include the terminal ileum and cecum. Chest x-rays should be repeated if new symptoms arise. In some cases, pulmonary function studies may be necessary. A diminished carbon monoxide diffusing capacity may indicate a restrictive lung disease such as sarcoidosis, even with a normal chest x-ray. In such cases, transbronchial biopsy may prove diagnostic. Flexible colonoscopy may be advisable, since colon
carcinoma is a cause of FUO and easily escapes detection by ultrasound andCT. CTof the chest and abdomen should be performed. If a spinal or paraspinal lesion is suspected, however,MRI is preferred. MRI may be superior to CT in demonstrating intraabdominal abscesses and aortic dissection, but the relative utility of MRI and CT in the diagnosis of FUOis unknown. At present, it appears that abdominal CT, with oral and intravenous contrast, should be used unless MRI is specifically indicated. Arteriography may be useful for patients in whom systemic necrotizing vasculitis is suspected. Saccular aneurysms may be seen, most commonly in renal or hepatic vessels, and may permit diagnosis of arteritis when biopsy is difficult.Figure 125-1 shows a renal angiogram of a patient with polyarteritis nodosa. Ultrasonography of the abdomen is useful for the investigation of the hepatobiliary tract, kidneys, spleen, and pelvis. Echocardiography may be helpful in an evaluation for bacterial endocarditis, pericarditis, nonbacterial thrombotic endocarditis, and atrial myxomas. Transesophageal echocardiography is especially sensitive for these lesions. Radionuclide scanning procedures using technetium (Tc) 99m sulfur colloid, gallium (Ga) 67 citrate, or indium (In) 111-labeled leukocytes or immunoglobulin may be useful in identifying and/or localizing inflammatory processes. In a recent study, Ga scintigraphy yielded useful diagnostic information in almost one-third of cases, and it was suggested that this procedure might actually be used before other imaging techniques if no specific organ is suspected of being abnormal. Tc bone scan should be undertaken to look for osteomyelitis or bony metastases;67Ga scan may be used to identify sarcoidosis (Chap. 318) or Pneumocystis carinii (Chap. 209) in the lungs or Crohn's disease (Chap. 287) in the abdomen.111In-labeled white blood cell (WBC) scan may be used to locate abscesses;111In-labeled immunoglobulin scan also shows promise in this regard. With67Ga,111In-WBC, and111In-immunoglobulin scans, false-positive and false-negative findings are common. Biopsy of the liver and bone marrow should be considered routine in the workup ofFUO if the studies mentioned above are unrevealing or if fever is prolonged. It goes without saying that areas of suspected abnormality should be sampled for pathologic examination whenever practical. When possible, a section of the tissue block should be retained for further sections or stains.PCRtechnology makes it possible to identify and speciate mycobacterial DNA in paraffin-embedded, fixed tissues. Thus, in some cases, it is possible to make a retrospective diagnosis based on studies of long-fixed pathologic tissues. In a patient over age 50 (or occasionally in a younger patient) with the appropriate symptoms and laboratory findings, "blind biopsy" of one or both temporal arteries may yield a diagnosis of arteritis. If noted, tenderness or decreased pulsation should guide the selection of a site for biopsy. Lymph node biopsy may be helpful if nodes are enlarged, but inguinal nodes are often palpable and are seldom diagnostically useful. Exploratory laparotomy has been performed when all other diagnostic procedures fail but has largely been replaced by modern imaging and guided-biopsy techniques. Laparoscopic biopsy may provide more adequate guided sampling of lymph nodes or liver. Nosocomial FUO The primary considerations in diagnosing nosocomialFUO are the
underlying susceptibility of the patient coupled with the potential complications of hospitalization. The original surgical or procedural field is the place to begin a directed physical and laboratory examination for abscesses, hematomas, or infected foreign bodies. More than 50% of patients with nosocomial FUO are infected, and intravascular lines, septic phlebitis, and prostheses are all suspect. In this setting, the approach is to focus on sites where occult infections may be sequestered, such as the sinuses of intubated patients or a prostatic abscess in a man with a urinary catheter. Clostridium difficile colitis may be associated with fever and leukocytosis before the onset of diarrhea. In approximately 25% of patients with nosocomial FUO, the fever has a noninfectious cause. Among these causes are acalculous cholecystitis, deep vein thrombophlebitis, and pulmonary embolism. Drug fever, transfusion reactions, alcohol/drug withdrawal, adrenal insufficiency, thyroiditis, pancreatitis, gout, and pseudogout are among the many possible causes to consider. As in classic FUO, repeated meticulous physical examinations, coupled with focused diagnostic techniques, are imperative. Multiple blood, wound, and fluid cultures are mandatory. The pace of diagnostic tests is accelerated, and the threshold for procedures -CTscans, ultrasonography,111In-WBC scans, noninvasive venous studies -- is low. Even so, 20% of cases of nosocomial FUO may go undiagnosed. Like diagnostic measures, therapeutic maneuvers must be swift and decisive, as many patients are already critically ill. Intravenous lines must be changed (and cultured), drugs stopped for 72 h, and empirical therapy started if bacteremia is a threat. In many hospital settings, empirical antibiotic coverage for nosocomial FUO now includes vancomycin for methicillin-resistant Staphylococcus aureus as well as broad-spectrum gram-negative coverage with piperacillin/tazobactam, ticarcillin/clavulanate, imipenem, or meropenem. Practice guidelines covering many of these issues have been published jointly by the Infectious Diseases Society of America (IDSA) and the Society for Critical Care Medicine and can be accessed on the IDSA website (http://www.idsociety.org/practice/index.html). Neutropenic FUO (See also Chap. 85) Neutropenic patients are susceptible to focal bacterial and fungal infections, to bacteremic infections, to infections involving catheters (including septic thrombophlebitis), and to perianal infections. Candida and Aspergillus infections are common. Infections due to herpes simplex virus orCMV are sometimes causes ofFUO in this group. While the duration of illness may be short in these patients, the consequences of untreated infection may be catastrophic, with 50 to 60% infected, and 20% bacteremic. TheIDSA has published extensive practice guidelines covering these critically ill neutropenic patients; these guidelines appear on the website cited in the previous section. In these patients, severe mucositis, quinolone prophylaxis, colonization with methicillin-resistant S. aureus, obvious catheter-related infection, or hypotension would dictate the use of vancomycin plus ceftazidime or imipenem to provide empirical coverage for bacterial sepsis. HIV-Associated FUO HIV infection alone may be a cause of fever. Infection due to Mycobacterium avium or Mycobacterium intracellulare, tuberculosis, toxoplasmosis,CMVinfection, P. carinii infection, salmonellosis, cryptococcosis, histoplasmosis, non-Hodgkin's lymphoma, and (of particular importance) drug fever are all possible causes ofFUO. Mycobacterial infection can be diagnosed by blood cultures and by liver, bone marrow, and lymph node biopsies. ChestCTshould be performed to
identify enlarged mediastinal nodes. Serologic studies may reveal cryptococcal antigen, and67Ga scan may help identify P. carinii pulmonary infection. More than 80% of HIV patients with FUO are infected, but drug fever and lymphoma remain important considerations.*Treatment of HIV-associated FUO depends on many factors and is discussed in Chap. 309. TREATMENT The emphasis in patients with classicFUO is on continued observation and examination, with the avoidance of "shotgun" empirical therapy. Empirical treatment for endocarditis, for example, should be avoided unless there are specific reasons beyond fever to invoke this diagnosis. Every patient with FUO should undergo an exhaustive examination for tuberculosis. If thePPD skin test is positive or if granulomatous hepatitis or other granulomatous disease is present with anergy (and sarcoid seems unlikely), then a therapeutic trial with isoniazid and rifampin (and possibly a third drug) should be undertaken, with treatment usually continued for up to 6 weeks. A failure of the fever to respond over this period suggests an alternative diagnosis. The response of rheumatic fever and Still's disease to aspirin and nonsteroidal anti-inflammatory agents (NSAIDs) may be dramatic. The effects of glucocorticoids on temporal arteritis, polymyalgia rheumatica, and granulomatous hepatitis are equally dramatic. Colchicine is highly effective in preventing attacks of familial Mediterranean fever but is of little use once an attack is well under way. The ability of glucocorticoids and NSAIDs to mask fever while permitting the spread of infection dictates that their use be avoided unless infection has been largely ruled out and unless inflammatory disease is both probable and debilitating or threatening. When no underlying source ofFUO is identified after prolonged observation (>6 months), the prognosis is generally good, however vexing the fever may be to the patient. Under such circumstances, debilitating symptoms are treated withNSAIDs, and glucocorticoids are the last resort. The initiation of empirical therapy does not mark the end of the diagnostic workup; rather, it commits the physician to continued thoughtful reexamination and evaluation. Patience, compassion, equanimity, and intellectual flexibility are indispensable attributes for the clinician in dealing successfully with FUO. ACKNOWLEDGEMENT Sheldon M. Wolff, MD, now deceased, was an author of a previous version of this chapter. It is to his memory that the chapter is dedicated. (Bibliography omitted in Palm version) Back to Table of Contents
126. INFECTIVE ENDOCARDITIS - Adolf W. Karchmer INTRODUCTION The proliferation of microorganisms on the endothelium of the heart results in infective endocarditis. The prototypic lesion at the site of infection, the vegetation (see Plate IID-59,Fig. 126-CD1), is a mass of platelets, fibrin, microcolonies of microorganisms, and scant inflammatory cells. Infection most commonly involves heart valves (either native or prosthetic) but may also occur on the low-pressure side of the ventricular septum at the site of a defect, on the mural endocardium where it is damaged by aberrant jets of blood or foreign bodies, or on intracardiac devices themselves. The analogous process involving arteriovenous shunts, arterioarterial shunts (patent ductus arteriosus), or a coarctation of the aorta is called infective endarteritis. Endocarditis may be classified according to the temporal evolution of disease, the site of infection, the cause of infection, or a predisposing risk factor such as injection drug use. While each classification criterion provides therapeutic and prognostic insight, the methods overlap and none is sufficient alone. The classification of endocarditis as acute and subacute was initially used to describe the illness and the time elapsed until death; presently it is applied to the features and progression of infection until diagnosis. Acute endocarditis is a hectically febrile illness, rapidly damages cardiac structures, hematogenously seeds extracardiac sites, and, if untreated, progresses to death within weeks. Subacute endocarditis follows an indolent course; causes structural cardiac damage only slowly, if at all; rarely causes metastatic infection; and is gradually progressive unless complicated by a major embolic event or ruptured mycotic aneurysm. In developed countries, the incidence of endocarditis ranges from 1.5 to 6.2 cases per 100,000 population per year. In the late 1980s in a metropolitan area of the United States (Philadelphia), endocarditis occurred in 9.3 persons per 100,000 population per year. However, half of these cases arose as a consequence of injection drug use. The incidence of endocarditis is notably increased among the elderly. The cumulative rate of prosthetic valve endocarditis is 1.5 to 3.0% at 1 year after valve replacement and 3 to 6% at 5 years; the risk is greatest during the first 6 months after valve replacement. ETIOLOGY A vast array of microorganisms, including many species of bacteria and fungi, have been reported to cause sporadic episodes of endocarditis. Nevertheless, a small number of bacterial species cause the majority of cases (Table 126-1). The causative microorganisms vary somewhat among the major clinical types of endocarditis, in part because of the different portals of entry. The oral cavity, skin, and upper respiratory tract are the respective primary portals for the viridans streptococci, staphylococci, and HACEK organisms (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella) causing community-acquired native valve endocarditis. Streptococcus bovis originates from the gastrointestinal tract, where it is associated with polyps and colonic tumors, and enterococci enter the bloodstream from the genitourinary tract. Nosocomial native valve endocarditis is largely the consequence of bacteremia arising from intravascular catheters and less commonly from nosocomial wound and urinary tract
infection. Endocarditis complicates 6 to 25% of episodes of catheter-associated Staphylococcus aureus bacteremia; higher rates are detected by careful transesophageal echocardiography (TEE) screening (see "Echocardiography," below). Prosthetic valve endocarditis arising within 2 months of valve surgery is generally the result of intraoperative contamination of the prosthesis or a bacteremic postoperative complication. The nosocomial nature of these infections is reflected in their primary microbial causes: coagulase-negative staphylococci, S. aureus, facultative gram-negative bacilli, diphtheroids, and fungi. The portals of entry and organisms causing cases beginning >12 months after surgery are similar to those in community-acquired native valve endocarditis. Epidemiologic evidence suggests that prosthetic valve endocarditis due to coagulase-negative staphylococci that presents between 2 and 12 months after surgery is often nosocomial in origin but with a delayed onset. At least 85% of coagulase-negative staphylococci that cause prosthetic valve endocarditis within 12 months of surgery are methicillin-resistant; the rate of methicillin resistance decreases to 25% among coagulase-negative staphylococci causing prosthetic endocarditis that presents >1 year after valve surgery. Transvenous pacemaker lead and/or implanted defibrillator-associated endocarditis is usually a nosocomial infection. The majority of episodes occur within weeks of implantation or generator change and are caused by S. aureus or coagulase-negative staphylococci. Endocarditis occurring among injection drug users, especially when infection involves the tricuspid valve, is commonly caused by S. aureus strains, many of which are methicillin-resistant. The causes of left-sided valve infection in addicts are more varied, and the involved valves have often been damaged by prior episodes of endocarditis. A number of these cases are caused by Pseudomonas aeruginosa and Candida species, and sporadic cases are due to unusual organisms such as Bacillus, Lactobacillus, and Corynebacterium species. Polymicrobial endocarditis occurs more frequently in injection drug users than in patients who do not inject drugs. The presence of HIV in this population does not significantly impact the causes of endocarditis. From 5 to 15% of patients with endocarditis have negative blood cultures; in one-third to one-half of these cases, cultures are negative because of prior antibiotic exposure. The remainder of these patients are infected by fastidious organisms, such as pyridoxal-requiring streptococci (now designated Abiotrophia species), the gram-negative coccobacillary HACEK organisms, Bartonella henselae, or Bartonella quintana. Some fastidious organisms that cause endocarditis have characteristic epidemiologic settings (e.g., Coxiella burnetii in Europe, Brucella species in the Middle East). Tropheryma whippelii causes an indolent, culture-negative, afebrile form of endocarditis. PATHOGENESIS Unless it is injured, the normal endothelium is resistant to infection by most bacteria and to thrombus formation. Endothelial injury (e.g., at the site of impact of high-velocity jets or on the low-pressure side of a cardiac structural lesion) causes aberrant flow and allows either direct infection by virulent organisms or the development of an uninfected
platelet-fibrin thrombus -- a condition called nonbacterial thrombotic endocarditis (NBTE). The thrombus subsequently serves as a site of bacterial attachment during transient bacteremia. The cardiac lesions most commonly resulting in NBTE are mitral regurgitation, aortic stenosis, aortic regurgitation, ventricular septal defects, and complex congenital heart disease. These lesions result from rheumatic heart disease (particularly in the developing world, where rheumatic fever remains prevalent), mitral valve prolapse, degenerative heart disease, and congenital malformations. NBTE also arises as a result of a hypercoagulable state; this phenomenon gives rise to the clinical entity of marantic endocarditis (seen in patients with malignancy) and to bland vegetations complicating systemic lupus erythematosus and the antiphospholipid antibody syndrome. Organisms that cause endocarditis generally enter the bloodstream from mucosal surfaces, the skin, or sites of focal infection. Except for more virulent bacteria (e.g., S. aureus) that can adhere directly to intact endothelium or exposed subendothelial tissue, microorganisms in the blood adhere to thrombi. If resistant to the bactericidal activity of serum and the microbicidal peptides released by platelets, the organisms proliferate and either stimulate tissues to produce a procoagulant or themselves exert procoagulant activity leading to further platelet-fibrin deposition and vegetation formation. Although an enormous variety of microorganisms circulate transiently in the bloodstream, only a limited number commonly cause endocarditis. The etiologic organisms of endocarditis bear surface components that facilitate adherence to injured endothelium and host proteins or to thrombi. Experiments suggest that fibronectin receptors present on many gram-positive bacteria, clumping factor (a fibrinogen-binding surface protein) on S. aureus, and dextrans on streptococci facilitate adherence. Organisms become enmeshed in the growing platelet-fibrin vegetation and, in the absence of host defenses, proliferate to form dense microcolonies. More than 90% of the organisms in vegetations are metabolically inactive (nongrowing) and thus are relatively resistant to killing by antimicrobial agents. Proliferating surface organisms are shed into the bloodstream continuously, whereupon some are cleared by the reticuloendothelial system and others are redeposited on the vegetation and stimulate further vegetation growth. The pathophysiologic consequences and clinical manifestations of endocarditis -- other than constitutional symptoms, which are probably a result of cytokine production -- arise from damage to intracardiac structures; embolization of vegetation fragments, leading to infection or infarction of remote tissues; hematogenous infection of sites during bacteremia; and tissue injury due to the deposition of circulating immune complexes or immune responses to deposited bacterial antigens. CLINICAL MANIFESTATIONS The clinical syndrome of infective endocarditis is highly variable, may involve multiple organs, and spans a continuum between acute and subacute presentations. Native valve endocarditis (whether acquired in the community or nosocomially), prosthetic valve endocarditis, and endocarditis due to injection drug use share clinical and laboratory manifestations (Table 126-2). Although the relationship is not absolute, the causative microorganism is primarily responsible for the temporal course of endocarditis. b-Hemolytic streptococci, S. aureus, and pneumococci typically result in an acute course, although S. aureus occasionally causes subacute disease.
Endocarditis caused by Staphylococcus lugdunensis (a coagulase-negative species) or by enterococci may present acutely. Subacute endocarditis is typically caused by viridans streptococci, enterococci, coagulase-negative staphylococci, and the HACEK group. Endocarditis caused by Bartonella species and the agent of Q fever, C. burnetii, is exceptionally indolent. The clinical features of endocarditis are nonspecific. However, these symptoms in a febrile patient with valvular abnormalities or a behavior pattern (injection drug use) that predisposes to endocarditis suggest the diagnosis, as do bacteremia with organisms that frequently cause endocarditis, otherwise unexplained arterial emboli, and progressive cardiac valvular incompetence. In patients with subacute presentations, fever is typically low-grade and rarely exceeds 39.4°C (103°F); in contrast, temperatures between 39.4 and 40°C (103 and 104°F) are often noted in acute endocarditis. Fever may be blunted or absent in patients who are elderly or severely debilitated or who have marked cardiac or renal failure. Cardiac Manifestations Although heart murmurs are usually indicative of the predisposing cardiac pathology rather than of endocarditis, valvular damage and ruptured chordae may result in new regurgitant murmurs. In acute endocarditis involving a normal valve, murmurs are heard on presentation in only 30 to 45% of patients but ultimately are detected in 85%. Congestive heart failure develops in 30 to 40% of patients; it is usually a consequence of valvular dysfunction but occasionally is due to endocarditis-associated myocarditis or an intracardiac fistula. The temporal progression of heart failure is variable and depends upon the severity of valvular dysfunction; failure due to aortic valve dysfunction progresses more rapidly than that due to mitral valve dysfunction. Extension of infection beyond valve leaflets into adjacent annular or myocardial tissue results in perivalvular abscesses, which in turn may cause fistulae (from the root of the aorta into cardiac chambers or between cardiac chambers) with new murmurs. Abscesses may burrow from the aortic valve annulus through the epicardium, causing pericarditis. Extension of infection into paravalvular tissue adjacent to either the right or the noncoronary cusp of the aortic valve may interrupt the conduction system in the upper interventricular septum, leading to varying degrees of heart block. Although perivalvular abscesses arising from the mitral valve may potentially interrupt conduction pathways near the atrioventricular node or in the proximal bundle of His, such interruption occurs infrequently. Emboli to a coronary artery may result in myocardial infarction; nevertheless, embolic transmural infarcts are rare. Noncardiac Manifestations (Figs. 19-CD2,124-CD2, and126-CD2) The classic nonsuppurative peripheral manifestations of subacute endocarditis are related to the duration of infection and, with early diagnosis and treatment, have become infrequent. In contrast, septic embolization mimicking some of these lesions (subungual hemorrhage, (Fig. 19-CD4). Osler's nodes,Fig. 126-CD3) is common in patients with acute S. aureus endocarditis (see Plate IID-58). Musculoskeletal symptoms, including nonspecific inflammatory arthritis and back pain, usually remit promptly with treatment but must be distinguished from the symptoms of focal metastatic infection. Hematogenously seeded focal infection may involve any organ but most often is clinically evident in the skin, spleen, kidneys, skeletal system, and meninges. Arterial emboli, which may be asymptomatic and discovered only at autopsy, are clinically
apparent in up to 50% of patients. Vegetations>10 mm in diameter (as measured by echocardiography) and those located on the mitral valve are more likely to embolize than are smaller or nonmitral vegetations. Embolic events -- often with infarction -involving the extremities, spleen, kidneys (Fig. 126-1), bowel, or brain are often noted at presentation. With antibiotic treatment, the frequency of embolic events decreases from 13 per 1000 patient-days during the initial week to 1.2 per 1000 patient-days after the third week. Emboli occurring late during or after effective therapy do not in themselves constitute evidence of failed antimicrobial treatment. Neurologic symptoms, most often resulting from embolic strokes, occur in up to 40% of patients. Other neurologic complications include aseptic or purulent meningitis, intracranial hemorrhage due to hemorrhagic infarcts or ruptured mycotic aneurysms, seizures, and encephalopathy. Microabscesses in brain and meninges occur commonly in S. aureus endocarditis; surgically drainable abscesses are infrequent. Immune complex deposition on the glomerular basement membrane causes diffuse hypocomplementemic glomerulonephritis and renal dysfunction, which typically improve with effective antimicrobial therapy. Embolic renal infarcts cause flank pain and hematuria but rarely cause renal dysfunction. Manifestations of Specific Predisposing Conditions Among injection drug users, infection involving valves on the left side of the heart presents with the typical clinical features of endocarditis. In almost 50% of patients with endocarditis associated with injection drug use, infection is limited to the tricuspid valve. These patients present with fever, faint or no murmur, and (in 75% of cases) prominent pulmonary findings, including cough, pleuritic chest pain, nodular pulmonary infiltrates, and occasionally pyopneumothorax. Nosocomial endocarditis (defined as that which results from hospital care within the prior month and most commonly presenting as intravascular catheter-associated bacteremia), if not associated with a retained intracardiac device, has typical manifestations. Endocarditis associated with flow-directed pulmonary artery catheters is often cryptic, with symptoms masked by comorbid critical illness, and is commonly diagnosed at autopsy. Transvenous pacemaker lead and/or implanted defibrillator-associated endocarditis commonly follows initial implantation or a generator unit change; may be associated with obvious or cryptic generator pocket infection; and results in fever, minimal murmur, and pulmonary symptoms similar to those encountered in addicts with tricuspid endocarditis. Prosthetic valve endocarditis presents with typical clinical features. Cases arising within 60 days of valve surgery (early onset) lack peripheral vascular manifestations and may be obscured by comorbidity associated with recent surgery. In both early-onset and more delayed presentations, paravalvular infection is common and often results in partial valve dehiscence, regurgitant murmurs, congestive heart failure, or disruption of the conduction system. DIAGNOSIS The Duke Criteria The diagnosis of infective endocarditis is established with certainty only when vegetations obtained at cardiac surgery, at autopsy, or from an artery (an
embolus) are examined histologically and microbiologically. Nevertheless, a highly sensitive and specific diagnostic schema -- known as the Duke criteria -- has been developed on the basis of clinical, laboratory, and echocardiographic findings (Table 126-3). Documentation of two major criteria, of one major and three minor criteria, or of five minor criteria allows a clinical diagnosis of definite endocarditis. The diagnosis of endocarditis is rejected if an alternative diagnosis is established, if symptoms resolve and do not recur with £4 days of antibiotic therapy, or if surgery or autopsy after £4 days of antimicrobial therapy yields no histologic evidence of endocarditis. Illnesses not classified as definite endocarditis or rejected are considered cases of possible infective endocarditis. When pathologically confirmed cases have been scored retrospectively by these criteria, 90% fulfill the definition of definite or possible endocarditis; 10% are rejected (primarily because of an incomplete echocardiographic evaluation). In comparison with expert opinion, the Duke criteria identify cases considered to be endocarditis but also accept a small percentage of cases rejected by the experts. This potential for a false-positive diagnosis is the major deficiency in this schema when used clinically. If all patients with a diagnosis of definite or possible endocarditis are fully treated for endocarditis, this reduced specificity results in excess treatment for some patients. A modification of the schema has been proposed in order to increase its specificity without significantly reducing its sensitivity. This modification would require documentation of at least one major or three minor criteria for cases to be categorized as possible endocarditis. The roles of bacteremia and echocardiographic findings in the diagnosis of endocarditis are appropriately emphasized in the Duke criteria. That multiple blood cultures obtained over time are positive is consistent with the known continuous low-density nature of bacteremia in patients with endocarditis (£100 organisms per milliliter). Among untreated endocarditis patients who ultimately have a positive blood culture, 95% of all blood cultures are positive, and in 98% of cases one of the initial two sets of cultures yields the microorganism. The diagnostic criteria attach significance to the species of organism isolated from blood cultures. To fulfill a major criterion, the isolation of an organism that causes both endocarditis and bacteremia in the absence of endocarditis (e.g., S. aureus, enterococci) must take place repeatedly (i.e., persistent bacteremia) and in the absence of a primary focus of infection. Organisms that rarely cause endocarditis but commonly contaminate blood cultures (e.g., diphtheroids, coagulase-negative species) must be isolated repeatedly if their isolation is to serve as a major criterion. Blood Cultures Isolation of the causative microorganism from blood cultures is critical not only for diagnosis but also for determination of antimicrobial susceptibility and planning of treatment. In the absence of prior antibiotic therapy, a total of three blood culture sets, ideally with the first separated from the last by at least 1 h, should be obtained from different venipuncture sites over 24 h. If the cultures remain negative after 48 to 72 h, two or three additional blood cultures, including a lysis-centrifugation culture, should be obtained, and the laboratory should be asked to pursue fastidious microorganisms by prolonging incubation time and performing special subcultures. Empirical antimicrobial therapy should not be administered initially to hemodynamically stable patients with subacute endocarditis, especially those who have received antibiotics within the preceding 2 weeks; thus, if necessary, additional blood cultures can be obtained without the confounding effect of empirical treatment. Patients with
acute endocarditis or with deteriorating hemodynamics that may require urgent surgery should be treated empirically immediately after obtaining the initial three sets of blood cultures. Non-Blood-Culture Tests for the Etiologic Agent Serologic tests can be used to identify some organisms causing endocarditis that are difficult to recover by blood culture: Brucella, Bartonella, Legionella, and C. burnetii. Pathogens can also be identified in vegetations by culture, by microscopic examination with special stains, and by use of polymerase chain reaction to recover unique microbial DNA or 16S rRNA. Echocardiography Cardiac imaging with echocardiography allows anatomic confirmation of infective endocarditis, sizing of vegetations, detection of intracardiac complications, and assessment of cardiac function. A two-dimensional study with color flow and continuous as well as pulsed Doppler is optimal. Transthoracic echocardiography (TTE) is noninvasive and exceptionally specific; however, it cannot image vegetations90% of patients with definite endocarditis; nevertheless, false-negative studies are noted in 6 to 18% of endocarditis patients. TEE is the optimal method for the diagnosis of prosthetic endocarditis or the detection of myocardial abscess, valve perforation, or intracardiac fistulae. Experts favor echocardiographic evaluation of all patients with a clinical diagnosis of endocarditis; however, the test should not be used to screen patients with otherwise explained positive blood cultures or patients with unexplained fever. In patients with a low pretest likelihood of endocarditis (95% of cases; thus, a negative result helps to exclude this diagnosis. TREATMENT While drainage -- either percutaneous (with a pigtail catheter kept in place) or surgical -remains the mainstay of therapy for intraabdominal abscesses (including liver abscesses), there is growing interest in medical management alone for pyogenic liver abscesses. The drugs used in empirical broad-spectrum antibiotic therapy include the same ones used in intraabdominal sepsis. Usually, a diagnostic aspirate of abscess contents should be obtained before the initiation of empirical therapy, with antibiotic choices adjusted when the results of Gram's staining and culture become available. Cases treated without definitive drainage generally require longer courses of antibiotic therapy. When percutaneous drainage was compared with open surgical drainage, the average length of hospital stay for the former was almost twice that for the latter, although both the time required for fever to resolve and mortality were the same for the two procedures. Mortality was appreciable despite treatment, averaging 15%. Several factors may predict the failure of percutaneous drainage and therefore may favor primary surgical intervention. These factors include the presence of multiple, sizable abscesses; viscous abscess contents that tend to plug the catheter; associated disease (e.g., disease of the biliary tract) that requires surgery; or the lack of a clinical response to percutaneous drainage in 4 to 7 days. Treatment of candidal liver abscesses usually entails lengthy administration of amphotericin B, although reports have described successful maintenance therapy with fluconazole after an initial course of amphotericin (Chap. 205). Splenic Abscesses Splenic abscesses are much less common than liver abscesses. In fact, no splenic abscesses were observed in Altemeier's series of 540 intraabdominal
abscesses. The incidence of splenic abscesses has ranged from 0.14 to 0.7% in various autopsy series. The clinical setting and the organisms isolated usually differ from those for liver abscesses. The degree of clinical suspicion for splenic abscess needs to be high, as this condition is frequently fatal if left untreated. Even in the most recently published series, diagnosis was made only at autopsy in 37% of cases. While splenic abscesses may arise occasionally from contiguous spread of infection or from direct trauma to the spleen, hematogenous spread of infection is the usual mode of development. Bacterial endocarditis is the most common associated infection. Splenic abscesses can develop in patients who have received extensive immunosuppressive therapy (particularly those with malignancy involving the spleen) and in patients with hemoglobinopathies or other hematologic disorders (especially sickle cell anemia). While ~50% of patients with splenic abscesses have abdominal pain, the pain is localized to the left upper quadrant in only half of these cases. Splenomegaly is found in ~50% of cases. Fever and leukocytosis are generally present; the development of fever preceded diagnosis by an average of 20 days in one series. Left-sided chest findings may include abnormalities to auscultation, and chest radiographic findings may include an infiltrate or a left-sided pleural effusion. When splenic abscesses are being considered in a differential diagnosis,CT scan of the abdomen has been the most sensitive diagnostic tool. Ultrasonography can yield the diagnosis, but cases have been missed with this modality. Liver-spleen scan or gallium scan may also be useful. Streptococcal species are the most common bacterial isolates from splenic abscesses, and S. aureus is the next most common; presumably these prevalences reflect the bacterial cause of the associated endocarditis. An increase in the frequency of isolation of gram-negative aerobic organisms from splenic abscesses has been reported; these organisms often derive from a urinary tract focus, with associated bacteremia, or from another intraabdominal source. Salmonella species are seen fairly commonly, especially in patients with sickle cell hemoglobinopathy. Anaerobic species accounted for only 5% of isolates in the largest collected series, but the reporting of a number of "sterile abscesses" may indicate that optimal techniques for the isolation of anaerobes were not employed. TREATMENT Because of the high mortality figures reported for splenic abscesses, the treatment of choice is splenectomy with adjunctive antibiotics. However, percutaneous drainage has been successful. The most important factor in successful treatment of splenic abscesses is early consideration of the diagnosis. Perinephric and Renal Abscesses Perinephric and renal abscesses are not common: The former accounted for only ~0.02% of hospital admissions and the latter for ~0.2% in Altemeier's series of 540 intraabdominal abscesses. While liver abscesses generally arise from contiguous foci of infection or track from other intraabdominal sources and splenic abscesses usually arise from hematogenous spread (e.g., spread from bacterial endocarditis), perinephric and renal abscesses have a different pathogenesis. Before antibiotics became available, most renal and perinephric abscesses were hematogenous in origin, with S. aureus most commonly recovered. Now, in contrast,>75% of perinephric and renal abscesses arise from an initial urinary tract infection. Infection ascends from the bladder to the kidney, with pyelonephritis occurring
first. Bacteria may directly invade the renal parenchyma from medulla to cortex. Local vascular channels within the kidney may also facilitate the transport of organisms. Areas of abscess developing within the parenchyma may rupture into the perinephric space. The kidneys and adrenal glands are surrounded by a layer of perirenal fat that, in turn, is surrounded by Gerota's fascia, which extends superiorly to the diaphragm and inferiorly to the pelvic fat. When abscesses extend into the perinephric space, tracking may occur through Gerota's fascia into the psoas or transversalis muscles, into the anterior peritoneal cavity, superiorly to the subdiaphragmatic space, or inferiorly to the pelvis. Of the several risk factors that have been associated with the development of perinephric abscesses, the most important is the presence of concomitant nephrolithiasis producing local obstruction to urinary flow. Of patients with perinephric abscess, 20 to 60% have renal stones. In addition, other structural abnormalities of the urinary tract, a history of urologic surgery, trauma, and diabetes mellitus have all been identified as risk factors. The organisms most frequently encountered in perinephric and renal abscesses are E. coli, Proteus species, and Klebsiella species. E. coli, the aerobic species most commonly found in colonic flora, seems to have unique virulence properties in the urinary tract, including factors promoting adherence to uroepithelial cells. The urease of Proteus species splits urea, thereby creating a more alkaline and hospitable environment for bacterial proliferation. Proteus species are frequently found in association with large struvite stones caused by the precipitation of magnesium ammonium sulfate in an alkaline environment. These stones serve as a nidus for recurrent urinary tract infection. While a single bacterial species is usually recovered from a perinephric or renal abscess, multiple species may also be found. If a urine culture is not contaminated with periurethral flora and is found to contain more than one organism, a perinephric abscess or renal abscess should be considered in the differential diagnosis. Urine cultures may also be polymicrobial in cases of bladder diverticulum. Candida species should be considered in the etiology of renal abscesses. This fungus may spread to the kidney via the hematogenous route or by ascension from the bladder. The hallmark of the latter route of infection is ureteral obstruction with large fungal balls. The presentation of perinephric and renal abscesses is quite nonspecific. Flank pain and abdominal pain are common. At least 50% of patients are febrile. Pain may be referred to the groin or leg, particularly with extension of infection. The diagnosis of perinephric abscess, like that of splenic abscess, is frequently delayed, and mortality in some series is appreciable, although lower than in the past. Perinephric or renal abscess should be most seriously considered when a patient presents with symptoms and signs of pyelonephritis and remains febrile after 4 or 5 days, by which time the fever should have resolved. Moreover, when a urine culture yields a polymicrobial flora, when a patient is known to have renal stone disease, or when fever and pyuria coexist with a sterile urine culture, the diagnosis of perinephric or renal abscess should be entertained. Renal ultrasonography and abdominalCT are the most useful diagnostic modalities. If a renal abscess or perinephric abscess is diagnosed, nephrolithiasis should be excluded, especially when a high urinary pH suggests the presence of a urea-splitting organism.
TREATMENT Treatment for perinephric or renal abscesses, like that for other intraabdominal abscesses, includes drainage of pus and antibiotic therapy directed at the organism(s) recovered. For perinephric abscesses, percutaneous drainage is usually successful. Pancreatic Abscesses*See Chap. 304. (Bibliography omitted in Palm version) Back to Table of Contents
131. ACUTE INFECTIOUS DIARRHEAL DISEASES AND BACTERIAL FOOD POISONING - Joan R. Butterton, Stephen B. Calderwood Ranging from mild annoyances during vacations to devastating dehydrating illnesses that can kill within hours, acute gastrointestinal illnesses rank second only to acute upper respiratory illnesses as the most common diseases worldwide. In children 6 months was seen in 18% of patients, and infertility due to tubal occlusion in 17%; 4% of the pregnancies that did occur were ectopic, representing approximately a sixfold increase over the expected rate of ectopic pregnancies. The rate of infertility after salpingitis was found to be related to the age of the patient, the duration of symptoms when treatment was started, the severity of salpingitis (as determined by laparoscopy) at the time of diagnosis, and the number of episodes of salpingitis. The postsalpingitis risk of infertility due to tubal occlusion among sexually active women not using contraceptives was 14% at 15 to 24 years of age and 26% at 25 to 34 years of age; the risk for women of all ages combined was 11% after one episode of salpingitis, 23% after two episodes, and 54% after three or more episodes. Women with chlamydial salpingitis who developed more severe inflammatory damage to the reproductive tract had significantly increased titers of antibody to the chlamydial heat-shock protein HSP60, as did women with infertility or ectopic pregnancy following chlamydial salpingitis. The risk of infertility after treated gonococcal salpingitis appeared lower than that after chlamydial salpingitis or polymicrobial PID. A study of outcomes of PID at the University of Washington found a sevenfold increase in the risk of ectopic pregnancy and an eightfold increase in the rate of hysterectomy after PID. In several countries, a striking relationship has also been found between infertility due to tubal occlusion and the prevalence and titer of antibody to C. trachomatis. Recurrent salpingitis has been seen in ~15 to 25% of women treated for salpingitis in various studies. PREVENTION Prevention of PIDdepends first on the effective control of gonococcal and chlamydial infection in the general population. Effective methods include the promotion of changes in sexual behavior and the use of barrier contraceptives together with ensuring ready access to modern methods of diagnosis of these infections and effective treatment of sex partners to control further spread. The decline in popularity of the IUD, particularly among nulliparous women, has undoubtedly helped to reduce the incidence of PID. It is also possible, but not proven, that the use of oral contraceptives and the declining proportion of women who have practiced vaginal douching since the link between douching and PID became known have contributed to lower rates of PID. A randomized controlled trial designed to determine whether selective screening for chlamydial infection reduced the risk of subsequent PID showed that women randomized to undergo screening had a 56% lower rate of PID over the following year than did women
receiving the usual care without screening. This report strongly supports risk-based screening for Chlamydia as a highly effective way to reduce the incidence of PID and the prevalence of post-PID sequelae. The complications and sequelae of salpingitis are minimized by early diagnosis and prompt effective treatment. It seems logical, but is unproven, that broad-spectrum therapy effective against all of the common causes ofPIDoffers the best outcome. Although few methodologically sound clinical trials (especially with prolonged follow-up) have been conducted, one meta-analysis showed a benefit of providing good coverage against anaerobes. One placebo-controlled study showed that concurrent anti-inflammatory therapy with prednisolone hastened the reduction of acute inflammatory changes but did not improve the end results, as measured by fertility, hysterosalpingographic findings, or chronic pain. The potential value of anti-inflammatory therapy remains to be evaluated adequately. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 3 -CLINICAL SYNDROMES: NOSOCOMIAL INFECTIONS 134. INFECTION CONTROL IN THE HOSPITAL - Robert A. Weinstein The costs of nosocomial (hospital-acquired) infections are great. It is estimated that nosocomial infections cost $4.5 billion and contribute to 88,000 deaths annually. Although infection-control and hospital epidemiology activities have been the subjects of increasing scientific study over the past 30 years, efforts to lower infection risks have been continually challenged by the growing numbers of immunocompromised patients, antibiotic-resistant bacteria, fungal and viral superinfections, and invasive devices and procedures. Four international decennial conferences on infection control, organized by the Centers for Disease Control and Prevention (CDC), have clearly documented these formidable trends. This chapter reviews the basic surveillance and prevention activities that have been developed to deal with these problems and that form the foundation for current hospital epidemiology programs. ORGANIZATION AND RESPONSIBILITIES OF INFECTION-CONTROL PROGRAMS The standards of the Joint Commission on Accreditation of Healthcare Organizations require all accredited hospitals to have an active program for surveillance, prevention, and control of nosocomial infections; a multidisciplinary infection-control committee usually oversees the program. The agents of the committee are the chairperson, who is preferably an infectious disease physician, and the infection-control practitioners, who are usually trained in nursing or medical technology and in epidemiology and public health. Education of physicians in infection control and hospital epidemiology is required in infectious disease fellowship programs and is available in courses provided by professional societies, primarily the Society for Healthcare Epidemiology of America. In the 1970s, theCDC's extensive Study on the Efficacy of Nosocomial Infection Control found that nosocomial infection rates fell by 32% in hospitals that established programs with organized surveillance and control activities; a trained, effectual infection-control physician; and one infection-control practitioner per 250 beds. In contrast, rates in hospitals without effective programs increased by 18%. Since that study, the responsibilities and roles of hospital epidemiology programs have expanded in several directions. Diagnosis-related reimbursement has led hospital administrators to place increased emphasis on cost containment and on documentation of the cost-effectiveness of infection control. The quality-improvement movements and the Joint Commission have redirected infection-control attention, in part, beyond the mere writing of policies and procedures to improvement of the actual processes and optimization of outcomes. In a few hospitals, epidemiology programs have taken on additional pharmacoepidemiologic and antibiotic-use review responsibilities. Finally, all programs must now respond to increasing governmental regulation of hospital waste and to standards mandated by the Occupational Safety and Health Administration for protecting health care workers from occupational exposure to bloodborne pathogens and tuberculosis. SURVEILLANCE Traditionally, infection-control practitioners survey inpatients for nosocomial infections
(defined as those neither present nor incubating at the time of admission). Surveillance involves a review of microbiology laboratory results, "shoe-leather" epidemiology on the nursing wards, application of standardized definitions of infection, ongoing dialogue with hospital workers, and common sense. Some innovative infection-control programs have taken advantage of the increased use of computerized pharmacy, microbiology, and other databases in hospitals to create algorithm-driven surveillance activities. Most hospitals aim surveillance at infections that (1) are associated with a high level of morbidity, e.g., intensive care unit (ICU)-related infections and nosocomial pneumonia; (2) are costly, e.g., cardiac surgical wound infections; (3) are difficult to treat, e.g., infections due to antibiotic-resistant bacteria; (4) pose recurring epidemic problems, e.g., Clostridium difficile-related diarrhea; and (5) are potentially preventable, e.g., vascular access-related infections. Quality-assurance activities in infection control have led to increased surveillance of the compliance of personnel with infection-control policies (e.g., monitoring of actual adherence to hand-washing recommendations). The results of surveillance are expressed as rates; for example, 5 to 10% of patients develop nosocomial infections. Although such overall statistics are often requested of hospitals by administrators or surveyors, they have little value unless qualified by site of infection, by patient population, and by exposure to risk factors. Meaningful denominators for infection rates include the number of patients exposed to a specific risk (e.g., rates of pneumonia among patients using mechanical ventilators) and the number of intervention days (e.g., rates of pneumonia per 1000 patient-days on a ventilator). Temporal trends in rates should be reviewed, and rates should be compared with regional and national norms. However, even comparison rates generated by theCDC's ongoing National Nosocomial Infections Surveillance System, which collects data from more than 270 hospitals that use standardized definitions of nosocomial infections, have not been validated independently and represent a nonrandom sample of hospitals. Interhospital comparisons are easily confounded by the wide range in risk factors and in severity of underlying illnesses; unless rates are adjusted for these factors, comparisons may be misleading. Unfortunately, systems for making such adjustments either are rudimentary or have not been well validated. The ongoing analysis of an individual hospital's infection rates helps to determine whether control efforts are succeeding and where increased education and control measures should be focused. Knowledge of infection rates is also useful in discussions with the hospital administration regarding areas to which additional resources should be directed. PREVENTION AND CONTROL MEASURES Epidemiologic Basis and General Measures Nosocomial infections follow basic epidemiologic patterns that can help to direct prevention and control measures. Nosocomial pathogens have reservoirs, are transmitted by predictable routes, and require susceptible hosts. Reservoirs and sources exist in the inanimate environment (e.g., tap water contaminated with Legionella) and in the animate environment (e.g., infected or colonized health care workers, patients, and hospital visitors). The mode of
transmission most often is either cross-infection (e.g., indirect spread of pathogens from one patient to another on the inadequately washed hands of hospital personnel) or autoinoculation (e.g., aspiration of oropharyngeal flora into the lung along an endotracheal tube). Occasionally, pathogens (e.g., group A streptococci and many respiratory viruses) are spread indirectly from person to person via infectious droplets released by coughing or sneezing. Much less common -- but often devastating in terms of epidemic risk -- is true airborne spread of droplet nuclei (as in nosocomial chickenpox) or common-source spread by contaminated materials (e.g., iodophors contaminated with Pseudomonas). Factors that increase host susceptibility include underlying conditions and the many medical-surgical interventions and procedures that bypass or compromise normal host defenses. Through its program, the hospital's infection-control committee must determine the general and specific measures used to control infections and must review and recommend specific antiseptics and disinfectants for hospital use. Given the prominence of cross-infection, hand washing is the single most important preventive measure in hospitals. Many studies have examined the antimicrobial activity of a wide variety of antiseptic-containing hand-washing agents. The use of such medicated agents is important before invasive procedures and possibly inICUsettings. In light of the poor general compliance with hand-washing recommendations, the importance of using any hand cleanser between patient contacts cannot be overemphasized (Table 134-1). The fact that 25 to 50% of nosocomial infections are due to the combined effect of the patient's own flora and invasive devices highlights the importance of improvements in the use and design of such devices (Chap. 135). Intensive educational programs can be associated with at least a temporary reduction in infection rates through improved asepsis in handling and earlier removal of invasive devices, but the maintenance of such gains is often difficult. Epidemiologic studies are used increasingly to assess the value of newer devices and site-specific control measures and to debunk some traditional yet ineffective and costly measures, such as routine culturing of the environment and personnel for "pathogens." Urinary Tract Infections Approaches to the prevention of urinary tract infections have included the use of topical meatal antimicrobials, drainage bag disinfectants, antimicrobial-coated catheters, and sealed catheter-drainage tube junctions to eliminate inadvertent breaks in the system. Because of conflicting study results, none of these measures is considered routine. Systemic antimicrobials given for other purposes decrease the risk of urinary tract infection during the first 4 days of catheterization, after which resistant bacteria or yeasts emerge as pathogens. Selective decontamination of the gut is also associated with a reduced risk. Again, however, neither approach is routine. Irrigation of catheters, with or without antimicrobials, may actually increase the risk of infection. Pneumonia Control measures for pneumonia are aimed at the remediation of risk factors in general patient care (e.g., minimizing aspiration-prone supine positioning) and at meticulous aseptic care of respirator equipment (e.g., disinfecting or sterilizing all in-line reusable components such as nebulizers, replacing tubing circuits at intervals of >48 h -- rather than more frequently -- to lessen the number of breaks in the system,
and teaching aseptic technique for suctioning). In a large multicenter trial, sucralfate, which provides stress-ulcer prophylaxis without altering gastric pH, did not reduce the risk of ventilator-associated pneumonia, despite the theoretical advantage of lessened risk for gastric colonization by gram-negative bacilli. The benefit of selective decontamination of the oropharynx and gut with nonabsorbable antimicrobials has been controversial. Surgical Wound Infections The most important control measures for surgical wound infections include the use of antimicrobial prophylaxis at the start of high-risk procedures, attention to technical surgical issues and operating-room asepsis (e.g., not shaving the operative site until surgery and avoiding open or prophylactic drains), and preoperative therapy for active infection. In one study, rates of postoperative infection were lower among patients who had normothermia maintained during colorectal surgery. Reporting of surveillance results to surgeons has been associated with reductions in infection rates. The increasingly extensive review of infection rates by regulatory agencies and third-party payers emphasizes the importance of stratifying rates by patient-related risk factors and of developing meaningful systems for interhospital comparisons and for wound surveillance after the patient's discharge from the hospital or clinic (when more than 50% of infections first become apparent). Infections Related to Vascular Access and Monitoring (See alsoChap. 135) Control measures for infections associated with vascular access and monitoring include the moving of peripheral or arterial catheters to a new site at specified intervals (e.g., every 72 h for peripheral intravenous catheters), which may be facilitated by use of an intravenous team; application of disposable transducers and aseptic technique for the accessing of transducers or other vascular ports; removal of "idle" catheters; and consideration of use of central venous catheters impregnated with anti-infective agents. Unresolved issues include the best frequency for the rotation of central venous catheter sites (guidewire-assisted catheter changes at the same site do not lessen infection risk); the best antiseptics for site preparation and for catheter dressing; the appropriate role for mupirocin ointment, a topical antibiotic with excellent antistaphylococcal activity, in site care; and the relative degrees of risk posed by percutaneous central catheters and by newer designs -- tunneled, totally implanted, or peripherally inserted central catheters (PICC lines). Improvements in composition of semitransparent access-site dressings and potential nursing benefits (ease of bathing and site inspection and protection of the site from secretions) favor use of such coverings. Isolation Techniques Written policies for the isolation of infectious patients are a standard component of infection-control programs. In 1996, theCDCrevised its isolation guidelines to be simpler; to recognize the importance of all body fluids, secretions, and excretions in the transmission of nosocomial pathogens; and to focus precautions on the major routes of infection transmission. The revised guidelines contain two tiers of precautions. Standard precautions are designed for the care of all patients in hospitals to reduce the risk of transmission of microorganisms from both recognized and unrecognized sources of infection. These precautions include gloving, as well as hand washing, for potential contact with blood; with all other body fluids, secretions, and excretions, regardless of whether they contain visible blood; with nonintact skin; and with mucous membranes. Depending on exposure
risks, standard precautions also include use of masks, eye protection, and gowns. In the second tier are precautions for the care of patients with suspected or diagnosed colonization or infection with transmissible pathogens. These transmission-based guidelines collapse the older category- and disease-specific isolation guidelines into three sets of precautions based on probable routes of transmission: airborne precautions, droplet precautions, and contact precautions. Sets of precautions may be combined for diseases that have more than one route of transmission (e.g., varicella). Potentially contagious clinical syndromes, such as acute diarrhea, are included in the revised guidelines. Because some prevalent antibiotic-resistant pathogens, particularly vancomycin-resistant enterococci (VRE), may be present on intact skin of patients in hospitals, some experts recommend gloving for all contact with patients who are acutely ill and/or from high-risk units, such asICUs. In recent trials, wearing gloves did not replace the need for hand washing because hands occasionally became contaminated during wearing or removal of gloves. Some studies have suggested that use of gowns and gloves compared with routine care of patients (i.e., using neither of these barriers) decreases the risk of nosocomial infection; however, more recent evaluation suggests that gowning by personnel does not add benefit beyond that conferred by gloving and hand washing. Nevertheless, requiring increased precaution levels can improve the compliance of health care workers with isolation recommendations by 30%. EPIDEMIC PROBLEMS Outbreaks are always big news but probably account for fewer than 5% of nosocomial infections. The investigation and control of epidemics in hospitals require that infection-control personnel develop a case definition, confirm that an outbreak really exists (since many apparent epidemics are actually pseudo-outbreaks due to surveillance or laboratory artifacts), review aseptic practices and disinfectant use, determine the extent of the outbreak, perform an epidemiologic investigation to determine modes of transmission, work closely with microbiology personnel to culture for common sources or personnel carriers as appropriate and to type epidemiologically important isolates, and heighten surveillance to judge the effect of control measures. Control measures generally include the early reinforcement of routine aseptic practices during a search for compliance problems that may have fostered the outbreak, the ensuring of the appropriate isolation of cases (and the institution of cohort isolation and nursing if needed), and the implementation of further controls on the basis of the findings of the investigation. Examples of some potential epidemic problems follow. Chickenpox When health care workers are exposed to chickenpox in the community or through patients with initially unrecognized infections, or when these employees work during the 24 h before developing chickenpox, infection-control practitioners institute a varicella exposure investigation and control plan. The names of exposed workers and patients are obtained; medical histories are reviewed, and (if necessary) serologic tests for immunity are conducted; physicians are notified of susceptible exposed patients; postexposure prophylaxis with varicella-zoster immune globulin (VZIG) is considered for immunocompromised or pregnant contacts (see Table 183-1); preemptive use of acyclovir is considered as an alternative strategy in some susceptible persons; and
susceptible exposed employees are furloughed during the at-risk period for disease (8 to 21 days, or 28 days if VZIG has been administered). Preexposure varicella vaccination can markedly decrease risk for susceptible employees. Tuberculosis The resurgence of pulmonary tuberculosis in the United States since 1987 and a series of nosocomial outbreaks of infection with multidrug-resistant strains -primarily involving patients with AIDS and their caregivers -- have led to a reevaluation of tuberculosis control. Important control measures include prompt recognition, isolation, and treatment of cases; recognition of atypical presentations (e.g., lower-lobe infiltrates without cavitation); use of negative pressure, 100% exhaust, private isolation rooms with closed doors, and six air changes per hour; use of face masks (approved by the National Institute for Occupational Safety and Health) by caregivers entering isolation rooms; possible use of high-efficiency particulate air filter units and/or ultraviolet lights for disinfecting air when other engineering controls are not feasible or reliable; and follow-up skin-testing of susceptible personnel who have been exposed to infectious patients before isolation. Group A Streptococci The potential for a group A streptococcal outbreak should be considered when even a single nosocomial case occurs. Most outbreaks involve surgical wounds and are due to the presence of an asymptomatic carrier in the operating room. Investigation can be confounded by carriage at extrapharyngeal sites such as the rectum and vagina. Health care workers in whom carriage has been linked to nosocomial transmission of group A streptococci are removed from the patient-care setting and are not permitted to return until carriage has been eliminated by antimicrobial therapy. Aspergillus Aspergillus spores are common in the environment, particularly on dusty surfaces. When hospital ceiling tiles are removed to provide access for electrical wiring or plumbing or when dusty areas are disturbed during hospital renovation, the spores become airborne. Inhalation of spores by immunosuppressed (particularly neutropenic) patients creates a risk of pulmonary and/or paranasal sinus infection and disseminated aspergillosis. Routine surveillance among neutropenic patients for infections with filamentous fungi, such as Aspergillus and Fusarium, helps hospitals to determine whether they have unduly large environmental loads of these organisms. To lower the risk, hospitals should inspect and clean air-handling equipment on a routine schedule, review all planned hospital renovations with infection-control personnel and subsequently construct appropriate barriers, remove immunosuppressed patients from renovation sites, and consider the use of high-efficiency particulate air filters for rooms housing immunosuppressed patients. Legionella Sporadic and epidemic cases of nosocomial Legionella pneumonia are most often due to the contamination of potable water and predominantly affect immunosuppressed patients, particularly those receiving glucocorticoid medication. The risk varies greatly within and among geographic regions, depending on the extent of hospital hot-water contamination, on the presence or absence of high-risk patient populations, and on specific hospital practices (e.g., inappropriate use of nonsterile water in respiratory therapy equipment). Laboratory-based surveillance for nosocomial Legionella should be performed, and a diagnosis of legionellosis should probably be considered more often than it is. If cases are detected, environmental samples (e.g., tap
water) should be cultured. If cultures yield Legionella and if typing of clinical and environmental isolates reveals a correlation, eradication measures should be pursued (Chap. 151). An alternative approach is to periodically culture tap water on wards housing high-risk patients. If Legionella is found, a concerted effort should be made to culture samples from all patients with nosocomial pneumonia for Legionella. Antibiotic-Resistant Bacteria Outbreaks of antibiotic resistance can depend on any of the following events: Darwinian selection of bacterial chromosomal mutations, spread of plasmid- and/or transposon-borne resistance among bacterial species, and (re)admission to the hospital of patients chronically infected with resistant bacteria. After the introduction of resistant strains, dissemination occurs by cross-infection on unwashed hands of caregivers or, occasionally, via personnel carriage and/or environmental contamination. Outbreak control depends on close laboratory surveillance, with early detection of problems; on the reinforcement of routine asepsis (e.g., hand washing); on the implementation of barrier precautions for all colonized and/or infected patients; on the use of patient-surveillance cultures to more fully ascertain the extent of patient colonization; and on the timely initiation of an epidemiologic investigation when rates increase. Colonized personnel who are implicated in nosocomial transmission and patients who pose a threat may be decontaminated; for example, colonization with methicillin-resistant Staphylococcus aureus may be controlled with oral antibiotics, including trimethoprim-sulfamethoxazole and rifampin, and with topical agents, including hexachlorophene or chlorhexidine and mupirocin. In a fewICUs, selective decontamination has been used successfully as a temporary emergency control measure for outbreaks of infection due to gram-negative bacilli. The most recent bacterial-resistance problem to plague hospitals is the emergence ofVRE. Initially anICUproblem, VRE have now spread onto general wards in many hospitals. VRE are particularly problematic because of a substantial "iceberg" effect (i.e., the fact that, for each individual with a clinical infection, many other patients are colonized); the occurrence of both gastrointestinal and skin colonization (reflecting fecal contamination on the skin of ill, hospitalized patients); and the propensity for these organisms to contaminate the patient's environment, which may increase the risk of cross-infection. Control of VRE requires strict attention to hand washing by personnel, concerted use of barrier precautions or cohort nursing for patients known to be colonized or infected, and emphasis on thorough cleaning of the rooms of these patients. Spread of vancomycin resistance to S. aureus is a major concern. Clinical infections with methicillin-resistant S. aureus strains that exhibit reduced susceptibility to vancomycin have been reported in a few patients, usually in the setting of prolonged or repeated treatment with vancomycin. The detection of these strains requires augmented laboratory activities, and their identification should trigger an aggressive epidemiologic investigation and aggressive infection-control measures. Because the excessive use of broad-spectrum antibiotics underlies many resistance problems, antibiotic-control policies (Table 134-2) must be considered a cornerstone of resistance-control efforts. Although the efficacy of antibiotic-control measures in reducing rates of antimicrobial resistance has not been proved in prospective controlled
trials, it seems worthwhile to restrict the use of particular agents to narrowly defined indications or possibly to cycle the use of antibiotic classes to limit selective pressure on the nosocomial flora. EMPLOYEE HEALTH SERVICE ISSUES An institution's employee health service is a critical component of its infection-control efforts. New employees should be processed through the service, where a contagious-disease history can be taken; evidence of immunity to a variety of diseases, such as hepatitis B, chickenpox, measles, and rubella, can be sought; immunizations for hepatitis B, measles, rubella, and varicella can be given as needed and a reminder about the need for yearly influenza immunization can be imparted; baseline and "booster" purified protein derivative of tuberculin skin-testing can be performed; and education about personal responsibility for infection control can be initiated. Evaluations of employees should be codified to meet the requirements of accrediting and regulatory agencies. The employee health service must have protocols for dealing with workers who have been exposed to contagious diseases, such as those percutaneously or mucosally exposed to the blood of patients infected with HIV. Postexposure HIV prophylaxis with a combination of antiretroviral agents (e.g., zidovudine and lamivudine, with or without indinavir or nelfinavir) is recommended. Protocols are also needed for dealing with caregivers who have common contagious diseases, such as chickenpox, group A streptococcal infections, respiratory infections, and infectious diarrhea, and for those who have less common but high-visibility public health problems, such as chronic hepatitis B or C or HIV infection, for which exposure-control guidelines have been published by theCDC and by the Society for Healthcare Epidemiology of America. (Bibliography omitted in Palm version) Back to Table of Contents
135. HOSPITAL-ACQUIRED AND INTRAVASCULAR DEVICE-RELATED INFECTIONS - Dori F. Zaleznik Nosocomial infections are defined as infections acquired during or as a result of hospitalization. Generally, a patient who has been in the hospital for2 million nosocomial infections per year and an annual cost of >$2 billion. Some authorities estimate that the odds of death are doubled for patients who develop a nosocomial infection, although clearly such factors as underlying disease and severity of illness also play an important role in outcome. Although immunosuppressed hosts are especially vulnerable to infections acquired in a hospital, common nosocomial infections occur even in immunocompetent hosts. The National Nosocomial Infections Surveillance (NNIS) Registry has been monitoring nosocomial infection rates since 1970. Its most recent report covers the period from October 1996 through April 1998 and includes both teaching and nonteaching hospitals and both small and large facilities. The most common nosocomial infections have remained the same. Urinary tract infections (UTIs), pneumonia, and surgical-site infections (SSIs, formerly termed wound infections) are most frequent. However, primary bloodstream infections, especially those associated with intravascular devices, have increased in frequency, as have infections in medical and surgical intensive care units (ICUs) and infections caused by antimicrobial-resistant pathogens. The potential impact of nosocomial infections is considerable when assessed in terms of incidence, morbidity, mortality, and financial burden. Analyses of these factors examine nosocomial infections as both medical and economic issues. The clinical problem facing the physician is the development of a new fever in a patient in the hospital. In the evaluation of such a patient, information about the most common categories of infection may not be sufficient. Rather, the clinician must also use clinical clues from the patient's presentation and hospitalization to diagnose a nosocomial infection. Approach to the Patient The evaluation of a hospitalized patient with new fever should include a careful history (Chap. 17). Particular attention should be paid to symptoms of headache, cough, abdominal pain, diarrhea, flank pain, dysuria, urinary frequency, and leg pain. Other features related to the patient's hospitalization are also important, such as the presence and type of intravenous devices, the past or current use of a urinary catheter, the
surgical procedure conducted (if any), and the new medications administered, including those for surgical prophylaxis. The physical examination should be directed at possible sources of infection and should focus particularly on the skin (with a search for rash or embolic lesions); the lungs; the abdomen (especially the right upper quadrant); the costovertebral angles; surgical wounds; the calves; and current and old intravenous access sites (for signs of phlebitis). The laboratory evaluation of all hospitalized patients with new fever should include a complete blood count with differential, a chest radiograph, and blood and urine cultures. Other diagnostic tests to consider include liver function tests, plain-film or other studies of the abdomen, routine aerobic cultures of sputum or other relevant body fluids, and (in cases of diarrhea) testing of stool for Clostridium difficile toxin. CATEGORIES OF INFECTION Pneumonia Certainly the astute clinician will question the patient thoroughly and perform a rapid comprehensive physical examination. One way to continue the approach to the hospitalized patient with a new fever is to consider potential infections that may be life-threatening, such as pneumonia. Most at risk for developing nosocomial pneumonia are patients in anICU, especially those who are intubated; patients with an altered level of consciousness, especially those with nasogastric tubes; elderly patients; patients with chronic lung disease; postoperative patients; and any of the above patients taking H2blockers or antacids. Nosocomial pneumonia in theNNISRegistry is diagnosed 4 to 7 times per 1000 hospitalizations. Among patients on ventilators, the occurrence of pneumonia is estimated at 15 cases per 1000 ventilator days in medical and surgical ICUs. Mortality figures for nosocomial pneumonia are as high as 50%. Oropharyngeal and gastric colonization plays a critical role in the pathogenesis of pneumonia in hospitalized patients. The oropharynx can become colonized by many species of aerobic gram-negative organisms within 48 h of the patient's hospitalization; aspiration occurs commonly during sleep and is increased by such factors as a nasogastric tube, altered consciousness, decreased gag reflex, or delayed gastric emptying. As for gastric colonization, bacterial counts in the stomach rise in the presence of medications that raise gastric pH, such as H2blockers and antacids, as well as in malnourished, achlorhydric, and some elderly patients. The prevalence of pneumonia is reportedly two to three times higher among intubated patients receiving H2blockers or antacids for stress-ulcer prophylaxis than among intubated patients receiving sucralfate, a medication that heals ulcers without altering gastric pH. Gastric colonization is believed to influence the development of pneumonia by retrograde colonization of the oropharynx. Ventilated patients are also at risk of developing pneumonia by exposure to bacteria leaking around the cuff of the endotracheal tube or to bacteria from nebulizers, condensate within ventilator circuits, or humidifiers. Outside the ICU, pneumonia should be suspected when a patient develops a new cough, fever, leukocytosis, sputum production, and a new infiltrate on chest x-ray. Diagnosis can be complicated in patients with congestive heart failure who have concomitant chest x-ray abnormalities or in patients with chronic sputum production. Some organisms, such as Legionella spp., may not be associated with peripheral leukocytosis.
InICUpatients, especially those who are intubated, the signs of pneumonia are relatively subtle, and thus the diagnosis is often relatively complex. In particular, the chest x-rays are difficult to interpret, because fluid overload, congestive heart failure, and acute respiratory distress syndrome (ARDS) are all common findings in intubated patients. Polymorphonuclear leukocytes (PMNs) are often present on Gram-stained preparations of purulent secretions from these patients. An important clue to pneumonia is a change in the output or character of these secretions. If their volume or thickness increases or their color changes, a sputum Gram's stain should be performed and pneumonia seriously considered in the differential diagnosis. Serial Gram's stains are useful, as the number of PMNs may increase substantially and the type(s) of organisms may shift with the development of pneumonia. For example, the baseline sputum sample from an intubated patient may contain about 25 PMNs per high-power field and have mixed gram-positive and gram-negative organisms of several morphologic types in moderate numbers. On the day of a new fever, the same patient may have copious amounts of more tenacious sputum with more PMNs and a predominance of enteric-appearing gram-negative rods. Even without distinct changes in the chest x-ray, this patient would be considered to have developed pneumonia. Another subtle sign of pneumonia in the intubated patient is a requirement for change in ventilator settings in the absence of fluid overload, a mechanical alteration (e.g., a shift in endotracheal tube placement), or a pneumothorax. The major organisms of concern in nosocomial pneumonia are gram-negative aerobic bacteria. Pseudomonas aeruginosa was the most common isolate in theNNISsurvey ofICUs, with a frequency of 21%; Staphylococcus aureus was next most common at 20%. Acinetobacter has become a more common pathogen in ventilator-associated pneumonia. While surveys of organisms are useful, it is essential to know which pathogens are common in a given institution, as hospitals and especially ICUs differ in their resident flora. In some institutions, methicillin-resistant S. aureus, Stenotrophomonas (formerly Xanthomonas) maltophilia, Flavobacterium spp., and even Legionella spp. may be of particular concern. Viruses such as respiratory syncytial virus and adenovirus are receiving increased attention as etiologic agents of nosocomial pneumonia in both adults and children. In the past, viruses have been underrepresented in statistics on the agents of nosocomial pneumonia because the diagnosis of viral infection is more difficult and because many microbiology laboratories do not have the capability to isolate viruses. Antibiotic resistance is another important issue to address in the management of a hospitalized patient. An outgrowth of theNNISsurveys is Project ICARE, which tracks antibiotic usage patterns and resistance rates in a subset of NNIS institutions. Rates of resistance are generally higher in ICUsand track with increased use of antibiotics. P. aeruginosa, Enterobacter spp., and enterococci are the pathogens of greatest concern in the development of antibiotic resistance. In addition to knowing the sensitivity patterns of the hospital flora, one must consider whether a patient has received continuous or multiple courses of antibiotic therapy. To reduce the likelihood of altering the sensitivity patterns of the patient's flora, antibiotic courses for pneumonia should be kept as short as possible, with coverage as narrow as possible for the organism(s) involved. Bacteremia Another potentially life-threatening nosocomial infection to consider in the evaluation of the patient with a new fever is bacteremia, which is usually related to the
presence of an intravascular device (Chap. 134). While many common nosocomial infections such as pneumonia orUTI can be accompanied by bacteremia, primary bacteremia is defined by isolation of a recognized pathogen from the blood without an infection at another site. One carefully controlled study reported bloodstream infection in 2.7% of admissions to a surgicalICU, with 50% mortality and a prolongation of hospitalization by 24 days in survivors. One difficulty in assessing the significance of bacteremia is to distinguish true pathogens from contaminating skin flora. This distinction is especially important in establishing an infection of an indwelling intravascular catheter because organisms that inhabit the skin, such as coagulase-negative staphylococci, also frequently cause infection. The most common point of entry for infection related to intravascular devices is the insertion site, with spread of the infection along the outside of the device initially. Other means of entry for infecting organisms include introduction via contaminated infusates or tubing, ports, or leaking connections and hematogenous seeding of a catheter during bacteremia. While gram-negative aerobic bacilli are probably the most feared nosocomial bloodstream pathogens, the NNISdata for 1980 through 1989 showed that the isolation of these organisms had not increased in frequency over the decade. The frequency of bloodstream isolation increased the most for coagulase-negative staphylococci, with the next highest increase for Candida spp. Other leading causes of line-related bacteremia were S. aureus and enterococci. Subsequent studies confirmed these findings. Nosocomial endocarditis is an important, newly recognized entity that develops largely as a complication of invasive procedures or intravascular devices and may account for as many as 10% of cases of infective endocarditis. Establishing an infection of an intravascular device or primary bacteremia as the cause of fever in a hospitalized patient is a diagnosis of exclusion. If a patient has a fever and signs of cutaneous involvement (erythema, induration, tenderness, or purulent drainage) at the insertion site of a catheter, full cultures should be performed, the vascular-access line removed, and the catheter tip sent for quantitative culture. Studies have correlated the growth of³15 colonies from a catheter tip with infection of the line. More commonly, the exit site does not show signs of infection, and there is considerable debate about the necessity of removing a line from a febrile patient at that point. Although line changes over a guidewire have been shown to be safe, unless another site of infection is obvious, it is generally advisable to remove the line and to change the site when a patient develops a new fever. The traditional teaching is that an infected intravenous device should be removed. In current practice, however, especially with surgically implanted intravenous catheters, a decision may be made to attempt treatment with antibiotics while leaving the catheter in place. This practice is often successful when the infecting organism is a coagulase-negative Staphylococcus species but is less often effective with other organisms, particularly Candida spp. and gram-negative bacilli. Salvage of catheters used for hemodialysis is especially important and has been successfully accomplished with infections caused by a variety of organisms. Another controversial management issue is whether to draw blood for culture through a line. While some studies report a correlation in the 90% range between culture results for blood drawn through vascular-access lines and those for peripheral blood, the former cultures can be either false-positive or false-negative. If the line culture is positive and
no peripheral blood has been drawn, it is impossible to determine whether the patient has true bacteremia or the culture merely reflects bacteria associated with the line. Whether bacteremia is high- or low-grade and whether it is sustained or transient may influence the duration of antibiotic therapy and cannot be determined from cultures of blood specimens obtained through a line. An area of considerable interest and controversy is the prevention of catheter-related infections through the use of intravenous devices impregnated with chlorhexidine/silver sulfadiazine or minocycline/rifampin. A meta-analysis of 11 studies found a decrease in both catheter colonization and catheter-related bacteremia with chlorhexidine/silver sulfadiazine-impregnated catheters, with associated cost savings. One multicenter prospective randomized trial directly compared the two types of catheters and found the minocycline/rifampin-impregnated catheter to be superior. Antibiotic-resistant strains were not recovered in this trial, although concern has been raised that antibiotic impregnation may increase the development of resistance. The several other studies that do not support these findings include one in which the incidence of bacteremia did not decrease with the use of chlorhexidine/silver sulfadiazine-impregnated catheters. Surgical-Site Infection Evaluation of fever in the postoperative patient must include careful evaluation of the surgical wound. AlthoughSSIreportedly accounts for 19% of nosocomial infections, the true incidence of postoperative wound infection is difficult to assess, particularly at a time when many patients are hospitalized for relatively short periods. In a number of studies, careful follow-up for the development of SSI after discharge -- especially observation of the wound by a trained observer, such as a nurse -- has shown the actual rates of SSI in all categories of surgery to be greater than the reported rates. SSI rates vary from 4.6 to 8.2% for nonteaching and large teaching hospitals, respectively. Rates also vary by procedure, with abdominal surgery resulting in the highest rates. Risk factors for the development of postoperative wound infection include the presence of a drain; a long preoperative length of stay, with the rates doubling for each week of preoperative hospitalization; preoperative shaving of the field, especially if performed ³24 h beforehand; a long duration of surgery; and the presence of an untreated remote infection. Infection rates also vary with the surgeon. Perioperative antibiotic prophylaxis has been shown to decrease rates of wound infection in a number of careful studies, including those of clean surgical procedures. Antibiotic coverage after the surgical wound is closed has not been shown to provide additional benefit. A surgical wound should be examined for localized tenderness and induration, fluctuance, drainage of purulent material, and dehiscence of sutures. Mechanical factors, as well as infection, can cause wound dehiscence. Sternal wounds following cardiac surgery are of special concern because the consequences of infection can be severe. The surface of the wound may not present an obvious cause for concern, but ongoing fevers, serous drainage, and especially the development of rocking or instability of the sternum may be sufficient cause for surgical exploration of the wound in some cases. Mediastinitis or sternal osteomyelitis is a severe complication of cardiac surgery. Wounds associated with the placement of prosthetic devices, such as mechanical joints, are also of special concern. Infection of these wounds can lead to infection of the prosthesis, and clearance of prosthetic joint infections generally requires surgical
removal of the device. The most common pathogens causingSSI are coagulase-negative Staphylococcus and S. aureus, but antibiotic-resistant bacteria and fungi are also becoming more frequent etiologies. Early infections may be associated with organisms that produce rapid, progressive skin infection, such as group A Streptococcus and Clostridium spp. Group A Streptococcus has been identified in some cases of recurrent infection of saphenous-vein graft harvest sites. Urinary Tract Infection UTI, the most common type of nosocomial infection, is generally the easiest to treat and has the least severe sequelae. Four principal risk factors have been associated repeatedly with the development of UTI in hospitalized patients: female sex, prolonged urinary catheterization, lack of systemic antibiotic therapy, and breach of appropriate catheter care. The administration of systemic antibiotics to patients with urinary catheters in place for 1 to 5 days has been associated with a decrease in rates of bacteriuria. For patients with catheters in place for ³6 days, however, this benefit is not observed. The pathogenesis of catheter-associatedUTIappears to differ in men and women. In women, the typical mechanism involves periurethral colonization with fecal flora and tracking of organisms up the catheter to the bladder; thus the pathogenesis resembles that of UTI in noncatheterized female patients, in whom bacteria track up the short female urethra. In contrast, periurethral colonization often cannot be demonstrated in men; most infections seem to arise from intraluminal spread of organisms to the bladder. Some organisms, such as Proteus and Pseudomonas spp., appear to facilitate the growth along the inside of the urinary catheter of a biofilm that encrusts and obstructs the flow of urine. UTIis certainly an extremely common nosocomial infection; however, it is important to define this type of infection precisely. Especially in the evaluation of a febrile hospitalized patient, it is crucial to think carefully about all possible sources of infection and not to assume that UTI is the probable cause. In patients who have had urinary catheters in place for a number of days, fever, dysuria, frequency, leukocytosis, and especially flank pain or costovertebral angle tenderness are highly suggestive of bladder infection or pyelonephritis. In patients with fever but no other symptoms or signs referable to the urinary tract, one should look for ancillary findings suggestive of urinary tract involvement, such as white blood cells without epithelial cells in the urine sediment or leukocyte esterase or nitrite on urinalysis. A urine culture positive for a single organism should not be accepted as definitive evidence of UTI in an asymptomatic patient. While one might treat the febrile patient who has a positive urine culture with antibiotics, it is prudent to repeat the culture before the institution of therapy. Inability to recover any organism or the same organism on repeat culture, particularly if the patient does not respond to antibiotics, should raise questions about the validity of the diagnosis of UTI. In addition, isolation of two or more bacteria from a single specimen is most likely due to contamination unless there is reason to suspect a bladder diverticulum or a perinephric abscess. Other Infectious Sources of Fever Several other types of infection may cause fever in the hospitalized patient and should be considered in the differential diagnosis of new
fever. In patients who have received antibiotics (even a single dose as surgical prophylaxis), antibiotic-associated diarrhea may develop. This condition is usually caused by the spore-forming organism C. difficile, which produces toxins that cause diarrhea. Some patients may appear quite toxic with this infection, with high fevers, leukocytosis, and profuse diarrhea. The organism is quite hardy and is difficult to eradicate from the hospital environment. The hands of hospital personnel have been implicated as a mode of transmission of this organism, as have electronic rectal thermometers. The colon may become colonized with C. difficile while the patient is in the hospital, but -- particularly if the patient is still taking antibiotics when sent home -diarrhea may not develop until after discharge. Unless patients are consuming foods from outside the hospital, food-borne diarrheal illness is uncommon among hospitalized patients. Thus, an extensive stool evaluation is generally not cost-effective in the management of these patients. Other infections to consider in the hospitalized patient include decubitus ulcers, particularly in patients in chronic-care wards or confined to bed rest for prolonged periods, and sinusitis, especially in intubated patients. NONINFECTIOUS SOURCES OF FEVER A consideration of several common noninfectious causes of fever in hospitalized patients is part of a thorough evaluation of new fever. Drug treatment is the foremost noninfectious cause of fever. Drug fever may occur with or without an accompanying rash or eosinophilia and can be caused by a new medication or by medications the patient has been receiving for some time. Particular agents associated with drug fever include phenytoin, H2blockers, procainamide, and antibiotics, most notably sulfonamides. Even drug-associated fevers can be quite high in some patients and may take up to 5 days to resolve after discontinuation of treatment with the offending agent. Other noninfectious causes of fever include phlebitis, often at the site of an old intravenous line and sometimes followed by suppurative thrombophlebitis with clots or septic emboli, and pulmonary emboli, especially in patients undergoing prolonged bed rest; prophylactic heparin or mechanical boots are often used to reduce the risk of pulmonary embolism in the latter patients. Other entities to consider include tissue necrosis following surgery, trauma, or burns; hematomas; pancreatitis; atelectasis; and acalculous cholecystitis. CONCLUSION The range of possibilities for the etiology of a new fever in a hospitalized patient is quite broad. An attention to detail, a careful history and physical examination, and a knowledge of the infections and organisms likely to cause nosocomial problems usually lead to an accurate diagnosis. (Bibliography omitted in Palm version) Back to Table of Contents
136. INFECTIONS IN TRANSPLANT RECIPIENTS - Robert Finberg, Joyce Fingeroth The evaluation of infections in transplant recipients involves consideration of both the donor and the recipient of the transplanted organ. Infections following transplantation are complicated by the use of drugs that are necessary to enhance the likelihood of survival of the transplanted organ but that also cause the host to be immunocompromised. Thus what might have been a latent or asymptomatic infection in an immunocompetent donor or in the recipient prior to therapy becomes a life-threatening problem when the recipient becomes immunosuppressed. A variety of organisms have been transmitted by organ transplantation (Table 136-1). Careful attention to the sterility of the medium used to process the organ combined with meticulous microbiologic evaluation reduces rates of transmission of bacteria that may be present or grow in the organ culture medium. From 2% to >20% of donor kidneys are estimated to be contaminated with bacteria -- in most cases, with the organisms that colonize the skin or grow in the tissue culture medium used to bathe the donor kidney while it awaits implantation. The reported rate of bacterial contamination of transplanted bone marrow is as high as 17% but is most commonly ~1%. The use of enrichment columns and monoclonal-antibody depletion procedures results in a higher incidence of contamination. Approximately 2% of cryopreserved marrow and peripheral blood stem cells transfused as part of treatment for cancer are contaminated. In one series of patients receiving contaminated products, 14% had fever or bacteremia, but none died. Results of cultures performed at the time of cryopreservation and at the time of thawing were helpful in guiding therapy for the recipient. In many transplantation centers, transmission of infections that may be latent or clinically inapparent in the donor organ has resulted in the development of specific donor-screening protocols. In addition to ordering serologic studies focusing on viruses such as herpes-group viruses [herpes simplex virus (HSV) 1, HSV-2], varicella-zoster virus (VZV), cytomegalovirus (CMV), human herpesvirus (HHV) 6, Epstein-Barr virus (EBV), HHV-8, hepatitis B and C viruses, and HIV and on parasites such as Toxoplasma gondii, clinicians caring for organ donors should consider assessing stool (for parasites) and skin testing for Mycobacterium tuberculosis. It is expected that the recipient will have been likewise assessed. This chapter considers aspects of infection unique to various transplantation settings. INFECTIONS IN BONE MARROW AND HEMATOPOIETIC STEM CELL TRANSPLANT RECIPIENTS Bone marrow or hematopoietic stem cell transplantation for either immunodeficiency or cancer results in a transient state of complete immune incompetence. Immediately after transplantation, both phagocytes and immune cells (T and B cells) are absent, and the host is extremely susceptible to infection. The reconstitution that follows transplantation has been likened to maturation of the immune system in neonates. The analogy does not entirely predict infections seen in bone marrow transplant (BMT) and hematopoietic stem cell transplant (HSCT) recipients, however, because the new marrow matures in an old host who has several latent infections already.
TIMING OF INFECTIONS In the first month after bone marrow or hematopoietic stem cell transplantation, infectious complications are similar to those in granulocytopenic patients receiving chemotherapy for acute leukemia (Chap. 85). Because of the anticipated 1- to 4-week duration of neutropenia in this population, many centers give prophylactic antibiotics to patients upon initiation of chemotherapy. Prophylactic trimethoprim-sulfamethoxazole or ciprofloxacin decreases the incidence of gram-negative bacteremia among these patients. In the second month after transplantation, a major concern (particularly in allogeneicBMT/HSCTrecipients) isCMVdisease (Chap. 185), which rarely has its onset earlier than 14 days after transplantation and may become evident up to 4 months after the procedure (most commonly at 1 to 3 months). In cases in which the donor marrow is depleted of T cells [to prevent graft-versus-host disease (GVHD) or eliminate a T cell tumor], the disease may be manifested earlier. Patients who receive ganciclovir (for prophylaxis, preemptive treatment, or treatment; see below) may develop CMV infection even later than 4 months after transplantation; treatment appears to delay the development of the normal immune response to CMV infection. Although CMV disease may present as isolated fever, cytopenia, or gastrointestinal disease, the foremost cause of death from CMV infection in this setting is pneumonia. The diagnosis of pneumonia inBMT/HSCTrecipients poses some special problems (Table 85-5). Because patients have undergone treatment with multiple chemotherapeutic agents and sometimes radiation, their differential diagnosis should include -- in addition to bacterial pneumonia -- CMVpneumonitis, pneumonia of other viral or fungal etiology, parasitic pneumonia, diffuse alveolar hemorrhage, and chemicalor radiation-associated pneumonitis. Since fungal disease and viruses such as respiratory syncytial virus (RSV), parainfluenza virus (types 1, 2, and 3), influenza A and B viruses, and adenovirus are also causes of pneumonia in this setting, it is important to diagnose CMV specifically (see below). M. tuberculosis has been an uncommon cause of pneumonia among BMT/HSCT recipients in western countries (120 days after transplantation). The use of quantitative viral load assays, which are not dependent on circulating polymorphonuclear leukocytes, should permit early accurate diagnosis in the future. Treatment ofCMVpneumonia inBMT/HSCTrecipients requires both intravenous immune globulin (IVIG) and ganciclovir. In patients who cannot tolerate ganciclovir, foscarnet is a useful alternative, although it may produce nephrotoxicity and electrolyte imbalance. Transfusion of CMV-specific T cells from the donor decreased viral load in a small series of patients; this result suggests that immunotherapy may play a role in the treatment of this disease in the future.*For further discussion, see Chap. 185. Human Herpesviruses 6 and 7HHV-6, the cause of exanthem subitum in children (Chap. 185), is a ubiquitous herpesvirus that reactivates (as determined by culture of the virus from the blood) in ~50% of transplant recipients between 2 and 4 weeks after surgery. In some cases, reactivation of HHV-6 appears to be associated with neutropenia; since, likeCMV, this virus can be found in marrow cells, it is possible that HHV-6 reactivation is responsible for some of the neutropenia that follows bone marrow transplantation. Although encephalitis developing after transplantation has been associated with HHV-6 in cerebrospinal fluid (CSF), the causality of the association is not well defined. HHV-6 DNA is sometimes found in lung samples after transplantation. However, its role in pneumonitis is unclear. While HHV-6 has been shown to be sensitive to foscarnet (and in some instances to ganciclovir) in vitro, the efficacy of
antiviral treatment has not been well studied. Little is known about the related herpesvirus HHV-7 or its role in posttransplantation infection.*For further discussion, see Chap. 185. Epstein-Barr Virus PrimaryEBVinfection can be fatal to transplant recipients; EBV reactivation can cause EBV-B cell lymphoproliferative disease (LPD), which may also be fatal to patients taking immunosuppressive drugs. The localization of EBV to B cells leads to several interesting phenomena inBMT/HSCTrecipients. The marrow ablation that occurs as part of the BMT/HSCT procedure may eliminate latent EBV from the host. Infection can then be reacquired immediately after transplantation by transfer of infected donor B cells. Alternatively, transplantation from a seronegative donor may result in cure. The recipient is then at risk for a second primary infection. EBV-LPDcan develop in the recipient's B cells (if any should survive marrow ablation) but is more likely to be a consequence of outgrowth of infected donor cells. Both lytic and latent EBV replication are more likely during immunosuppression (e.g., they are associated withGVHD and the use of antibodies to T cells). Although less likely in autologous transplantation, reactivation can occur in T cell-depleted autologous recipients (e.g., patients being treated for a T cell lymphoma with marrow depletion using antibodies to T cells). EBV-LPD, which usually becomes apparent 1 to 3 months after engraftment, can cause high fevers and cervical adenopathy resembling the symptoms of infectious mononucleosis but more commonly presents as an extranodal mass. The incidence of 0.6% among allogeneicBMT/HSCTrecipients contrasts with figures of ~5% for renal transplant recipients and up to 20% for cardiac transplant patients. In all cases, EBV-LPD is more likely to occur with continued immunosuppression (especially that caused by the use of antibodies to T cells and cyclosporine or tacrolimus). EBV-specific T cells generated from the donor have been used experimentally to prevent and to treat EBV-LPDin the allogeneic recipient. Some studies indicate that EBV-LPD can be treated with antibodies to B cell surface antigens. Use of an anti-CD20 monoclonal antibody (Rituximab) to treat B cell lymphomas that express this surface protein has elicited some dramatic responses. Studies are in progress to assess efficacy in EBV-LPD, in which the involved B cells commonly bear CD20. The role of antivirals is uncertain because no available agents have been documented to have activity against latent EBV infection. Ganciclovir has been postulated to have activity on the basis of its ability to inhibit proliferation of B cells, but this activity is associated with toxicity. Both interferona and retinoic acid have been used in the treatment of EBV-LPD, as hasIVIG, but no large studies have assessed the efficacy of these agents. Chemotherapeutic regimens have been used as a last resort, even though patients' tolerance and long-term results have been disappointing in this setting.*For further discussion, see Chap. 184. Human Herpesvirus 8 TheEBV-related gamma herpesvirusHHV-8, which is causally associated with Kaposi's sarcoma, with primary effusion lymphoma, and sometimes with multicentric Castleman's disease, has rarely resulted in disease inBMT/HSCTrecipients. The reasons may be a relatively low seroprevalence in the population and the limited duration of profound T cell suppression after bone marrow/hematopoietic stem cell transplantation.*For further discussion, see Chap. 185.
Other (Nonherpes) Viruses Both RSVand parainfluenza viruses, particularly type 3, can cause severe or even fatal pneumonia inBMTrecipients. Infections with both of these agents sometimes occur as disastrous nosocomial epidemics. Therapy with aerosolized ribavirin as well as RSV immunoglobulin or monoclonal antibody to RSV (Palivizumab) has been reported to lessen the severity of RSV disease, but there are no large studies to prove efficacy. Influenza is also seen in BMT recipients and generally mirrors the presence of infection in the community. Several drugs are available for the treatment of influenza (amantadine/rimantadine, ribavirin?) but have limited effects, primarily reducing symptoms and shortening the duration of illness. The newly approved neuraminidase inhibitors are active against both influenza A virus and influenza B virus. Their role in ameliorating disease in this patient population is unknown. Adenovirus can be isolated from BMT recipients at rates varying from 5 to 18%. Although hemorrhagic cystitis, pneumonia, and fatal disseminated infection have been reported, adenovirus infection, which (likeCMVinfection) usually occurs in the first or second month after transplantation, is often asymptomatic. Therapy with intravenous ribavirin is questionably effective. Cidofovir has proved effective in animal models and in case reports. Infections with parvovirus B19 (presenting as anemia or occasionally pancytopenia) and enteroviruses (sometimes fatal) can occur. Pleconaril, a newly developed capsid-binding agent, is being studied for treatment of enterovirus infection. Rotaviruses are a common cause of gastroenteritis in BMT/HSCTrecipients. BK and, to a lesser extent, JC virus (polyomavirus hominis 1 and 2, respectively) are found in the urine of some transplant recipients. BK viruria may be associated with hemorrhagic cystitis. Progressive multifocal leukoencephalopathy caused by JC virus is rare among BMT/HSCT recipients compared with the rate among patients with impaired T cell function due to HIV infection. There is no known treatment for this disease; however, cidofovir and other agents are under study. INFECTIONS IN SOLID ORGAN TRANSPLANT RECIPIENTS Morbidity and mortality among solid organ transplant recipients have been reduced by the use of more effective antibiotics. The organisms that cause infections in recipients of solid organ transplants are different from those that infectBMT/HSCTrecipients because solid organ recipients do not go through a period of neutropenia. As the transplantation procedure involves surgery, however, solid organ recipients are subject to infections at anastomotic sites and to wound infections. Compared with BMT/HSCT recipients, organ transplant patients are immunosuppressed for more prolonged periods (often permanently). Thus they are susceptible to the same organisms as patients with chronically impaired T cell immunity (Chap. 85, especiallyTable 85-1). During the early period (6 months after transplantation do not seem to be associated with the high rate of pyelonephritis or relapse seen with infections that occur in the first 3
months and may be treated for shorter periods. Prophylaxis with trimethoprim-sulfamethoxazole [1 double-strength tablet (800 mg sulfamethoxazole, 160 mg trimethoprim) per day] for the first 4 months after transplantation decreases the incidence of early and middle-period infections (see below and Table 136-4). Middle-Period Infections Because of continuing immunosuppression, kidney transplant recipients are predisposed to lung infections characteristic of those in patients with T cell deficiency (i.e., infections with intracellular bacteria, mycobacteria, nocardiae, fungi, viruses, and parasites). The high mortality associated with Legionella pneumophila infection (Chap. 151) led to the closing of renal transplant units in hospitals with endemic legionellosis. About 50% of all renal transplant recipients presenting with fever 1 to 4 months after transplantation have evidence ofCMVdisease; CMV itself accounts for the fever in over two-thirds of cases and thus is the predominant pathogen during this period. CMV infection (Chap. 185) may also present as arthralgias or myalgias. During this period, this infection may represent primary disease (in the case of a seronegative recipient of a kidney from a seropositive donor) or may present as reactivation disease or superinfection. Patients may have atypical lymphocytosis. Unlike immunocompetent patients, however, they often do not have lymphadenopathy or splenomegaly. Therefore, clinical suspicion and laboratory confirmation are necessary for diagnosis. The clinical syndrome may be accompanied by bone marrow suppression (particularly leukopenia). CMV also causes glomerulopathy and is associated with an increased incidence of other opportunistic infections. Because of the frequency and severity of CMV disease, a considerable effort has been made to prevent and treat it in renal transplant recipients. Administration of an immune globulin preparation enriched with antibodies to CMV (CMV-Ig) decreases the incidence in the group at highest risk for severe infections (seronegative recipients of seropositive kidneys). Ganciclovir is useful for the treatment of serious CMV disease. One study showed a significant (50%) reduction in CMV disease and rejection at 6 months in patients who received prophylactic valacyclovir (an acyclovir congener) for the first 90 days after renal transplantation. If confirmed, these results will likely change practice. Infection with the other herpes-group viruses may become evident within 6 months after transplantation or later. Early after transplantation,HSV may cause either oral or anogenital lesions that are usually responsive to acyclovir. Large ulcerating lesions in the anogenital area may lead to bladder and rectal dysfunction as well as predisposing to bacterial infection.VZV may cause fatal disseminated infection in nonimmune kidney transplant recipients, but in immune patients reactivation zoster usually does not disseminate outside the dermatome; thus disseminated VZV infection is a less fearsome complication in kidney transplantation than in bone marrow transplantation.HHV-6 may reactivate and (although usually asymptomatic) may be associated with fever, rash, marrow suppression, or encephalitis. EBVreactivation disease is more serious; it may present as an extranodal proliferation of B cells that invade theCNS, nasopharynx, liver, small bowel, heart, and transplanted kidney. The disease is diagnosed by the finding of a proliferation of EBV-positive B
cells. The incidence of EBV-LPDis higher among patients given high doses of cyclosporine, tacrolimus, or other immunosuppressive agents (including anti-T cell antibodies). Fortunately, disease often regresses once immunocompetence is restored.HHV-8 infection can be transmitted with the donor kidney and is associated with the development of Kaposi's sarcoma in the recipient. Kaposi's sarcoma (primary vs. reactivation of HHV-8) often appears within 1 year after transplantation, although the range of onset times is wide (1 month to ~20 years). The papovaviruses BK and JC (polyomaviruses hominis 1 and 2) have been cultured from the urine of kidney transplant recipients (as they have from that ofBMTrecipients). The excretion of BK virus is associated with ureteral strictures and that of JC virus with progressive multifocal leukoencephalopathy (rare). Adenoviruses may persist with continued immunosuppression in these patients. Kidney transplant recipients are also subject to infections with other intracellular organisms. These patients may develop pulmonary infections with Nocardia, Aspergillus, and Mucor as well as infections with other pathogens in which the T cell/macrophage axis plays an important role. In patients without intravenous catheters, L. monocytogenes is the most common cause of bacteremia³1 month after renal transplantation. Kidney transplant recipients may develop Salmonella bacteremia, which can lead to endovascular infections and require prolonged therapy. Pulmonary infections with P. carinii are common unless the patient is maintained on trimethoprim-sulfamethoxazole prophylaxis. Nocardia infection (Chap. 165) may present in the skin, bones, lungs, orCNS(where it usually takes the form of single or multiple brain abscesses). Nocardia infection generally occurs ³1 month after transplantation and may follow immunosuppressive treatment for an episode of rejection. Pulmonary findings are nonspecific: localized disease with or without cavities is most common, but the disease may disseminate. The diagnosis is made by culture of the organism from sputum or from the involved nodule. As with P. carinii, prophylaxis with trimethoprim-sulfamethoxazole appears to be efficacious in the prevention of disease. The occurrence of Nocardia infections>2 years after transplantation suggests that a long-term prophylactic regimen may be justified. Toxoplasmosis can occur in seropositive patients, usually developing in the first few months after kidney transplantation. Again, trimethoprim-sulfamethoxazole is helpful in prevention. In endemic areas, histoplasmosis, coccidioidomycosis, and blastomycosis may cause pulmonary infiltrates or disseminated disease. Late Infections Late infections (>6 months after kidney transplantation) includeCMVretinitis and a variety ofCNScomplications. Patients (particularly those whose immunosuppression has been increased) are at risk for subacute meningitis due to Cryptococcus neoformans. Cryptococcal disease may present in an insidious manner (sometimes as a skin infection before the development of clear CNS findings). Listeria meningitis may have an acute presentation and requires prompt therapy to avoid a fatal outcome. Patients who continue to take glucocorticoids are predisposed to infection. "Transplant elbow" is a recurrent bacterial infection in and around the elbow that is thought to result from a combination of poor tensile strength of the skin of steroid-treated patients and
steroid-induced proximal myopathy that requires patients to push themselves up with their elbows to get out of chairs. Bouts of cellulitis (usually caused by Staphylococcus aureus) recur until patients are provided with elbow protection. Kidney transplant recipients are susceptible to invasive fungal infections -- such as those due to Aspergillus and Rhizopus, which may present as superficial lesions before dissemination. Mycobacterial infection (particularly that with M. marinum) can be diagnosed by skin examination. Infection with Prototheca wickerhamii (an achlorophyllic alga) has been diagnosed by skin biopsy. Warts caused by human papillomaviruses (HPVs) are a late consequence of persistent immunosuppression; local therapy is usually satisfactory. HEART TRANSPLANTATION Early Infections Sternal wound infection and mediastinitis are early complications of heart transplantation. An indolent course is common, with fever or a mildly elevated white blood cell count preceding the development of site tenderness or drainage. Clinical suspicion based on evidence of sternal instability and failure to heal may lead to the diagnosis. Although common residents of the skin (e.g., S. aureus and S. epidermidis) as well as gram-negative organisms (e.g., Pseudomonas aeruginosa) and fungi (e.g., Candida) are often involved, mediastinitis in these patients (in rare cases) can also be due to Mycoplasma hominis (Chap. 178). Since this organism requires an anaerobic environment for growth and may be difficult to see on conventional medium, the laboratory should be alerted that M. hominis infection is suspected. M. hominis mediastinitis has been cured with a combination of surgical debridement (sometimes requiring muscle-flap placement) plus clindamycin and tetracycline. Organisms associated with mediastinitis may be cultured from accompanying pericardial fluid. Middle-Period Infections T. gondii (Chap. 217) resident in the heart of a seropositive donor may be transmitted to a seronegative recipient. Thus serologic screening for T. gondii infection is important before and in the months after cardiac transplantation. Rarely, active disease can be introduced at the time of transplantation. The overall incidence of toxoplasmosis is so high in this setting that some prophylaxis is warranted. Although alternatives are available, the most frequently used agent is trimethoprim-sulfamethoxazole, which prevents infection with Pneumocystis, Nocardia, and other bacterial pathogens.CMV has also been transmitted by heart transplantation.CNSinfections can be caused by Toxoplasma, Nocardia, and Aspergillus. L. monocytogenes meningitis should be considered in heart transplant recipients with fever and headache. CMVinfection is associated with poor outcomes after heart transplantation. The virus is usually cultivable 1 to 2 months after transplantation, causes manifestations (usually fever and atypical lymphocytosis, often associated with leukopenia and thrombocytopenia) at 2 to 3 months, and produces severe disease (e.g., pneumonia) at 3 to 4 months. Seropositive recipients usually develop cultivable virus faster than patients whose primary CMV infection is a consequence of transplantation. Between 40 and 70% of patients develop symptomatic CMV disease in the form of (1) CMV pneumonia, the most likely form of CMV disease to be fatal; (2) CMV esophagitis and gastritis, sometimes accompanied by abdominal pain with or without ulcerations and
bleeding; and (3) the CMV syndrome consisting of CMV in the blood with fever, leukopenia, thrombocytopenia, and hepatic enzyme abnormalities. Ganciclovir is efficacious in the treatment of CMV infection; prophylaxis with ganciclovir or possibly with other antivirals, as described for renal transplantation, may reduce the incidence of CMV-related disease. Late Infections EBVinfection usually presents as a lymphoma-like proliferation of B cells late after heart transplantation, particularly in patients maintained on heavy immunosuppression. A subset of heart and heart-lung transplant recipients may develop early (within 2 months) fulminant EBV-LPD. Treatment includes the reduction of immunosuppression if possible and the consideration of B cell antibodies (Rituximab), immunomodulatory agents, or chemotherapy, as discussed earlier under bone marrow/hematopoietic stem cell transplantation.HHV-8-associated disease, including primary effusion lymphoma, has been reported in heart transplant recipients. Prophylaxis for P. carinii infection is required for these patients (see below). LUNG TRANSPLANTATION Early Infections It is not surprising that lung transplants are predisposed to the development of pneumonia. The combination of ischemia and the resulting mucosal damage together with accompanying denervation and lack of lymph drainage probably contributes to the high rate of pneumonia (66% in one series). The prophylactic use of high doses of broad-spectrum antibiotics for the first 3 or 4 days after surgery decreases the incidence of pneumonia. Gram-negative pathogens (Enterobacteriaceae and Pseudomonas species) are troublesome in the first 2 weeks after surgery (the period of maximal vulnerability). Pneumonia can also be caused by Candida (possibly as a result of colonization of the donor lung), Aspergillus, and Cryptococcus. Mediastinitis may occur at an even higher rate among lung transplant recipients than among heart transplant recipients and most commonly develops within 2 weeks of surgery. Pneumonitis due to CMV(which may be transmitted as a consequence of transplantation) usually presents between 2 weeks and 3 months after surgery, with primary disease occurring later than reactivation disease. Middle-Period Infections The incidence ofCMVinfection, either reactivated or primary, is between 75 and 100% if either the donor or the recipient is seropositive for CMV. CMV-induced disease appears to be most severe in recipients of lung and heart-lung transplants. Whether this severity relates to the mismatch in lung antigen-presenting and host immune cells or is attributable to other (nonimmune) factors is not known. More than half of lung transplant recipients with symptomatic CMV disease have pneumonia. Difficulty in distinguishing the radiographic picture of CMV infection from organ rejection further complicates therapy. CMV can also cause bronchiolitis obliterans in lung transplants. The development of pneumonitis related toHSV has led to the prophylactic use of acyclovir. Such prophylaxis may also decrease rates of CMV disease, but ganciclovir is more active against CMV and is also active against HSV. Ganciclovir prophylaxis for CMV disease in lung transplant recipients is recommended. Late Infections The incidence of P. carinii infection (which may present with a paucity of findings) is high among lung and heart-lung transplant recipients. Some form of
prophylaxis for P. carinii pneumonia is indicated in all organ transplant situations (Tables 136-4 and136-5). Trimethoprim-sulfamethoxazole prophylaxis for 12 months after transplantation may be sufficient to prevent P. carinii disease in patients whose degree of immunosuppression is not increased. As in other transplant recipients, infection withEBV may cause either a mononucleosis-like syndrome orLPD. The tendency of the B cell blasts to present in the lung appears to be greater after lung transplantation than after the transplantation of other organs. Reduction of immunosuppression causes remission in some cases, but airway compression can be fatal and more rapid intervention may therefore become necessary. The approach to EBV-LPD is similar to that described in other sections. LIVER TRANSPLANTATION Early Infections As in other types of transplantation, early bacterial infections are a major problem after liver transplantation. Many centers administer systemic broad-spectrum antibiotics for the first 5 days after surgery, even in the absence of documented infection. However, despite prophylaxis, infectious complications are common and are correlated with the duration of the surgical procedure and the type of biliary drainage. An operation lasting>12 h is associated with an increased likelihood of infection. Patients who have a choledochojejunostomy with drainage of the biliary duct to a Roux-en-Y jejunal bowel loop have more fungal infections than those whose bile is drained via a choledochocholedochostomy with anastomosis of the donor common bile duct to the recipient common bile duct. Peritonitis and intraabdominal abscesses are common complications of liver transplantation. Bacterial peritonitis may result from biliary leaks and primary or secondary infection after leakage of bile. Peritonitis in liver transplant recipients is often polymicrobial, commonly involving enterococci, aerobic gram-negative bacteria, staphylococci, anaerobes, or Candida. Only one-third of patients with intraabdominal abscesses have bacteremia. Abscesses within the first month after surgery may occur not only over the liver but also in the spleen, pericolic area, and pelvis. Treatment includes antibiotic administration and drainage as necessary. Liver transplant patients have a high incidence of fungal infections, and the occurrence of fungal infection (often candidiasis) correlates with preoperative use of glucocorticoids, a long duration of treatment with antibacterial agents, and posttransplantation use of immunosuppressive agents. Middle-Period Infections The development of postsurgical biliary stricture predisposes patients to cholangitis. These patients may lack the characteristic signs and symptoms of cholangitis: fever, abdominal pain, and jaundice. Alternatively, these findings may be present but may suggest graft rejection. The diagnosis of cholangitis in liver transplant recipients therefore requires documentation of bacteremia or demonstration of aggregated neutrophils in bile duct biopsy specimens. Unfortunately, invasive studies of the biliary tract (either T-tube cholangiography or endoscopic retrograde cholangiopancreatography) may themselves lead to cholangitis. For this reason, many clinicians recommend prophylaxis with antibiotics covering gram-negative organisms and anaerobes when these procedures are performed in liver transplant recipients.
Viral hepatitis is a common complication of liver transplantation (Chap. 295). Reactivation of hepatitis B and C infections, for which transplantation may be performed, is problematic. To prevent hepatitis B infection, high-dose intravenous hepatitis B immune globulin is often administered. The long-term efficacy of lamivudine (3TC) in inhibiting hepatitis B viral replication after transplantation is being studied. A combination of interferon a and ribavirin is being tested for treatment/prophylaxis of hepatitis C infection. As in other transplantation settings, reactivation disease with herpes-group viruses is common (Table 136-2). Herpesviruses can be transmitted in donor organs. AlthoughCMVhepatitis occurs in ~4% of liver transplant recipients, it is usually not so severe as to require retransplantation. CMV disease develops in the majority of seronegative recipients of organs from CMV-positive donors, but fatality rates are lower in liver transplant recipients than in lung or heart-lung transplant recipients. Disease due to CMV is associated with the vanishing bile duct syndrome after liver transplantation. Patients respond to treatment with ganciclovir; prophylaxis with CMV immune globulin and acyclovir or oral ganciclovir may modify disease. A role forHHV-6 in posttransplantation fever and leukopenia has been proposed.EBV-LPDafter liver transplantation shows a propensity for involvement of the liver, and such disease may be of donor origin. PANCREAS TRANSPLANTATION Transplantation of the pancreas is complicated by early abdominal infection in ~20 to 40% of cases. To prevent contamination of the allograft with enteric bacteria and yeasts, some surgeons, instead of draining the pancreas through the bowel, drain secretions into the urinary tract or bladder. A cuff of duodenum is often used in the anastomosis between the pancreatic graft and the bladder. In addition to bicarbonate loss, this technique causes a high rate of urinary tract infection (30 to 40%) and sterile cystitis. Over the long term, bowel drainage is better tolerated. An alternative method -- the transplantation of islet cells only -- may eliminate the problems characteristically posed by wound and urinary tract sepsis in pancreas transplant recipients. Issues related to the development ofCMVinfection,EBV-LPD, and infections with opportunistic pathogens in patients receiving a pancreas are similar to those in other solid organ transplant recipients. MISCELLANEOUS INFECTIONS IN SOLID ORGAN TRANSPLANTATION Indwelling Intravenous Catheter Infections The prolonged use of indwelling intravenous catheters for administration of medication, blood products, and nutrition is common in diverse transplantation settings and poses a risk of local and bloodstream infection. Significant insertion-site infection is most commonly caused by S. aureus. Bloodstream infection most frequently develops within a week of catheter placement or in patients who become neutropenic. Coagulase-negative staphylococci are the most common isolates from the blood.*For further discussion of differential diagnosis and therapeutic options, see Chap. 85.
Tuberculosis The incidence of tuberculosis occurring within 12 months after solid organ transplantation ranges broadly worldwide (0.35 to 15%), reflecting prevalences in local populations. Nonrenal transplantation,GVHDwithin 6 months, and intensity of immunosuppression are predictive of tuberculosis reactivation and development of disseminated disease in a host with latent disease. Tuberculosis has rarely been transmitted from the donor organ. In contrast to the low mortality inBMT/HSCTrecipients, mortality in solid organ transplant patients is reported to be 29%. Isoniazid toxicity has not been a significant problem except in the liver transplantation setting. Virus-Associated Malignancies In addition to malignancy associated with gammaherpesvirus infection (EBV,HHV-8) and simple warts (HPV), transplant recipients, particularly those who require long-term immunosuppression, are more likely than the general population to develop tumors that are virus-associated or suspected of being virus-associated. The interval to tumor development is usually >1 year. Transplant recipients develop nonmelanoma skin or lip cancers that, in contrast to de novo skin cancers, have a high squamous cell-to-basal cell ratio. Whether HPV plays a major role in these lesions is being investigated. Cervical and vulvar carcinomas, quite clearly associated with HPV, develop with increased frequency in female transplant recipients. In renal transplant recipients, rates of melanoma are modestly increased and rates of cancers of the kidney and bladder are increased. VACCINATION OF TRANSPLANT RECIPIENTS In addition to receiving antibiotic prophylaxis, transplant recipients should be vaccinated against likely pathogens (Table 136-6). In the case ofBMTrecipients, optimal responses cannot be achieved until after reconstitution, despite previous immunization of both donor and recipient. Recipients of allogeneic BMTs must be reimmunized if they are to be protected against pathogens. The situation is less clear-cut in the case of autologous transplantation. T and B cells in the peripheral blood may reconstitute the response if they are transferred in adequate numbers. However, cancer patients (particularly those with Hodgkin's disease, in whom vaccination has been extensively studied) who are undergoing chemotherapy do not respond normally to immunization, and titers of antibodies to infectious agents fall more rapidly than in healthy individuals. Therefore, even immunosuppressed patients who have not had marrow transplants may need booster vaccine injections. If memory cells are specifically eliminated as part of a marrow "cleanup" procedure, it will be necessary to reimmunize the recipient with a new primary series. Optimal times for immunizations of different transplant populations are being evaluated. Immunization of household and other contacts (including health care personnel) against influenza every season is likely to benefit the patient by preventing local spread. In the absence of compelling data as to optimal timing, it is reasonable to administer the pneumococcal and H. influenzae type b conjugate vaccines to both autologous and allogeneicBMTrecipients 12 months after transplantation and again 12 months later (since the response to the initial vaccine dose is weak in the early posttransplantation period). These two vaccines are particularly important for patients who have undergone splenectomy. In addition, Neisseria meningitidis polysaccharide vaccine, diphtheria vaccine, tetanus vaccine, and inactivated polio vaccine can all be given at these same
intervals (12 and 24 months after transplantation). Some authorities recommend a new primary series for tetanus/diphtheria and inactivated polio vaccine (vaccination 12, 14, and 16 months after transplantation). Because of the risk of spread, household contacts of BMT recipients (or of patients immunosuppressed as a result of chemotherapy) should receive only inactivated polio vaccine. Live-virus measles/mumps/rubella (MMR) vaccine can be given to autologous BMT recipients 24 months after transplantation and to most allogeneic BMT recipients at the same point if they are not receiving maintenance therapy with immunosuppressive drugs and do not have ongoingGVHD. The risk of spread from a household contact is lower for MMR than for polio vaccine. In patients who have active GVHD and/or are taking high maintenance doses of glucocorticoids, it may be prudent to avoid all live-virus vaccines. In the absence of detectable antibody titers, vaccination to prevent hepatitis B and hepatitis A also seems advisable. In the case of solid organ transplant recipients, administration of all the usual vaccines and of the indicated booster doses should be completed before immunosuppression, if possible, to maximize responses. For patients taking immunosuppressive agents, the administration of pneumococcal vaccine should be repeated every 5 years. No data are available for meningococcal polysaccharide vaccine, but it is probably reasonable to administer it along with the pneumococcal vaccine or more frequently (every 3 years for persons with significant exposure risk). H. influenzae conjugate vaccine is safe and should be efficacious in this population; therefore, its administration before transplantation is recommended. Booster doses of this vaccine are not recommended for adults. Solid organ transplant recipients who continue to receive immunosuppressive drugs (glucocorticoids, cyclosporine) should not receive live-virus vaccines. A person in this group exposed to measles should be given immune globulin. Similarly, an immunocompromised patient who is seronegative for varicella and who comes into contact with a person who has chickenpox should be given varicella-zoster immune globulin as soon as possible (and certainly within 96 h) or, if this is not possible, started immediately on a 10- to 14-day course of acyclovir therapy. Susceptible household contacts of transplant recipients should receive live attenuatedVZVvaccine, but vaccinees should avoid direct contact with the patient if a rash develops. Immunocompromised patients who travel may benefit from some but not all vaccines. In general, they should receive any killed or inactivated vaccine preparation appropriate to the area they are visiting; this recommendation includes the vaccines for Japanese encephalitis, hepatitis A and B, poliomyelitis, meningococcal infection, and typhoid. The live typhoid vaccines are not recommended for use in most immunocompromised patients, but inactivated typhoid or the purified polysaccharide vaccine can be used. Live yellow fever vaccine should not be administered. Phenol-inactivated cholera vaccine is probably of little use in this setting. On the other hand, immunization with the purified-protein hepatitis B vaccine is indicated if patients are likely to be exposed. Patients who will reside for>6 months in areas where hepatitis B is common (Africa, Southeast Asia, the Middle East, Eastern Europe, parts of South America, and the Caribbean) should receive hepatitis B vaccine. Inactivated hepatitis A vaccine should be used in the appropriate setting (Chap. 122). If hepatitis A vaccine is not administered, travelers should consider receiving passive protection with immune globulin (the dose depending on the duration of travel in the high-risk area).
(Bibliography omitted in Palm version) Back to Table of Contents
SECTION 4 -APPROACH TO THERAPY FOR BACTERIAL DISEASES 137. TREATMENT AND PROPHYLAXIS OF BACTERIAL INFECTIONS - Gordon L. Archer, Ronald E. Polk The development of drugs that prevent and cure bacterial infections is one of the twentieth century's major contributions to human longevity and quality of life. Antibacterial agents are among the most commonly prescribed drugs of any kind worldwide. Used appropriately, these drugs are lifesaving. However, their indiscriminate use drives up the cost of health care, leads to a plethora of side effects and drug interactions, and fosters the emergence of bacterial resistance, rendering previously valuable drugs useless. The rational use of antibacterial agents depends on an understanding of their mechanisms of action, pharmacokinetics, toxicities, and interactions; bacterial strategies for resistance; and bacterial susceptibility in vitro. In addition, patient-associated parameters, such as the site of infection and the immune and excretory status of the host, are critically important to appropriate therapeutic decisions. This chapter provides specific data required for making an informed choice of antibacterial agent. MECHANISMS OF ACTION Antibacterial agents, like all antimicrobial drugs, are directed against unique targets not present in mammalian cells. The goal is to limit toxicity to the host and maximize chemotherapeutic activity affecting invading microbes only. The mechanisms of action of the antibacterial agents to be discussed in this section are summarized in Table 137-1 and are depicted inFig. 137-1. INHIBITION OF CELL-WALL SYNTHESIS One major difference between bacterial and mammalian cells is the presence in bacteria of a rigid wall external to the cell membrane. The wall protects bacterial cells from osmotic rupture, which would result from the cell's usual marked hyperosmolarity (by up to 20 atm) relative to the host environment. The structure conferring cell-wall rigidity and resistance to osmotic lysis in both gram-positive and -negative bacteria is peptidoglycan, a large, covalently linked sacculus that surrounds the bacterium. In gram-positive bacteria, peptidoglycan is the only layered structure external to the cell membrane and is thick (20 to 80 nm); in gram-negative bacteria, there is an outer membrane external to a very thin (1-nm) peptidoglycan layer. Chemotherapeutic agents directed at any stage of the synthesis, export, assembly, or cross-linking of peptidoglycan lead to inhibition of bacterial cell growth and, in most cases, to cell death. Peptidoglycan is composed of (1) a backbone of two alternating sugars, N-acetylglucosamine and N-acetylmuramic acid; (2) a chain of four amino acids that extends down from the backbone (stem peptides); and (3) a peptide bridge that cross-links the peptide chains. Peptidoglycan is formed by the addition of subunits (a sugar with its five attached amino acids) that are assembled in the cytoplasm and transported through the cytoplasmic membrane to the cell surface. Subsequent cross-linking is driven by cleavage of the terminal stem-peptide amino acid. Antibacterial agents act to inhibit cell-wall synthesis in several ways, as described below.
Bacitracin, a cyclic peptide antibiotic, inhibits the conversion to its active form of the lipid carrier that moves the water-soluble cytoplasmic peptidoglycan subunits through the cell membrane to the cell exterior. Cell-wall subunits accumulate in the cytoplasm and cannot be added to the growing peptidoglycan chain. Glycopeptides (vancomycin and teicoplanin) are high-molecular-weight antibiotics that bind to the terminal D-alanine-D-alanine component of the stem peptide while the subunits are external to the cell membrane but still linked to the lipid carrier. This binding sterically inhibits the addition of subunits to the peptidoglycan backbone. b-Lactam antibiotics (penicillins, cephalosporins, carbapenems, and monobactams;Table 137-2), characterized by a four-memberedb-lactam ring, prevent the cross-linking reaction called transpeptidation. The energy for attaching a peptide cross-bridge from the stem peptide of one peptidoglycan subunit to another is derived from the cleavage of a terminal D-alanine residue from the subunit stem peptide. The cross-bridge amino acid is then attached to the penultimate D-alanine by transpeptidase enzymes. The b-lactam ring of the antibiotic forms an irreversible covalent acyl bond with the transpeptidase enzyme (probably because of the antibiotic's steric similarity to the enzyme's D-alanine-D-alanine target), preventing the cross-linking reaction. Transpeptidases and similar enzymes involved in cross-linking are called penicillin-binding proteins (PBPs) because they all have active sites that bind b-lactam antibiotics. Virtually all the antibiotics that inhibit bacterial cell-wall synthesis are bactericidal. That is, they eventually result in the cell's death due to osmotic lysis. However, much of the loss of cell-wall integrity following treatment with cell wall-active agents is due to the bacteria's own cell-wall remodeling enzymes (autolysins) that cleave peptidoglycan bonds in the normal course of cell growth. In the presence of antibacterial agents that inhibit cell-wall growth, autolysis proceeds without normal cell-wall repair; weakness and eventual cellular lysis occur. INHIBITION OF PROTEIN SYNTHESIS Most of the antibacterial agents that inhibit protein synthesis interact with the bacterial ribosome. The difference between the composition of bacterial and mammalian ribosomes gives these compounds their selectivity. Aminoglycosides (gentamicin, kanamycin, tobramycin, streptomycin, netilmicin, neomycin, and amikacin) are a group of structurally related compounds containing three linked hexose sugars. They exert a bactericidal effect by binding irreversibly to the 30S subunit of the bacterial ribosome and blocking initiation of protein synthesis. The reason for the lethal effect of aminoglycosides (as opposed to the largely bacteriostatic effect of other protein synthesis-inhibiting antibacterial drugs, including the macrolides, the lincosamides, chloramphenicol, and tetracycline) is not completely understood. Uptake of aminoglycosides and their penetration through the cell membrane constitute an aerobic, energy-dependent process. Thus, aminoglycoside activity is markedly reduced in an anaerobic environment. Spectinomycin, an aminocyclitol antibiotic, also acts on the 30S ribosomal subunit but has a different mechanism of action from the
aminoglycosides and is bacteriostatic rather than bactericidal. Macrolides (erythromycin, clarithromycin, and azithromycin) consist of a large lactone ring to which sugars are attached. They bind specifically to the 50S portion of the bacterial ribosome. After attachment of mRNA to the initiation site of the 30S ribosomal subunit (the process blocked by aminoglycosides), the 50S subunit becomes bound to the 30S component to form the 70S ribosomal complex, and protein chain elongation proceeds. Binding of macrolides to the 50S ribosomal subunit inhibits protein chain elongation. Lincosamides (clindamycin and lincomycin), although structurally unrelated to macrolides, bind to a site on the 50S ribosome nearly identical to the binding site for macrolides. Although the mechanism and site of action of macrolides and lincosamides are similar, the number and types of bacteria against which these two groups of agents are active differ. Chloramphenicol consists of a single aromatic ring and a short side chain. This antibiotic binds reversibly to the 50S portion of the bacterial ribosome at a site close to but not identical with the binding sites of the macrolides and lincosamides. The ribosomal binding of chloramphenicol inhibits peptide bond formation. Tetracyclines (tetracycline, doxycycline, and minocycline) consist of four aromatic rings with various substituent groups. They interact reversibly with the bacterial 30S ribosomal subunit, blocking the binding of aminoacyl tRNA to the mRNA-ribosome complex. This mechanism is markedly different from that of the aminoglycosides, which also bind to the 30S subunit. The specificity of tetracyclines for bacteria depends both on their selectivity for bacterial ribosomes and on their requirement for active, energy-dependent transport into the bacterial cell by a system not found in mammalian cell membranes. Mupirocin (pseudomonic acid) is produced by the bacterium Pseudomonas fluorescens. Its mechanism of action is unique in that it inhibits the enzyme isoleucine tRNA synthetase by competing with bacterial isoleucine for its binding site on the enzyme. Inhibition of this enzyme depletes cellular stores of isoleucine-charged tRNA and therefore leads to a cessation of protein synthesis. The antibiotic is selective for bacteria because mammalian isoleucine tRNA synthetase lacks affinity for the compound. INHIBITION OF BACTERIAL METABOLISM The antimetabolites are all synthetic compounds that interfere with bacterial synthesis of folic acid. Products of the folic acid synthesis pathway function as coenzymes for the one-carbon transfer reactions that are essential for the synthesis of thymidine, all purines, and several amino acids. Inhibition of folate synthesis leads to cessation of bacterial cell growth and, in some cases, to bacterial cell death. The principal antibacterial antimetabolites are sulfonamides (sulfisoxazole, sulfadiazine, and sulfamethoxazole) and trimethoprim. Sulfonamides are structural analogues of p-aminobenzoic acid (PABA), one of the three structural components of folic acid (the other two being pteridine and glutamate). The
first step in the synthesis of folic acid is the addition of PABA to pteridine by the enzyme dihydropteroic acid synthetase. Sulfonamides compete with PABA as substrates for the enzyme. The selective effect of sulfonamides is due to the fact that bacteria synthesize folic acid, while mammalian cells cannot synthesize the cofactor and must have exogenous supplies. However, the activity of sulfonamides can be greatly reduced by the presence of excess PABA or by the exogenous addition of end products of one-carbon transfer reactions (e.g., thymidine and purines). High concentrations of the latter substances may be present in some infections as a result of tissue and white cell breakdown, compromising sulfonamide activity. Trimethoprim is a diaminopyrimidine, a structural analogue of the pteridine moiety of folic acid. It is a competitive inhibitor of dihydrofolate reductase; this enzyme is responsible for reduction of dihydrofolic acid to tetrahydrofolic acid, the essential final component in the folic acid synthesis pathway that is necessary for all one-carbon transfer reactions. Like the sulfonamides, trimethoprim is bactericidal in the absence of thymine but is only bacteriostatic when this pyrimidine is present in high concentration. The selective antibacterial activity of trimethoprim is based on the extreme sensitivity of bacterial dihydrofolate reductase to inhibition by this drug in comparison with the mammalian enzyme. The bacterial enzyme is approximately 50,000 times more sensitive to such inhibition. INHIBITION OF NUCLEIC ACID SYNTHESIS OR ACTIVITY Numerous antibacterial compounds have disparate effects on nucleic acids. The quinolones, including nalidixic acid and its fluorinated derivatives (norfloxacin, ciprofloxacin, ofloxacin, levofloxacin, sparfloxacin, grepafloxacin, and trovafloxacin), are synthetic compounds that inhibit the activity of the A subunit of the bacterial enzyme DNA gyrase as well as topoisomerase IV. DNA gyrase and topoisomerases are responsible for negative supercoiling of DNA, an essential conformation for DNA replication in the intact cell. Inhibition of the activity of DNA gyrase and topoisomerase IV is lethal to bacterial cells. The antibiotic novobiocin also interferes with the activity of DNA gyrase, but it interferes with the B subunit. Rifampin, used primarily against Mycobacterium tuberculosis, is also active against a variety of other bacteria. Rifampin binds tightly to the B subunit of bacterial DNA-dependent RNA polymerase, thus inhibiting transcription of DNA into RNA. Mammalian-cell RNA polymerase is not sensitive to the compound. Nitrofurantoin, a synthetic compound, causes DNA damage. The nitrofurans, compounds containing a single five-membered ring, are reduced by a bacterial enzyme to highly reactive, short-lived intermediates that are thought to cause DNA strand breakage, either directly or indirectly. Metronidazole, a synthetic imidazole, is active against a wide range of anaerobic bacteria and protozoa. This activity is totally dependent on the organism's anaerobic electron-transport system for energy production. In the presence of this system, the nitro group of metronidazole is reduced to a series of transiently produced, reactive intermediates that are thought to cause DNA damage. The unique redox system of anaerobes accounts for the selective antibacterial activity of metronidazole. This
compound is also a mutagen and a radiosensitizer of hypoxic mammalian cells. ALTERATION OF CELL-MEMBRANE PERMEABILITY The polymyxins (polymyxin B and colistin, or polymyxin E) are cyclic, basic polypeptides. They behave as cationic, surface-active compounds that disrupt the permeability of both the outer and the cytoplasmic membranes of gram-negative bacteria. Gramicidin A is a polypeptide of 15 amino acids that acts as an ionophore, forming pores or channels in lipid bilayers. MECHANISMS OF RESISTANCE Some bacteria have intrinsic resistance to certain classes of antibacterial agents (e.g., obligate anaerobic bacteria to aminoglycosides and gram-negative bacteria to vancomycin). Clearly these agents can never be used alone in the treatment of infections caused by resistant bacteria. In addition, bacteria that are ordinarily susceptible to antibacterial agents can acquire resistance. Acquired resistance is one of the major limitations to effective antibacterial chemotherapy. Resistance can develop by mutation of resident genes or by acquisition of new genes. New genes mediating resistance are usually spread from cell to cell by way of mobile genetic elements such as plasmids, transposons, and bacteriophages. The resistant bacterial populations flourish in areas of high antimicrobial use, where they enjoy a selective advantage over susceptible populations. The major mechanisms used by bacteria to resist the action of antimicrobial agents are inactivation of the compound, alteration or overproduction of the antibacterial target through mutation of the target protein's gene, acquisition of a new gene that encodes a drug-insensitive target, decreased permeability of the cell envelope to the agent, and active elimination of the compound from the periplasm or interior of the cell. Specific mechanisms of bacterial resistance to the major antibacterial agents are outlined below, summarized inTable 137-1, and depicted inFig. 137-1. b-LACTAMS Bacteria develop resistance to b-lactam antibiotics by a variety of mechanisms. Most common is the destruction of the drug byb-lactamases. Theb-lactamases of gram-negative bacteria are confined to the periplasm, between the inner and outer membranes, while gram-positive bacteria secrete their b-lactamases into the surrounding medium. These enzymes have a higher affinity for the antibiotic than the antibiotic has for its target. Binding results in hydrolysis of theb-lactam ring. Genes encodingb-lactamases have been found in both chromosomal and extrachromosomal locations and in both gram-positive and -negative bacteria; these genes are often on mobile genetic elements. One strategy that has been devised for circumventing resistance mediated byb-lactamases is to combine the susceptibleb-lactam with an inhibitor that avidly binds the inactivating enzyme, preventing its attack on the antibiotic. Unfortunately, the inhibitors (e.g., clavulanic acid, sulbactam, and tazobactam) do not bind all classes of b-lactamase and thus cannot be depended on to prevent the
inactivation of b-lactam antibiotics by such enzymes. No b-lactam antibiotic or inhibitor has been produced that can resist all of the manyb-lactamases that have been identified. A second mechanism of bacterial resistance to b-lactam antibiotics is an alteration inPBPtargets so that the PBPs have a markedly reduced affinity for the drug. While this alteration may occur by mutation of existing genes, the acquisition of new PBP genes (as in staphylococcal resistance to methicillin) or of new pieces of PBP genes (as in streptococcal, gonococcal, and meningococcal resistance to penicillin) is more important. A final resistance mechanism is the coupling, in gram-negative bacteria, of a decrease in outer-membrane permeability with rapid efflux of the antibiotic from the periplasm to the cell exterior. Mutations of genes encoding outer-membrane proteins called porins decrease the entry ofb-lactam antibiotics into the cell, while additional proteins form channels that actively pump b-lactams out of the cell. Resistance of Enterobacteriaceae to some cephalosporins and resistance of Pseudomonas spp. to cephalosporins and ureidopenicillins are the best examples of this mechanism. VANCOMYCIN Clinically important resistance to vancomycin was first described among enterococci in France in 1988. Vancomycin-resistant enterococci have subsequently become disseminated worldwide. The genes encoding resistance are carried on plasmids that can transfer themselves from cell to cell. Resistance is mediated by enzymes that substitute D-lactate for D-alanine on the peptidoglycan stem peptide so that there is no longer an appropriate target for vancomycin binding. This alteration does not appear to affect cell-wall integrity, however. This type of acquired vancomycin resistance is so far confined to enterococci and is seen in Enterococcus faecium rather than in the more common pathogen E. faecalis. Most clinically important staphylococci (i.e., Staphylococcus aureus and S. epidermidis) remain susceptible. However, in 1996, an isolate of S. aureus recovered from an infected patient in Japan was shown to be eight times less susceptible to vancomycin than were usual isolates. Since that report, an additional three S. aureus isolates with intermediate susceptibility to vancomycin have been recovered from infected patients in the United States, as have numerous coagulase-negative staphylococci with reduced vancomycin susceptibility. These isolates have not acquired the genes that mediate vancomycin resistance in enterococci but are mutant bacteria with markedly thickened cell walls. These mutants were apparently selected in patients who were undergoing prolonged vancomycin therapy. AMINOGLYCOSIDES The most common aminoglycoside resistance mechanism is inactivation of the antibiotic. Aminoglycoside-modifying enzymes, usually encoded on plasmids, transfer phosphate, adenyl, or acetyl residues from intracellular molecules to hydroxyl or amino side groups on the antibiotic. The modified antibiotic is less active because of diminished binding to its ribosomal target. Modifying enzymes that can inactivate any of the available aminoglycosides have been found in both gram-positive and -negative bacteria.
A second aminoglycoside resistance mechanism that has been identified predominantly in clinical isolates of Pseudomonas aeruginosa is decreased antibiotic uptake, presumably due to alterations in the bacterial outer membrane. MACROLIDES AND LINCOSAMIDES Resistance in gram-positive bacteria, the usual target organisms for macrolides and lincosamides, is due to the production of an enzyme -- most commonly plasmid-encoded -- that methylates ribosomal RNA, interfering with binding of the antibiotics to their target. Methylation mediates resistance to erythromycin, clarithromycin, azithromycin, and clindamycin. Streptococci can also actively efflux these compounds. CHLORAMPHENICOL Most bacteria resistant to chloramphenicol produce a plasmid-encoded enzyme, chloramphenicol acetyltransferase, that inactivates the compound by acetylation. TETRACYCLINES The most common mechanism of tetracycline resistance in gram-negative bacteria is a plasmid-encoded active-efflux pump that is inserted into the cytoplasmic membrane and extrudes antibiotic from the cell. Resistance in gram-positive bacteria is due either to active efflux or to ribosomal alterations that diminish binding of the antibiotic to its target. Genes involved in ribosomal protection are found on mobile genetic elements. MUPIROCIN Although the topical compound mupirocin was relatively recently introduced into clinical use, resistance is already becoming widespread in some areas. The mechanism appears to be either mutation of the target isoleucine tRNA synthetase so that it is no longer inhibited by the antibiotic or plasmid-encoded production of a form of the target enzyme that binds mupirocin poorly. TRIMETHOPRIM AND SULFONAMIDES The most prevalent mechanism of resistance to trimethoprim and the sulfonamides in both gram-positive and -negative bacteria is the acquisition of plasmid-encoded genes that produce a new, drug-insensitive target -- specifically, an insensitive dihydrofolate reductase for trimethoprim and an altered dihydropteroate synthetase for sulfonamides. QUINOLONES Resistance to the newer fluoroquinolones emerged rapidly among Staphylococcus and Pseudomonas spp. after the introduction of these agents. The most common mechanism is the development of one or more mutations in target DNA gyrases and topoisomerase IV that prevent the antibacterial agent from interfering with the activity of the enzyme. Some gram-negative bacteria develop mutations that both decrease
outer-membrane porin permeability and cause active drug efflux from the cytoplasm. Mutations that result in active quinolone efflux are also found in gram-positive bacteria. RIFAMPIN Bacteria rapidly become resistant to rifampin by developing mutations in the B subunit of RNA polymerase that render the enzyme unable to bind the antibiotic. The rapid selection of resistant mutants is the major limitation to the use of this antibiotic against otherwise-susceptible staphylococci and requires that it be used in combination with another antistaphylococcal agent. MULTIPLE ANTIBIOTIC RESISTANCE The acquisition by one bacterium of resistance to multiple antibacterial agents is becoming increasingly common. The two major mechanisms are the acquisition of multiple unrelated resistance genes and the development of mutations in a single gene or gene complex that mediate resistance to a series of unrelated compounds. The construction of multiresistant strains by acquisition of multiple genes occurs by sequential steps of gene transfer and environmental selection in areas of high-level antimicrobial use. In contrast, mutations in a single gene can conceivably be selected in a single step. Bacteria that are multiresistant by virtue of the acquisition of new genes include hospital-associated gram-negative bacteria, enterococci, and staphylococci and community-acquired strains of salmonellae, gonococci, and pneumococci. Mutations that confer resistance to multiple unrelated antimicrobial agents occur in the genes encoding outer-membrane porins and efflux proteins of gram-negative bacteria. These mutations decrease bacterial intracellular and periplasmic accumulation ofb-lactams, quinolones, tetracycline, chloramphenicol, and trimethoprim. Multiresistant bacterial isolates pose increasing problems in U.S. hospitals; strains resistant to all available antibacterial chemotherapy have already been identified. PHARMACOKINETICS The pharmacokinetic profile of an antibacterial agent refers to concentrations in serum and tissue versus time and reflects the processes of absorption, distribution, metabolism, and excretion. Important characteristics include peak and trough serum concentrations and mathematically derived parameters such as half-life, clearance, and distribution volume. Pharmacokinetic information is useful for estimating the appropriate antibacterial dose and frequency of administration, for adjusting dosages in patients with impaired excretory capacity, and for comparing one drug with another.*For further discussion of basic pharmacokinetic principles, see Chap. 70. ABSORPTION Data on absorption can refer to oral, intramuscular, or intravenous administration. Oral Administration Most patients with infection are treated with oral antibacterial agents in the outpatient setting. Advantages of oral therapy over parenteral therapy include lower cost, generally fewer adverse effects (including complications of indwelling lines), and greater acceptance by patients. The percentage of an orally administered
antibacterial agent that is absorbed (i.e., the agent's bioavailability) ranges from as little as 10 to 20% (erythromycin and penicillin G) to nearly 100% (clindamycin, metronidazole, doxycycline, and trimethoprim-sulfamethoxazole). These differences in bioavailability are not clinically important as long as concentrations at the site of infection are sufficient to inhibit or kill the pathogen. However, therapeutic efficacy may be compromised when absorption is reduced as a result of physiologic or pathologic conditions (such as the presence of food for some drugs or the shunting of blood away from the gastrointestinal tract in patients with hypotension), drug interactions (such as that of quinolones and metal cations), or noncompliance. The oral route is usually used for patients with relatively mild infections in whom absorption is not thought to be compromised by the preceding conditions. In addition, the oral route can be used in more severely ill patients after they have responded to parenteral therapy. Intramuscular Administration Although the intramuscular route of administration usually results in 100% bioavailability, it is not as widely used in the United States as the oral and intravenous routes, in part because of the pain often associated with intramuscular injections and the relative ease of intravenous access in the hospitalized patient. Intramuscular injection may be suitable for specific indications requiring an "immediate" and reliable effect (e.g., with long-acting forms of penicillin, including benzathine and procaine, and with single doses of ceftriaxone for uncomplicated gonococcal infection). Intravenous Administration The intravenous route is appropriate when oral antibacterial agents are not effective against a particular pathogen, when bioavailability is uncertain, or when larger doses are required than are feasible with the oral route. After intravenous administration, bioavailability is 100%; serum concentrations are maximal at the end of the infusion. For many patients requiring long-term antimicrobial therapy, outpatient intravenous administration with the use of convenient portable pumps may be cost-effective and safe when oral therapy is not feasible. Alternatively, some oral antibacterial drugs such as fluoroquinolones are sufficiently active against Enterobacteriaceae to rival parenteral therapy; their use may allow the patient to return home from the hospital earlier or to avoid hospitalization entirely. DISTRIBUTION To be effective, an antibacterial agent must exceed the minimal concentration required to inhibit bacterial growth (MIC;Chap. 121). Serum concentrations usually exceed the MIC for susceptible bacteria, but since most infections are extravascular, the antibiotic must also distribute to the site of the infection. Concentrations of most antibacterials in interstitial fluid are similar to free drug concentrations in serum. However, when the infection is located in a "protected" site where penetration is poor, such as cerebrospinal fluid (CSF), the eye, the prostate, or infected cardiac vegetations, high parenteral doses or local administration for prolonged periods may be required for cure. In addition, even though an antibacterial agent may penetrate to the site of infection, its activity may be antagonized by factors in the local environment, such as an unfavorable pH or inactivation by cellular degradation products. For example, since the activity of aminoglycosides is reduced at acidic pH, the acidic environment in many infected tissues may be partly responsible for the relatively poor efficacy of aminoglycoside monotherapy. In addition, the abscess milieu reduces the activity of many antibacterial
compounds, so that surgical drainage may be required for cure. Most bacteria that cause human infections are located extracellularly. Intracellular pathogens such as Legionella, Chlamydia, Brucella, and Salmonella may persist or cause relapse if the antibacterial agent does not enter the cell. In general, b-lactams, vancomycin, and aminoglycosides penetrate cells poorly, whereas macrolides, tetracyclines, metronidazole, chloramphenicol, rifampin, trimethoprim-sulfamethoxazole, and quinolones penetrate cells well. METABOLISM AND ELIMINATION Like other drugs, antibacterial agents are disposed of by hepatic elimination (metabolism or biliary elimination), by renal excretion in unchanged or metabolized form, or by a combination of the two processes. For most antibacterial drugs, metabolism leads to loss of in vitro activity, although some agents, such as cefotaxime, rifampin, and clarithromycin, have bioactive metabolites that may contribute to their overall efficacy. The most practical application of knowing the mode of excretion of an antibacterial agent is adjustment of the dosage when elimination capability is impaired. Direct, nonidiosyncratic toxicity from antibacterial drugs most often results from failure to reduce the dosage appropriately in a patient with impaired elimination. For agents that are primarily cleared intact by glomerular filtration, drug clearance is linearly correlated with creatinine clearance. Unfortunately, for drugs whose elimination is primarily hepatic, no simple marker (such as serum creatinine) is useful for dosage adjustment in subjects with liver disease. Even in patients with severe hepatic disease, residual metabolic capability is usually sufficient to preclude accumulation and toxic effects. However, for drugs that undergo hepatic metabolism and have a narrow therapeutic index (such as chloramphenicol), alternative therapy may be warranted in patients with liver disease, since the technology for the monitoring of serum levels is not widely available. PRINCIPLES OF ANTIBACTERIAL CHEMOTHERAPY The choice of an antibacterial compound for a particular patient and a specific infection involves more than just a knowledge of the agent's mechanism of action and pharmacokinetic profile. The basic tenets of chemotherapy, to be elaborated below, include the following: First, material containing the infecting organism(s) should be obtained before the start of treatment so that presumptive identification can be made by microscopic examination of stained specimens and the organism can be grown for definitive identification and susceptibility testing. Second, once the organism is identified and its susceptibility to antibacterial agents is determined, the regimen with the narrowest effective spectrum should be chosen. Third, the choice of antibacterial agent is guided by the pharmacokinetic and adverse-reaction profile of active compounds, the site of infection, the immune status of the host, and evidence of efficacy from well-performed clinical trials. Finally, if all other factors are equal, the least expensive antibacterial regimen should be chosen. SUSCEPTIBILITY OF BACTERIA TO ANTIBACTERIAL DRUGS IN VITRO
The determination of the susceptibility of the patient's infecting organism to a panel of appropriate antibacterial agents is an essential first step in devising a chemotherapeutic regimen. The details of susceptibility testing are discussed elsewhere (Chap. 121). Such testing is designed to estimate the susceptibility of a bacterial isolate to an antibacterial drug under standardized conditions that favor rapidly growing aerobic or facultative organisms and to assess bacteriostasis only. Specialized testing is required for the assessment of bactericidal antimicrobial activity; for the detection of resistance among such fastidious organisms as obligate anaerobes, Haemophilus spp., and pneumococci; and for the determination of resistance phenotypes with variable expression, such as resistance to methicillin or oxacillin among staphylococci. RELATIONSHIP OF PHARMACOKINETICS AND IN VITRO SUSCEPTIBILITY TO CLINICAL RESPONSE The relationship between the report of susceptibility in vitro and the clinical pharmacokinetics of the antibacterial agent helps predict clinical response. Bacteria are usually considered to be susceptible to a drug if the achievable peak serum concentration exceeds theMIC by at least fourfold. The breakpoint is the concentration of the antibiotic that separates susceptible from resistant bacteria (Fig. 137-2). When a majority of the isolates of a given bacterial species are inhibited at concentrations below the breakpoint, the species is considered to be within the spectrum of the antibiotic (see "Choice of Antibacterial Therapy," below). The pharmacodynamic profile of an antibiotic is the quantitative relationship among the time course of antibiotic concentrations in serum and tissue, in vitro susceptibility, and microbial response. Three pharmacodynamic parameters quantify these relationships: the ratio of the area under the curve (AUC) for the plasma concentration vs. time curve toMIC(AUC/MIC), the ratio of the maximal serum concentration to the MIC (Cmax/MIC), and the time during a dosing interval that plasma concentrations exceed the MIC (t > MIC). The pharmacodynamic profile of an antibiotic class is characterized as either concentration dependent (fluoroquinolones, aminoglycosides), such that the increase in antibiotic concentration leads to a more rapid rate of bacterial death, or time dependent (b-lactams, vancomycin), such that the reduction in bacterial density is proportional to the time that concentrations exceed the MIC. For concentration-dependent antibiotics, the Cmax/MIC or AUC/MIC ratio correlates best with the reduction in microbial density in vitro and in animal investigations. Dosing strategies attempt to maximize these ratios by the administration of a "large" dose relative to the MIC for anticipated pathogens, often at "long" intervals (relative to the serum half-life). Once-daily dosing of aminoglycoside antibiotics is the practical consequence of these relationships. In contrast, dosage strategies for time-dependent antibiotics emphasize the administration of sufficient doses at appropriate intervals to maintain serum concentrations above the MIC, typically for at least 40 to 50% of the dosing interval. The clinical implications of these relationships are in the early stages of investigation, but their elucidation should eventually result in more rational antibacterial regimens. STATUS OF THE HOST Various host factors must be considered in the devising of antibacterial chemotherapy.
The host's antibacterial immune function is of importance, particularly as it relates to opsonophagocytic function. Since the major host defense against acute, overwhelming bacterial infection is the polymorphonuclear leukocyte, patients with neutropenia must be treated aggressively and empirically with bactericidal drugs for suspected infection (Chap. 85). Likewise, patients who have deficient humoral immunity (e.g., those with chronic lymphocytic leukemia and multiple myeloma) and individuals with surgical or functional asplenia (e.g., those with sickle cell disease) should be treated empirically for infections with encapsulated organisms, especially the pneumococcus. Pregnancy increases the risk of toxicity of certain antibacterial drugs for the mother (e.g., the hepatic toxicity of tetracycline), affects drug disposition and pharmacokinetics, and -- because of the risk of fetal toxicity -- severely limits the choice of agents for treating infections. Certain antibacterials are contraindicated in pregnancy either because their safety has not been established or because they are known to be toxic. These agents include all fluoroquinolones, clarithromycin, erythromycin estolate (but not erythromycin base), and tetracyclines. Data on the safety of many other antibacterial drugs are limited, but these drugs may be used cautiously when there is no suitable alternative and the perceived benefit outweighs the risk. These agents include the aminoglycosides, azithromycin, clindamycin, imipenem, metronidazole, trimethoprim, and vancomycin. Chloramphenicol, nitrofurantoin, and the sulfonamides are contraindicated in the third trimester but can be used cautiously in the first two trimesters. In patients with concomitant viral infections, the incidence of adverse reactions to antibacterial drugs may be unusually high. For example, persons with infectious mononucleosis and those infected with HIV may react more often to ampicillin and folic acid synthesis inhibitors, respectively. In addition, the patient's age, sex, racial heritage, and excretory status all determine the incidence and type of side effects that can be expected with certain antibacterial agents. SITE OF INFECTION The location of the infected site may play a major role in the choice and dose of antimicrobial drug. Patients with suspected meningitis should receive drugs that can cross the blood-CSFbarrier; in addition, because of the relative paucity of phagocytes and opsonins at the site of infection, the agents should be bactericidal. Chloramphenicol, one of the standard drugs used in the treatment of meningitis, is bactericidal for common organisms causing meningitis (i.e., meningococci, pneumococci, and Haemophilus influenzae, but not enteric gram-negative bacilli), is highly lipid-soluble, and enters the CSF well. However,b-lactams, the mainstay of therapy for most of these infections, do not normally reach high levels in CSF. Their efficacy is based on the increased permeability of the blood-brain and blood-CSF barriers to hydrophilic molecules during inflammation and the extreme susceptibility of most infectious organisms to even small amounts of b-lactam drug. The vegetation, which is the major site of infection in bacterial endocarditis, is also a focus that is protected from normal host-defense mechanisms. Antibacterial therapy needs to be bactericidal, with the selected agent administered parenterally over a long
period and at a dose that produces serum levels at least eight times higher than the minimal bactericidal concentration (MBC) for the infecting organism. Likewise, osteomyelitis involves a site that is somewhat resistant to opsonophagocytic removal of infecting bacteria; furthermore, avascular bone (sequestrum) represents a foreign body that thwarts normal host-defense mechanisms. Chronic prostatitis is exceedingly difficult to cure because most antibiotics do not penetrate the nonfenestrated capillaries serving the prostate, especially when acute inflammation is absent. Drugs that are "ion trapped" after entering prostatic tissue, such as trimethoprim and fluoroquinolones, may be uniquely effective because of this mechanism. Intraocular infections, especially endophthalmitis, are difficult to treat because drug penetration into the vitreous from blood is hindered by retinal capillaries lacking fenestration. Inflammation does little to disrupt this barrier. Thus, direct injection into the vitreous is necessary in many cases. Antibiotic penetration into abscesses is usually poor. Even when an antibiotic does penetrate into the abscess, local conditions, such as low pH or the presence of enzymes that hydrolyze the drug, may antagonize its activity. In contrast, urinary tract infections, when confined to the bladder, are relatively easy to cure, in part because of the higher concentration of most antibiotics in urine than in blood. Since blood is the usual reference fluid in defining susceptibility, even organisms found to be "resistant" to achievable serum concentrations may be susceptible to achievable urine concentrations. For drugs that are used only for the treatment of urinary tract infections, such as nitrofurantoin and methenamine salts, achievable urine concentrations are used to determine susceptibility. COMBINATION CHEMOTHERAPY One of the tenets of antibacterial chemotherapy is that if the infecting bacterium has been identified, the most specific chemotherapy possible should be used. The use of a single agent with a narrow spectrum of activity against the pathogen diminishes the alteration of normal flora and thus limits the overgrowth of resistant nosocomial organisms (e.g., Candida albicans, enterococci, Clostridium difficile, or methicillin-resistant staphylococci), avoids the potential toxicity of multiple-drug regimens, and reduces cost. However, certain circumstances call for the use of more than one antibacterial agent. These are summarized below. 1. Prevention of the emergence of resistant mutants. Spontaneous mutations occur at a detectable frequency in certain genes encoding the target proteins for some antibacterial agents. The use of these agents can eliminate the susceptible population, select out resistant mutants at the site of infection, and result in the failure of chemotherapy. Resistant mutants are usually selected when theMIC of the antibacterial agent for the infecting bacterium is close to achievable levels in serum or tissues and/or when the site of infection limits the access or activity of the agent. Among the most common examples are rifampin for staphylococci, imipenem for Pseudomonas, and ciprofloxacin for staphylococci and Pseudomonas. Small-colony variants of staphylococci resistant to aminoglycosides also emerge during monotherapy with these antibiotics. A second antibacterial agent with a mechanism of action different from that of the first is added to prevent the emergence of these resistant mutants (e.g., imipenem plus an aminoglycoside for systemic Pseudomonas infections). However, since resistant mutants have emerged following combination chemotherapy, this
approach is not uniformly successful. 2. Synergistic or additive activity. Against some bacteria, two or more agents are clearly more active than one; whether or not this is the case is usually judged on the basis of testing in vitro. Synergistic or additive activity involves a lowering of theMIC orMBC of each or all of the drugs tested in combination against a specific bacterium. In synergy, each agent is more active when combined with a second drug than it would be alone, and the drugs' combined activity is therefore greater than the sum of the individual activities of each drug. In an additive relationship, the combined activity of the drugs is equal to the sum of their individual activities. Among the best examples of a synergistic or additive effect, confirmed both in vitro and by animal studies, are the enhanced bactericidal activities of certainb-lactam/aminoglycoside combinations against enterococci, viridans streptococci, and P. aeruginosa. The synergistic or additive activity of these combinations has also been demonstrated for selected isolates of enteric gram-negative bacteria and staphylococci. The combination of trimethoprim and sulfamethoxazole has synergistic or additive activity against many enteric gram-negative bacteria. Most other antimicrobial combinations display indifferent activity (i.e., the combination is no better than the more active of the two agents alone), and some combinations (e.g., penicillin plus tetracycline against pneumococci) may be antagonistic (i.e., the combination is worse than either drug alone). 3. Therapy directed against multiple potential pathogens. For certain infections, either a mixture of pathogens is suspected or the patient is desperately ill with an as-yet-unidentified infection (see "Empirical Therapy," below). In these situations, the most important of the likely infecting bacteria must be covered by therapy until culture and susceptibility results become available. Examples of the former infections are intraabdominal or brain abscesses and infections of limbs in diabetic patients with microvascular disease. The latter situations include fevers in neutropenic patients, acute pneumonia from aspiration of oral flora by hospitalized patients, and septic shock or sepsis syndrome. EMPIRICAL THERAPY In certain situations, antibacterial therapy is begun before a specific bacterial pathogen has been identified. The choice of agent is guided by the results of studies identifying the usual pathogens at that site or in that clinical setting, by pharmacodynamic considerations, and by the resistance profile of the expected pathogens in a particular hospital or geographic area. Situations in which empirical therapy is appropriate include the following: 1. Life-threatening infection. Any suspected bacterial infection in a patient with a life-threatening illness should be treated presumptively. Therapy is usually begun with more than one agent and is later tailored to a specific pathogen if one is eventually identified. 2. Treatment of infections in unhospitalized patients with no cultures performed. In many situations, it is appropriate to treat non-life-threatening infections without obtaining cultures. These situations include outpatient infections such as community-acquired cases of pneumonia, cystitis, cellulitis or local wound infection, sinusitis, otitis, urethritis,
and prostatitis. However, if any of these infections recurs or fails to respond to initial therapy, every effort should be made to obtain cultures to guide re-treatment. CHOICE OF ANTIBACTERIAL THERAPY The antibacterial spectrum of specific agents and the infections for which they represent the treatment of choice are detailed below. No attempt has been made to include all the potential situations in which antibacterial agents may be used. A more detailed discussion of specific bacteria and infections that they cause can be found elsewhere in this volume. b-LACTAMS (Table 137-2) All penicillins (except for the semisynthetic, penicillinase-resistant antistaphylococcal agents) are hydrolyzed byb-lactamases and are ineffective against isolates that produce these enzymes. Penicillin G has a spectrum that includes spirochetes (Treponema pallidum, Borrelia, and Leptospira), streptococci (groups A and B, viridans, and Streptococcus pneumoniae), enterococci, most Neisseria spp., a few staphylococci, many fastidious oral bacteria (including many Porphyromonas and Prevotella spp., streptococci, Actinomyces, and Fusobacterium), Clostridium spp. (except C. difficile), Pasteurella multocida, Erysipelothrix rhusiopathiae, and Streptobacillus moniliformis. However, penicillin G resistance is widespread among staphylococci; is increasing rapidly among gonococci, enterococci, and pneumococci; and is emerging among meningococci, viridans streptococci, and oral anaerobes such as Porphyromonas and Prevotella. Penicillin G is the drug of choice for syphilis, yaws, leptospirosis, group A and B streptococcal infections, actinomycosis, oral and periodontal infections, meningococcal meningitis and meningococcemia, viridans streptococcal endocarditis, clostridial myonecrosis, tetanus, anthrax, rat-bite fever, P. multocida infections, and erysipeloid (E. rhusiopathiae). Ampicillin extends the spectrum of penicillin G to some gram-negative rods. It is active against some isolates of Escherichia coli, Proteus mirabilis, Salmonella, Shigella, and H. influenzae and is one of the drugs of choice for susceptible organisms causing urinary tract infections, salmonellosis, H. influenzae meningitis and epiglottitis, and Listeria monocytogenes meningitis. High rates of resistance have lessened its value as empirical therapy in some situations. For example, more than 80% of isolates of E. coli and P. mirabilis are resistant in some hospitals, as are 10 to 30% of isolates of H. influenzae; moreover, in some outbreaks of infection due to salmonellae, all isolates are ampicillin-resistant. The penicillinase-resistant penicillins are used solely for the treatment of staphylococcal infections and are the drugs of choice for systemic or deep staphylococcal infections caused by susceptible organisms. Unfortunately, on average, approximately 30% of S. aureus isolates and more than 60% of coagulase-negative staphylococcal isolates acquired in U.S. hospitals are resistant to these agents (i.e., methicillin-resistant). The spectrum of these agents also includes most of the same gram-positive bacteria that are susceptible to penicillin G. The spectrum of the antipseudomonal penicillins includes the bacteria covered by
ampicillin as well as some additional nonpseudomonal enteric gram-negative bacilli. For example, piperacillin is active against many indole-positive Proteus, Enterobacter, Klebsiella, Providencia, and Serratia spp. However, the susceptibility of these penicillins tob-lactamase markedly limits their utility as empirical therapy when infections caused by gram-negative enteric organisms are suspected. The major use of these compounds is in the treatment of proven or suspected infections with P. aeruginosa and Acinetobacter, for which they are among the drugs of choice. Their relative antipseudomonal activities can be ranked as follows: piperacillin >mezlocillin/ticarcillin >carbenicillin. The addition ofb-lactamase inhibitors (clavulanic acid, sulbactam, or tazobactam) to ampicillin, amoxicillin, ticarcillin, or piperacillin extends the spectrum of these agents to include many organisms that are resistant by virtue of b-lactamase production. These organisms include E. coli, Klebsiella spp., all Proteus spp., H. influenzae, Moraxella catarrhalis, Providencia spp., and Bacteroides spp. Such combinations are also active against staphylococci that produce b-lactamase but are not methicillin-resistant. However, the efficacy of these combinations in serious staphylococcal infections has not been adequately proven. Furthermore, Enterobacter, Pseudomonas, Acinetobacter, and various enteric gram-negative isolates either produceb-lactamases not inhibited by these compounds or develop resistance attributable to non-b-lactamase-mediated mechanisms. The first-generation cephalosporins have a spectrum that includes penicillinase-producing, methicillin-susceptible staphylococci and streptococci. While these drugs may be used when infections with gram-positive bacteria are suspected, they are not the drugs of choice for such infections. They have excellent activity against many isolates of E. coli, Klebsiella pneumoniae, and P. mirabilis and are among the drugs of choice in presumptive therapy for community-acquired urinary tract infections. They have no activity against Bacteroides fragilis, enterococci, methicillin-resistant staphylococci, Pseudomonas, Acinetobacter, Enterobacter, indole-positive Proteus, and Serratia and poor activity against H. influenzae. The parenteral second-generation cephalosporins extend the gram-negative spectrum of first-generation compounds. The various second-generation agents have differing activities. Cefuroxime and cefamandole retain activity against gram-positive cocci and are also active against H. influenzae, Neisseria, some Enterobacter isolates, and indole-positive Proteus but exhibit poor activity against B. fragilis. Cefoxitin and cefotetan have reasonably good activity against B. fragilis, but cefotetan is less effective against some other Bacteroides spp. (Chaps. 130 and167). Both of the latter drugs display poor activity against gram-positive cocci and Enterobacter. No second-generation cephalosporin is active against Pseudomonas or Acinetobacter. Oral second-generation cephalosporins have fair activity against gram-positive cocci and H. influenzae and are widely used in outpatient therapy for otitis media, sinusitis, and lower respiratory tract infections, although cheaper agents that are equally effective are preferable. Cefixime, cefuroxime axetil, and cefpodoxime are among the drugs of choice for single-dose treatment of gonococcal urethritis. Third-generation cephalosporins all have a broad spectrum of activity against enteric
gram-negative rods and are especially useful for treating hospital-acquired infections caused by multiresistant organisms. In addition, ceftazidime and cefepime have excellent antipseudomonal activity. The other third-generation cephalosporins have poor antipseudomonal activity. Since resistance to third-generation cephalosporins is increasing among all nosocomial gram-negative rods, the use of these agents should be guided by susceptibility testing. The gram-positive spectrum of the third-generation cephalosporins is variable. All are less active than first-generation cephalosporins against methicillin-susceptible staphylococci; ceftazidime has the least antistaphylococcal activity of this group. However, ceftriaxone, ceftizoxime, and cefotaxime have excellent activity against streptococci, especially S. pneumoniae. Ceftazidime is not recommended for treatment of streptococcal infections. Because of its excellent gram-negative spectrum; its activity against Haemophilus, many S. pneumoniae strains, and penicillin-resistant Neisseria; its long serum half-life; and its high serum andCSFlevels, ceftriaxone has become one of the drugs of choice for empirical therapy for bacterial meningitis (except that caused by Listeria and by highly penicillin-resistant pneumococcal strains), all gonococcal infections, salmonellosis, and typhoid fever. The third-generation cephalosporins are among the drugs of choice for nonpseudomonal hospital-acquired pneumonia. Cefepime is more resistant to chromosomal b-lactamase produced by Enterobacter spp. than are other third-generation cephalosporins. Third-generation cephalosporins have poor activity against Bacteroides and no activity against methicillin-resistant staphylococci, Enterococcus, Acinetobacter, or Stenotrophomonas. The carbapenems currently available in the United States are imipenem and meropenem. Imipenem is marketed in combination with the renal dipeptidase inhibitor cilastatin, which enables imipenem to escape renal inactivation and thus to reach higher urinary levels. Imipenem and meropenem have excellent activity in vitro against virtually all bacterial pathogens except Stenotrophomonas, methicillin-resistant staphylococci, and E. faecium. Limitations to their use are their relatively low blood levels, short serum half-life, and high cost. Imipenem has dose-related central nervous system side effects that appear to be less frequent with meropenem. Resistance to imipenem and meropenem is a problem only among nosocomial isolates of P. aeruginosa, approximately 20% of which are resistant. Because of their broad spectrum, carbapenems can be used as empirical therapy for serious nosocomial infections thought to be caused by multiple bacterial species or multiresistant organisms. Imipenem and meropenem are often used to treat hospital-acquired infections caused by Enterobacter spp. because these organisms produce inducibleb-lactamases that inactivate third-generation cephalosporins but not the carbapenems. The latter antibiotics are often held in reserve as therapy for nosocomial infections due to gram-negative pathogens resistant to third-generation cephalosporins. The only monobactam currently available is aztreonam. This antibiotic has a spectrum limited to facultative gram-negative enteric bacilli. It has no activity against any gram-positive or anaerobic bacterium. Its gram-negative spectrum is similar to that of ceftazidime, with equally good activity against Pseudomonas. Its primary advantages are its theoretical ability to preserve the normal gram-positive and anaerobic flora and the lack of cross-reactive immediate hypersensitivity in patients who have had this type of reaction to otherb-lactam antibiotics.
VANCOMYCIN The spectrum of vancomycin is limited to gram-positive cocci, especially enterococci, streptococci, and staphylococci. Vancomycin serves as second-line therapy for most gram-positive bacterial infections but is the drug of choice for infections caused by methicillin-resistant staphylococci or Corynebacterium jeikeium and for serious infections in penicillin-allergic patients. Given orally (a route by which it is not absorbed), vancomycin can be used to treat antibiotic-associated pseudomembranous colitis caused by C. difficile in patients who have failed to respond to metronidazole, the drug of choice. Vancomycin has also been recommended as initial empirical therapy for presumed pneumococcal meningitis because of increasing pneumococcal resistance to penicillins and cephalosporins. Resistance to vancomycin is increasing rapidly among isolates of E. faecium in large hospitals, particularly in areas of high vancomycin use. In addition, S. aureus isolates with reduced susceptibility to vancomycin have now been detected. Because of the growing threat of vancomycin-resistant enterococci and the potential for increasing resistance among staphylococci, a national advisory committee has established guidelines for appropriate and limited use of this antibiotic. AMINOGLYCOSIDES The aminoglycosides are rapidly bactericidal in vitro at low concentrations, with activity limited to facultative gram-negative bacteria and staphylococci. They have no activity against anaerobic bacteria and are not effective in environments that are acidic or have a low oxygen tension. However, their spectrum includes virtually all gram-negative bacteria that are not strict anaerobes, and they are among the drugs of choice for any suspected gram-negative bacteremic infection, particularly in neutropenic patients. Aminoglycosides are synergistically bactericidal in combination with a penicillin for the treatment of staphylococcal, enterococcal, or viridans streptococcal endocarditis and are usually combined with ab-lactam antibiotic for the treatment of gram-negative bacteremia. Aminoglycosides are also among the drugs of choice for severe infections of the upper urinary tract. The major limitations to use of aminoglycosides are their renal and otic toxicity, their diminished activity at certain sites of infection (e.g., abscesses and the central nervous system), and the resistance of target bacteria. Among the available agents, gentamicin is generally preferred because of its low cost; however, tobramycin has slightly greater activity against P. aeruginosa, and amikacin retains activity against many tobramycin- and gentamicin-resistant gram-negative bacteria because it is inactivated by fewer aminoglycoside-modifying enzymes. Streptomycin is still one of the drugs of choice in initial therapy for tularemia, plague, glanders, and brucellosis and is a second-line agent for the treatment of tuberculosis. MACROLIDES Erythromycin has broad-spectrum activity against gram-positive bacteria, with additional activity against Legionella, Mycoplasma, Campylobacter, and some Chlamydia isolates. It is the drug of choice for infections due to Legionella, Campylobacter, and Mycoplasma and is among the drugs of choice for community-acquired pneumococcal pneumonia and group A streptococcal pharyngitis in penicillin-allergic patients. However, resistance to erythromycin among group A streptococci and pneumococci is increasing
dramatically in some areas. Erythromycin also appears to be one of the drugs of choice for infections caused by the agent of bacillary angiomatosis (Bartonella henselae) in immunocompromised patients. The newer macrolides clarithromycin and azithromycin have an antibacterial spectrum similar to that of erythromycin in vitro. However, azithromycin has greater activity against Chlamydia. Clarithromycin, in combination with a proton pump inhibitor, has been designated a drug of choice for the treatment of gastric infections due to Helicobacter pylori (gastritis, gastric and duodenal ulcers). Both azithromycin and clarithromycin are active against nontuberculous mycobacteria, and both appear to have fewer gastrointestinal side effects than does erythromycin. LINCOSAMIDES The only lincosamide used in the United States is clindamycin. It shares the gram-positive coccal spectrum of erythromycin but is more active, in some cases showing bactericidal activity, against susceptible staphylococci. However, resistance among staphylococci and some streptococci, mediated by the same genes responsible for macrolide resistance, limits clindamycin's usefulness against gram-positive cocci. In general, all staphylococci resistant to erythromycin should be considered resistant to clindamycin regardless of the results of in vitro susceptibility testing. However, at least half of the streptococci resistant to erythromycin are truly susceptible to clindamycin. In these bacteria, resistance is mediated by a drug-efflux pump that removes macrolides but not lincosamides. Despite increasing resistance, clindamycin remains useful for most anaerobic infections because of its broad spectrum of activity against both gram-positive and -negative strict anaerobes. It is also a drug of choice for the treatment of severe, invasive group A streptococcal infections. In contrast, clindamycin, like erythromycin, has no clinically significant activity against facultative gram-negative enteric bacilli. Appropriate use is limited only by resistance or the development of pseudomembranous colitis, the major serious side effect of this drug. CHLORAMPHENICOL Chloramphenicol has a broad spectrum of activity against gram-positive and -negative bacteria, although plasmid-mediated resistance has diminished its effective spectrum. This antibiotic is rarely used in adult infections because of the rare idiosyncratic side effect of irreversible bone-marrow aplasia and the availability of other agents with similar activity. It remains one of the drugs of choice for the treatment of typhoid fever and plague and is still useful for the treatment of brucellosis and both pneumococcal and meningococcal meningitis in penicillin-allergic patients. TETRACYCLINES Tetracyclines have a broad spectrum of bacteriostatic activity against gram-positive and -negative bacteria and are widely used in a variety of community-acquired infections. These agents are among the drugs of choice for chronic bronchitis, granuloma inguinale, brucellosis (with streptomycin), tularemia, glanders, melioidosis, spirochetal infections caused by Borrelia (Lyme disease and relapsing fever; doxycycline), infections caused by Vibrio vulnificus, some Aeromonas infections, infections due to Stenotrophomonas (minocycline), plague, and ehrlichiosis (doxycycline). The tetracyclines are also used in penicillin-allergic patients for the treatment of
leptospirosis, syphilis, actinomycosis, and skin and soft-tissue infections caused by gram-positive cocci. They are among the drugs of choice for infections due to chlamydiae (doxycycline), rickettsiae, and ehrlichiae and for granulomatous skin infection due to Mycobacterium marinum (minocycline). Doxycycline is also among the drugs recommended for the treatment of community-acquired pneumonia. SULFONAMIDES AND TRIMETHOPRIM The folic acid synthesis inhibitors have a broad spectrum of bacteriostatic activity individually; in combination, they can be bactericidal against facultative gram-negative bacteria and staphylococci. The fixed combination of sulfamethoxazole and trimethoprim, the major folic acid synthesis inhibitors used in therapy for bacterial infections, has modest activity against some streptococci and no activity against strict anaerobes. However, resistance to the combination of sulfamethoxazole and trimethoprim is common among methicillin-resistant staphylococci and penicillin-resistant pneumococci and is increasing among E. coli strains that cause urinary tract infections. The individual sulfonamides are rarely used in the treatment of bacterial infections but are among the drugs of choice for the treatment of nocardial infections, leprosy (dapsone, a sulfone), and toxoplasmosis (sulfadiazine). Although increasing resistance has been reported among gram-negative organisms, trimethoprim-sulfamethoxazole remains one of the drugs of choice for the treatment of uncomplicated urinary tract infections (except for those caused by enterococci) and is widely used in the treatment of otitis media. It can be used in therapy for upper respiratory tract infections in which S. pneumoniae, H. influenzae, or M. catarrhalis is suspected; for gonococcal and meningococcal infections; for chancroid; and for infections thought to be caused by Aeromonas, Stenotrophomonas, Burkholderia cepacia, Acinetobacter, and Yersinia enterocolitica. For nosocomial infections due to Stenotrophomonas, trimethoprim-sulfamethoxazole is the drug of choice. FLUOROQUINOLONES The fluoroquinolones have excellent activity against most facultative gram-negative rods and variable activity against gram-positive cocci; only trovafloxacin is active against obligate anaerobes. The quinolones are the oral agents with greatest activity against P. aeruginosa; ciprofloxacin is the most active against this species. All the quinolones except norfloxacin are well absorbed orally; ciprofloxacin, levofloxacin, trovafloxacin, and ofloxacin are also administered as intravenous formulations. The quinolones are among the drugs of choice for urinary tract infections, bacterial gastroenteritis, community-acquired pneumonia, and enteric fever and may be useful in therapy for serious hospital-acquired infections caused by gram-negative organisms. While older quinolones (ciprofloxacin, ofloxacin, and norfloxacin) have limited activity against gram-positive bacteria, the newer quinolones have an expanded spectrum of activity against gram-positive cocci, including staphylococci (both methicillin-susceptible and methicillin-resistant) and streptococci (especially S. pneumoniae). However, because of the potential for development of severe liver toxicity, it has been recommended that trovafloxacin use be limited to hospitalized patients with serious or life-threatening infections. Quinolones can also be used as prophylaxis for persons at risk for meningococcal meningitis. However, rapid expansion in the use of quinolones should be coupled with a consideration of the potential for development of resistance among all
bacteria targeted by these drugs. RIFAMPIN Rifampin has been used in combinations for the treatment of serious infections due to methicillin-resistant staphylococci (e.g., coagulase-negative staphylococcal foreign-body infections). Because the spontaneous selection of rifampin-resistant mutants occurs rapidly, rifampin should never be used alone in the treatment of staphylococcal infections. Rifampin is also used for chemoprophylaxis in persons at risk of meningococcal meningitis and for the treatment of Legionella pneumonia. METRONIDAZOLE Metronidazole has a spectrum limited to anaerobic bacteria. It is one of the drugs of choice for the treatment of any abscess in which the involvement of obligate anaerobes is suspected (e.g., lung, brain, or intraabdominal abscesses) because of its spectrum and its ability to penetrate into the area of infection. Other antibacterial agents should be used in combination with metronidazole if facultative and aerobic pathogens are also thought to be involved. Metronidazole is the drug of choice for the treatment of bacterial vaginosis and antibiotic-associated pseudomembranous colitis. URINARY TRACT ANTISEPTICS Urinary tract antiseptics are active only in the lower urinary tract and cannot be used for the treatment of upper urinary tract or systemic infections. Their activity is limited to susceptible gram-negative enteric bacteria. The available agents in this category include nitrofurantoin and methenamine salts. TOPICAL ANTIBACTERIAL AGENTS Mupirocin is available only as a topical preparation for use against staphylococci and streptococci. Its major applications are for impetigo and eradication of the staphylococcal carrier state. It is the drug of choice for the elimination of nasal carriage of both methicillin-susceptible and methicillin-resistant staphylococci. Unfortunately, the emergence of resistance is limiting its usefulness in some hospitals. Although their efficacy has never been well documented, topical preparations that include sulfonamides, polymyxin B, neomycin, bacitracin, gramicidin, and novobiocin in a variety of combinations are widely used as eye drops, irrigation solutions, and ointments for superficial skin infections. ADVERSE REACTIONS Adverse drug reactions are frequently classified by mechanism as either dose-related ("toxic") effects or unpredictable reactions. Unpredictable reactions are further categorized as either idiosyncratic or allergic. Dose-related reactions include aminoglycoside-induced nephrotoxicity, penicillin-induced seizures, and vancomycin-induced anaphylactoid reactions. Many of these reactions can be avoided by reducing dosage, limiting the duration of therapy, or reducing the frequency or rate of
administration. Adverse reactions to antibacterial agents are a common cause of morbidity, requiring alteration in therapy and additional expense, and they occasionally result in death. The elderly, often those with the more severe infections, may be especially prone to certain adverse reactions.*For further discussion of adverse drug reactions, see Chap. 71. b-LACTAMS The therapeutic index for b-lactam antibiotics is broad, and dose-related adverse reactions are uncommon and largely preventable. The greatest concern is allergic reactions. All types can occur, including anaphylaxis (type 1, hypersensitivity reactions), nephritis and Coombs-positive hemolytic anemia (type 2, cytotoxic reactions), drug fever and serum sickness (type 3, immune-complex formation), contact dermatitis (type 4, cell-mediated effects), and maculopapular eruption (type 5, idiopathic reactions). Approximately 1 to 4% of treatment courses result in an allergic reaction, and approximately 0.004 to 0.015% of treatment courses result in anaphylaxis. Fewer than half the patients who claim an allergy to penicillin react to skin testing with the major and minor determinants (penicilloyl-polylysine and benzylpenicillin degradation products, respectively); those with negative skin tests only rarely react adversely to subsequent therapeutic doses. Generally, a suitable alternative to b-lactams is available for patients who have a severe allergy, and penicillin desensitization can be carefully undertaken if there is no suitable alternative. A small proportion (24 h before delivery), prolonged labor, fever, or chorioamnionitis. Because the usual source of the organisms infecting a neonate is the mother's birth canal, efforts have been made to prevent group B streptococcal infections by the identification of high-risk carrier mothers and their treatment with various forms of antibiotic or immunoprophylaxis. Prophylactic administration of ampicillin or penicillin to such patients during delivery has been shown to reduce the risk of infection in the newborn. This approach has been hampered by the logistical difficulties of identifying colonized women before delivery, since the results of vaginal cultures early in pregnancy are poor predictors of carrier status at delivery. TheCDC has suggested two alternative approaches to the prevention of neonatal group B streptococcal infection: In the first approach, women are screened for anogenital colonization at 35 to 37 weeks of pregnancy by means of a swab culture of the lower vagina and anorectum; intrapartum chemoprophylaxis is offered to carriers and is recommended for those carriers with any of the risk factors noted above, those anticipating multiple births, and those who have previously given birth to an infant with group B streptococcal infection. In the second approach, screening cultures need not be performed, but intrapartum chemoprophylaxis is recommended for all women with one or more of the risk factors noted above. The recommended regimen for chemoprophylaxis is 5 million units of penicillin G followed by 2.5 million units every 4 h until delivery. Clindamycin or erythromycin may be substituted in women allergic to penicillin. Treatment of all pregnant women who are colonized or who have risk factors for neonatal infection will result in exposure of 15 to 25% of pregnant women and newborns to antibiotics, with the attendant risks of allergic reactions and selection for resistant organisms. Although still in the developmental stages, a group B streptococcal vaccine may ultimately offer a better solution to prevention. Because transplacental passage of maternal antibodies produces protective antibody levels in the newborn, efforts are under way to develop a vaccine against group B streptococci that can be given to
childbearing women before or during pregnancy. Results of phase 1 clinical trials of group B streptococcal capsular polysaccharide-protein conjugate vaccines suggest that a multivalent conjugate vaccine would be safe and highly immunogenic. INFECTION IN ADULTS The majority of group B streptococcal infections in adults are related to pregnancy and parturition. Peripartum fever, the most common manifestation, is sometimes accompanied by symptoms and signs of endometritis or chorioamnionitis (abdominal distention and uterine or adnexal tenderness). Blood cultures are often positive, as are cultures of vaginal swabs. Bacteremia is usually transitory but occasionally results in meningitis or endocarditis. Infections in adults that are not associated with the peripartum period generally involve individuals who are elderly or have some underlying chronic illness, such as diabetes mellitus or a malignancy. Among the infections that develop with some frequency in adults are cellulitis and soft tissue infection (including infected diabetic skin ulcers), urinary tract infection, pneumonia, endocarditis, and septic arthritis. Other reported infections include meningitis, osteomyelitis, and intraabdominal or pelvic abscesses. TREATMENT Group B streptococci are less sensitive to penicillin than group A organisms, requiring somewhat higher doses. Adults with serious localized infections (pneumonia, pyelonephritis, abscess) should receive doses in the range of 12 million units of penicillin G daily, while patients with endocarditis or meningitis should receive 18 to 24 million units per day in divided doses. Vancomycin is an acceptable alternative for patients allergic to penicillin. ENTEROCOCCI, GROUP D STREPTOCOCCI ENTEROCOCCI Lancefield group D includes the enterococci, organisms now classified in a separate genus from other streptococci, and nonenterococcal group D streptococci. Enterococci are distinguished from nonenterococcal group D streptococci by their ability to grow in the presence of 6.5% sodium chloride and by the results of other biochemical tests. The enterococcal species that are significant pathogens for humans are E. faecalis and E. faecium. These organisms tend to produce infection in patients who are elderly or debilitated or in whom mucosal or epithelial barriers have been disrupted or the balance of the normal flora altered by antibiotic treatment. Urinary tract infections due to enterococci are quite common, particularly among patients who have received antibiotic treatment or undergone instrumentation of the urinary tract. Enterococci are a frequent cause of nosocomial bacteremia in patients with intravascular catheters. These organisms account for 10 to 20% of cases of bacterial endocarditis on both native and prosthetic valves. The presentation of enterococcal endocarditis is usually subacute but may be acute, with rapidly progressive valve destruction. Enterococci are frequently cultured from bile and are involved in infectious complications of biliary surgery and in liver abscesses. Moreover, enterococci are often isolated from polymicrobial infections arising from the bowel flora (e.g., intraabdominal abscesses), from abdominal surgical
wounds, and from diabetic foot ulcers. While such mixed infections are frequently cured by antimicrobials not active against enterococci, specific therapy directed against enterococci is warranted when these organisms are the predominant species or are isolated from blood cultures. TREATMENT Unlike other streptococci, enterococci are not reliably killed by penicillin or ampicillin alone at concentrations achieved clinically in the blood or tissues. Ampicillin reaches sufficiently high urinary concentrations to constitute adequate monotherapy for uncomplicated urinary tract infections. Because in vitro testing has shown evidence of synergistic killing of most enterococcal strains by the combination of penicillin or ampicillin with an aminoglycoside, combined therapy is recommended for enterococcal endocarditis and meningitis; the regimen is penicillin (3 to 4 million units every 4 h) or ampicillin (2 g every 4 h) plus moderate-dose gentamicin (1 mg/kg every 8 h for patients with normal renal function). Enterococcal endocarditis should be treated for a minimum of 4 weeks and for 6 weeks if symptoms have been present for³3 months or if the infection involves a prosthetic heart valve. For nonendocarditis bacteremia and other serious enterococcal infections, it is not known whether the efficacy of single-agentb-lactam therapy is improved by the addition of gentamicin, but many infectious disease specialists use combination therapy for such infections, especially in critically ill patients. Vancomycin, in combination with gentamicin, may be substituted for penicillin in allergic patients. Enterococci are resistant to all cephalosporins; therefore, this class of antibiotics should not be used for treatment of enterococcal infections. Antimicrobial susceptibility testing should be performed routinely on enterococcal isolates from patients with serious infections, and therapy should be adjusted according to the results (Table 140-4). Most enterococci are resistant to streptomycin, and this drug should not be used for treatment of enterococcal infection unless in vitro testing of the strain indicates susceptibility. Though less widespread than streptomycin resistance, high-level resistance to gentamicin -- with a minimum inhibitory concentration (MIC) of >2000 ug/mL -- has become common. Gentamicin-resistant enterococci should be tested for susceptibility to streptomycin; occasional gentamicin-resistant enterococci are sensitive to streptomycin. If the isolate is resistant to all aminoglycosides, treatment with penicillin or ampicillin alone may be successful. The prolonged administration (i.e., for at least 6 weeks) of high-dose ampicillin (e.g., 12 g/d) is recommended for endocarditis due to these highly resistant enterococci. Enterococci may be resistant to penicillins via two distinct mechanisms. The first is the production ofb-lactamase (mediating resistance to penicillin and ampicillin), which has been reported for E. faecalis isolates from several locations in the United States and other countries. Because the amount of b-lactamase produced by enterococci may be insufficient for detection by routine antibiotic susceptibility testing, isolates from serious infections should be screened specifically forb-lactamase production with use of a chromogenic cephalosporin or by another method. For the treatment ofb-lactamase-producing strains, vancomycin, ampicillin/sulbactam, amoxicillin/clavulanate, or imipenem may be used in combination with gentamicin. The second mechanism of penicillin resistance is not mediated byb-lactamase and may
be due to altered penicillin-binding proteins. This intrinsic penicillin resistance is common among E. faecium isolates, which routinely are more resistant tob-lactam antibiotics than are isolates of E. faecalis. Moderately resistant enterococci (MICs of penicillin and ampicillin, 16 to 64 ug/mL) may be susceptible to high-dose penicillin or ampicillin plus gentamicin, but strains with MICs of³200 ug/mL must be considered resistant to clinically achievable levels of b-lactam antibiotics, including imipenem. Vancomycin plus gentamicin is the recommended regimen for infections due to enterococci with high-level intrinsic resistance tob-lactams. Vancomycin-resistant enterococci, first reported from clinical sources in the late 1980s, have become common in many hospitals. Three major vancomycin resistance phenotypes have been described: VanA, VanB, and VanC. The VanA phenotype is associated with high-level resistance to vancomycin and to teicoplanin, a related glycopeptide antibiotic not currently available in the United States. VanB and VanC strains are resistant to vancomycin but susceptible to teicoplanin, although teicoplanin resistance may develop during treatment in VanB strains. For enterococci resistant to both vancomycin and b-lactams, there are no established therapies. Regimens that have been tried with some success in individual cases or experimentally include ciprofloxacin plus rifampin plus gentamicin; ampicillin plus vancomycin (particularly if in vitro testing shows synergistic bacteriostatic activity); and chloramphenicol or tetracycline (if the strain is susceptible in vitro). Quinupristin/dalfopristin (Synercid) is a streptogramin combination with in vitro activity against E. faecium, including vancomycin-resistant isolates, but not against E. faecalis or other enterococcal species. The evidence for clinical efficacy of this agent in serious enterococcal infections is limited, and, as of this writing, quinupristin/dalfopristin is not yet licensed for use in the United States. OTHER GROUP D STREPTOCOCCI The main nonenterococcal group D streptococcal species that causes human infections is S. bovis. S. bovis endocarditis is often associated with neoplasms of the gastrointestinal tract -- most frequently a colon carcinoma or polyp -- but is also reported in association with other bowel lesions. When occult gastrointestinal lesions are carefully sought, abnormalities are found in ³60% of patients with S. bovis endocarditis. In contrast to the enterococci, nonenterococcal group D streptococci like S. bovis are reliably killed by penicillin as a single agent, and penicillin is the treatment of choice for S. bovis infections. VIRIDANS AND OTHER STREPTOCOCCI VIRIDANS STREPTOCOCCI Consisting of multiple species of a-hemolytic streptococci, the viridans streptococci are a heterogeneous group of organisms that are important as agents of bacterial endocarditis (Chap. 126). Several species of viridans streptococci, including S. salivarius, S. mitis, S. sanguis, and S. mutans, are part of the normal flora of the mouth, where they live in close association with the teeth and gingiva. Some species contribute to the development of dental caries. The transient viridans streptococcal bacteremia induced by eating, tooth-brushing, flossing, and other sources of minor trauma, together
with adherence to biologic surfaces, is thought to account for the predilection of these organisms to cause endocarditis. Viridans streptococci are also isolated, often as part of a mixed flora, from sites of sinusitis, brain abscess, and liver abscess. Viridans streptococcal bacteremia occurs relatively frequently in neutropenic patients, particularly after bone marrow transplantation or high-dose chemotherapy for cancer. Some of these patients develop a sepsis syndrome with high fever and shock. Risk factors for viridans streptococcal bacteremia include chemotherapy with high-dose cytosine arabinoside, prior treatment with trimethoprim-sulfamethoxazole or a fluoroquinolone, treatment with antacids or histamine antagonists, mucositis, and profound neutropenia. The S. milleri group (also referred to as the S. intermedius or S. anginosus group) includes three species that cause human disease: S. intermedius, S. anginosus, and S. constellatus. These organisms are often are considered viridans streptococci, although they differ somewhat from other viridans streptococci in both their hemolytic pattern (they may bea-, b-, or nonhemolytic) and the disease syndromes they cause. This group commonly produces suppurative infections, particularly abscesses of brain and abdominal viscera, and infections related to the oral cavity or respiratory tract, such as peritonsillar abscess, lung abscess, and empyema. TREATMENT Isolates from neutropenic patients with bacteremia often are resistant to penicillin; thus these patients should be treated presumptively with vancomycin until the results of susceptibility testing become available. Viridans streptococci isolated in other clinical settings usually are sensitive to penicillin. NUTRITIONALLY VARIANT STREPTOCOCCI Occasional isolates cultured from the blood of patients with endocarditis fail to grow when subcultured on solid media. These nutritionally variant streptococci require supplemental thiol compounds or active forms of vitamin B6(pyridoxal or pyridoxamine) for growth in the laboratory. The nutritionally variant streptococci are generally grouped with the viridans streptococci because they cause similar types of infections. However, they have been reclassified on the basis of 16S RNA sequence comparisons into a separate genus, Abiotrophia, with two species: A. defectivus and A. adjacens. TREATMENT Because treatment failure and relapse appear to be more common for cases of endocarditis due to nutritionally variant streptococci than for usual viridans streptococci, the addition of gentamicin (1 mg/kg every 8 h for patients with normal renal function) to the penicillin regimen is recommended in therapy for endocarditis due to these organisms. OTHER STREPTOCOCCI S. suis is an important pathogen in swine and has been reported to cause meningitis in
humans, usually in individuals with occupational exposure to pigs. Strains of S. suis associated with human infections have generally reacted with Lancefield group R typing serum and sometimes with group D typing serum as well. Isolates may bea- or b-hemolytic and are sensitive to penicillin. S. iniae, a pathogen of fish, has been associated with infections in humans who have handled live or freshly killed fish. Cellulitis of the hand is the most common form of human infection, although bacteremia and endocarditis have been reported. Anaerobic streptococci, or peptostreptococci, are part of the normal flora of the oral cavity, bowel, and vagina. Infections caused by the anaerobic streptococci are discussed in Chap. 167. (Bibliography omitted in Palm version) Back to Table of Contents
141. DIPHTHERIA, OTHER CORYNEBACTERIAL INFECTIONS, AND ANTHRAX Randall K. Holmes DIPHTHERIA DEFINITION Diphtheria is a localized infection of mucous membranes or skin caused by Corynebacterium diphtheriae. A characteristic pseudomembrane may be present at the site of infection (Fig. 141-CD1). Some strains of C. diphtheriae produce diphtheria toxin, a protein that can cause myocarditis, polyneuritis, and other systemic toxic effects. Respiratory diphtheria is usually caused by toxinogenic (tox+) C. diphtheriae, but cutaneous diphtheria is frequently caused by nontoxinogenic (tox-) strains. ETIOLOGY C. diphtheriae is an aerobic, nonmotile, nonsporulating, irregularly staining, gram-positive rod. The bacteria are club-shaped and are often arranged in clusters (Chinese letters) or parallel arrays (palisades). C. diphtheriae forms gray to black colonies on selective media containing tellurite. The gravis, mitis, and intermedius biotypes are distinguished by colonial morphology and laboratory tests. Both tox+ and tox- strains cause infections, and tox+ strains of all three biotypes can cause severe disease. The gene for diphtheria toxin is present in specific corynephages, and tox- C. diphtheriae can acquire the ability to produce diphtheria toxin by infection with tox+phages (phage conversion). Growth of C. diphtheriae under low-iron conditions that mimic the environment of host tissues induces production of diphtheria toxin and expression of systems for siderophore-dependent iron uptake and utilization of iron from heme. IMMUNOLOGY Treatment of diphtheria toxin with formaldehyde converts it to a nontoxic but immunogenic product (diphtheria toxoid). Immunization with toxoid elicits antibody (antitoxin) that neutralizes the toxin and prevents diphtheria. The attack rate and mortality rate for diphtheria are low in immune individuals with antitoxin titers of >0.01 unit per milliliter. Antitoxin neither prevents colonization by C. diphtheriae nor eradicates the carrier state. When most individuals in a population have protective levels of antitoxin (herd immunity), the carrier rate for tox+ strains of C. diphtheriae falls to a low level, and the risk that susceptible individuals will be exposed to tox+ C. diphtheriae decreases dramatically. Susceptible individuals may contract diphtheria if they travel to regions where the disease is present or if tox+ strains of C. diphtheriae are introduced into their community. EPIDEMIOLOGY AND IMMUNITY Humans are the principal reservoir for C. diphtheriae. Transmission occurs primarily by close personal contact. The risk is greater that C. diphtheriae will be transmitted to susceptible individuals from patients with diphtheria than from carriers. The incubation period for respiratory diphtheria is typically 2 to 5 days and rarely up to 8 days.
Cutaneous diphtheria is usually a secondary infection whose signs develop an average of 7 days (range, 1 to >21 days) after the appearance of primary dermatologic lesions of other etiologies. In temperate climates, diphtheria primarily involves the respiratory tract; occurs throughout the year, with a peak incidence in colder months; and is usually caused by tox+ C. diphtheriae. Before immunization was introduced, diphtheria was primarily a disease of children; it affected up to 10% of individuals in this group and sometimes caused devastating epidemics. Most young infants were immune because of transplacental transfer of maternal IgG antitoxin, but children became susceptible by 6 to 12 months of age. Approximately 75% of individuals became immune by age 10 as a result of contact with C. diphtheriae. Mortality rates of 30 to 40% were common in untreated disease and were sometimes >50% in epidemics. Treatment with antitoxin reduced the case-fatality rate to 5 to 10%. Routine immunization of children in the United States resulted in a progressive decrease of diphtheria from the peak of 206,939 cases (incidence rate, 191 cases per 100,000 population) in 1921 to 96%), but immunization rates for younger children are substantially lower. Among adults >20 years of age, 19 to 77% are susceptible as a consequence of failure to receive periodic booster immunizations and lack of contact with C. diphtheriae. The most recent large diphtheria outbreak in the United States (about 1100 cases) occurred in Seattle, Washington, between 1972 and 1982. Alcoholism, low socioeconomic status, crowded living conditions, and Native-American ethnic background were significant risk factors in this outbreak. A massive diphtheria epidemic (>157,000 cases and 5000 deaths) occurred recently in the states of the former Soviet Union and accounted for>80% of diphtheria cases reported worldwide during that interval. The epidemic began in 1990 with 1436 cases (0.49 per 100,000 population), peaked in 1995 with 50,425 cases (17.29 per 100,000 population), and waned by 1998 with 2720 cases (0.93 per 100,000 population) as the result of a mass immunization program. The progression of the epidemic was associated with emergence of a previously uncommon clonal group of tox+ C. diphtheriae strains that accounted for>80% of isolates by 1994; molecular analysis of the tox gene from these strains demonstrated that the existing diphtheria toxoid vaccine remained appropriate for control of the epidemic by vaccination. A majority of cases throughout this epidemic occurred in persons ³15 years old, and adults from 40 to 49 years old had very high incidence and death rates. In 1994, case-fatality rates varied from 2.8%in the Russian Federation to 23% in Lithuania and Turkmenistan. Factors that facilitated the spread of this epidemic included large-scale population movements, socioeconomic instability, deteriorating health infrastructure,
delayed implementation of aggressive control measures in response to the epidemic, inadequate information for physicians and the public, and frequent shortages of supplies for prevention and treatment of the disease. The most important risk factor for diphtheria in the Republic of Georgia was lack of vaccination (matched odds ratio, 19.2), but household diphtheria exposure, exposure to skin lesions, the presence of tonsils, a history of eczema, preceding fever with myalgia, sharing a bed, sharing glasses and cups, and taking a bath less often than weekly were also significant risk factors. Although small numbers of imported cases from this epidemic occurred in western European countries, none resulted in secondary transmission of diphtheria, notwithstanding a high proportion of susceptible adults in countries with imported cases. Inadequate primary immunization of children in the states of the former Soviet Union in the years preceding the epidemic, along with failure to maintain adequate immunity in adults by booster immunization, may have synergistically facilitated transmission of diphtheria and emergence of the massive epidemic in this region. In the tropics, cutaneous diphtheria is more common than respiratory diphtheria, occurs throughout the year, and often develops as a secondary infection complicating other dermatoses. Isolates of C. diphtheriae from skin lesions are more often tox-than tox+ . Cutaneous diphtheria is increasingly recognized in temperate climates and accounted for 86% of the 1100 cases in the Seattle epidemic of 1972 to 1982. Since 1980, cutaneous diphtheria has not been a reportable disease in the United States, and recent health statistics include only respiratory diphtheria. During the 1990s, tox- strains of C. diphtheriae were associated with new types of infections. In the United Kingdom, these strains caused symptomatic pharyngitis, predominantly among homosexual men, that was sometimes accompanied by tonsillar exudate. In Switzerland, strains with a high potential for invasiveness were isolated from 38 intravenous drug users and shown by ribotyping to be clonally related. The latter strains caused infections of the skin (15 cases), respiratory tract (10 cases), and blood (13 cases). Among the patients with bloodstream infections, 9 had endocarditis, and 4 of these 9 patients died. PATHOLOGY AND PATHOGENESIS C. diphtheriae infects mucous membranes, most commonly in the respiratory tract, and also invades open skin lesions resulting from insect bites or trauma. In infections caused by tox+ C. diphtheriae, initial edema and hyperemia are often followed by epithelial necrosis and acute inflammation. Coagulation of the dense fibrinopurulent exudate produces a pseudomembrane, and the inflammatory reaction accompanied by vascular congestion extends into the underlying tissues. The pseudomembrane contains large numbers of C. diphtheriae organisms, but the bacteria are rarely isolated from the blood or internal organs. Diphtheria toxin acts both locally and systemically, and the lethal dose for humans is ~0.1 ug/kg. Toxin contributes locally to pseudomembrane formation; systemically, it can cause myocarditis, neuritis, and focal necrosis in various organs, including the kidneys, liver, and adrenal glands. Changes in the myocardium include cloudy swelling of muscle fibers and interstitial edema. These changes are followed within weeks by hyaline and granular degeneration (sometimes with fatty degeneration), progressing to myolysis and
finally to the replacement of lost muscle by fibrosis. Thus, diphtheria can cause permanent cardiac damage. In diphtheritic polyneuritis, pathologic changes include patchy breakdown of myelin sheaths in peripheral and autonomic nerves, but recovery of nerve damage is the rule if the patient survives. Diphtheria toxin is produced by C. diphtheriae as an extracellular polypeptide. Proteolytic cleavage forms nicked toxin consisting of fragments A and B. Fragment B binds to a plasma-membrane receptor (a precursor of a heparin-binding growth factor resembling epidermal growth factor), and the bound toxin is internalized by receptor-mediated endocytosis. Fragment A is translocated across the endosomal membrane and released into the cytoplasm, where it catalyzes the transfer of the adenosine diphosphate ribose moiety from nicotinamide adenine dinucleotide (NAD) to a modified histidine residue (diphthamide) on elongation factor 2 (EF-2), thereby inactivating EF-2 and inhibiting protein synthesis. One molecule of fragment A in the cytoplasm can kill a cell. Other metabolic alterations are secondary to inhibition of protein synthesis. CLINICAL MANIFESTATIONS Patients with C. diphtheriae in the respiratory tract are classified as diphtheria cases if pseudomembranes are present and as diphtheria carriers if pseudomembranes are absent. The disease is graded as tonsillar if pseudomembranes are localized to the tonsils, as combined types or delayed diagnosis if more extensive pseudomembranes are present, and as severe if cervical adenopathy or cervical edema is also present. Onset is often gradual, but most patients seek medical care within a few days of becoming ill. Fever of 37.8° to 38.9°C (100° to 102°F), sore throat, and weakness are the most common symptoms, while dysphagia, headache, and change of voice occur in fewer than half of patients. Neck edema and difficulty breathing are noted in£10% of patients and are associated with an increased risk of death. Systemic manifestations are due primarily to toxic effects of diphtheria toxin. Patients without toxicity exhibit discomfort and malaise associated with local infection, whereas severely toxic patients may develop listlessness, pallor, and tachycardia that can progress rapidly to vascular collapse. Primary infection in the respiratory tract is most often tonsillopharyngeal but may also be (in decreasing order) laryngeal, nasal, and tracheobronchial. Multiple sites are frequently involved, and secondary spread of pharyngeal infection upward to the nasal mucosa or downward to the larynx and tracheobronchial tree is much more common than primary infection at those sites. Systemic toxicity is usually most severe when extensive pseudomembrane extends from the tonsils and pharynx into contiguous regions. A small percentage of patients present with malignant or "bull-neck" diphtheria, with extensive pseudomembrane formation, foul breath, massive swelling of the tonsils and uvula, thick speech, cervical lymphadenopathy, striking edematous swelling of the submandibular region and anterior neck, and severe toxicity. In tonsillopharyngeal diphtheria, isolated spots of gray or white exudate may appear first. These spots often extend and coalesce within a day to form a confluent, sharply demarcated pseudomembrane that becomes progressively thicker, more tightly adherent to the underlying tissue, and darker gray in color. Unlike the exudate in
streptococcal pharyngitis, the diphtheritic pseudomembrane often extends beyond the margin of the tonsils onto the tonsillar pillars, palate, or uvula. Dislodging the membrane is likely to cause bleeding. Laryngeal diphtheria often presents as hoarseness and cough. Demonstration of laryngeal pseudomembrane by laryngoscopy helps distinguish diphtheria from other infectious forms of laryngitis. Patients with nasal diphtheria may present with unilateral or bilateral serosanguineous nasal discharge associated with irritation of the nares or lip. Primary or secondary diphtheritic infection occasionally involves other mucous membranes, including the conjunctiva and the membranes of the genitourinary and gastrointestinal tracts. Cutaneous diphtheria usually presents as an infection by C. diphtheriae of preexisting dermatoses involving the lower extremities, upper extremities, head, or trunk. The clinical features are similar to those of other secondary cutaneous bacterial infections. In the tropics, cutaneous diphtheria may present as a primary cutaneous lesion, typically with morphologically distinct "punched-out" ulcers that are covered by necrotic slough or membrane and have well-demarcated edges. C. diphtheriae is an occasional cause of invasive infections, including endocarditis and septic arthritis. Risk factors for such infections include preexisting cardiac abnormalities, abuse of intravenous drugs, and alcoholic cirrhosis. COMPLICATIONS Obstruction of the respiratory tract can be caused by extensive pseudomembrane formation and swelling early in the disease or by sloughed pseudomembrane that becomes lodged in the airways later in the disease. The risk is greater when infection involves the larynx or the tracheobronchial tree and in children because of the small size of the airways. Myocarditis and polyneuritis are the most prominent toxic manifestations of diphtheria. The risk of each is proportional to the severity of local disease. Myocarditis occurred in 22% and neuritis in 5% of 656 hospitalized patients (54% female, 70% ³15 years old) with diphtheria in the Kyrgyz Republic in 1995; 7% of patients with myocarditis and 2% of patients without myocarditis died. The median interval from hospitalization to death was 4.5 days (range, 0 to 13 days). Bulbar dysfunction in diphtheritic neuritis typically develops during the first 2 weeks. Palatal and pharyngeal paralysis usually develops first. Swallowing is difficult, the voice is nasal, and ingested fluids may be regurgitated through the nose. Additional bulbar signs may develop over several weeks, with oculomotor and ciliary paralysis more common than facial or laryngeal paralysis. Peripheral polyneuritis typically begins from 1 to 3 months after the onset of diphtheria with proximal weakness of the extremities, which spreads distally. Paresthesia may occur, most often in a glove-and-stocking distribution. Polyneuritis usually resolves completely, with the time needed for improvement approximately equal to that elapsing from exposure to the development of symptoms. Pneumonia occurs in more than one-half of fatal cases of diphtheria. Less common complications include renal failure, encephalitis, cerebral infarction, pulmonary
embolism, and bacteremia or endocarditis due to invasive infection by C. diphtheriae. Serum sickness may result from antitoxin therapy. COURSE AND PROGNOSIS Most cases of diphtheria develop in nonimmunized patients. The attack rate, severity of disease, and risk of complications are much lower in immunized patients. The pseudomembrane may continue to increase in size during the first day after administration of antitoxin. During the next several days to a week, it becomes softer, less adherent, and nonconfluent and eventually disappears. In the preantibiotic era, C. diphtheriae persisted in the throat for ~2 weeks in one-half of patients and for ³1 month in about one-fifth. Mortality increases with the severity of local disease, the extent of pseudomembrane formation, and the delay between onset of local disease and administration of antitoxin. The death rate is highest during the first week of illness; among patients with bull-neck diphtheria; among patients with myocarditis who develop ventricular tachycardia, atrial fibrillation, or complete heart block; among patients with laryngeal or tracheobronchial involvement; among infants and patients >60 years of age; and among alcoholics. Both the mortality rate and the risk of myocarditis or peripheral neuropathy are significantly lower in cutaneous diphtheria than in respiratory diphtheria. DIAGNOSIS A characteristic pseudomembrane on the mucosa of the tonsils, palate, oropharynx, nasopharynx, nose, or larynx suggests diphtheria but is not uniformly present. Diphtheritic pseudomembrane must be distinguished from other pharyngeal exudates, including those of group A b-hemolytic streptococcal infections, infectious mononucleosis, viral pharyngitis, fusospirochetal infection, and candidiasis. Diphtheria should be considered in patients with sore throat, cervical adenopathy or swelling, and low-grade fever, especially when these manifestations are accompanied by systemic toxicity, hoarseness, stridor, palatal paralysis, or serosanguineous nasal discharge with or without demonstrable pseudomembrane. Treatment with diphtheria antitoxin should begin as soon as the clinical diagnosis of diphtheria is made. Definitive diagnosis of diphtheria depends on the isolation of C. diphtheriae from local lesions. The laboratory should be notified that diphtheria is suspected to ensure the use of selective tellurite medium appropriate for the isolation of C. diphtheriae. All isolates of C. diphtheriae should be subjected to toxicity testing. Primary isolates can be screened rapidly for the presence of the tox gene by the polymerase chain reaction, although occasional strains of C. diphtheriae that carry an inactive toxin gene give false-positive results. Biochemical tests needed to differentiate C. diphtheriae from corynebacteria of the normal flora (diphtheroids) require several days. Group A b-hemolytic streptococci and Staphylococcus aureus are also isolated frequently from patients with diphtheria. Cutaneous diphtheria may present as a characteristic "punched-out" ulcer with a membrane, but it is more often indistinguishable from other inflammatory dermatoses. Diagnosis depends on a high degree of suspicion and on culture of cutaneous lesions on laboratory media appropriate for isolation of C. diphtheriae. Throat samples from all patients with cutaneous diphtheria should be cultured for C. diphtheriae.
TREATMENT The decision to administer diphtheria antitoxin must be based on the clinical diagnosis of diphtheria without definitive laboratory confirmation, since each day of delay in treatment is associated with increased mortality. Because diphtheria antitoxin is produced in horses, it is necessary to inquire about possible allergy to horse serum and to perform a conjunctival or intracutaneous test with diluted antitoxin for immediate hypersensitivity. Epinephrine must be available for immediate administration to patients with severe allergic reactions. Patients with immediate hypersensitivity should be desensitized before a full therapeutic dose of antitoxin is given. The dose of diphtheria antitoxin currently recommended by the Committee on Infectious Diseases of the American Academy of Pediatrics is based on the site of the primary infection and the duration and severity of disease: 20,000 to 40,000 units for disease that has been present for £48 h and involves the pharynx or larynx; 40,000 to 60,000 units for nasopharyngeal infections; and 80,000 to 100,000 units for disease that is extensive, has been present for³3 days, or is accompanied by diffuse swelling of the neck. Antitoxin is administered intravenously by infusion in saline over 60 min to neutralize unbound toxin rapidly. The ~10% risk of serum sickness is acceptable because of the established therapeutic value of antitoxin in decreasing mortality from respiratory diphtheria. The risk of systemic toxicity is lower in cutaneous diphtheria than in respiratory diphtheria and must be weighed against the potential adverse effects of antitoxin treatment; authorities are not unanimous in recommending antitoxin therapy for cutaneous diphtheria. Antibiotics have little demonstrated effect on the healing of local infection in diphtheria patients treated with antitoxin. The primary goal of antibiotic therapy for patients or carriers is therefore to eradicate C. diphtheriae and prevent its transmission from the patient to susceptible contacts. Erythromycin, penicillin G, rifampin, or clindamycin is recommended by most authorities. Commonly recommended regimens for the treatment of adults with respiratory diphtheria are erythromycin (500 mg four times daily, given parenterally or orally) or intramuscular procaine penicillin G (600,000 units at 12-h intervals) for 14 days. Patients with cutaneous diphtheria and carriers can be treated orally with erythromycin (500 mg four times daily) or rifampin (600 mg once daily) for 7 days. If compliance is in question, a single dose of benzathine penicillin G (1.2 to 2.4 million units intramuscularly) can be substituted. Eradication of C. diphtheriae should be documented by negative cultures of samples taken on two or three successive days, beginning at least 24 h after the completion of antibiotic therapy. Some authorities also recommend a repeat throat culture 2 weeks later. The small percentage of patients who continue to be infected with C. diphtheriae after treatment should receive an additional 10-day course of oral erythromycin or rifampin. Plasmid-mediated resistance to erythromycin of the MLS type emerged transiently in C. diphtheriae during the Seattle epidemic, but its frequency declined dramatically after the routine use of erythromycin was discontinued. Patients with respiratory or cutaneous diphtheria caused by tox+ C. diphtheriae or by strains of unknown toxinogenicity should be hospitalized, kept in bed initially, handled with isolation procedures appropriate for the site of infection, and given supportive care as needed. Respiratory and cardiac function must be monitored closely. Early intubation
or tracheostomy is recommended when the larynx is involved or signs of impending airway obstruction are detected. Tracheobronchial membrane can sometimes be removed mechanically via the endotracheal tube or tracheostomy. Primary or secondary pneumonia should be diagnosed and treated promptly. Sedative or hypnotic drugs that may mask respiratory symptoms are contraindicated. Close electrocardiographic monitoring, treatment of arrhythmias, and electrical pacing for heart block are essential. Congestive heart failure should be treated as described inChap. 232. Glucocorticoids do not reduce the risk of diphtheritic myocarditis or polyneuritis. Ulcerative or ecthymatous cutaneous lesions should be treated with Burow's solution applied on wet compresses after debridement of necrotic areas, and treatment for associated conditions such as pediculosis, scabies, or underlying dermatoses should be instituted. Recovery from diphtheria does not always confer active immunity, and initiation of an immunization regimen for diphtheria that is appropriate for the patient's age should be an integral part of the treatment plan. PREVENTION Vaccines available in the United States for immunization against diphtheria include diphtheria and tetanus toxoids and pertussis vaccine adsorbed (DTP), diphtheria and tetanus toxoids and acellular pertussis vaccine adsorbed (DTaP), diphtheria and tetanus toxoids adsorbed (DT; for pediatric use), and tetanus and diphtheria toxoids adsorbed (Td; for adult use). DTaP is preferred over DTP for primary immunization of children without contraindications, and use of DTP is no longer recommended. Td contains less diphtheria toxoid than DTP, DTaP, or DT and causes fewer adverse reactions in adults. Current guidelines for primary immunization of children and adults against diphtheria and for maintaining immunity by periodic booster doses of appropriate vaccines throughout life are summarized inChap. 122. Close contacts of diphtheria patients should be cultured for C. diphtheriae, kept under surveillance for 1 week, and treated with appropriate antibiotics if cultures are positive. Previously immunized close contacts should receive an appropriate booster containing diphtheria toxoid if their last booster was given>5 years previously. If immunization status is uncertain, close contacts should receive an antibiotic regimen appropriate for carriers and a primary immunization series appropriate for their age. OTHER CORYNEBACTERIAL INFECTIONS DEFINITION Medically important coryneform bacteria (formerly called diphtheroids) include members of the normal flora that cause opportunistic infections, human pathogens of relatively low virulence, and animal pathogens that cause occasional zoonotic infections. Reported infections caused by coryneform bacteria have increased substantially in number over the past two decades. Isolates of C. jeikeium and C. urealyticum are often resistant to multiple antibiotics. ETIOLOGY AND LABORATORY DIAGNOSIS Because coryneform bacteria are potential pathogens, it is important not to dismiss
them as constituents of the normal flora or as contaminants when they are found in clinical specimens. Laboratory differentiation of coryneform bacteria is important when they are isolated repeatedly, when they are recovered in pure culture or in large numbers, or when they form pigmented or hemolytic colonies. The coryneform bacteria are a large, heterogeneous group of gram-positive, pleomorphic, irregularly staining bacilli or coccobacilli that superficially resemble C. diphtheriae and are difficult to identify and classify. The genus Corynebacterium is currently divided into three groups of species: the nonlipophilic, fermentative corynebacteria (including C. diphtheriae, C. xerosis, C. striatum, C. minutissimum, and others); the nonlipophilic, nonfermentative corynebacteria (including C. pseudodiphtheriticum and others); and the lipophilic corynebacteria (including C. jeikeium, C. urealyticum, and others). Coryneform bacteria also belong to many other genera (including Actinomyces, Arcanobacterium, and Rhodococcus) as well as to several groups that have not yet been assigned to genera and species by the U.S. Centers for Disease Control and Prevention (CDC). ECOLOGY AND EPIDEMIOLOGY Humans are the probable natural reservoir for C. xerosis, C. pseudodiphtheriticum (formerly C. hofmannii), C. striatum, C. minutissimum, C. jeikeium (formerlyCDCgroup JK), C. urealyticum (formerly CDC group D2), and Arcanobacterium haemolyticum (formerly C. haemolyticum). Animals are the probable natural reservoir for Actinomyces pyogenes (formerly C. pyogenes; cows, sheep, pigs), C. ulcerans (cows, horses), and C. pseudotuberculosis (sheep, horses, goats, cattle). The natural reservoir for Rhodococcus equi (formerly C. equi) is soil. The ecologic niches for many other coryneform bacteria of medical importance are not well defined. The coryneform bacteria found most frequently as components of the normal flora include C. pseudodiphtheriticum (pharynx, skin), C. xerosis (conjunctival sac, nasopharynx, skin), and C. striatum (anterior nares, skin). Coryneform bacteria that commonly colonize the skin of hospitalized patients include C. jeikeium (axilla, groin, perineum) and C. urealyticum. C. jeikeium most often colonizes patients with malignancies or severe immunodeficiency; it is also isolated from environmental sources (surfaces, air) in hospitals and from the hands of ward staff. C. ulcerans infections are acquired by consumption of raw milk. C. pseudotuberculosis infections are acquired by contact with animals or animal products or by consumption of raw milk. PATHOGENESIS AND CLINICAL MANIFESTATIONS C. jeikeium was recognized in 1976 as a cause of infections in immunocompromised hosts. This organism also causes infections in immunocompetent hosts, but severe infections continue to be most frequent in patients with hematologic malignancies and neutropenia. Skin colonization precedes clinical infection. Additional risk factors for nosocomial C. jeikeium sepsis include prolonged hospitalization, breaks in the integument, chronic intravascular catheterization, and prior treatment with broad-spectrum antibiotics. Other presentations of C. jeikeium infection include endocarditis, device-related infections, pulmonary infiltrates, cutaneous septic emboli, soft tissue infections, and rashes. Endocarditis due to C. jeikeium occurs primarily in
patients with prosthetic heart valves. C. jeikeium is a rare cause of central nervous system infections in patients with ventricular shunts. C. urealyticum (formerlyCDCgroup D2) was identified in 1985 as a significant cause of nosocomial urinary tract infections, including acute and chronic cystitis and pyelonephritis. The organism closely resembles C. jeikeium but differs from the latter by producing urease and failing to convert glucose to acidic metabolites. Hydrolysis of urea by urease causes alkalinization of the urine and formation of ammonium magnesium phosphate (struvite) stones. C. urealyticum is a cause of alkaline-encrusted cystitis in patients with preexisting bladder lesions that serve as foci for precipitation of struvite crystals. Risk factors associated with symptomatic urinary tract infections include preexisting immunosuppression, recent urologic procedures (including renal transplantation), underlying disorders of the genitourinary tract, and a history of urinary tract infections. A. haemolyticum causes pharyngitis and chronic skin ulcers; less frequently, it causes a variety of deep tissue infections, septicemia, and endocarditis. Some 90% of A. haemolyticum infections occur in patients between 10 and 30 years old. A. haemolyticum pharyngitis in this age group is 5 to 13% as frequent as Streptococcus pyogenes pharyngitis. An erythematous rash is present in 30 to 67% of cases. The rash is usually scarlatiniform and most pronounced on the trunk and proximal extremities, but it sometimes resembles urticaria or erythema multiforme. Because rash is more frequent in A. haemolyticum infections than in S. pyogenes infections, A. haemolyticum should be considered as a possible etiology in older children and adults who present with the scarlet fever syndrome. Infection due to A. haemolyticum can also present as extensive pharyngeal exudate and can mimic diphtheria. A. haemolyticum occasionally causes peritonsillar abscess, sepsis, endocarditis, or meningitis. C. minutissimum is frequently isolated from the lesions of erythrasma (Fig. 141-CD2), a common superficial skin infection characterized by the presence in intertriginous areas of reddish-brown, scaly, pruritic, macular patches that exhibit coral-red fluorescence under a Wood's light. The etiology of erythrasma appears to be polymicrobial; infection of the skin by C. minutissimum has been shown to follow the onset of maceration and scaling. Deep infections caused by C. minutissimum are rare and include abscesses, bacteremia, endocarditis, peritonitis, pyelonephritis, and infection of central venous catheters. Among coryneform bacteria that cause disease in animals and occasionally in humans, R. equi has emerged as an important intracellular opportunistic pathogen in immunocompromised patients. Most reported cases are necrotizing pulmonary infections that resemble tuberculosis or nocardiosis in patients with severely defective cell-mediated immunity. Cases of R. equi infection are being diagnosed with increasing frequency in patients with AIDS. A. pyogenes causes bovine mastitis, a disease transmitted by flies. Yearly epidemics of leg ulcers infected with A. pyogenes occurred among schoolchildren in Thailand between 1979 and 1984 and were postulated to have resulted from introduction of the organism into traumatic skin lesions by flies. Reported A. pyogenes infections in adults in Denmark have included abscesses, cystitis, intraabdominal infections, and mastoiditis
with bacteremia. C. ulcerans infections in humans usually present as pharyngitis and can mimic respiratory diphtheria, whereas infections caused by C. pseudotuberculosis typically present as suppurative granulomatous lymphadenitis. Some strains of C. ulcerans and C. pseudotuberculosis produce diphtheria toxin. Human infections by tox+ strains of C. ulcerans -- but not by tox+ strains of C. pseudotuberculosis -- have been reported, and administration of diphtheria antitoxin is therefore warranted in infections by C. ulcerans that are presumed on clinical grounds to be caused by toxinogenic strains. C. pseudodiphtheriticum, a commensal of low virulence, is an uncommon cause of pneumonia in men with AIDS and of endocarditis, necrotizing tracheitis, tracheobronchitis, and urinary tract infection in patients without known immune deficiencies. Likewise, C. xerosis and C. striatum only occasionally cause human infections. DIAGNOSIS The clinical features of C. jeikeium infections are not pathognomonic. The diagnosis of these infections is based on a high index of suspicion, identification of the organism by culture in appropriate clinical specimens, and exclusion of other likely causes of infection. C. urealyticum often goes undetected by routine urine cultures; rather, it is necessary to incubate the cultures for 24 to 48 h on blood agar or on special media. Cultivation should be prolonged in selected cases -- i.e., those involving patients (especially elderly men with preexisting genitourinary abnormalities) with alkaline urine, ammonium magnesium phosphate stones, gram-positive bacilli in the urine, or negative standard urine cultures despite clinical evidence of bacteriuria. Other microbes that can cause urinary tract infections with alkaline urine include Proteus, Ureaplasma, and some staphylococci and streptococci. Alkaline-encrusted cystitis is an anatomic diagnosis made by cystoscopy. The differential diagnosis of A. haemolyticum pharyngitis with rash includes scarlet fever; rubella; staphylococcal and streptococcal toxic shock syndromes; infections caused by Epstein-Barr virus, cytomegalovirus, and enteroviruses (especially coxsackieviruses); disseminated gonococcal infection; secondary syphilis; and drug allergy. Routine diagnostic methods for throat cultures are not ideal for the detection of A. haemolyticum, nor is this organism detected by the rapid tests for S. pyogenes that are sometimes substituted for throat cultures. Pharyngitis caused by A. haemolyticum in adolescents and adults is likely to be underdiagnosed until improved tests for the organism are used by diagnostic laboratories. Erythrasma is diagnosed clinically. Because of uncertainty about the etiologic role of C. minutissimum, culture of erythrasma lesions is not currently recommended. Pharyngitis caused by tox+ strains of C. ulcerans may be clinically indistinguishable from diphtheria. The presentations of infections caused by other coryneform bacteria are not usually diagnostic; cultures are required for identification of the causal organisms.
TREATMENT Strains of C. jeikeium are typically resistant to most antibiotics. Vancomycin is the drug of choice for empirical treatment of infections caused by this organism, although antimicrobial susceptibility testing may reveal other antibiotic options for some isolates. For device-related C. jeikeium infections, removal of the infected device is usually required in addition to appropriate antibiotic therapy. C. urealyticum is often resistant to the antibiotics used commonly for the treatment of urinary tract infections. Empirical treatment with vancomycin is appropriate pending the results of antimicrobial susceptibility testing. Several courses of antibiotic therapy may be necessary for bacteriologic cure. Patients with alkaline-encrusted cystitis require resection of the encrusted lesions in addition to antibiotic therapy. No controlled trials of treatment for A. haemolyticum infections have been performed. In vitro tests usually demonstrate susceptibility to penicillins, erythromycin, azithromycin, clindamycin, doxycycline, ciprofloxacin, and vancomycin, but treatment failures have been reported with appropriate doses of penicillins. Limited data suggest that the clinical course of A. haemolyticum pharyngitis may be shortened by treatment with erythromycin. Infections with C. ulcerans that present like diphtheria or are known to be caused by tox+ strains should be treated like diphtheria. Oral erythromycin is usually effective for treatment of erythrasma. For infections caused by R. equi, vancomycin is the drug of choice. Possible alternatives include erythromycin, rifampin, aminoglycosides, and chloramphenicol; the combination of erythromycin and rifampin is attractive because of possible synergy. Penicillins should not be used, because R. equi rapidly develops resistance. Many weeks of antibiotic treatment, sometimes supplemented by surgical intervention, are often needed for infections caused by R. equi. Suppressive therapy with antibiotics should be continued indefinitely in patients with AIDS after initial treatment of infections caused by R. equi. Initial treatment of infections caused by other coryneform bacteria should be based on the identity of the organism and published data regarding antibiotic susceptibility. Therapy should be modified, when necessary, in light of the results of antibiotic susceptibility tests. ANTHRAX DEFINITION Anthrax is an infection caused by Bacillus anthracis that occurs primarily in herbivores. Humans become infected when B. anthracis spores are introduced into the body by contact with infected animals or contaminated animal products, insect bites, ingestion, or inhalation. Aerosolized spores of B. anthracis have the potential for use in biological warfare or bioterrorism. Cutaneous anthrax is most common and is characterized by the development of a localized skin lesion with a central eschar surrounded by marked nonpitting edema. Inhalation anthrax (woolsorters' disease) typically involves hemorrhagic mediastinitis, rapidly progressive systemic infection, and a very high mortality rate. Gastrointestinal anthrax is rare and is associated with a high mortality rate.
ETIOLOGIC AGENT AND EPIDEMIOLOGY B. anthracis is a large, aerobic, spore-forming, gram-positive rod that is encapsulated and nonmotile and grows in chains. Sporulation does not take place in living animals. The rectangular shape of the individual bacteria gives chains of B. anthracis a boxcar-like appearance. Virulent strains of B. anthracis are pathogenic for animals, including mice and guinea pigs. Spores of B. anthracis can survive for years in dry earth but are destroyed by boiling for 10 min, by treatment with oxidizing agents such as potassium permanganate or hydrogen peroxide, or by dilute formaldehyde. Most strains of B. anthracis are susceptible to penicillin. Anthrax occurs worldwide and is most prevalent among domestic herbivores (including cattle, sheep, horses, and goats) and wild herbivores. Grazing animals become infected when they forage for food in areas contaminated with spores of B. anthracis. Anthrax in herbivores tends to be severe, with high mortality. Terminally ill animals with overwhelming bacteremic infections often bleed from the nose, mouth, and bowel, thereby contaminating soil or water with vegetative B. anthracis that can sporulate and persist in the environment. The carcasses of infected animals provide additional potential foci of contamination. Humans are more resistant to anthrax than are herbivorous animals. The estimated number of human cases worldwide is 20,000 to 100,000 per year. Human cases are classified as agricultural or industrial. Agricultural cases result most often from contact with animals that have anthrax (e.g., during skinning, butchering, or dissecting), from bites of contaminated or infected flies, and (in rare instances) from consumption of contaminated meat. Industrial cases are associated with exposure to contaminated hides, goat hair, wool, or bones. Only three cases of cutaneous anthrax were reported to theCDC from 1984 through 1993, and gastrointestinal anthrax has never been documented in the United States. In an epidemic in the former Soviet Union at Sverdlovsk in 1979, cases were initially reported as cutaneous and gastrointestinal anthrax associated with contaminated meat; however, subsequent analysis of epidemiologic data and autopsy findings for most of the fatal cases established that the disease was inhalational anthrax associated with accidental airborne release of B. anthracis from a nearby military biological weapons facility. A massive outbreak in Zimbabwe between 1978 and the early 1980s involved more than 9700 cases of agricultural anthrax in humans. This outbreak occurred during wartime and was associated with disruption of the veterinary and medical infrastructure and cessation of veterinary anthrax vaccination programs. PATHOGENESIS B. anthracis can evade phagocytosis, invade the bloodstream, multiply rapidly to a high population density in vivo, and kill quickly. The poly-D-glutamic acid capsule of B. anthracis confers resistance to phagocytosis. Anthrax toxin consists of three different proteins called protective antigen (PA), edema factor (EF), and lethal factor (LF). The toxin was discovered in studies demonstrating that transfer of sterile blood from guinea pigs dying of anthrax to uninfected guinea pigs killed the recipients. PA binds to plasma membranes of target cells and is cleaved by a cellular protease into two fragments. The
larger fragment remains on the cell surface, displays a binding site for a domain that is present in both EF and LF, and serves as a specific receptor that mediates endocytic entry of EF or LF into the target cells. The catalytic activity of EF, a calmodulin-dependent adenylate cyclase, is expressed in the cytoplasm of human or animal cells that contain both calmodulin and ATP. The biologic effects of EF, which include formation of edema in anthrax lesions and inhibition of polymorphonuclear leukocyte functions, are mediated by the intracellular cyclic AMP that is produced by the enzymatic action of EF. In contrast, LF is a highly specific endopeptidase that cleaves several members of the MAP-kinase-kinase protein family and inactivates their functions in signal transduction pathways. Macrophages appear to be the principal targets of LF in animals, and intoxication of macrophages by LF is associated with production of reactive oxygen species, release of cytokines (including tumor necrosis factor a and interleukin 1b), shock, and death. Cutaneous anthrax is initiated when spores of B. anthracis are introduced into the skin through cuts or abrasions or by biting flies. The spores germinate within hours, and the vegetative cells multiply and produce anthrax toxin. The cutaneous anthrax lesion is characterized by necrosis, vascular congestion, hemorrhage, and gelatinous edema, but few leukocytes are present. In inhalational anthrax, B. anthracis spores in airborne particles 7 years of age. Wound Management Proper wound management requires consideration of the need for (1) passive immunization withTIG and (2) active immunization with vaccine, preferably Td in persons over age 7. For clean minor wounds, Td is administered to persons who (1) have unknown tetanus immunization histories; (2) have received fewer than three doses of adsorbed tetanus toxoid; (3) have received three or more doses of adsorbed vaccine, with the last dose given >10 years previously; and (4) have received three doses of fluid (nonadsorbed) vaccine. The recommendations for contaminated or severe wounds are identical, except that vaccine should be given to those who have received three or more doses of adsorbed tetanus toxoid if >5 years have elapsed since the last dose. Passive immunization with TIG is not recommended for clean minor wounds but is given for all other wounds if the patient's vaccination history indicates unknown or partial immunization. The dose of TIG for passive immunization of persons with wounds of average severity is 250 units intramuscularly, which produces a protective antibody level in the serum for at least 4 to 6 weeks; the appropriate dose ofTAT is 3000 to 6000 units. Vaccine and tetanus antitoxin should be administered at separate sites in separate syringes. Neonatal Tetanus Measures aimed at preventing neonatal tetanus include maternal vaccination, even during pregnancy; efforts to increase the proportion of births that take place in the hospital; and the provision of training for nonmedical birth attendants. PROGNOSIS The application of methods to monitor and support oxygenation has markedly improved the prognosis in tetanus; mortality rates as low as 10% have been reported from units accustomed to handling such cases. In the United States during the period 1995 through 1997, the case-fatality rate was 11%; 11 deaths from tetanus were reported in 1990, 11 in 1991, and 9 in 1992. The outcome is poor in neonates and the elderly and in patients with a short incubation period, a short interval from the onset of symptoms to admission, or a short period from onset of symptoms to the first spasm (period of onset). Outcome is also related to the extent of prior vaccination. The course of tetanus extends over 4 to 6 weeks, and patients may require ventilatory support for 3 weeks during this period. Increased tone and minor spasms can last for months, but recovery is usually complete. (Bibliography omitted in Palm version) Back to Table of Contents
144. BOTULISM - Elias Abrutyn DEFINITION Botulism is a paralytic disease that begins with cranial nerve involvement and progresses caudally to involve the extremities. It is caused by potent protein neurotoxins elaborated by Clostridium botulinum. The toxins' high potency has led to consideration of their use in bioterrorism or biological warfare. Cases may be classified as (1) food-borne botulism, from ingestion of preformed toxin in food contaminated with C. botulinum; (2) wound botulism, from toxin produced in wounds contaminated with the organism; (3) infant botulism, from ingestion of spores and production of toxin in the intestine of infants; or (4) adult infectious botulism, a group that includes some cases in older children and adults in which disease is produced by a mechanism similar to that described for infant botulism. ETIOLOGIC AGENT C. botulinum, a species encompassing a heterogeneous group of anaerobic gram-positive organisms that form subterminal spores, is found in soil and marine environments throughout the world and elaborates the most potent bacterial toxin known. Organisms of types A through G have been distinguished by the antigenic specificities of their toxins; a classification system based on physiologic characteristics has also been described. Rare strains of other clostridial species -- C. butyricum and C. baratii -- have also been found to produce toxin. C. botulinum strains with proteolytic activity can digest food and produce a spoiled appearance; nonproteolytic types leave the appearance of food unchanged. Of the eight distinct toxin types described (A, B, C1, C2, D, E, F, and G), all except C2 are neurotoxins; C2 is a cytotoxin of unknown clinical significance. Botulinum neurotoxin, whether ingested or produced in the intestine or a wound, enters the vascular system and is transported to peripheral cholinergic nerve terminals, including neuromuscular junctions, postganglionic parasympathetic nerve endings, and peripheral ganglia. The central nervous system is not involved. Active neurotoxin (150 kDa) is composed of a heavy chain (a 100-kDa fragment responsible for neurospecific binding and translocation into the nerve cell) and a light chain (a 50-kDa fragment responsible for intracellular catalytic activity). The steps involved in neurotoxin activity include (1) specific binding to presynaptic nerve cells at the myoneural junction, (2) internalization of the toxin inside the nerve cell in endocytic vesicles, (3) translocation of the toxin into the cytosol, and (4) proteolysis by toxin (a zinc endopeptidase) of components of the neuroexocytosis apparatus curtailing release of the neurotransmitter acetylcholine. Cure follows sprouting of new nerve terminals. Toxin is heat-labile, but spores are highly heat-resistant; both can be inactivated under appropriate conditions (see "Prevention," below). In the gastrointestinal tract, toxin is complexed with nontoxin proteins and resists degradation. Toxin types A, B, E, and (in rare instances) F cause human disease; type G (now called C. argentinense) has been associated with sudden death, but not with neuroparalytic illness, in a few patients in Switzerland; and types C and D cause animal disease.
EPIDEMIOLOGY Human botulism occurs worldwide. In the United States, the geographic distribution of cases by toxin type parallels the distribution of organism types found in the environment. Type A predominates west of the Rocky Mountains; type B is generally distributed but is more common in the East; and type E is found in the Pacific Northwest, Alaska, and the Great Lakes area. In the United States, food-borne botulism has been associated primarily with home-canned food, particularly vegetables, fruit, and condiments, and less commonly with meat and fish. Type E outbreaks are frequently associated with fish products. Commercial products occasionally cause outbreaks, but some of these outbreaks have resulted from improper handling after purchase. Outbreaks in restaurants, schools, and private homes have been traced to uncommon sources (commercial potpies, beef stew, turkey loaf, sauteed onions, baked potatoes, and chopped garlic in oil). Food-borne botulism can occur when (1) a food to be preserved is contaminated with spores, (2) preservation does not inactivate the spores but kills other putrefactive bacteria that might inhibit the growth of C. botulinum and provides anaerobic conditions at a pH and temperature that allow germination and toxin production, and (3) food is not heated to a temperature that destroys toxin before being eaten. CLINICAL MANIFESTATIONS Food-Borne Botulism Following ingestion of food containing toxin, illness varies from a mild condition for which no medical advice is sought to very severe disease that can result in death within 24 h. The incubation period is usually 18 to 36 h but, depending on toxin dose, can extend from a few hours to several days. Symmetric descending paralysis is characteristic and can lead to respiratory failure and death. Cranial nerve involvement, which almost always marks the onset of symptoms, usually produces diplopia, dysarthria, and/or dysphagia. Weakness progresses, often rapidly, from the head to involve the neck, arms, thorax, and legs; the weakness is occasionally asymmetric. Nausea, vomiting, and abdominal pain may precede or follow the onset of paralysis. Dizziness, blurred vision, dry mouth, and very dry, occasionally sore throat are common. Patients are generally alert and oriented, but they may be drowsy, agitated, and anxious. Typically, they have no fever. Ptosis is frequent; the pupillary reflexes may be depressed, and fixed or dilated pupils are noted in half of patients. The gag reflex may be suppressed, and deep tendon reflexes may be normal or decreased. Paralytic ileus, severe constipation, and urinary retention are common. Wound Botulism When wounds are contaminated with C. botulinum spores, the spores may germinate into vegetative organisms that produce toxin. This rare condition resembles food-borne illness except that the incubation period is longer, averaging about 10 days, and gastrointestinal symptoms are lacking. Wound botulism has been documented after traumatic injury involving contamination with soil; in injection drug users, for whom black-tar heroin use has been identified as a risk factor; and after cesarean delivery. The illness has occurred even after antibiotics have been given to prevent wound infection. When present, fever is probably attributable to concurrent infection with other bacteria. The wound may appear benign.
Infant Botulism In infant botulism, the most common form of the disease, toxin is produced in and absorbed from the intestine after the germination of ingested spores. The severity ranges from mild illness with failure to thrive to fulminant severe paralysis with respiratory failure and may be one cause of sudden infant death. The identification of contaminated honey as one source of spores has led to the recommendation that honey not be fed to children60 days before onset of symptoms or diagnosis, the patient's most recent sex partner should be treated. Patients should be instructed to abstain from sexual intercourse until therapy is completed and until they and their sex partners no longer have symptoms. Greater emphasis must be placed on prevention by public health education, individual patient counseling, and behavior modification. Preventing the spread of gonorrhea may help reduce the transmission of HIV. No effective vaccine for gonorrhea is yet available, but efforts to test a porin vaccine candidate are under way. ACKNOWLEDGEMENT The authors acknowledge the contributions of Dr. King K. Holmes and Dr. Stephen A. Morse to the chapter on this subject in the earlier editions. (Bibliography omitted in Palm version) Back to Table of Contents
148. MORAXELLA CATARRHALIS AND OTHER MORAXELLA SPECIES - Daniel M. Musher MORAXELLA CATARRHALIS The gram-negative coccus now known as Moraxella catarrhalis has undergone three changes of name in as many decades. Originally called Micrococcus catarrhalis, it was renamed Neisseria catarrhalis in the 1960s because of its morphologic similarity to Neisseria spp. Then, in 1970, it was elevated to the status of a distinct genus, Branhamella, on the basis of DNA homology. In 1979 this organism was placed into the genus Moraxella, of which Branhamella may be a subgenus. A component of the normal bacterial flora of the upper airways, M. catarrhalis has been increasingly recognized as a cause of otitis media, sinusitis, and bronchopulmonary infection. BACTERIOLOGY AND IMMUNITY On Gram's staining, M. catarrhalis organisms appear as gram-negative cocci, sometimes occurring in pairs and retaining the side-by-side kidney-bean configuration of Neisseria (Plate VI-1). These cocci tend to retain crystal violet during the decolorizing step and may be confused with Staphylococcus aureus. Moraxella colonies grow well on blood or chocolate agar but may be overlooked because of their resemblance to Neisseria spp. (a major component of the normal pharyngeal flora). Moraxella is readily distinguishable from Neisseria spp. by biochemical tests. Strains of M. catarrhalis show a surprising degree of homogeneity in terms of their outer-membrane proteins. Antibody to some of these proteins is generally present in serum of children over the age of 4 years; however, colonizing or disease-causing isolates may survive in serum despite this naturally present antibody and complement. Bactericidal antibody emerges following natural infection and may be directed against one or more conserved outer-membrane proteins -- a property of potential value in vaccine development. The presence of certain outer-membrane proteins is associated with virulence in mice, and antibody may be protective. These proteins are under investigation for use as vaccines. EPIDEMIOLOGY With repeated cultures and the use of selective media, M. catarrhalis can be isolated from the upper respiratory tract or saliva of 50% of healthy schoolchildren and of up to 7% of healthy adults. When conventional microbiologic techniques are used, Moraxella can be isolated from sputum of about 10% of persons who have chronic bronchitis and 25% of those who have bronchiectasis in the absence of acute infection. Investigators in both the northern and southern hemispheres have reported a striking seasonal variation in the isolation of this organism from clinical specimens, with a peak in late winter/early spring and a nadir in late summer/early fall. Direct contact has not been shown to contribute to community-acquired infection, but nosocomial spread of infection has been documented occasionally. OTITIS MEDIA AND SINUSITIS
M. catarrhalis has repeatedly been shown to be the third most common bacterial isolate from middle-ear fluid of children who have otitis media, being surpassed only by Streptococcus pneumoniae and nontypable Haemophilus influenzae. Recent studies have shown that this organism is also a prominent isolate from sinus cavities in acute and chronic sinusitis. PURULENT TRACHEOBRONCHITIS AND PNEUMONIA M. catarrhalis causes acute exacerbations of chronic bronchitis (increased production and/or purulence of sputum), purulent tracheobronchitis (the latter also involving fever and leukocytosis), and pneumonia. The great majority of infected persons are>50 years old and have a long history of cigarette smoking and underlying chronic obstructive pulmonary disease (COPD); many have lung cancer as well. In one study, 76% of affected persons had COPD (severe in many cases), and one-third of those with COPD had lung cancer; most patients also had clinical evidence of malnutrition. In one extensive series of cases, M. catarrhalis pneumonia did not occur in otherwise-healthy hosts. Symptoms of M. catarrhalis infection have been regarded as modest in severity. Both cough and the amount and purulence of sputum are usually increased above baseline. Chills are reported in one-quarter of patients, pleuritic pain in one-third, and malaise in 40%. Most patients have peak temperatures of80% in most developing countries. In the United States, prevalence varies with age; around 50% of 60-year-old persons as opposed to 25% of 30-year-old persons are colonized. Spontaneous acquisition or loss of the bacterium in adulthood is uncommon. H. pylori is usually acquired in childhood. (The age association is mostly due to a birth-cohort effect.) Other than age, the main risk factor for colonization is low socioeconomic status; crowding and low family income in childhood are especially strong correlates of colonization. Humans are the only important reservoir of H. pylori. Members of a family may carry the same strain, and colonization is particularly common in childhood institutions. These findings imply direct person-to-person spread, but whether transmission takes place by the fecal-oral or oral-oral route is unknown. H. pylori DNA has been found in water sources, and indirect epidemiological evidence indicates that contaminated water may lead to human colonization in developing countries. Much research is focused on determining which of these possible routes of acquisition is most important. CLINICAL MANIFESTATIONS Essentially all H. pylori-colonized persons have gastric inflammation, but this condition
in itself is asymptomatic (Fig. 154-1). Symptoms are due to illnesses such as peptic ulceration or gastric malignancy, which develop in fewer than 10% of individuals colonized with H. pylori. More than 80% of peptic ulcers are related to H. pylori colonization, most of the remainder being due to damage caused by aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs). The main lines of evidence for an ulcer-promoting role of H. pylori are (1) that the presence of the organism is a risk factor for the development of ulcers, (2) that (non-NSAID-induced) ulcers rarely develop in the absence of H. pylori, (3) that eradication of H. pylori results in a dramatic drop in the rate of ulcer relapse (from about 80% to 15% in the first year), and (4) that experimental infection of gerbils and mice causes gastroduodenal injury. Prospective case-control studies have shown that H. pylori colonization is a risk factor for adenocarcinomas of the stomach (other than those arising in the gastric cardia). However, persons who have had documented duodenal ulcers are less likely than other persons to develop gastric adenocarcinoma later in life; the implication is that, whereas H. pylori colonization increases risk for both duodenal ulcerogenesis and gastric carcinogenesis, other factors determine which disease path is taken. The presence of H. pylori is strongly associated with gastric lymphoma. Low-grade B-cell mucosa-associated lymphoid tissue (MALT) lymphomas, which are antigen driven, often regress following H. pylori eradication. Whether all such diagnosed cases represent true malignancies remains to be determined. Most H. pylori colonization is asymptomatic. Whether colonization occasionally causes symptoms (nonulcer dyspepsia) in the absence of ulcers or malignancy is controversial. Some but not all trials of H. pylori eradication in nonulcer dyspepsia have shown a reduction of symptoms in a small proportion of patients. As there is no prospective method for identifying this small group, eradication of H. pylori in patients with nonulcer dyspepsia is not currently indicated. Much interest has focused on a possible protective role for H. pylori in gastroesophageal reflux disease (GERD) and adenocarcinoma of the esophagus and gastric cardia. The main lines of evidence for this role are that (1) there is a temporal relationship between a falling prevalence of H. pylori colonization and a rising incidence of these conditions; (2) in most studies, the prevalence of H. pylori colonization, especially with cagA+strains, is lower among patients with these esophageal diseases than among control subjects; and (3) eradication of H. pylori often leads to the development or worsening of GERD or its symptoms. Although there are plausible mechanisms for a protective effect of H. pylori against these diseases, none has yet been definitively identified. Thus a causal link remains probable but unproven. Several extra-gastrointestinal pathologies have been linked epidemiologically with H. pylori colonization. The most notable are ischemic heart disease and cerebrovascular disease. The associations have been found more commonly in small than in large studies, and most authorities consider them to be noncausal and due to confounding factors. PATHOLOGY AND PATHOGENESIS H. pylori colonization induces chronic superficial gastritis, which includes both
mononuclear and polymorphonuclear cell infiltration of the mucosa. (The term gastritis should be used specifically to describe histologic features; it also has been used to describe endoscopic appearances and even symptoms, neither of which have been linked to microscopic findings or to the presence of H. pylori.) The immune response to H. pylori includes both the production of antibody (local and systemic) and a cell-mediated response but is ineffective in clearing the bacterium. H. pylori and associated inflammation are most evident in the stomach but are also found in areas of gastric metaplasia and heterotopia (e.g., the duodenal bulb).The pattern of gastric inflammation is associated with disease risk: antral-predominant gastritis is most closely linked with duodenal ulceration and is common in the United States and other developed countries, whereas the predominant form in developing countries is pangastritis, which is epidemiologically linked with gastric ulceration and adenocarcinoma. Longitudinal analyses of gastric biopsy specimens taken years apart from the same patient show that inflammation may progress to atrophy, intestinal metaplasia, and dysplasia and then (by implication) to carcinoma. Patients with atrophic gastritis are at risk for vitamin B 12deficiency and its associated hematologic and neurologic sequelae. Continuous omeprazole therapy (for example, forGERD) may speed progression to atrophy when H. pylori is present. Most H. pylori-colonized persons do not develop clinical sequelae. That some persons develop overt disease whereas others do not is probably due to a combination of bacterial strain differences, host susceptibility to disease, and environmental factors; of these, bacterial factors are best studied. The two major disease-associated H. pylori virulence factors described so far are a vacuolating cytotoxin, VacA, and a group of genes termed the cag pathogenicity island (cag PaI). VacA occurs in several forms, and its level of production varies between strains; thus, although all strains have the gene (vacA) encoding the protein, not all exhibit vacuolating activity in vitro. Cytotoxic strains are more commonly isolated from patients with peptic ulcer disease than from persons without ulcers. The cag PaI includes genes that confer enhanced virulence on H. pylori strains, at least partly by inducing epithelial cells to produce proinflammatory cytokines. The gene cagA, an imperfect marker for the cag PaI, is useful for epidemiologic studies because it encodes a highly immunogenic protein, CagA. Patients with peptic ulcers or gastric adenocarcinoma are more likely to have CagA antibodies than persons without these conditions. However, patients with esophageal dysplasia or adenocarcinoma or with the premalignant condition Barrett's esophagus are less likely to harbor cagA+strains than are H. pylori-positive controls. Thus, eradication of cagA+strains from asymptomatic persons to prevent disease is not recommended. How does gastric H. pylori colonization increase risk for duodenal ulceration? One explanation is that antral H. pylori colonization diminishes the number of somatostatin-producing cells; somatostatin-mediated inhibition of gastrin release leads to hypergastrinemia. Individuals with antral-predominant gastritis (and thus a normally functioning acid-producing gastric corpus) develop increased acid secretion, which may increase the risk of duodenal ulceration per se or may induce gastric metaplasia in the duodenum, which becomes colonized by H. pylori, then inflamed, and finally ulcerated. After eradication of H. pylori from patients with duodenal ulcer disease, the level of acid secretion often falls.
DIAGNOSIS Tests for H. pylori can be divided into two groups: invasive tests, which require upper gastrointestinal endoscopy and are based on the analysis of gastric biopsy specimens, and noninvasive tests (Table 154-1). Invasive tests are preferred for (1) the initial management of dyspeptic patients, because the decision of whether or not to eradicate H. pylori depends on ulcer disease status, and (2) follow-up after treatment of patients with gastric ulceration to be certain that the ulcer was not malignant. Follow-up endoscopy should be performed at least 4 weeks after cessation of all anti-Helicobacter drugs, since at earlier points the H. pylori load may be low and tests may be falsely negative. The most convenient endoscopy-based test is the biopsy urease test, in which two antral biopsy specimens are put into a gel containing urea and an indicator. The presence of H. pylori urease elicits a color change, which often takes place within minutes but can require up to 24 h. Histologic examination of biopsy specimens is accurate, provided that a special stain (e.g., a modified Giemsa or silver stain) permitting optimal visualization of H. pylori is used. Histologic study yields additional information, including the degree and pattern of inflammation, atrophy, metaplasia, and dysplasia, although these details are rarely of clinical use. Microbiologic culture is most specific but may be insensitive due to difficulty with H. pylori isolation. Once cultured, the identity of H. pylori can be confirmed by its typical appearance on Gram's stain and its positive reactions in oxidase, catalase, and urease tests. Antibiotic sensitivities also can be determined. Specimens containing H. heilmanii are only weakly positive in the biopsy urease test. The diagnosis is based on visualization of the characteristic long, tight spiral bacteria in histologic sections. The simplest tests for H. pylori infection are serologic, involving the assessment of specific IgG levels in serum. The best of these tests are as accurate as other diagnostic methods, but many commercial tests, especially rapid office tests, perform poorly. In quantitative tests, a defined drop in antibody titer between matched serum samples taken before and 6 months after treatment (no sooner because of the slow decline in antibody titer) accurately indicates that H. pylori infection has been eradicated. The other major noninvasive tests are the13C and14C urea breath tests. In these simple tests, the patient drinks a labeled urea solution and then blows into a tube. The urea is labeled with either the nonradioactive isotope13C or a minute dose of the radioactive isotope14C (which exposes the patient to less radiation than a standard chest x-ray). If H. pylori urease is present, the urea is hydrolyzed and labeled carbon dioxide is detected in breath samples. Unlike serologic tests, urea breath tests can be used to assess the outcome of treatment 1 month after its completion and thus may replace endoscopy for this purpose in the follow-up of duodenal ulcer patients. As for endoscopic tests, all anti-Helicobacter drugs should be avoided in this period or the test may be falsely negative. TREATMENT At present, the only clear indications for treatment are H. pylori-related duodenal and gastric ulceration and the rare low-grade B-cellMALTlymphoma. H. pylori should be eradicated in patients with documented ulcer disease, whether or not the ulcers are currently active, to reduce the likelihood of relapse. At present, treatment is not
recommended for nonulcer dyspepsia or for prophylaxis against ulcers or gastric adenocarcinoma (although it may be reasonable to eradicate H. pylori in persons with a strong family history of gastric cancer). Reasons for avoiding treatment for these other potential indications include the expense, the induction of morbidity in otherwise healthy people, the risk of inducing widespread antibiotic resistance in H. pylori and in other colonizing bacteria, and the risk of inducing or worseningGERD. H. pylori is susceptible to a wide range of antibiotics in vitro, but monotherapy has been disappointing in vivo, probably because of inadequate antibiotic delivery to the full locus of colonization. Failure of monotherapy has led to the development of multidrug regimens, the most successful of which are triple and quadruple combinations that achieve H. pylori eradication rates of >90% in many trials and>75% in clinical practice. Current 7- to 14-day drug regimens consisting of a proton pump inhibitor and two or three antimicrobial agents often require only twice-daily dosing (Table 154-2). The two most important goals in H. pylori eradication are to obtain the patient's compliance with the dosing regimen and to use drugs to which H. pylori has not acquired resistance. Treatment failure following minor lapses in compliance is common and often leads to acquired resistance to metronidazole or clarithromycin. To stress the importance of compliance, written instructions should be given to the patient, and minor side effects of the regimen should be explained. Resistance to metronidazole and clarithromycin is of growing concern; however, in multidrug regimens, the clinical significance of single-drug resistance is diminished. Assessment of antibiotic susceptibilities before treatment would be optimal but is not usually undertaken. In the absence of susceptibility information, a history of antibiotic use should be obtained, and, if resistance is likely, metronidazole-containing regimens should be avoided. Metronidazole resistance is common among persons who have taken the agent previously, even years earlier, for other conditions such as giardiasis or trichomoniasis. If initial H. pylori treatment fails, compliance should be checked and re-treatment should be based on known antibiotic susceptibilities. When this information cannot be obtained, the recommended course is quadruple therapy without clarithromycin (if a clarithromycin-containing regimen was given first) or triple therapy with omeprazole/clarithromycin/amoxicillin (if clarithromycin has not been used) (Table 154-2). Given the high efficacy of treatment regimens, it is unclear whether the success of attempted H. pylori eradication should be checked. For gastric ulceration, the opportunity to retest for H. pylori is present at the repeat endoscopy, which is performed to evaluate healing. For duodenal ulceration, although many clinicians prefer to retest only if symptoms recur, a urea breath test or endoscopy should be performed no sooner than 1 month after treatment. This test will provide reassurance if treatment has been successful and will prompt re-treatment in cases of persistence. Clearance of H. heilmanii has been described following the use of bismuth compounds alone or triple-therapy regimens. However, in the absence of trials, it is unclear whether this result represents successful treatment or natural clearance of the bacterium. PREVENTION
Carriage of H. pylori has public health significance in developing countries, where gastric adenocarcinoma is a common cause of cancer death. However, H. pylori has co-evolved with its human host over millennia, and there may be disadvantages in preventing or eliminating colonization. For example, as has been mentioned, the absence of H. pylori appears to increase the risk of developingGERD and esophageal adenocarcinoma. If mass prevention were contemplated, vaccination would be preferred, and experimental immunization of animals has given promising results. However, in the United States and other developed countries, the incidences of H. pylori carriage, peptic ulceration, and gastric adenocarcinoma are dropping. Thus, prevention of colonization in these countries may be unnecessary or even unwise. (Bibliography omitted in Palm version) Back to Table of Contents
155. INFECTIONS DUE TO PSEUDOMONAS SPECIES AND RELATED ORGANISMS - Christopher A. Ohl, Matthew Pollack Pseudomonas species and phylogenetically related bacteria are ubiquitous, free-living, opportunistic gram-negative pathogens. P. aeruginosa, the most common human pathogen in this group, is the primary subject of this chapter. Also discussed are two pathogens of increasing importance: Burkholderia cepacia (formerly P. cepacia), primarily an opportunistic pathogen, and Stenotrophomonas maltophilia (formerly Xanthomonas maltophilia), which principally infects hospitalized patients. In addition, melioidosis, a tropical systemic disease with acute and chronic manifestations caused by B. pseudomallei (formerly P. pseudomallei), will be considered. INFECTIONS DUE TO P. AERUGINOSA P. aeruginosa is a small, nonsporulating, aerobic gram-negative rod belonging to the family Pseudomonadaceae. It is motile by virtue of its single polar flagellum. More than half of all clinical isolates produce the blue-green pigment pyocyanin; this pigment is helpful in the identification of the organism and accounts for the species name aeruginosa, which refers to the distinctive color of copper oxide. EPIDEMIOLOGY P. aeruginosa is widespread in nature, inhabiting soil, water, plants, and animals (including humans). It has a predilection for moist environments. This organism occasionally colonizes the skin, external ear, upper respiratory tract, or large bowel of healthy humans. Rates of carriage are relatively low, however, except among patients who have serious underlying disease, whose host defenses have been naturally or iatrogenically compromised, who have previously received antibiotic therapy, and/or who have been exposed to the hospital environment. Under these circumstances, colonization with P. aeruginosa frequently precedes infection, and factors that predispose to the former also increase the likelihood of the latter. Most P. aeruginosa infections are acquired in the hospital, where intensive care units account for the highest rates of infection. According to the National Nosocomial Infections Surveillance (NNIS) system, between 1992 and 1999, P. aeruginosa was the second most common cause of pneumonia, the fourth most common cause of urinary tract infection, and the sixth most frequent bloodstream isolate in intensive care units. Many potential reservoirs of infection have been identified in the hospital environment, including respiratory equipment, cleaning solutions, disinfectants, sinks, vegetables, flowers, endoscopes, and physiotherapy pools. Most reservoirs are associated with moisture. It is assumed that the organism is transmitted to patients via the hands of hospital personnel or via fomites. While some infecting strains of P. aeruginosa appear to be endemic within the hospital, others are traced to a common source associated with a specific outbreak or epidemic. Epidemiologic investigation is facilitated by serotyping (immunotyping) of strains on the basis of differences in lipopolysaccharide (LPS) structure and by the use of molecular techniques such as pulsed-field gel electrophoresis. PATHOGENESIS
That the pathogenesis of infections due to P. aeruginosa is complex is evidenced by the clinical diversity of the diseases related to this organism and by the multiplicity of virulence factors it produces. P. aeruginosa rarely causes disease in the healthy host. Relative risk for infection is greatly increased, however, when normal cutaneous or mucosal barriers have been breached or bypassed, when immunologic defense mechanisms have been compromised, or when the protective function of the normal bacterial flora has been disrupted (Table 155-1). The ubiquity of the organism, its flexible nutritional and metabolic requirements, its environmental resiliency, and its relative resistance to antibiotics help account for the frequency and success with which it acts as an opportunistic pathogen. Infections caused by P. aeruginosa usually begin with bacterial attachment and superficial colonization of cutaneous or mucosal surfaces and progress to localized bacterial invasion and damage to underlying tissues. The infection may remain anatomically localized or may spread by direct extension to contiguous structures. This process may continue with bloodstream invasion, dissemination, the systemic inflammatory-response syndrome (SIRS), multiple-organ dysfunction, and ultimately death. Not only is local infection more likely to occur in immunocompromised hosts, such as those with profound neutropenia, but it is more likely to culminate in bloodstream invasion and dissemination.LPS(endotoxin), which is a structural component of the bacterial outer membrane, is thought to play a pivotal role in the pathogenesis of the sepsis syndrome or SIRS. The initial attachment of P. aeruginosa to the respiratory epithelium and other epithelial surfaces appears to be mediated by bacterial organelles called pili or fimbriae and facilitated by alginate, a mucoid exopolysaccharide produced by most strains of the bacterium under appropriate environmental conditions. Alginate plays an important role in colonization and infection of the respiratory tract in patients with cystic fibrosis and in the formation of biofilms within which sessile colonies of P. aeruginosa enjoy relative protection from host defenses and antimicrobial agents. P. aeruginosa produces a number of extracellular virulence factors, including alkaline protease, elastase, phospholipase, cytotoxin, and exoenzymes (or exotoxins) A and S. The breakdown of host tissues by these bacterial products creates conditions conducive to enhanced bacterial proliferation, invasion, and tissue injury. Production and secretion of many of these extracellular virulence factors are under the regulatory control of a cell-to-cell signaling system that has been termed quorum sensing. Through lactones and other signal molecules secreted by individual P. aeruginosa bacteria, the entire bacterial population is able to sense its environment, communicate, and discern its own cell density. This regulatory system conceivably allows P. aeruginosa to produce extracellular virulence factors in a coordinated manner dependent on cell density and may give the pathogen an appreciable advantage over host defense mechanisms. The extracellular enzyme exotoxin A -- a diphtheria-like toxin -- is produced by most clinical isolates of P. aeruginosa. Exotoxin A inhibits mammalian protein synthesis by transferring the adenosine diphosphate (ADP) ribose moiety of the nicotinamide adenine dinucleotide into covalent linkage with elongation factor 2, inactivating this factor's ability to catalyze the elongation step in polypeptide assembly. Another
extracellular cytotoxin, exoenzyme S, is also an adenosine diphosphate ribosyltransferase but, unlike exotoxin A, preferentially ribosylates guanosine triphosphate-binding proteins, resulting in disruption of host cell actin cytoskeletons. Exoenzyme S is one of several extracellular virulence factors of P. aeruginosa that is directly introduced from the bacterial cytosol into the host cell cytoplasm via a complex array of transmembrane proteins termed the type III secretion apparatus. This process requires direct cell contact and allows the injection of virulence factors from the bacterium into host cells without interference from humoral immune defenses. CLINICAL MANIFESTATIONS AND DIAGNOSIS Respiratory Tract Infections Primary pneumonia, or nonbacteremic pneumonia, results from aspiration of upper respiratory tract secretions; often develops in patients with chronic lung disease, congestive heart failure, or AIDS; and is most common in an intensive care setting in association with mechanical ventilator use. Fever, chills, severe dyspnea, cyanosis, productive cough, apprehension, confusion, and other signs of severe systemic toxicity characterize this acute, often life-threatening infection. Chest roentgenograms typically show bilateral bronchopneumonia with nodular infiltrates and small areas of radiolucency; pleural effusions are common; empyema is relatively uncommon; and lobar consolidation is occasionally seen. Cavitary lesions are particularly common in AIDS patients with P. aeruginosa pneumonia. Pathologic lesions include alveolar necrosis, focal hemorrhages, and microabscesses. Bacteremic pneumonia due to P. aeruginosa begins as a respiratory infection but, in contrast to primary pneumonia, is typically associated with neutropenia, subsequent bloodstream invasion, and metastatic spread that produces characteristic lesions in the lungs and other viscera. Alveolar hemorrhage and necrosis are common. The signs and symptoms of this fulminant disease include those described for nonbacteremic pneumonia caused by this organism as well as those associated with gram-negative sepsis. Chest roentgenograms characteristically demonstrate a rapid progression from pulmonary vascular congestion to interstitial edema, then to pulmonary edema, and finally to diffuse necrotizing bronchopneumonia with cavity formation. The patient typically dies 3 or 4 days after initial presentation. Chronic infection of the lower respiratory tract with P. aeruginosa is caused almost exclusively by mucoid strains, which produce alginate. Such infection is prevalent among older children and young adults with cystic fibrosis and also develops in some patients with AIDS. In patients with cystic fibrosis, mucoid strains invariably colonize and infect patients with increasing prevalence over time, contributing to the acute exacerbations and chronic progression that characterize pulmonary disease in these individuals. Airway obstruction appears to begin with bronchiolitis, which causes mucus plugging and predisposes to P. aeruginosa infection. The infection produces more mucus plugging, chronic suppuration, bronchiectasis, atelectasis, and ultimately fibrosis. This process progresses to pulmonary insufficiency, hypoxemia, and alterations in cardiopulmonary dynamics resulting in pulmonary hypertension and cor pulmonale. Clinical manifestations of lower respiratory tract infections due to P. aeruginosa in patients with cystic fibrosis vary with the severity and duration of underlying lung disease, the frequency and intensity of acute episodes, and the presence of coinfecting
pathogens such as B. cepacia (Chap. 257). Soon after colonization, patients may experience recurrent upper respiratory symptoms followed by a lingering cough. Episodes of pneumonia develop later, with persistent cough between acute episodes. Eventually, patients exhibit a chronic productive cough, wheezing, diminished appetite, weight loss, growth retardation, and decreased activity. Acute exacerbations are typically accompanied by low-grade fever and heightened respiratory symptoms. Physical signs include evidence of malnutrition, an increase in anteroposterior diameter, intercostal retractions, cyanosis, inspiratory and expiratory wheezing, rhonchi, moist rales, abdominal distention, and clubbing of the fingers and toes. Laboratory abnormalities include leukocytosis with a left shift and hypoxemia with or without hypercarbia. Tests of pulmonary function demonstrate obstructive and restrictive defects. Chest roentgenograms reveal overaeration, patchy atelectasis, peribronchial fibrosis, and patchy infiltrates associated with pneumonia. In more advanced disease, there may be evidence of severe overaeration, depressed diaphragm, further increased anteroposterior diameter, extensive peribronchial infiltration, generalized bronchiectasis, and cyst formation. Bacteremia P. aeruginosa remains an important cause of life-threatening bloodstream infection in immunocompromised patients. Bacteremia is frequently iatrogenic and is usually seen in hospitalized patients with various comorbid conditions (Table 155-1). Bloodstream infection can be either primary (with no identifiable source) or secondary to a discrete focus of infection. The clinical features of P. aeruginosa bacteremia are similar to those of other forms of bacteremia. Common primary sites of infection include the urinary tract, gastrointestinal tract, lungs, skin and soft tissues, and intravascular foci, including indwelling central venous catheters. Fever, tachypnea, tachycardia, and prostration are common. Disorientation, confusion, or obtundation may be evident. Hypotension can progress to refractory shock. Renal failure, adult respiratory distress syndrome, and disseminated intravascular coagulation occur as complications. Pathognomonic skin lesions termed ecthyma gangrenosum (Fig. 19-CD1) develop in a relatively small minority of patients with P. aeruginosa bacteremia. The lesions begin as small hemorrhagic vesicles surrounded by a rim of erythema and undergo central necrosis with subsequent ulceration (seePlate IID-57C). They occur singly or in small numbers on the perineum, buttocks, and extremities; in the axillae; or elsewhere. Histologically, these lesions contain numerous bacteria invading blood vessels but few inflammatory cells. Bacteria are readily visible on Gram's staining and may be cultured from aspirated material. Endocarditis P. aeruginosa infects native heart valves in injection drug users as well as prosthetic heart valves. The source of P. aeruginosa strains infecting drug users appears to be standing water contaminating drug paraphernalia. Foreign materials mixed with heroin may cause injury to valve leaflets or mural endocardium, with resulting fibrosis and an increased risk for valve infection. Exposure of the tricuspid valve to both trauma and bacteria apparently accounts for the high incidence of tricuspid involvement in association with injection drug use. The pulmonic, mitral, or aortic valve and the mural endocardium of either atrium may be
affected in P. aeruginosa endocarditis. Multiple-valve infections are common. Tricuspid or right-sided involvement is often associated with septic pulmonary emboli. Right-sided P. aeruginosa endocarditis usually presents subacutely, while the appearance of left-sided disease is likely to be more acute or even fulminant. Fever is virtually invariable, and murmurs are usually detectable at initial presentation or shortly thereafter. Septic pulmonary emboli associated with right-sided disease result in cough, pleuritic chest pain, sputum production, pulmonary infiltration (with or without abscess formation), and pleural effusion. Left-sided infections may present as intractable heart failure or large systemic emboli. Mycotic aneurysms, cerebritis, or brain abscess may occur; septic infarcts are occasionally found in the spleen. Skin and soft tissue manifestations, including Janeway lesions, Osler's nodes, and ecthyma gangrenosum, are relatively uncommon. The diagnosis of P. aeruginosa endocarditis is based on positive blood culture in the absence of an extracardiac source; an indication of valvular dysfunction or vegetation on an echocardiogram; evidence of septic pulmonary lesions on a chest roentgenogram (in right-sided disease); and the actual demonstration of infected heart valves at the time of surgery. Central Nervous System Infections P. aeruginosa infections of the central nervous system include meningitis and brain abscess. These infections follow extension from a contiguous parameningeal structure such as the ear, mastoid, or paranasal sinus; direct inoculation into the subarachnoid space or brain through head trauma, surgery, or diagnostic procedures; or bacteremic spread from infection at a distant site. Like P. aeruginosa infections at other anatomic sites, central nervous system infections are documented almost exclusively in patients with compromised local or systemic immune-defense mechanisms. The clinical signs of P. aeruginosa meningitis, like those of other forms of acute bacterial meningitis, include fever, headache, stiff neck, confusion, and obtundation. The onset of illness may be acute or even fulminant, particularly in bacteremic patients, with a precipitous downhill course, shock, coma, and early death. In nonbacteremic patients, P. aeruginosa meningitis or brain abscess may present more insidiously, with a paucity of systemic symptoms. This presentation is especially common in infections resulting from recent neurosurgery, cancer of the head and neck, or direct extension from a parameningeal focus of chronic infection. Occasionally, P. aeruginosa meningitis runs a subacute or relapsing course that is thought to be related to the intermittent release of bacteria from a loculated site of infection. Ear Infections P. aeruginosa is often found in the external auditory canal, particularly under moist conditions and in the presence of inflammation or maceration (as in "swimmer's ear"). Moreover, this organism is the predominant pathogen associated with external otitis, a usually benign inflammatory process affecting the external auditory canal. The ear is painful or merely itchy, there is a purulent discharge, and pain is elicited by pulling on the pinna. The external canal appears edematous and is filled with detritus that often prevents visualization of the tympanic membrane. P. aeruginosa occasionally penetrates the epithelium overlying the floor of the external auditory canal at the junction between bone and cartilage and invades underlying soft
tissue. The ensuing invasive process, which involves soft tissue, cartilage, and cortical bone, is typically slow but destructive. Termed malignant external otitis, this condition occurs predominantly in elderly diabetic patients but is reported occasionally in infants with other underlying diseases and rarely in elderly nondiabetic patients. Virtually all cases of malignant external otitis are caused by P. aeruginosa. From the external ear, the infection advances to the retromandibular area or parotid space and enters the mastoid air cells and temporal bone. Advancing osteomyelitis at the base of the skull often involves the seventh, ninth, tenth, and eleventh cranial nerves. The cavernous sinus can become involved, as can the contralateral petrous apex. The middle ear is commonly spared; meningitis and brain abscess are relatively rare complications. Otorrhea and severe otalgia are common presenting symptoms of malignant external otitis. Facial-nerve paralysis tends to occur early, while other cranial-nerve palsies appear later. There may be a loss of hearing. Constitutional symptoms such as fever and weight loss are relatively uncommon. Physical examination almost always reveals remarkable tenderness of the pinna of the ear and abnormalities of the external auditory canal, including swelling, erythema, purulent discharge, debris, and granulation tissue in the canal wall. The tympanic membrane is often hidden from view and is sometimes perforated. Inflammation may involve the pinna as well as the periauricular, retromandibular, and mastoid areas. Peripheral leukocytosis is relatively infrequent in malignant external otitis, while the erythrocyte sedimentation rate is usually markedly elevated. Cerebrospinal fluid occasionally exhibits pleocytosis and an elevation in the protein level. Computed tomography (CT) or magnetic resonance imaging (MRI) of the mastoid or temporal bone typically reveals bony erosions and new bone formation, while the floor of the skull may have soft tissue densities associated with areas of cellulitis. In addition, technetium 99m bone scans and gallium 67 scans frequently give positive results. Cultures of samples from the external auditory canal and of surgical specimens are almost always positive for P. aeruginosa. Eye Infections (See alsoChap. 28) P. aeruginosa causes bacterial keratitis or corneal ulcer and endophthalmitis in the human eye. Keratitis due to P. aeruginosa may result from even minor corneal injury, which interrupts the integrity of the superficial epithelial surface and permits bacterial access to the underlying stroma. Corneal ulcer may complicate contact lens use, particularly when extended-wear soft contact lenses are involved. Contact lens solutions or the lenses themselves may be the source of the organism, which is probably inoculated into the eye at sites of minor lens-induced corneal damage. Patients who have sustained serious burns, have undergone ocular irradiation or tracheostomy, have been exposed to the intensive care environment, and/or are in a coma are also susceptible to P. aeruginosa-associated corneal ulcers. P. aeruginosa keratitis usually starts as a small central ulcer; spreads concentrically to involve a large portion of the cornea, sclera, and underlying stroma; and in some cases progresses to posterior corneal perforation. The clinical manifestations of P. aeruginosa keratitis include a rapidly expanding, necrotic stromal infiltrate in the bed of an epithelial injury; surrounding epithelial edema; an anterior chamber reaction; and mucopurulent discharge adherent to the ulcer's surface. Corneal ulcer due to P. aeruginosa may advance rapidly to involve the entire
cornea in £2 days or may evolve subacutely over several days. Systemic symptoms are uncommon. Complications include corneal perforation, anterior chamber involvement, and endophthalmitis. P. aeruginosa endophthalmitis is typically a rapidly progressive, sight-threatening condition that demands immediate therapeutic intervention. It may complicate penetrating injuries of the eye, intraocular surgery, hematogenous spread from other sites of Pseudomonas infection, or posterior perforation of corneal ulcers. Clinical manifestations may include eye pain, conjunctival hyperemia, chemosis, lid edema, decreased visual acuity, hypopyon, severe anterior uveitis, and signs of possible vitreous involvement. Panophthalmitis may result from this intraocular infection. Bone and Joint Infections Vertebral osteomyelitis due to P. aeruginosa is associated with complicated urinary tract infection, genitourinary instrumentation or surgery, and injection drug use. Vertebral infections that are associated with a urinary tract source most often develop in the elderly and usually affect the lumbosacral spine. Presumably the route of infection in these patients is a shared venous plexus between the pelvis and spine. Injection drug use-related infections typically occur in younger patients and may affect the cervical or lumbosacral spine. P. aeruginosa vertebral osteomyelitis is usually an indolent disease. Accordingly, symptoms may develop weeks or even months before diagnosis. Back or neck pain is generally reported, while fever and systemic symptoms are relatively uncommon. Local tenderness and decreased range of motion of the affected spine are typical. Leukocytosis may be noted, the erythrocyte sedimentation rate is almost always markedly elevated, and blood cultures are sometimes positive. Roentgenograms reveal loss of bone density, narrowed intervertebral space, destruction of vertebral end plates, lytic lesions of vertebral bodies, sclerosis, and occasionally osteophyte formation.CT andMRI are the most sensitive and specific means of defining lesions. Technetium bone scans and gallium scans usually yield positive results. An etiologic diagnosis requires the culture of material obtained by needle aspiration or biopsy of the affected spine under fluoroscopic guidance; open biopsy is sometimes needed. Sternoclavicular pyarthrosis caused by P. aeruginosa is another complication of injection drug use; in some cases it is associated with P. aeruginosa endocarditis, but more often it is not. Joint involvement is usually monoarticular, with the sternoclavicular joint more often affected than sternochondral joints. Patients present with acute or chronic pain in the anterior chest wall, often associated with fever and restricted movement of the homolateral shoulder. Physical examination reveals tenderness, erythema, and swelling over the affected joint. Leukocytosis is common, and the erythrocyte sedimentation rate is almost invariably elevated. Roentgenograms show soft tissue edema; bone demineralization; lytic lesions; and periosteal elevation of the clavicular head, rib, or sternum. Material obtained by arthrocentesis or synovial biopsy yields P. aeruginosa in culture. P. aeruginosa infections of the symphysis pubis are associated with pelvic surgery and injection drug use. The symphysis pubis, like other fibrocartilaginous joints, exhibits a peculiar susceptibility to bloodborne infection with P. aeruginosa. Affected patients report pain in the groin, hip, thigh, and/or lower abdomen that is made worse by walking. Fever is variable, and the duration of symptoms before diagnosis ranges from days to
months. The erythrocyte sedimentation rate is markedly elevated. Roentgenography orCT shows irregularities of the pubic margins, separation of the symphysis pubis, and osteomyelitic abnormalities of the pubic rami that may be extensive. Bone scans are usually positive. Needle aspiration or biopsy is necessary to obtain material for culture. A positive culture is particularly important for the discrimination of P. aeruginosa infections and other pyogenic infections from osteitis pubis, which is thought to be a noninfectious condition complicating pelvic surgery, childbirth, or trauma. P. aeruginosa osteochondritis of the foot follows puncture wounds of the foot, primarily in children. The organism infects the small joints and bones, including the proximal phalanges, metatarsals, metatarsophalangeal joints, tarsal bones, and calcaneus. On average, local pain and swelling last for several weeks, and systemic symptoms are usually lacking. There may be plantar cellulitis over the involved area or tenderness upon deep palpation. Results of roentgenograms and bone scans are generally positive. Aspiration of the affected joint frequently yields purulent material in which P. aeruginosa can be demonstrated by Gram's staining and by culture. P. aeruginosa is one of the most common causative agents in a variety of other, less specific syndromes involving nonhematogenous infections of bones and joints and collectively referred to as chronic contiguous osteomyelitis. These infections may result, for example, from compound fractures, contamination associated with open reduction and fixation of closed fractures, sternotomy performed in conjunction with cardiac surgery, contiguous spread from infected ischemic ulcers related to peripheral vascular disease or diabetes mellitus, and cellulitis in general. The chronicity, indolence, and heterogeneity of these infections explain their varied clinical manifestations and the frequent need for complicated long-term management. Urinary Tract Infections P. aeruginosa is one of the most common causes of complicated and nosocomial infections of the urinary tract. These infections may result from urinary tract catheterization, instrumentation, surgery, or obstruction; they may arise from persistent foci (e.g., the prostate or stones) and may be chronic or recurrent. The urinary tract may be a target for bloodborne infection in patients with P. aeruginosa bacteremia but more often is the source of bacteremia. Chronic P. aeruginosa infections of the urinary tract are relatively common among patients with indwelling urinary catheters, altered urinary tract anatomy secondary to diversionary procedures, and paraplegia. The clinical features of urinary tract infections due to P. aeruginosa are usually indistinguishable from those of other bacterial infections. However, P. aeruginosa infections exhibit a propensity for persistence, chronicity, resistance to antibiotic therapy, and recurrence. More unusual forms of urinary tract involvement peculiar to P. aeruginosa include (1) ulcerative lesions of the renal pelvis, ureters, and bladder that cause sloughing of vesical membranes in the urine; and (2) ecthyma-like lesions of the renal cortex that are seen in association with Pseudomonas sepsis. Skin and Soft Tissue Infections As indicated above, P. aeruginosa bacteremia may be associated with the disseminated skin lesions of ecthyma gangrenosum (seePlate IID-57C). Less common skin manifestations of P. aeruginosa sepsis include vesicular or pustular lesions, bullae, subcutaneous nodules, deep abscesses, and cellulitis.
Metastatic lesions of the skin or mucous membranes complicate Pseudomonas sepsis and occasionally produce massive necrosis or gangrene of the extremities, perineum, face, or oropharynx. Primary P. aeruginosa pyoderma occurs when the skin breaks down secondary to trauma, burn injury, dermatitis, or ulcers related to peripheral vascular disease or pressure sores. Moist conditions and neutropenia may predispose to this condition. The clinical appearance of primary P. aeruginosa pyoderma, which frequently includes hemorrhage and necrosis, resembles that of metastatic P. aeruginosa skin lesions. Histologic studies document vascular invasion by bacteria in both diseases. A rare distinguishing feature of P. aeruginosa pyoderma is its association with a blue-green exudate and a characteristic fruity odor. P. aeruginosa wound sepsis complicating extensive third-degree burn injuries is associated with an extremely high mortality rate. This infection results from colonization of the burn site or burn eschar, invasion of the subeschar space and underlying dermis, vascular invasion, and systemic spread. The development and progression of P. aeruginosa burn wound sepsis are facilitated by the injury-associated breakdown of normal skin, selection of empirical antibiotics with inadequate coverage for this pathogen, and burn-related immune defects. Local manifestations include black, dark brown, or violaceous discoloration of the burn eschar; degeneration of underlying granulation tissue, hemorrhage, and premature eschar separation; edema, hemorrhage, and necrosis of skin adjacent to the burn site; and erythematous nodular lesions in unburned skin. Systemic manifestations include fever or hypothermia and other signs of sepsis,SIRS, or multiple-organ system failure. The diagnosis of P. aeruginosa burn sepsis is based on these local and systemic clinical manifestations and on a burn wound biopsy that reveals both>105colony-forming units of P. aeruginosa per gram of tissue and histologic evidence of bacterial invasion of unburned tissue, vasculitis, or intense inflammation at the burn margin. P. aeruginosa causes diffuse pruritic maculopapular and vesiculopustular rashes associated with exposure to contaminated hot tubs (Fig. 128-CD4), spas, whirlpools, and swimming pools. Many cases of P. aeruginosa dermatitis have occurred in conjunction with a common-source outbreak. At least two nosocomial common-source outbreaks -- one related to a physiotherapy pool -- have been reported. Skin rashes may be limited to areas covered by swimsuits or may be more diffuse, sparing only the head and neck. Low-grade fever or other associated symptoms are uncommon. The illness is usually self-limited, and the rash resolves without specific therapy after cessation of exposure. P. aeruginosa Infections in Patients with AIDS During the 1980s and 1990s, P. aeruginosa infections were increasingly associated with AIDS. The vast majority of these infections are currently seen in patients with advanced AIDS, previous opportunistic infections, and CD4+ lymphocyte counts50% of three or more blood cultures are positive, an endovascular infection should be suspected. Preexisting valvular heart disease is a strong risk factor for the development of endocarditis, while atherosclerotic plaque, prosthetic grafts, and aortic aneurysms are associated with arteritis. Arteritis should be suspected in elderly patients who have a history of prolonged fever with associated back, chest, or abdominal pain preceded by gastroenteritis. Endocarditis and arteritis are rare (5 years but was not well sustained in younger vaccinees. The second approach is that of a live attenuated vaccine strain developed, for example, by the isolation or creation of mutants lacking activeCTX. Three criteria must be met in live vaccine design: The vaccine strain must induce protective immunity, it must be safe to administer, and it must be minimally reactogenic. Safety criteria include the vaccine strain's potential to regain virulence, either spontaneously or via horizontal gene transfer from environmental strains, as well as its potential to donate virulence genes to other strains. Reactogenicity refers to its potential to cause symptoms such as fever or diarrhea in vaccinees. Strain CVD 103-HgR, an oral live cholera vaccine licensed for immunization of travelers in Europe, is derived from a classical biotype strain of V. cholerae by the deletion of theCTX A subunit gene and the insertion in the hemolysin gene of a mercury resistance
marker. This strain has been extensively tested in volunteers; although it is poorly excreted in the stool of human vaccinees, a single dose produces a significant increase in the titer of vibriocidal antibody in ~75% of recipients, including children between the ages of 2 and 4 years, with almost no reactogenicity. Studies in volunteers demonstrate that this vaccine is more effective against classical than against El Tor cholera. Unfortunately, in a large field trial in Indonesian children, this vaccine failed to induce protection against clinical cholera. Other live attenuated vaccine candidate strains have been prepared from El Tor and O139 V. cholerae. In studies in volunteers, these vaccine strains have often exhibited significant reactogenicity whose cause (given the absence of activeCTX) is unclear. Reactogenicity may result from the production of another toxic moiety (e.g., the hemagglutinin/protease or the RTX toxin) by the live attenuated strain. Alternatively, intestinal colonization itself may result in reactogenicity. These El Tor- and O139-derived live vaccine strains are therefore at least several years away from potential licensing. Because of the minimal efficacy of existing parenteral vaccines, cholera immunization is recommended for U.S. travelers only if it is mandated by the countries they plan to visit. OTHER VIBRIO SPECIES In recent years, the taxonomic, epidemiologic, pathophysiologic, and clinical features of vibrios that do not cause clinical cholera have become increasingly well understood. Ten human pathogens are currently recognized in the genus Vibrio. Included are species associated primarily with gastrointestinal illness (V. parahaemolyticus, non-O1 V. cholerae, V. mimicus, V. fluvialis, V. hollisae, and V. furnissii) and species associated primarily with soft tissue infections (V. vulnificus, V. alginolyticus, and V. damsela). In addition, V. vulnificus has emerged as a cause of primary sepsis in certain compromised hosts. Vibrios are abundant in coastal waters the world over and tend to concentrate in the tissues of filter-feeding mollusks. Under optimal conditions, some can double in number in as little as 9 min. Consequently, seawater and raw or undercooked shellfish are important sources of human infection (Table 159-3). Vibrios grow best at temperatures of 28°C to 44°C but not at all below 4°C or above 60°C. Most can be cultured on blood or MacConkey agar, each of which contains enough salt to support the growth of the halophilic organisms (³0.5%). As with V. cholerae,TCBS is the best selective medium. The species can be differentiated in the laboratory by standard biochemical tests. The most important members of the group are V. parahaemolyticus and V. vulnificus. These and selected other species are considered below in greater detail. SPECIES ASSOCIATED PRIMARILY WITH GASTROINTESTINAL ILLNESS V. parahaemolyticus First implicated as a cause of enteritis by Japanese workers in 1953, V. parahaemolyticus is now recognized as an important intestinal pathogen in many parts of the world. In one study from Japan, 24% of reported cases of food poisoning were attributed to this organism, presumably owing to the widespread consumption of raw seafood there. In the United States, V. parahaemolyticus has been responsible for several well-documented common-source outbreaks of diarrhea, typically linked to ingestion of undercooked or improperly handled seafood or of other
foods that have been contaminated by seawater. Most reports have come from the Atlantic Coast, the Gulf of Mexico, and Hawaii. The organism is ubiquitous in marine environments and is able to grow in saline concentrations as high as 8 to 10%. The ability to cause hemolysis on Wagatsuma agar (known as the Kanagawa phenomenon) is closely linked to enteropathogenicity. In one study, 96.5% of isolates from patients with diarrhea were hemolytic versus only ~1% of isolates from seawater. Hemolysis is attributed to a 42-kDa heat-stable protein, the exact pathophysiologic role of which is uncertain. The mechanism by which V. parahaemolyticus causes diarrhea is not clear. V. parahaemolyticus has been associated with two distinct gastrointestinal presentations. The more common is a syndrome of watery diarrhea, accompanied in most cases by abdominal cramps, nausea, and vomiting and in about one-quarter of cases by fever and chills. The incubation period ranges from 4 h to 4 days, and the symptomatic period lasts for a median of 3 days. The vast majority of North American cases have been of this type. The less common syndrome is one of dysentery, described in India and Bangladesh and characterized by severe abdominal cramps, nausea, vomiting, and bloody or mucoid stools. Most cases of either type are self-limited and require neither antimicrobial treatment nor hospitalization. Severe infections are associated with underlying diseases, including diabetes, preexisting liver disease, iron-overload states, or immunosuppression. The occasional severe case should be treated with fluid replacement and antibiotics, as described above for cholera. Death is very rare. There are no reliable differential diagnostic features. V. parahaemolyticus should be considered as a possible cause in all cases of diarrhea that can be epidemiologically linked to seafood consumption or to the sea itself. In addition to gastrointestinal disease, V. parahaemolyticus is a rare cause of extraintestinal infections, including wound infections, otitis, and -- very rarely -- sepsis. Non-O1 V. cholerae The heterogeneous non-O1 V. cholerae organisms are biochemically indistinguishable from V. cholerae O1 on routine testing but fail to agglutinate in O1 antiserum. While technically a non-O1 vibrio, V. cholerae O139 Bengal is not grouped with these pathogens because of its potential to cause epidemic cholera, as detailed above. Non-O1 V. cholerae strains have been responsible for several well-described food-borne outbreaks of gastroenteritis as well as for sporadic cases of otitis media, wound infection, and bacteremia. About half of all U.S. isolates are obtained from stool specimens. Like other vibrios, non-O1 V. cholerae organisms are widely distributed in marine environments; unlike most other vibrios, however, they require only trace amounts of NaCl to survive (i.e., they are nonhalophilic). Recognized U.S. cases invariably have been associated either with the consumption of raw oysters or with recent travel, typically to Mexico. The clinical spectrum of diarrheal disease caused by non-O1 V. cholerae is broad and likely reflects the heterogeneous virulence attributes of the group. Occasional isolates make a protein enterotoxin very similar toCTX. Others produce cytotoxins, hemolysins, or invasins. Gastroenteritis due to non-O1 V. cholerae typically has an incubation period of10%. Most clinical isolates come from superinfected wounds, which presumably became contaminated at the beach. Infection varies in severity but is generally not serious and responds well to antibiotic therapy and drainage. A few reports have described otitis externa, otitis media, or conjunctivitis. Therapy with tetracycline is usually curative. V. alginolyticus is a rare cause of bacteremia in immunocompromised hosts. (Bibliography omitted in Palm version) Back to Table of Contents
160. BRUCELLOSIS - M. Monir Madkour, Dennis L. Kasper DEFINITION Brucellosis is a zoonosis transmitted to humans from infected animals. Its clinical features are not disease specific. Brucellosis has many synonyms derived from the geographical regions in which the disease occurs (e.g., Mediterranean fever, Malta fever, Gibraltar fever, Cyprus fever); from the remittent character of its fever (e.g., undulant fever); or from its resemblance to malaria and typhoid (e.g., typhomalarial fever, intermittent typhoid). ETIOLOGY Human brucellosis can be caused by any of four species: Brucella melitensis (the most common and most virulent cause of brucellosis worldwide) is acquired primarily from goats, sheep, and camels; B. abortus from cattle; B. suis from hogs; and B. canis from dogs. These small aerobic gram-negative bacilli are unencapsulated, nonmotile, non-spore-forming, facultative intracellular parasites that cause lifelong infection in animals. Brucellae are killed by boiling or pasteurization of milk and milk products. They survive for up to 8 weeks in unpasteurized, white, soft cheese made from goat's milk and are not killed by freezing. The organisms remain viable for up to 40 days in dried soil contaminated with infected-animal urine, stool, vaginal discharge, and products of conception and for longer periods in damp soil. EPIDEMIOLOGY The global incidence of human brucellosis is not known because of the variable quality of disease reporting and notification systems in many countries. Worldwide, the only countries believed to be free of brucellosis are Norway, Sweden, Finland, Denmark, Iceland, Switzerland, the Czech and Slovak republics, Romania, the United Kingdom (including the Channel Islands), the Netherlands, Japan, Luxembourg, Cyprus, and Bulgaria; the U.S. Virgin Islands are also free of the disease. Reports indicate that, even in developed nations, the true incidence of brucellosis may be up to 26 times higher than official figures suggest. In the United States, about 200 new cases are reported every year; however, it is estimated that only 4 to 10% of cases are recognized and reported. Consumption of imported cheese, travel abroad, and occupation-related exposures are the most frequently identified sources of infection. In communities where brucellosis is endemic, the disease occurs in children and the family members of infected persons are at risk. Even in countries where animal brucellosis is controlled, the disease occasionally develops among farmers, meat-processing workers, veterinarians, and laboratory workers. The Brucella organism is transmitted most commonly through the ingestion of untreated milk or milk products; raw meat (i.e., blood) and bone marrow have also been implicated. However, the organism can be contracted via inhalation during contact with animals, especially by children and by slaughterhouse, farm, and laboratory workers. Other routes of infection for at-risk workers include skin abrasion, autoinoculation, and conjunctival splashing. The organism has occasionally been transmitted from person to person through the placenta, during breast-feeding, and during sexual activity.
Aerosolized B. melitensis is a classic agent of biological warfare. PATHOGENESIS AND IMMUNITY Serum opsonizes Brucella organisms for ingestion by polymorphonuclear leukocytes and activated macrophages. Brucellae resist intracellular phagocytic killing by mechanisms such as the suppression of the myeloperoxide-hydrogen peroxide-halide system and the production of superoxide dismutase. The pathogen-phagocyte interaction plays a key role in determining the severity and outcome of brucellosis. The organisms surviving within and escaping from the phagocytes multiply and reach the bloodstream via the lymphatics, subsequently localizing in the liver, spleen, bones, kidneys, lymph nodes, heart valves, nervous system, and testes. In these organs, the bacteria are ingested by macrophages and survive by inhibition of phagosome-lysosome fusion. In infected tissues, inflammatory responses or noncaseating granulomas typically develop, and caseating granulomas and abscesses have been described. Cytokines, including interleukin (IL) 1, IL-12, and tumor necrosis factor, appear to be important in host defense against Brucella infection. The smooth lipopolysaccharide (LPS) of Brucella is the major known virulence factor. Strains with rough LPS are more likely than those with smooth LPS to be lysed by nonimmune serum. In virulent strains, the foremost target for specific antibodies is the LPS. Serum IgM antibodies to LPS appear within 1 week after infection and are followed later by IgG and IgA. Titers of both IgM and IgG antibody fall after treatment, and failure of these titers to decline should prompt an evaluation for relapse or persistent infection. CLASSIFICATION Brucellosis is classified according to whether or not the disease is active (i.e., symptoms or progressive tissue damage and significantly raised Brucella agglutinin levels with or without positive cultures) and whether or not there is localized infection. The state of activity and the site of localization have a significant impact on recommended treatment. Classification of brucellosis as acute, subacute, serologic, bacteremic, or of mixed types serves no purpose in diagnosis and management. CLINICAL MANIFESTATIONS AND COMPLICATIONS Brucellosis is a systemic disease with protean manifestations. Its features may mimic those of other febrile illnesses. The incubation period lasts for about 1 to 3 weeks but may be as long as several months, depending on the virulence of the organisms, their route of entry, the infecting dose, and the host's preexisting health status. The onset of symptoms may be either abrupt (over 1 to 2 days) or gradual (³1 week). The most common symptoms are fever, chills, diaphoresis, headaches, myalgia, fatigue, anorexia, joint and low-back pain, weight loss, constipation, sore throat, and dry cough. Physical examination often reveals no abnormalities, and patients can look deceptively well. Some patients, in contrast, are acutely ill, with pallor, lymphadenopathy, hepatosplenomegaly, arthritis, spinal tenderness, epididymoorchitis, rash, meningitis, cardiac murmurs, or pneumonia. The fever of brucellosis has no distinctive pattern but may exhibit diurnal variation, with normal temperatures in the morning and high
temperatures in the afternoon and evening. Localization to an organ or a system may be evident at the onset of the disease.Table 160-1 lists the frequencies of key historic features, symptoms, and signs among 500 patients with brucellosis due to B. melitensis. Bones and Joints Although monarticular septic arthritis occurs, 30 to 40% of patients have reactive asymmetric polyarthritis involving the knees, hips, shoulders, and sacroiliac and sternoclavicular joints. The total white cell count in synovial fluid ranges from 4000 to 40,000/mL, typically with about 60% polymorphonuclear leukocytes. The synovial fluid glucose concentration may be reduced and the protein concentration elevated; cultures of synovial fluid are positive in about 50% of cases. Infection with Brucella organisms commonly causes osteomyelitis of the lumbar vertebrae, starting at the superior end plate (an area with a rich blood supply) and occasionally progressing to involve the entire vertebra, disk space, and adjacent vertebrae. Extraspinal Brucella osteomyelitis is rare. In Brucella septic arthritis and osteomyelitis, the peripheral white cell count is typically normal, while the erythrocyte sedimentation rate may be either normal or elevated. Heart Cardiovascular complications of brucellosis include endocarditis, myocarditis, pericarditis, aortic root abscess, mycotic aneurysms, thrombophlebitis with pulmonary aneurysm, and pulmonary embolism. Brucella endocarditis may develop on valves previously damaged by rheumatic fever or congenital malformation but also occurs on previously normal valves. The clinical features are indistinguishable from those of endocarditis caused by other organisms (Chap. 126). Endocarditis is the leading cause of death in brucellosis, although the outcome of Brucella endocarditis has been more favorable in recent years because of advances in early diagnosis, antibiotic treatment, and cardiac surgery. Physicians who suspect brucellae as a cause of culture-negative endocarditis in patients with possible environmental exposure should notify the bacteriology laboratory performing the blood culture so that extended incubation, specific media, and biohazard precautions can be employed. Respiratory Tract Brucellae can produce respiratory symptoms. A flulike illness with sore throat, tonsillitis, and dry cough is common and usually mild. Hilar and paratracheal lymphadenopathy, pneumonia, solitary or multiple pulmonary nodules, lung abscess, and empyema have been reported. Gastrointestinal Tract and Hepatobiliary System Gastrointestinal manifestations of Brucella infection are generally mild and may include nausea, vomiting, constipation, acute abdominal pain, and/or diarrhea. Pathologic examination of the liver may reveal any of several changes, including noncaseating granulomas (Fig. 160-CD1), suppurative abscesses, mononuclear cell infiltration, or hyperemia of the intestinal mucosa. Acute ileitis with inflammation of Peyer's patches and colitis have been reported. Hepatic and splenic enlargement may be documented in 15 to 20% of cases, and abscesses may develop in the liver and spleen. Infected ascites, pancreatitis, and cholecystitis have been reported. Mild jaundice may be evident, with elevated levels of bilirubin and hepatic enzymes. Genitourinary Tract The various genitourinary infections attributed to brucellae include unilateral or bilateral epididymoorchitis, which is rarely associated with testicular
abscess. Prostatitis, seminal vesiculitis, dysmenorrhea, amenorrhea, tuboovarian abscess, salpingitis, cervicitis, acute pyelonephritis, glomerulonephritis, and massive proteinuria have also been documented. Brucella organisms have been cultured from the urine in up to 50% of cases of genitourinary tract infection. Central Nervous System Neurobrucellosis is uncommon but serious and includes meningitis, meningoencephalitis, multiple cerebral or cerebellar abscesses, ruptured mycotic aneurysms, myelitis, Guillain-Barre syndrome, cranial nerve lesions, hemiplegia, sciatica, myositis, and rhabdomyolysis. Papillitis, papilledema, retrobulbar neuritis, optic atrophy, and ophthalmoplegia due to lesions in cranial nerves III, IV, and VI may occur in Brucella meningoencephalitis. Cerebrospinal fluid (CSF) pressure is usually elevated; the fluid may appear clear, turbid, or hemorrhagic; the protein concentration and cell count (predominantly lymphocytes) are elevated; and the glucose concentration may be either reduced or normal. In Brucella meningitis, which can occur at any time during the course of the disease, the organism may be cultured from the CSF. Other Manifestations Conjunctival splashing with live attenuated B. abortus vaccine (S19) during animal vaccination may cause conjunctivitis, keratitis, and corneal ulcers, with progression to systemic disease in some cases. Uveitis, optic neuritis, retinopathy, retinal detachment, and endophthalmitis may result from hematogenous spread. Skin manifestations of brucellosis are uncommon. They include maculopapular eruptions, purpura and petechiae, chronic ulcerations, multiple cutaneous and subcutaneous abscesses, discharging sinuses, superficial thrombophlebitis, erythema nodosum, and pemphigus. Brucellosis during human pregnancy can cause abortion or intrauterine fetal death. Brucellae have been isolated from the human placenta, fetus, and newborn. The bone marrow of Brucella-infected patients frequently contains noncaseating granulomas. Among the hematologic complications of brucellosis are anemia, leukopenia, and thrombocytopenia. Endocrinologic findings reported in brucellosis include thyroiditis with abscess formation, adrenal insufficiency, and the syndrome of inappropriate secretion of antidiuretic hormone. DIAGNOSIS The combination of potential exposure, consistent clinical features, and significantly raised levels of Brucella agglutinin (with or without positive cultures of blood, body fluid, or tissues) confirms the diagnosis of active brucellosis. The organism's identity is confirmed by phage typing, DNA characterization, or metabolic profiling. Use of a CO2detection system (such as BACTEC; Becton Dickinson, Sparks, MD) for blood culture provides a more sensitive and rapid culture result than standard methods, with positivity usually apparent after only 2 to 5 days of incubation. Serum antibodies to Brucella can be detected by several methods, including standard tube agglutinins (STA), the 2-mercaptoethanol agglutination test, Coombs' test, enzyme-linked immunosorbent
assay, and polymerase chain reaction (PCR). B. abortus antigens, which are commonly used for serologic tests, cross-react with B. melitensis and B. suis but not with B. canis. The specific antigen required for assay of antibodies to B. canis is not commercially available. B. canis antibody titers can be determined in the United States at the Centers for Disease Control and Prevention in Atlanta. A false-negative result in the STA may be obtained because of the prozone phenomenon, which can be avoided by testing of sera at both low and high dilutions. In endemic areas a Brucella antibody titer of 1:320 or 1:640 is significant, while in nonendemic areas an antibody titer of 1:160 is considered significant. Detection of elevated levels of antibody to Brucella organisms in the absence of symptoms during the screening of potential blood donors is common in endemic areas. To establish a diagnosis in these regions, clinical and serologic evaluation should be repeated after 2 to 4 weeks and a further rise in titer sought. A high titer of specific IgM suggests recent exposure, while a high titer of specific IgG suggests active disease. Lower titers of IgG may indicate past exposure or treated infection. Cooperation and consultation with a clinical microbiology laboratory are important when brucellosis is suspected. It may be necessary to observe culture bottles for up to 6 weeks before organisms become detectable. Subcultures should be prepared on duplicate blood agar plates (with and without an atmosphere of 10% CO2) and special media (such as a blood- or serum-enriched peptone-based medium) or with a rapid CO2detection system. Patients whose blood or bone marrow is cultured are positive at one site or the other in 50 to 70% of cases. The peripheral white cell count is usually normal but may be low, with relative lymphocytosis. Thrombocytopenia and disseminated intravascular coagulation may be documented. Levels of hepatic enzymes and serum bilirubin may be raised. Radiologic investigations aimed at detecting skeletal involvement include plain radiography, bone scintigraphy, computed tomography (CT), and magnetic resonance imaging (MRI). Bone scintigraphy is more sensitive than conventional radiography in detecting areas of spinal and extraspinal involvement, particularly in the early stage of infection. CT is useful for further evaluation of spinal lesions and of the extension of infection into the spinal canal. MRI is the modality of choice for the assessment of Brucella spondylitis and is more sensitive than scintigraphy or CT for demonstration of the extent of disease. Plain lateral radiography of the spine may reveal bone sclerosis, with destruction and erosion of the superior end plate anteriorly. As the disease progresses, healing with osteophyte formation and reduction of disk space may take place. In Brucella septic monarthritis, plain radiography may show effusion and soft tissue swelling without bone or joint destruction. Scintigraphy may document increased uptake in sacroiliac joints or lumbar vertebrae, even when plain radiography gives normal results.MRIshows diffuse high-signal intensity of the affected vertebrae and may reveal narrowing of the spinal canal as well as loss of definition of the posterior aspect of the vertebrae. TREATMENT Single-agent therapy for brucellosis has now been abandoned because of the high rates
of failure and relapse and the potential development of antibiotic resistance. Relatively short courses (100 mm/h. Rheumatoid factor and antinuclear antibodies are usually absent. The diagnosis of Yersinia-induced reactive arthritis or other nonsuppurative inflammatory sequelae can be difficult, especially when triggering infections are asymptomatic or clinically mild or occur several weeks before the diagnosis is attempted. Because the isolation of a pathogenic Yersinia strain from feces is the most specific diagnostic test in such cases, it should be attempted. Since culture is of limited sensitivity in this clinical setting, a high index of suspicion and positive results of serologic tests for Y. enterocolitica or Y. pseudotuberculosis are usually required for diagnosis. TREATMENT The effectiveness of antimicrobial agents in the treatment of yersinial enteritis, enterocolitis, mesenteric adenitis, or terminal ileitis has not been established. These conditions are usually self-limited, and their treatment is symptom-based and supportive. In uncomplicated cases, diarrhea should be treated with fluid and electrolyte replacement, with the route of delivery dependent on clinical severity. Enteric precautions are advisable for patients hospitalized with yersinial diarrhea. In general, antimicrobial treatment should be reserved for patients with septicemia, metastatic focal infections, or immunosuppression and enterocolitis. Controlled clinical comparisons of antimicrobial agents in the treatment of severe cases of yersiniosis have not yet been conducted. In such cases, drug selection should ultimately be guided by clinical response and bacterial sensitivity patterns. Clinical isolates of Y. enterocolitica and Y. pseudotuberculosis are usually susceptible in vitro to aminoglycosides, third-generation cephalosporins, chloramphenicol, quinolones, tetracyclines, and trimethoprim-sulfamethoxazole. In laboratory animals infected with enteropathogenic yersiniae, the fluoroquinolones have exerted the strongest bactericidal effects in vivo; clinical experience with these drugs against these pathogens in humans is promising but limited. Because they produceb-lactamases, isolates typically are resistant to penicillin, ampicillin, carbenicillin, and first-generation and most second-generation cephalosporins. Optimal dosages and durations of therapy have not been established. Mortality from Y. enterocolitica septicemia currently is ~10% despite treatment. Focal extraintestinal infections may require at least 3 weeks of therapy. No role for antimicrobial agents in the management of the nonsuppurative inflammatory manifestations of yersiniosis has been established. Patients with reactive arthritis may benefit from treatment with nonsteroidal anti-inflammatory drugs, intraarticular steroid injections, and physical therapy. PREVENTION AND CONTROL The importance of safe food-handling and food-preparation practices in the prevention
of yersiniosis cannot be overemphasized. Caution is particularly warranted in the case of pork and other animal products. The consumption of raw or undercooked meats, especially pork, should be avoided. Increased efforts to prevent the spread of enteric pathogens in household, pet-care, day-care, and hospital settings and in the food industry would be likely to decrease the incidence of yersiniosis. Current regulations of the U.S. Food and Drug Administration require visual inspection of packed red cell units before transfusion, with the discarding of units in which bacterial contamination is suspected on the basis of darkening (reflecting decreased oxygen saturation and hemolysis). Since the risk is minimal, more specific measures to further decrease the likelihood of transfusion of Y. enterocolitica-contaminated blood products (e.g., limiting the period for which red cells can be stored before transfusion) are not considered cost-effective. Yersiniosis is not routinely reportable to public health authorities in most jurisdictions. However, clinicians who suspect a common-source outbreak (e.g., because they have documented a familial case cluster or have diagnosed the disease in several apparently unrelated patients over a short period) or some other public health threat (e.g., because they have found occult Y. enterocolitica bacteremia in a recent blood donor) should consult promptly with local public health officials. (Bibliography omitted in Palm version) Back to Table of Contents
163. BARTONELLA INFECTIONS, INCLUDING CAT-SCRATCH DISEASE - Lucy Stuart Tompkins Bartonella spp., including B. bacilliformis, B. henselae, B. quintana, and B. clarridgeiae, are tiny gram-negative bacilli that can adhere to and invade mammalian cells, including endothelial cells and erythrocytes. Previously classified as Rochalimaea spp. within the rickettsia group, Bartonella spp. have now been removed from the order Rickettsiales on the grounds that they are not obligate intracellular parasites. These agents cause a wide spectrum of clinical illnesses, including trench fever, cat-scratch disease (CSD), bacillary angiomatosis, endocarditis, Oroya fever, and verruga peruana (Fig. 163-CD1). The pathologic manifestations of Bartonella disease vary with the immune status of the host. OROYA FEVER AND VERRUGA PERUANA DEFINITION AND ETIOLOGY Oroya fever and verruga peruana are caused by B. bacilliformis. Oroya fever is characterized by fever, profound anemia, and -- unless antibiotic treatment is given -high mortality. The lesions referred to as verruga peruana may develop during the convalescent phase of Oroya fever or during chronic infection with B. bacilliformis. In 1885 Daniel Carrion, a Peruvian medical student, inoculated himself with blood from a patient with verruga peruana and subsequently died of Oroya fever, thus proving that both diseases are caused by a single agent. EPIDEMIOLOGY Infection with B. bacilliformis follows the bite of the sandfly vector Phlebotomus, an insect found in the river valleys of the Andes Mountains at altitudes of 600 to 2500 m. Oroya fever develops in nonimmune individuals who are not residents of the endemic region, whereas verruga peruana occurs in persons who apparently have been exposed in the past, including those who have recently had Oroya fever. The infection has not been acquired in the United States. PATHOLOGY During initial infection in the nonimmune host, B. bacilliformis cells adhere to erythrocytes and produce indentations in the cell membrane; the bacteria subsequently enter the erythrocytes and cause persistent deformation of the cytoskeleton. The parasitized erythrocytes are ultimately phagocytosed and destroyed. Although the life span of infected erythrocytes is markedly shortened, not all of this change can be attributed to the mechanical fragility induced by the internalization of bacteria. Decreased bone marrow erythropoiesis also contributes to anemia. CLINICAL MANIFESTATIONS The onset of symptoms in Oroya fever may be either insidious or abrupt, after an incubation period of approximately 3 weeks. The subacute presentation may include low-grade fever, malaise, headache, and anorexia. Sudden-onset disease commences
with high fever, chills, diaphoresis, headaches, and changes in mental status. These manifestations are followed by the sudden development of profound anemia, which is due to a marked decrease in erythrocyte numbers and is associated with macrocytic changes, poikilocytosis, Howell-Jolly bodies, nucleated erythrocytes, and immature myeloid cells. The leukocyte differential usually shifts to the left, although the total leukocyte count may be normal. The erythrocyte count may fall to extremely low levels. In eosin/thiazine-stained peripheral-blood smears, numerous microorganisms can be seen adhering to most erythrocytes. During the acute phase, muscle and joint pain and headache may be severe; central nervous system changes include insomnia, delirium, and a decreased level of consciousness. Thrombocytopenic purpura may develop. If the patient survives, a convalescent phase ensues, characterized by the sudden disappearance of bacteria from blood smears, declining fever, and an increase in the erythrocyte count. Although much of the mortality associated with Oroya fever is due to profound anemia and toxicity, secondary bacterial infections (including salmonellosis and other enteric infections, malaria, and tuberculosis) are often an important contributing factor. After convalescence from acute Oroya fever, verrugas may develop. These red or purple cutaneous lesions may be either tiny and sessile or large, pedunculated, and nodular. They bear a marked resemblance to the lesions of bacillary angiomatosis and to Kaposi's sarcoma. DIAGNOSIS During acute infection, bacteria can be cultured from the blood on agar containing rabbit blood, with incubation at 28°C. The hallmark of verruga peruana is the formation of new blood vessels (angiogenesis) at the sites of bacterial replication. TREATMENT Oroya fever responds to a variety of antimicrobial agents, including chloramphenicol, tetracyclines, penicillin, and streptomycin. Chloramphenicol is used most often because of its efficacy against most Salmonella infections (as salmonellosis may develop intercurrently). Verruga peruana may respond similarly; however, failure to respond to therapy and relapse are common and require the reinstitution of prolonged therapy. BACILLARY ANGIOMATOSIS DEFINITION AND ETIOLOGY Bacillary angiomatosis was initially described as a condition occurring primarily in patients with AIDS and characterized by vascular cutaneous lesions resembling Kaposi's sarcoma. The disease can disseminate to involve virtually any organ system. Immunocompromised individuals, especially those infected with HIV, are at particularly high risk for bacillary angiomatosis, although in rare instances the patient is not obviously immunosuppressed. Both B. henselae and B. quintana (the infectious agent initially associated with trench fever) produce bacillary angiomatosis in persons with immunodeficiency.
EPIDEMIOLOGY Acquisition of B. henselae has been significantly associated with exposure to young cats infested with fleas (Ctenocephalides felis). Because a high percentage of cats are seropositive, it has been suggested that patients with HIV infection avoid exposure to these animals. The finding that a large proportion of cats with fleas have persistent asymptomatic B. henselae bacteremia suggests that the domestic cat is the animal reservoir of this microorganism. The flea may serve as a transmitting vector in the cross-infection of cats, but its role in human infection is not clear. Tick-associated cases of B. henselae bacteremia have been reported in healthy immunocompetent individuals. Person-to-person transmission of B. quintana by the human body louse (Pediculus humanis corporis) was documented during World War I under conditions of poor personal hygiene and sanitation. Although lice are suspected of transmission, the reservoir of B. quintana has not been identified. A case-control study revealed that B. henselae and B. quintana differ significantly in terms of epidemiologic risk factors. All cases of B. henselae infection were associated with exposure to cats and their fleas and occurred sporadically, whereas the cases of B. quintana infection occurred in clusters and were associated with low socioeconomic status, homelessness, and exposure to body lice. Direct transmission of B. henselae from cats to their owners, presumably through cutaneous trauma, was supported by the matching DNA fingerprint patterns of isolates from the two sources. MICROBIOLOGY B. henselae can be demonstrated in tissue by Warthin-Starry staining. Clumps and clusters of pleomorphic bacilli appear as purple deposits in tissue stained with hematoxylin and eosin. Although the bacteria may be difficult to cultivate in the laboratory, they can eventually be isolated from cultures of blood and of material from other sites. Colonies develop after prolonged incubation (1 to 4 weeks) on blood-containing media and pit the agar; bacterial cells are gram-negative. B. quintana grows as a smooth, nonpitting colony on solid agar after prolonged incubation. Classification of B. henselae was first accomplished when molecular techniques were used to analyze bacterial ribosomal genes extracted from tissue samples. Definitive identification of Bartonella spp. is based on sequence analysis of 16S ribosomal DNA. PATHOGENESIS AND PATHOLOGY Bacillary angiomatosis is characterized by a lobular proliferation of new blood vessels (angiogenesis) and a neutrophilic inflammatory response to myriad bacilli located within collagen-rich microscopic and macroscopic nodules. The endothelial cells lining the vascular spaces have a typical epithelioid appearance, and the lesions may resemble Kaposi's sarcoma histopathologically, although the characteristic spindle cell of the latter disease is usually absent. The bacterial and eukaryotic host factors that elicit the pathologic response are unknown.
CLINICAL MANIFESTATIONS The skin lesions of bacillary angiomatosis (also called epithelioid angiomatosis) are vascular nodules, papules, or tumors (Fig. 163-CD2) that range from tiny lesions resembling cherry angiomas or pyogenic granulomas to large, pedunculated, exophytic masses (Fig. 163-1). Characteristically, the lesions are red or purple, resembling Kaposi's sarcoma; they may be surrounded by an epithelial collarette, may be located anywhere on the skin, and may involve mucous membranes. The overlying epidermis may be focally ulcerated, and the underlying bone may be invaded and destroyed. Dissemination of B. henselae infection occurs primarily in patients with cellular immune defects. Clinical manifestations accompanying dissemination are often nonspecific and include persistent fever, abdominal pain, weight loss, and malaise. Although the liver, spleen, bone marrow, and lymph nodes are primarily affected, HIV-infected patients may also develop central nervous system abnormalities (including psychiatric disorders and brain lesions), which are responsive to antibiotic therapy. Skin lesions usually are not evident in disseminated infection. Involvement of the liver or spleen may produce bacillary peliosis hepatis. Patients with the latter condition may report localized pain on palpation of the abdomen. Nodular lesions of variable size can be demonstrated by computed tomography or magnetic resonance imaging, with or without contrast agents. In a case-control study of bacillary angiomatosis (see "Epidemiology" above), only B. henselae was associated with hepatosplenic disease (peliosis hepatis) and displayed a predilection for the lymph nodes. B. quintana, in contrast, was associated with osseous and subcutaneous infection. DIAGNOSIS The diagnosis of bacillary angiomatosis is based primarily on the typical histopathologic findings of angiomas in association with clumps of tiny bacilli revealed by Warthin-Starry silver stain. Infection due to B. henselae can also be established by culture or by identification of specific DNA sequences. B. henselae is most easily isolated from blood through a lysis-centrifugation system. Colonies may be detected on blood-containing agar (rabbit blood is preferred) incubated with 5 to 10% CO2 at 37°C for 2 to 4 weeks. B. quintana may be isolated from BACTEC (Becton Dickinson, Sparks, MD) aerobic bottles containing resin. Isolation from skin lesions and other tissues is more difficult but should be attempted when feasible. Initial reports suggested that cocultivation with endothelial cell monolayers was necessary; however, isolation by direct plating onto freshly prepared agar media has also been successful. Bacilli picked from new colonies but not subcultured may not stain, even with acridine orange; they stain weakly with safranin. Identification of B. henselae and B. quintana is based primarily on cellular fatty-acid analysis and polymerase chain reaction (PCR)-based restriction fragment length polymorphism analysis. Definitive identification of Bartonella spp. depends on DNA sequence analysis of 16S ribosomal RNA genes. The diagnosis ofCSD (see next section) can be made by specific serologic testing that detects B. henselae-specific antibodies, but the sensitivity and specificity of this method in patients with cutaneous and disseminated bacillary angiomatosis have not been determined. DIFFERENTIAL DIAGNOSIS
The differential diagnosis of cutaneous bacillary angiomatosis includes Kaposi's sarcoma, angiomas, and pyogenic granulomas. These conditions can be distinguished by histopathologic examination of biopsied material. Cutaneous bacillary angiomatosis caused by B. henselae or B. quintana resembles verruga peruana, which is not seen outside of South America. In patients with AIDS, Kaposi's sarcoma lesions and bacillary angiomatosis may coexist. TREATMENT Cutaneous lesions have been treated with a wide variety of antimicrobial drugs, including macrolides, tetracyclines, and antituberculous agents; B. henselae is susceptible to most antibiotics in vitro. Erythromycin (2 g/d), given orally for 3 weeks, is usually effective, as are newer macrolides; however, relapse may require prolonged therapy (3 weeks to 2 months) with an antibiotic that reaches an intracellular compartment, such as a macrolide or doxycycline (200 mg/d). Patients with peliosis hepatis should be treated with intravenous antibiotics, and those with disseminated disease or bacteremia should be treated with a prolonged course (3 weeks to 2 months) of systemic antibiotic, such as a macrolide (e.g., erythromycin, 2 g/d). In a case-control study of bacillary angiomatosis, treatment with a macrolide was associated with a therapeutic response and sterile tissue samples and may have been protective, whereas treatment with trimethoprim-sulfamethoxazole, ciprofloxacin, penicillins, or cephalosporins had no protective effect. Cutaneous lesions may or may not regress spontaneously, perhaps depending on the status of the host's immunity. The safety of ciprofloxacin in pregnant or lactating women has not been established. No antimicrobial has been studied prospectively, and information on efficacy comes only from case reports. CAT-SCRATCH DISEASE DEFINITION AND ETIOLOGY TypicalCSD is manifested by painful regional lymphadenopathy persisting for several weeks or months after a cat scratch. Occasionally, infection may disseminate and produce more generalized lymphadenopathy and systemic manifestations, which may be confused with the manifestations of lymphoma. B. henselae is the causative agent of CSD. There is no evidence that B. quintana causes CSD, and this microbe is not carried by cats. The role of Afipia felis (originally proposed as the agent of CSD) is unclear inasmuch as only a few cases are associated with its isolation. B. henselae remains the predominant species causing typical CSD. Several reports suggest that B. clarridgeiae may also cause feline lymphadenopathy. EPIDEMIOLOGY Approximately 60% of cases ofCSD in the United States occur in children. Exposure to bacteremic young cats that either are flea-infested or have been in contact with another cat carrying fleas poses a significant risk of infection. Most infections are caused by a scratch and only rare cases by a bite or by licking. Most cases occur in the warmer
months, when fleas are active. Regions of the United States where fleas are endemic have higher rates of infection. The flea may serve to transmit infection between cats; it is not known whether humans can be infected through the bite of an infected flea. CLINICAL MANIFESTATIONS A localized papule, progressing to a pustule that often crusts over, develops 3 to 5 days after a cat scratch (Fig. 163-CD3). Tender regional lymphadenopathy develops within 1 to 2 weeks after inoculation; by this time, the papule may have healed spontaneously. Scratches are most often sustained on the hands or face, producing epitrochlear, axillary, pectoral, and cervical lymph node involvement. The involved nodes occasionally become suppurative; bacterial superinfection with staphylococci or other cutaneous pathogens may develop. Although most patients do not have fever, systemic symptoms are frequent and include malaise, anorexia, and weight loss. Without treatment, lymphadenopathy persists for weeks or even months and may be confused with lymphatic malignancy. Other manifestations in apparently immunocompetent patients include encephalitis, seizures and coma (especially in children), meningitis, transverse myelitis, granulomatous hepatitis and splenitis, osteomyelitis, and disseminated infection. Conjunctival inoculation may cause Parinaud's oculoglandular syndrome, with conjunctivitis and preauricular lymphadenopathy. PATHOLOGY The histopathologic hallmark ofCSD is granulomatous inflammation with stellate necrosis but no evidence of angiogenesis. Thus, infection by B. henselae can produce two entirely different pathologic reactions, depending on the immune status of the host: CSD or bacillary angiomatosis. DIAGNOSIS CSDshould be suspected if the patient has a history of exposure to cats and develops lymphadenopathy and a skin lesion. The diagnosis can be confirmed by pathologic examination of the involved nodes. Tiny bacilli in clusters can sometimes be seen in biopsy samples stained with Warthin-Starry silver. The CSD skin test, in which lymph node material obtained from patients with CSD serves as an antigen, is no longer used for diagnosis because of concerns about the transmission of viral agents. A specific serologic test has been developed and may produce a positive result in 70 to 90% of patients with intact immunity. The identification of B. henselae 16S ribosomal RNA genes in biopsy material by PCR amplification with specific oligonucleotide primers can also be diagnostically useful; however, these methods are not yet commercially available. Cultures of lymph nodes, cerebrospinal fluid, or other tissues are rarely positive. TREATMENT AlthoughCSD is generally self-limited, tender regional lymphadenopathy and systemic symptoms may be debilitating. Patients with encephalitis or other serious manifestations should be treated with antibiotics. A randomized, double-blind, placebo-controlled trial demonstrated significant clinical benefit of treatment with oral azithromycin for 5 days in
cases of typical CSD (regimen for adults weighing >100 lb: one dose of 500 mg on day 1, 250 mg on days 2 through 5). Several reports suggest that aminoglycoside treatment (e.g., intravenous gentamicin at standard doses calculated to result in therapeutic levels) is effective in patients with encephalitis and other systemic infections. The oral agents that appear to be useful are those that also are most effective for the treatment of bacillary angiomatosis; they include ciprofloxacin, doxycycline, and azithromycin. Unlike bacillary angiomatosis, CSD responds to treatment with ciprofloxacin. The necessary duration of therapy is variable. TRENCH FEVER DEFINITION AND ETIOLOGY Trench fever was first described as a debilitating febrile illness associated with prolonged B. quintana bacteremia in soldiers fighting in Europe during World War I. Although not usually fatal, the illness accounted for substantial morbidity. In recent years, trench fever has reemerged in the United States and has been caused by either B. henselae -- the agent ofCSD and bacillary angiomatosis -- or B. quintana. EPIDEMIOLOGY Although trench fever was thought to have disappeared from the United States, recent cases have been diagnosed in homeless persons (B. quintana) and in persons bitten by ticks (B. henselae). During World War I, trench fever was transmitted from person to person by the human body louse. Transmission by ectoparasites is suspected in the recent cases of B. quintana infection but has not been firmly documented. Patients with trench fever have apparently normal immune defenses. CLINICAL MANIFESTATIONS Trench fever is characterized by the sudden onset of headache, aseptic meningitis, persistent fever (which can be high-grade and is commonly paroxysmal), malaise, weight loss, and other nonspecific symptoms. Severe musculoskeletal pain is more common among immunocompetent than among immunocompromised patients. Bacteremia can persist for days or weeks, and relapses have followed short courses of antibiotic therapy. Localized findings are uncommon. DIAGNOSIS Trench fever is diagnosed by the finding of sustained bacteremia. B. henselae and B. quintana grow slowly. Colonies develop on rabbit blood agar after 1 to 4 weeks of incubation under conditions of increased CO2. Serologic tests for this disease have not yet been standardized. TREATMENT A prolonged course (4 weeks) of antimicrobial therapy may be required. Agents that can cross the mammalian cell membrane are most effective, including erythromycin (2 g/d) or azithromycin (500 mg/d). Data on the efficacy of these agents come from a limited
number of case reports. OTHER BARTONELLA INFECTIONS, INCLUDING CULTURE-NEGATIVE ENDOCARDITIS The application of molecular methods to the detection of microorganisms that are difficult to cultivate in the laboratory has revealed new Bartonella spp. and has established Bartonella spp. as a cause of endocarditis cases previously classified as being of unknown etiology. B. quintana is the most frequently isolated Bartonella species in these cases. Two new species, B. elizabethae and B. clarridgeiae, as well as B. henselae have also been identified as agents of subacute and chronic endocarditis. The diagnosis of Bartonella endocarditis is confirmed by blood cultures. Specific antibodies are produced; however, B. quintana infection may produce antibodies that cross-react with Chlamydia pneumoniae. (Bibliography omitted in Palm version) Back to Table of Contents
164. DONOVANOSIS - Gavin Hart Donovanosis is a chronic, progressively destructive bacterial infection of the genital region that is generally regarded as sexually transmitted. The disease has been known by many other names, the most common of which are granuloma inguinale and granuloma venereum. ETIOLOGY Donovanosis is caused by Calymmatobacterium granulomatis, an intracellular, gram-negative, pleomorphic, encapsulated (when mature) bacterium measuring 1.5 by 0.7 um. C. granulomatis shares many morphologic and serologic characteristics and >99% homology at the nucleotide level with Klebsiella species that are pathogenic to humans. Polymerase chain reaction amplification of the phoE gene shows it to be closely related to that in Klebsiella pneumoniae, K. rhinoscleromatis, and K. ozaenae. Electron microscopy shows typical gram-negative morphology and a large capsule but no flagella. Filiform or vesicular protrusions occur on a corrugated cell wall. EPIDEMIOLOGY Donovanosis is endemic among Aborigines in central Australia as well as in Papua New Guinea, southeastern India, southern Africa, and the Caribbean and adjacent areas of South America. In the first half of the twentieth century, the disease was endemic in parts of the United States (with an estimated 5000 to 10,000 cases in 1947); small epidemics still occur in this country and in other developed countries. Over 70% of cases involve persons 20 to 40 years of age. The infection is predominantly sexually transmitted, but extragenital skin lesions can follow transmission from concurrent genital lesions via the fingers or through other nonsexual contact, and autoinoculation may produce new lesions from contact with adjacent skin ("kissing" lesions). Infants born to infected mothers have acquired infection at birth. The classification of donovanosis as a sexually transmitted disease (STD) has been disputed because of cases in young children and occasionally in sexually inactive individuals, transmission by direct body contact and via inanimate intermediaries, and the low and variable prevalence of donovanosis among sexual partners (0.4 to 52%). The dominance of sexual transmission is suggested by the combined factors of lesions predominantly affecting the genitalia, the highest prevalence among persons in age and socioeconomic groups that are most often affected by STDs, and the predictable occurrence of disease in visitors to areas of endemicity following sexual exposure. CLINICAL MANIFESTATIONS The incubation period is usually 1 to 4 weeks but may extend to 1 year. Skin lesions have been detected in infants 6 weeks to 6 months after birth. The disease begins as one or more subcutaneous nodules that erode through the skin to produce clean, granulomatous, sharply defined, usually painless lesions (Fig. 164-1). These lesions, which bleed readily on contact, slowly enlarge. The genitalia are involved in 90% of cases, the inguinal region in 10%, and the anal region in 5 to 10%. Genital swelling, particularly of the labia, is a common feature and occasionally progresses to
pseudoelephantiasis. Phimosis and paraphimosis are common local complications, and progressive erosion of affected tissues may completely destroy the penis or other organs. Less common clinical variants include a hypertrophic form (cauliflower- or wartlike lesions), a necrotic form (destructive lesions with foul-smelling exudate, often resembling amebiasis), and a sclerotic or cicatricial form, which has a dry base with extensive scar tissue (Fig. 132-CD3). Extragenital lesions occur in at least 6% of cases. Oral donovanosis, the most common extragenital manifestation, presents as pain or bleeding in the mouth, lesions on the lips, or extensive swelling of the gums and palate. Donovanosis may affect most bones, and sometimes many bones are affected at the same time; the tibia is involved in over 50% of such cases. Bony lesions are associated with constitutional symptoms (weight loss, fever, night sweats, and malaise) and are usually found in women. More than 50% of women have primary lesions on the cervix. Prompt pelvic examinations and early diagnosis are likely to substantially decrease the morbidity and mortality (a likely outcome in misdiagnosed spinal lesions) associated with extragenital donovanosis in women. DIAGNOSIS Laboratory Diagnosis The preferred method involves demonstration of typical intracellular Donovan bodies within large mononuclear cells visualized in smears prepared from lesions or biopsy specimens. With typical beefy lesions, a small piece of tissue is removed with forceps and scalpel, and a crush impression of the deep surface is made on a glass slide. The smear is air-dried, heat-fixed, and stained with Giemsa, Leishman's, or Wright's stain. For dry, flat, or necrotic lesions, a punch-biopsy specimen should be obtained from the advancing edge. This specimen can be used to prepare a smear or embedded for histologic examination (with a silver stain). Histologic examination shows epithelial proliferation, often simulating neoplasia, with a heavy inflammatory infiltrate of plasma cells, some neutrophils, and few if any lymphocytes. The large mononuclear cells are 25 to 90 um in diameter, with a vesicular or pyknotic nucleus. Up to 20 intracytoplasmic vacuoles contain pleomorphic Donovan bodies in either young uncapsulated forms (which often resemble closed safety pins) or mature capsulated forms. C. granulomatis has never been grown on artificial solid media but has been cultured in chicken embryonic yolk sacs, on human monocytes, and on human epithelial (HEp-2) cells. A sensitive and specific serologic test, based on indirect immunofluorescence, has been developed. Differential Diagnosis Condylomata lata of secondary syphilis may be confused with donovanosis; however, these lesions usually appear as white or pale moist plaques in the anogenital area, whereas the lesions of donovanosis are usually bright red. Syphilis and donovanosis frequently coexist because syphilis is usually highly prevalent in areas where donovanosis is endemic; thus positive syphilis serology does not exclude a diagnosis of donovanosis. Condylomata lata subside within 1 week of treatment with benzathine penicillin (2.4 million units), whereas donovanosis lesions remain unchanged. The necrotic form of donovanosis may resemble squamous cell carcinoma; likewise, cervical and vulvar lesions may closely resemble carcinoma. Penile amebiasis may
resemble necrotic donovanosis but usually follows anal intercourse and is much less common than donovanosis in areas where the latter is endemic. Atypical clinical variants of chancroid, referred to as pseudogranuloma inguinale, have been described in patients seen at clinics in Atlanta. Disseminated donovanosis lesions of bones, particularly in the spine, can mimic tuberculosis. Lesions that produce draining sinuses near the jaw may simulate actinomycosis. The histologic findings of donovanosis must be distinguished from those of rhinoscleroma, leishmaniasis, and histoplasmosis. Genital ulcers are a risk factor for HIV acquisition in developing countries, and patients with donovanosis should be tested for HIV infection. TREATMENT Table 164-1 shows the most effective regimens for treating donovanosis. Doxycycline is the first choice for therapy in developed countries. Erythromycin provides an effective option for pregnant patients, and azithromycin is an effective alternative that is more convenient to administer. Extensive lesions have been cured with oral azithromycin at a dosage of 500 mg/d, but the more convenient dose of 1 g weekly is also effective. Although chloramphenicol is the drug of choice in some developing countries, it is unlikely to be acceptable in developed countries because of bone marrow toxicity. Penicillin is not effective for treating donovanosis. Patients should be examined weekly, and therapy should be continued until lesions have healed (3 to 5 weeks, except in severe cases). If antibiotic therapy is stopped earlier, lesions often continue to heal, but the relapse rate is higher. If the lesions are unchanged after 2 weeks of treatment, an alternative antibiotic regimen should be used. The treatment regimens just described are usually adequate in HIV-infected patients without immunosuppression, but an increasing failure rate has been reported in immunosuppressed patients, for whom daily administration of azithromycin is recommended if other regimens fail to elicit a response. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 7 -MISCELLANEOUS BACTERIAL INFECTIONS 165. NOCARDIOSIS - Gregory A. Filice The term nocardiosis refers to invasive disease associated with members of the genus Nocardia. Of the several distinctive syndromes, pneumonia and disseminated disease are most common. Others include cellulitis, lymphocutaneous syndrome, actinomycetoma, and keratitis. MICROBIOLOGY Nocardiae are saprophytic aerobic actinomycetes that are common worldwide in soil, where they contribute to decay of organic matter. Nocardial taxonomy is complex and incompletely understood. Seven species have been associated with human disease: N. asteroides, N. brasiliensis, N. otitidis-caviarum (formerly N. caviae), N. farcinica, N. nova, N. transvalensis, and N. pseudobrasiliensis. N. asteroides is the species most commonly associated with invasive disease. N. farcinica is less common but tends to be virulent and prone to dissemination. The new species N. pseudobrasiliensis accounts for most cases of invasive disease previously attributed to N. brasiliensis. True N. brasiliensis isolates are usually associated with disease limited to the skin. N. transvalensis is generally associated with mycetoma or, in immunosuppressed persons, with pulmonary or systemic disease. EPIDEMIOLOGY Approximately 1000 cases of nocardial infection are diagnosed annually in the United States, 85% of them pulmonary and/or systemic. The disease is more common among adults and males. Outbreaks, which are rare, have been associated with contamination of the hospital environment, solutions, or drug injection equipment. Person-to-person spread is not well documented. There is no known seasonality. The risk of pulmonary or disseminated disease is greater than usual among people with deficient cell-mediated immunity, especially that associated with lymphoma, transplantation, or AIDS. In persons with AIDS, nocardiosis usually presents at a CD4+lymphocyte concentration of 100/uL withHAART, it is reasonable to discontinue antimycobacterial treatment. In vitro and in experimental animals, cytokines such asIL-12, granulocyte-macrophage colony-stimulating factor, andIFN-gact synergistically with antibiotics againstMAC. In a small-scale pilot trial including seven HIV-negative patients, IFN-g was beneficial. Encapsulation of many drugs into liposomes enhances their effect in animal models because both liposomes andMAC are ingested by macrophages. Relevant data from studies of humans are still scarce, however. Disseminated infections caused byNTMother than MAC have been too rare for therapy to be evaluated in controlled trials. The presently recommended treatment for these infections is the same as that for disseminated MAC infections. In particular, M.
genavense seems to be sensitive to clarithromycin and rifabutin. Prevention As has been discussed, disseminated infection withMACoccurs almost exclusively in persons severely immunocompromised by HIV infection. Therefore, the best approach to the prevention of MAC infections is the prevention and reversal of immunodeficiency byHAART. In patients whose HIV is resistant to HAART and who are severely immunosuppressed, with CD4+ counts80%. In 1997, there were 8550 reported cases of primary and secondary syphilis and 46,540 cases of all stages. The populations at highest risk for acquiring syphilis have changed. Between 1977 and 1982, approximately half of all patients with early syphilis in the United States were homosexual or bisexual men. Largely because of changing sexual practices in this population due to the AIDS epidemic, this proportion has decreased. The most recent epidemic of syphilis predominantly involved African-American heterosexual men and women and occurred largely in urban areas, where infectious syphilis has been correlated significantly with the exchange of sex for "crack" cocaine. The incidence of syphilis peaks at 15 to 34 years of age. The reported incidence is much higher among African Americans than in other ethnic groups and is higher in urban than in rural areas; 80% of infectious syphilis cases are reported from 15% of the counties in the United States. The incidence of congenital syphilis roughly parallels that of infectious syphilis in females. The number of reported cases of congenital syphilis in infants£1 year of age was lowest (107 cases) in 1978, when infectious syphilis was most prevalent among homosexual and bisexual men. The dramatic increase in the incidence of primary and secondary syphilis among women from 1986 to 1990 resulted in a proportionate increase in the number of infants born with congenital syphilis -- to 3275 infants in 1991. The incidence of early syphilis among men and women has declined since 1991, as has the number of reported cases of congenital syphilis in infants (with 1049 cases in 1997).
It is important to note, however, that the case definition for congenital syphilis was broadened in 1989 and now includes all live or stillborn infants delivered to women with untreated or inadequately treated syphilis at delivery. Approximately one of every two individuals named as sexual contacts of persons with infectious syphilis becomes infected. Many sexual contacts will already have developed manifestations of syphilis when they are first seen, and about 30% of apparently uninfected contacts who are examined within 30 days of exposure actually have incubating infection and will later develop infectious syphilis if not treated. Thus, the identification and "epidemiologic" treatment of all recently exposed sexual contacts constitute an important aspect of syphilis control. Also important is the identification of infected persons by serologic testing of pregnant women, persons admitted to hospitals, military inductees, and persons undergoing examination in physicians' offices. Still controversial are laws and regulations requiring routine premarital serologic testing for syphilis, where -- though national data are not available -- the yield is undoubtedly lower. NATURAL COURSE AND PATHOGENESIS OF UNTREATED SYPHILIS T. pallidum rapidly penetrates intact mucous membranes or microscopic abrasions in skin and within a few hours enters the lymphatics and blood to produce systemic infection and metastatic foci long before the appearance of a primary lesion. Blood from a patient with incubating or early syphilis is infectious. The generation time of T. pallidum during early active disease in vivo is estimated to be 30 to 33 h, and the incubation period of syphilis is inversely proportional to the number of organisms inoculated. The concentration of treponemes generally reaches at least 107 per gram of tissue before the appearance of a clinical lesion. On the basis of intradermal injection of graded doses of T. pallidum into eight volunteers, the 50% infectious dose was calculated to be 57 organisms. The median incubation period in humans (about 21 days) suggests an average inoculum of 500 to 1000 infectious organisms for naturally acquired disease. The incubation period (from inoculation until the primary lesion becomes discernible) rarely exceeds 6 weeks. Subcurative therapy during the incubation period may delay the onset of the primary lesion, but it is not certain that such treatment reduces the probability that symptomatic disease will ultimately develop. The primary lesion appears at the site of inoculation, usually persists for 4 to 6 weeks, and then heals spontaneously. Histopathologic examination of primary lesions shows perivascular infiltration, chiefly by lymphocytes (including CD8+ and CD4+ cells), plasma cells, and macrophages, with capillary endothelial proliferation and subsequent obliteration of small blood vessels. The CD4+ infiltration displays a TH1-type cytokine profile consistent with the activation of macrophages. At this time T. pallidum is demonstrable in the chancre in spaces between epithelial cells; within invaginations or phagosomes of epithelial cells, fibroblasts, plasma cells, and the endothelial cells of small capillaries; within lymphatic channels; and in the regional lymph nodes. Phagocytosis of organisms by activated macrophages ultimately causes their destruction, which results in spontaneous resolution of the chancre. The generalized parenchymal, constitutional, and mucocutaneous manifestations of secondary syphilis usually appear about 6 to 8 weeks after healing of the chancre,
although 15% of patients with secondary syphilis still have persisting or healing chancres. In other patients, secondary lesions may appear several months after the chancre has healed, and some patients may enter the latent stage without ever recognizing secondary lesions. The histopathologic features of secondary maculopapular skin lesions are hyperkeratosis of the epidermis; capillary proliferation with endothelial swelling in the superficial corium; and dermal papillae with transmigration of polymorphonuclear leukocytes and, in the deeper corium, perivascular infiltration by monocytes, plasma cells, and lymphocytes. Treponemes are found in many tissues, including the aqueous humor of the eye and the cerebrospinal fluid (CSF). Invasion of theCNS by T. pallidum occurs during the first weeks or months of infection, and CSF abnormalities are detected in as many as 40% of patients during the secondary stage. Clinical hepatitis and immune complex-induced membranous glomerulonephritis are relatively rare but recognized manifestations of secondary syphilis; liver function tests may yield abnormal results in up to a quarter of patients with early syphilis. Generalized nontender lymphadenopathy is noted in 85% of patients with secondary syphilis. The paradoxical appearance of secondary manifestations despite high titers of antibody (including immobilizing antibody) to T. pallidum is unexplained but may result from changes in expression of surface antigens. Secondary lesions subside within 2 to 6 weeks, and the infection enters the latent stage, which is detectable only by serologic testing. In the preantibiotic era, up to 25% of untreated patients experienced at least one generalized or localized mucocutaneous relapse, usually during the first year; therefore, identification and examination of sexual contacts are most important for patients with syphilis of2 years' duration. Adequate treatment of the mother before the 16th week of pregnancy should prevent fetal damage. Untreated maternal infection may result in a rate of fetal loss of up to 40% (with stillbirth more common than abortion because of the late onset of fetal pathology), prematurity, neonatal death, or nonfatal congenital syphilis. Among infants born alive, only fulminant congenital syphilis is clinically apparent at birth, and these babies have a very poor prognosis. The most common clinical problem is the healthy-appearing baby born to a mother with a positive serologic test. Routine serologic testing in early pregnancy is considered cost-effective in virtually all populations, even in areas with a low prenatal prevalence of syphilis. Where the prevalence of syphilis is high and when the patient is at high risk, syphilis serology should be repeated in the third trimester and at delivery. The manifestations of congenital syphilis can be divided into three types according to their timing: (1) early manifestations, which appear within the first 2 years of life (often between 2 and 10 weeks of age), are infectious and resemble the manifestations of severe secondary syphilis in the adult; (2) late manifestations, which appear after 2 years and are noninfectious; and (3) residual stigmata. The earliest sign of congenital syphilis is usually rhinitis, or "snuffles" (23%), which is soon followed by other mucocutaneous lesions (35 to 41%). These may include bullae (syphilitic pemphigus), vesicles, superficial desquamation, petechiae, and (later) papulosquamous lesions, mucous patches, and condylomata lata. The most common early manifestations are bone changes (61%), including osteochondritis, osteitis, and periostitis. Hepatosplenomegaly (50%), lymphadenopathy (32%), anemia (34%), jaundice (30%), thrombocytopenia, and leukocytosis are common. Neonatal congenital syphilis must be differentiated from other generalized congenital infections, including rubella, cytomegalovirus or herpes simplex virus infection, and toxoplasmosis, as well as from erythroblastosis fetalis. Neonatal death is usually due to pulmonary hemorrhage, secondary bacterial infection, or severe hepatitis. Late congenital syphilis is that which remains untreated after 2 years of age. In perhaps 60% of cases, the infection remains subclinical; the clinical spectrum in the remainder of cases differs in certain respects from that of acquired late syphilis in the adult. For example, cardiovascular syphilis rarely develops in late congenital syphilis, whereas interstitial keratitis is much more common and occurs between the ages of 5 and 25. Other manifestations associated with interstitial keratitis are eighth-nerve deafness and recurrent arthropathy. Bilateral knee effusions are known as Clutton's joints. Asymptomatic neurosyphilis is present in about one-third of untreated patients, and clinical neurosyphilis occurs in one-quarter of untreated individuals over 6 years of age. Gummatous periostitis occurs between the ages of 5 and 20 and, as in nonvenereal
endemic syphilis, tends to cause destructive lesions of the palate and nasal septum. Characteristic stigmata include Hutchinson's teeth -- centrally notched, widely spaced, peg-shaped upper central incisors -- and "mulberry" molars -- sixth-year molars with multiple, poorly developed cusps. The abnormal facies of patients with congenital syphilis include frontal bossing, saddle nose, and poorly developed maxillae. Saber shins, characterized by anterior tibial bowing, are rare. Rhagades are linear scars at the angles of the mouth and nose that are caused by secondary bacterial infection of the early facial eruption. Other stigmata include unexplained nerve deafness, old chorioretinitis, optic atrophy, and corneal opacities due to past interstitial keratitis. LABORATORY EXAMINATIONS Demonstration of the Organism Dark-field microscopic examination of lesion exudate is useful in evaluating moist cutaneous lesions, such as the chancre of primary syphilis or the condylomata lata of secondary syphilis. The identification of a single characteristic motile organism by a trained observer is sufficient for diagnosis. Examination of oral lesions and anal ulcers by this method is not recommended, as it is difficult to differentiate T. pallidum from other spirochetes that may be present. Most syphilis is diagnosed in settings where dark-field microscopy is not available. The direct fluorescent antibody T. pallidum (DFA-TP) test, an alternative available at central laboratories, uses fluorescein-conjugated polyclonal antitreponemal antibody for the detection of T. pallidum in fixed smears prepared from suspect lesions. More sensitivePCRtests have been developed but are available only in research laboratories. T. pallidum can be found in tissue with appropriate silver stains, although these results should be interpreted with caution because artifacts resembling T. pallidum are often seen. Treponemes can be demonstrated more reliably in tissue by immunofluorescent or immunohistochemical methods using specific monoclonal or polyclonal antibodies to T. pallidum. Serologic Tests for Syphilis There are two types of serologic tests for syphilis: nontreponemal and treponemal. Both types of tests are reactive in persons with any treponemal infection, including yaws, pinta, and endemic syphilis. The nontreponemal tests measure IgG and IgM directed against a cardiolipin-lecithin-cholesterol antigen complex. The most widely used nontreponemal antibody tests for syphilis are the rapid plasma reagin (RPR) test, which can be automated (ART), and theVDRLslide test. In these tests, antibody is detected by the microscopic or macroscopic flocculation of the antigen suspension. The RPR test may be more expensive than the VDRL test, but it is easier to perform and uses unheated serum; it is the test of choice for rapid serologic diagnosis in a clinic or office setting. The VDRL test, however, remains the standard for use withCSF. TheRPR andVDRLtests are equally sensitive and may be used for initial screening or for quantitation of serum antibody. The titer reflects the activity of the disease. A fourfold or greater rise in titer may be seen during the evolution of early syphilis. VDRL titers usually reach 1:32 or higher in secondary syphilis. A persistent fall of two dilutions
(fourfold) or greater following treatment of early syphilis provides essential evidence of an adequate response to therapy. VDRL titers do not correspond directly to RPR titers, and sequential quantitative testing (as for response to therapy) must employ a single test. Two standard treponemal tests are used for confirmation of reactive nontreponemal results: the fluorescent treponemal antibody-absorbed (FTA-ABS) test and the agglutination assays for antibodies to T. pallidum. The microhemagglutination assay for T. pallidum (MHA-TP) has been replaced by the Serodia TP-PA test (Fujirebio, Tokyo), which is more sensitive for primary syphilis. The T. pallidum hemagglutination test (TPHA) is widely used in Europe but is not available in the United States. Both the agglutination assays and the FTA-ABS tests are very specific and, when used for confirmation of positive non-treponemal tests, have a very high positive predictive value for the diagnosis of syphilis. However, even these tests give false-positive results at rates as high as 1 to 2% when used for the screening of normal populations. New enzyme-linked immunosorbent assays have also been approved as confirmatory tests. The relative sensitivities of theVDRLtest, theFTA-ABStest, and theMHA-TPin the various stages of untreated syphilis are shown inTable 172-1. The nontreponemal tests are nonreactive in about one-quarter of patients presenting with primary syphilis. In early primary syphilis, the detection of antibody can be maximized either by the performance of an FTA-ABS test or simply by repetition of a VDRL test after 1 to 2 weeks if the initial VDRL result is negative. All treponemal and nontreponemal tests are reactive during secondary syphilis, and a nonreactive result virtually excludes syphilis in a patient with otherwise-compatible mucocutaneous lesions. (Fewer than 1% of patients with secondary syphilis have a VDRL test that is nonreactive or weakly reactive with undiluted serum but is positive at higher serum dilutions -- the prozone phenomenon.) While the nontreponemal tests will become nonreactive or will be reactive in lower titers following therapy for early syphilis, the treponemal tests often remain reactive after therapy and therefore are not helpful in determining the infection status of persons with past syphilis. For practical purposes, most clinicians need to be familiar with the three uses of serologic tests for syphilis: (1) testing of large numbers of sera for screening or diagnostic purposes (e.g., theRPR orVDRLtest), (2) quantitative measurement of the antibody titer to assess the clinical activity of syphilis or to monitor the response to therapy (e.g., the VDRL or RPR test), and (3) confirmation of the diagnosis of syphilis in a patient with a positive nontreponemal antibody test or with a suspected clinical diagnosis of syphilis (e.g., theFTA-ABStest or Serodia TP-PA). For measurement of IgM in neonates in whom congenital syphilis is suspected, the syphilis Captia-M test (Trinity Biotech, Jamestown, NY) and the 19S IgMFTA-ABStest are available. False-Positive Serologic Tests for Syphilis Because the antigen used in nontreponemal tests is found in other tissues, the tests may be reactive in persons without treponemal infection, although rarely do titers exceed 1:8 in such patients. In a population selected for screening because of clinical suspicion, history of exposure, or increased risk for sexually transmitted infections, fewer than 1% of reactive tests are
falsely positive. The modernVDRL andRPRtests are 97 to 99% specific, and false-positive reactions are now limited largely to those conditions listed inTable 172-2. False positivity is common among persons with autoimmune disorders. The prevalence of false-positive nontreponemal tests increases with advancing age; 10% of people over 70 years of age have false-positive reactions. In the patient with a false-positive nontreponemal test, syphilis is excluded by a nonreactive treponemal test. Evaluation for Neurosyphilis Asymptomatic involvement of theCNS is detected by examination ofCSF for pleocytosis, increased protein concentration, andVDRLactivity. CSF abnormalities can be demonstrated in up to 40% of cases of primary or secondary syphilis and in 25% of cases of latent syphilis. In older asymptomatic seropositive individuals, the yield of lumbar puncture is relatively low. T. pallidum has been recovered by CSF inoculation into rabbits from up to 30% of patients with primary or secondary syphilis but rarely from those with latent syphilis. The demonstration of T. pallidum in CSF is often associated with other CSF abnormalities; however, organisms can be recovered from patients with otherwise-normal CSF. Before the advent of penicillin, the risk of developing clinical neurosyphilis was roughly proportional to the intensity of CSF changes in early syphilis. CSF examination is essential in the evaluation of any seropositive patient with neurologic signs and symptoms and is recommended for all patients with untreated syphilis of unknown duration or of >1 year's duration. The possibility of asymptomatic neurosyphilis in some patients with early disease is not addressed by these recommendations. Because standard therapy with penicillin G benzathine (benzathine benzylpenicillin) for early syphilis fails to result in treponemicidal levels in the CSF, some experts advise lumbar puncture in secondary and early latent syphilis, particularly in patients with HIV infection. In short, lumbar puncture should be performed in the evaluation of latent syphilis of>1 year's duration, in suspected neurosyphilis, and in late complications other than symptomatic neurosyphilis (since asymptomatic neurosyphilis may coexist with other late complications).CSFexamination is most clearly indicated in the following situations: neurologic signs or symptoms, treatment failure, a serum reagin titer³ 1:32, HIV antibody positivity, other evidence of active syphilis (e.g., aortitis, gumma, visual or hearing changes), or plans to administer nonpenicillin therapy. TheCSFVDRLtest is highly specific but relatively insensitive and may be nonreactive even in cases of progressive symptomatic neurosyphilis. The degree of sensitivity is highest in meningovascular syphilis and paresis and is lower in asymptomatic neurosyphilis and tabes dorsalis. The unabsorbed FTA test on CSF is reactive far more often than the CSF VDRL test in all stages of syphilis, but FTA reactivity may reflect passive transfer of serum antibody into the CSF. A nonreactive CSF FTA test, however, may be used to rule out neurosyphilis. Evaluation for Syphilis in Patients Infected with HIV Because persons at highest risk for syphilis (inner-city populations, homosexually active men, and people in many developing countries) are also at increased risk for HIV infection, these two infections frequently coexist in the same patient. There is evidence that syphilis and other genital-ulcer diseases may be important risk factors for the acquisition and transmission of HIV infection.
The manifestations of syphilis may be altered in patients with concurrent HIV infection, and multiple cases of neurologic relapse following standard therapy have been reported in HIV-infected patients. T. pallidum has been isolated from theCSF of several patients after therapy for early syphilis with penicillin G benzathine. A recent multicenter U.S. study of early syphilis found similar therapeutic responses in persons with and without concurrent HIV infection, although the study lacked sufficient statistical power to exclude an effect of HIV and 41% of subjects were lost to follow-up. This investigation confirmed the high rate ofCNSinvasion in early syphilis and the persistence of T. pallidum after standard therapy: 11 of 43 HIV-infected patients and 21 of 88 HIV-uninfected patients had T. pallidum detectable in CSF before therapy; 7 of the 35 patients who underwent lumbar puncture after therapy (some HIV-infected and others uninfected) still had T. pallidum detectable in CSF. The frequency of unusual clinical and laboratory manifestations of syphilis among patients co-infected with HIV is unknown. Such changes may be dependent on the stage of HIV infection and the degree of immunosuppression. There is no clear evidence that the sensitivity of serologic tests for syphilis or the serologic response to therapy in the vast majority of HIV-infected patients with early syphilis differs from the corresponding findings in patients not infected with HIV. Interpretation of serologic results should be the same for the two groups. Persons with newly diagnosed HIV infection should be tested for syphilis. Some authorities, persuaded by reports of the persistence of T. pallidum in theCSF of HIV-infected persons after standard penicillin benzathine therapy for early syphilis, recommend examination of CSF for evidence of neurosyphilis for all co-infected patients, regardless of the clinical stage of syphilis, with treatment for neurosyphilis if CSF abnormalities are found or if CSF examination is not performed. Others do not recommend routine CSF examination for HIV-co-infected patients with early syphilis and believe that standard therapy is sufficient. Serologic testing after treatment is important for all patients with syphilis, particularly those also infected with HIV. TREATMENT Treatment of Acquired Syphilis Penicillin G is the drug of choice for all stages of syphilis. T. pallidum is killed by very low concentrations of penicillin G, although a long period of exposure to penicillin is required because of the unusually slow rate of multiplication of the organism. The efficacy of penicillin for syphilis remains undiminished after 50 years of use. Other antibiotics effective in syphilis include the tetracyclines, erythromycin, and the cephalosporins. Aminoglycosides and spectinomycin inhibit T. pallidum only in very large doses, and the sulfonamides and the quinolones are inactive. Serum levels of penicillin G of³0.03 ug/mL for at least 7 days are considered necessary for the cure of early syphilis. Recurrence rates for a given regimen increase as infection progresses from incubating to seronegative primary to seropositive primary to secondary to late syphilis. Therefore, it is probable, but unproven, that a longer duration of therapy is required to effect cure as the infection progresses. For these reasons, some authorities use more prolonged penicillin therapy than that recommended by the U.S. Public Health Service when treating secondary, latent, or late syphilis.
The treatment regimens recommended for syphilis are summarized in Table 172-3 and are discussed below. Early Syphilis Preventive (abortive, "epidemiologic") treatment is recommended for seronegative individuals without signs of syphilis who have been exposed to infectious syphilis within the previous 3 months. Before treatment is given, every effort should be made to establish a diagnosis by examination and serologic testing. The regimens recommended for prevention are the same as those recommended for early syphilis. Penicillin G benzathine is the most widely used agent for the treatment of early syphilis (including primary, secondary, and early latent syphilis), although it is more painful on injection than penicillin G procaine. A single dose of 2.4 million units cures more than 95% of cases of primary syphilis. Because the drug's efficacy in secondary syphilis may be slightly lower, some physicians administer a second dose of 2.4 million units 1 week after the initial dose at this stage of disease. Clinical relapse can follow treatment with penicillin G benzathine in patients with both HIV infection and early syphilis. Because the risk of neurorelapse may be higher in HIV-infected patients, examination ofCSF from HIV-seropositive individuals with syphilis of any stage is recommended by some experts; therapy appropriate for neurosyphilis should be given if there is any evidence ofCNSsyphilis. For penicillin-allergic patients with early syphilis, a 2-week course of therapy with doxycycline or tetracycline is recommended. These regimens appear to be effective, although no well-controlled studies have been performed and poor compliance may be problematic. Although ceftriaxone and azithromycin have shown activity against T. pallidum in animals, human trials have not been of sufficient scope to permit the recommendation of either drug for any stage of syphilis. Late latent and late syphilis (normal CSF) IfCSFabnormalities are found, the patient should be treated for neurosyphilis. The recommended treatment for late latent syphilis with normal CSF, for cardiovascular syphilis, and for late benign syphilis (gumma) is penicillin G benzathine, 2.4 million units intramuscularly once a week for 3 successive weeks (7.2 million units total). Doxycycline or tetracycline (given for 1 month) offers an untested alternative for penicillin-allergic patients with latent or late syphilis and normal CSF. The clinical response to treatment for benign tertiary syphilis is usually impressive; however, responses to therapy for cardiovascular syphilis are not dramatic because aortic aneurysm and aortic regurgitation cannot be reversed by antibiotic treatment. Neurosyphilis The 1998 neurosyphilis treatment guidelines of the Centers for Disease Control and Prevention (CDC) are presented inTable 172-3. Penicillin G benzathine, given in total doses of up to 7.2 million units to adults or 50,000 units per kilogram to infants, does not produce detectable concentrations of penicillin G inCSF, and asymptomatic neurosyphilis may relapse in patients treated with 2.4 million units; the risk may be higher in HIV-infected patients. Therefore, the use of penicillin G benzathine alone for the treatment of neurosyphilis is not recommended. On the other hand, administration of intravenous penicillin G in doses of ³12 million units per day for 10 days or longer is thought to ensure treponemicidal concentrations of penicillin G in CSF and occasionally cures infection in patients who fail to respond to other therapy. The
clinical response to penicillin therapy for meningeal syphilis is dramatic, but the response to treatment for parenchymal neurosyphilis is variable. In general, treatment of neurosyphilis in which damage has already been done may produce no clinical change but may arrest disease progression. Several recent publications have reported neurologic relapse after high-dose intravenous penicillin therapy for neurosyphilis in HIV-infected patients. No alternative therapies have been explored, but careful follow-up is essential, and re-treatment is warranted in such patients. No data support the use of antibiotics other than penicillin G for the treatment of neurosyphilis; however, some of the third-generation cephalosporins may deserve further evaluation. In patients with penicillin allergy demonstrated by skin testing, desensitization may be the best course (Chap. 126). Management of Syphilis in Pregnancy Every pregnant woman should undergo a nontreponemal test at her first prenatal visit, and women at high risk of exposure should have a repeat test in the third trimester and at delivery. In the pregnant patient with presumed syphilis (evidenced by a reactive serology, with or without clinical manifestations) and with no history of treatment for syphilis, expeditious evaluation and initiation of treatment are essential. Therapy should be administered according to the stage of the disease, as for nonpregnant patients. Patients should be warned of the risk of a Jarisch-Herxheimer reaction, which may be associated with mild premature contractions but rarely results in premature delivery. Penicillin is the only recommended therapy for syphilis in pregnancy. If the patient has a well-documented penicillin allergy, desensitization and penicillin treatment should be undertaken in a hospital according to the 1998 sexually transmitted diseases treatment guidelines issued by theCDC. After treatment, a quantitative nontreponemal test should be repeated monthly throughout pregnancy. Treated women whose titers rise fourfold or who do not show a fourfold decrease in titer in a 3-month period should be re-treated. Evaluation and Management of Congenital Syphilis Newborn infants of mothers with reactiveVDRL orFTA-ABStests may themselves have reactive tests, whether or not they have become infected, because of transplacental transfer of maternal IgG antibody. Rising or persistent titers indicate infection, and the infant should be treated. Neonatal IgM antibody can be detected in cord or neonatal serum with the syphilis Captia-M or 19S IgM FTA-ABS test. Alternatively, monthly quantitative nontreponemal tests may be performed on asymptomatic infants born to women treated adequately with penicillin during pregnancy. An infant should be treated at birth if the seropositive mother has received penicillin therapy in the third trimester, inadequate penicillin treatment, or therapy with a drug other than penicillin; if her treatment status is unknown; or if the infant may be difficult to follow. It is unwise to require proof of diagnosis before treatment in such cases. TheCSFshould be examined to obtain baseline values before treatment. Penicillin is the only recommended drug for syphilis in infants. The penicillin dosage used for the treatment of the patient with late congenital syphilis is calculated in the same way as for the infant, until dosage based on weight reaches that used for adult neurosyphilis.
Specific recommendations for the treatment of infants are included in theCDC's 1998 guidelines. Jarisch-Herxheimer Reaction A dramatic though usually mild reaction consisting of fever (average temperature elevation, 1.5°C), chills, myalgias, headache, tachycardia, increased respiratory rate, increased circulating neutrophil count, and vasodilation with mild hypotension may follow the initiation of treatment for syphilis. This reaction occurs in approximately 50% of patients with primary syphilis, 90% of those with secondary syphilis, and 25% of those with early latent syphilis. The onset comes within 2 h of treatment, the temperature peaks at about 7 h, and defervescence takes place within 12 to 24 h. The reaction is more delayed in neurosyphilis, with fever peaking after 12 to 14 h. In patients with secondary syphilis, erythema and edema of the mucocutaneous lesions increase; occasionally, subclinical or early mucocutaneous lesions may first become apparent during the reaction. The pathogenesis of this reaction is undefined, although recent studies have demonstrated the induction of inflammatory mediators such as tumor necrosis factor by treponemal lipoproteins. Patients should be warned to expect such symptoms, which can be managed by bed rest and aspirin. Steroid and other anti-inflammatory therapy is not required for this mild transient reaction. Follow-Up Evaluation of Responses to Therapy The response of early syphilis to treatment should be determined by monitoring of the quantitativeVDRL orRPRtiter 1, 3, 6, and 12 months after treatment. More frequent serologic examination (1, 2, 3, 6, 9, and 12 months) is recommended for patients concurrently infected with HIV. Because theFTA-ABSand agglutination tests remain positive in most patients treated for seropositive syphilis, these tests are not useful in following the response to therapy. After successful treatment of seropositive first-episode primary or secondary syphilis, the VDRL titer progressively declines, becoming negative by 12 months in 40 to 75% of seropositive primary cases and in 20 to 40% of secondary cases. Patients with a history of syphilis have less rapid declines in titer and are less likely to become VDRL- or RPR-negative. If the VDRL test becomes negative or if VDRL titers drop to a fixed low value within 1 or 2 years, lumbar puncture is unnecessary since theCSFexamination is almost invariably normal and there is little risk of subsequent neurosyphilis. However, if a VDRL titer ³1:8 fails to fall by at least fourfold within 12 months, if the VDRL titer rises by fourfold, or if clinical symptoms persist or recur, re-treatment is indicated. Every effort should be made to differentiate treatment failure from reinfection, and the CSF should be examined. Patients in whom treatment failure is suspected, especially those with abnormal CSF, should be treated for neurosyphilis. If the patient remains seropositive but asymptomatic after such re-treatment, no further therapy is necessary. Patients treated for late latent syphilis frequently have low initial VDRL titers and may not have a fourfold drop after therapy with penicillin; about half of these patients remain seropositive (with low titers) for years after therapy. Re-treatment is not warranted unless the titer rises or signs and symptoms of syphilis appear. The activity of neurosyphilis correlates best with the degree ofCSFpleocytosis, and this measure provides the most sensitive index of response to treatment. CSF should be examined every 6 months for 3 years after the treatment of asymptomatic or symptomatic neurosyphilis or until CSF findings return to normal. An elevated CSF cell count falls to£10/uL in 3 to 12 months in 95% of adequately treated cases and becomes normal in all cases within 2 to 4 years. Elevated levels of CSF protein fall more slowly,
and the CSF VDRL titer declines gradually over a period of several years. Persistence of T. pallidum The persistence of T. pallidum in the aqueous humor,CSF, lymph nodes, brain, inflamed temporal arteries, and other tissues after "adequate" penicillin treatment has been suggested by dark-field microscopy, immunofluorescent antibody and silver staining techniques, rabbit inoculation, and PCR. Because the data on persisting treponemes are scanty, no modification of the treatment recommendations seems warranted for HIV-uninfected persons. Adherence to recommendations regarding CSF examination before the selection of therapy should minimize the possibility that T. pallidum will persist in the CSF. IMMUNITY TO AND PREVENTION OF SYPHILIS About 60% of contacts of patients with primary and secondary syphilis become infected, with lower risk in contacts exposed to early latent syphilis. The rate of development of acquired resistance to T. pallidum after natural or experimental infection is related to the size of the antigenic stimulus, which depends on both the size of the infecting inoculum and the duration of infection before treatment. The role of serum antibody in conferring immunity to syphilis remains controversial. Passively administered antibody prevents or delays the appearance of clinical manifestations of syphilis in the rabbit model; it does not prevent infection. Cellular immunity is considered to be of major importance in the healing of early lesions and the control of syphilitic infection. The cellular infiltration of early lesions predominantly involves T lymphocytes and macrophages. The cytokine milieu of primary and secondary lesions is of the TH1 type, consistent with the clearance of organisms by activated macrophages. Specific antibody enhances phagocytosis and is required for macrophage-mediated killing of T. pallidum. Inability to cultivate pathogenic treponemes in vitro has hindered the analysis of treponemal antigens. Attempts to induce immunity to syphilis by vaccination have shown limited promise, although repeated injection of rabbits withg-irradiated motile treponemes has conferred immunity to rechallenge. The outer membrane of T. pallidum contains few integral membrane proteins, and none has been definitively identified. Several newly described antigens of T. pallidum, including TprK, have induced partial immunity to challenge in the rabbit model, and syphilis vaccine development is being actively pursued. ACKNOWLEDGEMENT The author wishes to acknowledge the substantive contributions of the former coauthor, Dr. King K. Holmes, to the content and organization of this chapter. His original framework continues to serve as the structure for this revision. (Bibliography omitted in Palm version) Back to Table of Contents
173. ENDEMIC TREPONEMATOSES - Sheila A. Lukehart The endemic, or nonvenereal, treponematoses are bacterial infections that are caused by close relatives of Treponema pallidum subspecies pallidum, the etiologic agent of venereal syphilis (Chap. 172). Yaws, pinta, and endemic syphilis are distinguished from venereal syphilis by mode of transmission, age of acquisition, geographic distribution, and clinical features. These infections are limited primarily to rural areas of developing nations and are seen in the United States and Europe only in recent immigrants from endemic regions. Much of our "knowledge" about the endemic treponematoses is based upon impressions and observations of health care workers who have visited endemic areas; virtually no well-designed studies of the natural history, diagnosis, or treatment of these infections have been conducted. A comparison of the treponemal infections is shown in Table 173-1. EPIDEMIOLOGY The endemic treponematoses are chronic diseases acquired during childhood and, like syphilis, can cause severe late manifestations years after initial infection. These infections were very common in Africa, Asia, and South America when the World Health Organization (WHO) and UNICEF embarked on a highly successful mass eradication campaign. From 1952 to 1969, it is estimated that over 160 million people were examined for treponemal infections and over 50 million cases, contacts, and latent infections were treated. This categorical program is one of WHO's outstanding successes in that the prevalence of active yaws was reduced from >20% to 10% in some regions of Mali, Niger, Burkina Faso, and Senegal. In Asia and the western Pacific, yaws is still prevalent in Indonesia, Papua New Guinea, and the Solomon Islands; cases have also been identified in Laos and Kampuchea. In the Americas, foci of yaws persist in Haiti and other Caribbean islands, Peru, Colombia, Ecuador, Brazil, Guyana, and Surinam. Pinta is limited to Central America and northern South America, where it is found rarely and only in remote villages. MICROBIOLOGY The etiologic agents of the endemic treponematoses are T. pallidum subspecies pertenue (yaws), T. pallidum subspecies endemicum (endemic syphilis), and T. carateum (pinta). These little-studied organisms are morphologically identical to T. pallidum subspecies pallidum, and no antigenic differences among the pathogenic treponemes have been identified to date. A controversy has existed about whether the treponematoses are caused by different organisms or by the same organism, with clinical manifestations and routes of transmission being defined by the climate of the
region and the culture of the population. Three of the four organisms have been placed in the same species because of their genetic similarity; the fourth (T. carateum) remains a separate species simply because no organisms have been available for genetic studies. However, a molecular signature has been defined that can be used to differentiate T. pallidum subspecies pallidum from T. pallidum subspecies pertenue and T. pallidum subspecies endemicum, and unpublished studies have identified a number of distinct differences in the tpr gene family between venereal and nonvenereal treponemes. Whether these differences are related to the different clinical courses has not yet been determined. CLINICAL FEATURES All of the treponemal infections are characterized by defined disease stages, with a localized primary lesion, disseminated secondary lesions, periods of latency, and possible late lesions. The stages are most clearly defined in venereal syphilis, while the primary and secondary manifestations are more frequently overlapping in yaws and endemic syphilis; the late manifestations of pinta are very mild relative to the destructive lesions of the other treponematoses. The current preference is to divide the clinical course of the endemic treponematoses into "early" and "late" stages. The major clinical features that are thought to differ between venereal syphilis and the nonvenereal infections are the lack of congenital transmission and lack of central nervous system (CNS) involvement in the nonvenereal infections. It is not known whether these distinctions are accurate. Because of the high degree of genetic relatedness among the organisms, there is little biologic reason to think that T. pallidum subspecies endemicum and T. pallidum subspecies pertenue would be unable to cross the blood-brain barrier or to invade the placenta. These organisms obviously can disseminate from the site of primary infection to other tissues, and they can persist for decades. In this respect, they are like T. pallidum subspecies pallidum. Even if invasion of the placenta or the CNS occurs in endemic treponemal infection, there are a number of reasons that these manifestations might not have been recognized. The lack of recognized congenital infection may be due to the fact that the nonvenereal treponematoses are usually acquired during childhood. The degree of spirochetemia (the presumed source of placental and fetal infection) is greatly diminished during the latent stage, and by the time an infected girl becomes sexually mature, she would be at low risk for transplacental transmission. Neurologic involvement may not have been recognized in nonvenereal treponemal infection because of the lack of trained medical personnel in endemic regions, the lag of years to decades between acquisition of infection and possible CNS manifestations, or a low rate of symptomatic CNS disease. The lack of longitudinal studies in endemic areas makes conclusions about the natural history of these infections tenuous. Some published evidence supports congenital transmission as well as cardiovascular, ophthalmologic, andCNSinvolvement in yaws. Although the case is strong, particularly for CNS involvement, most studies that have shown a relatively high incidence (average, 24.9%) of cerebrospinal fluid (CSF) abnormalities in patients with yaws were not controlled for other possible causes of CSF abnormalities, did not include treponeme-specific tests, or did not follow patients for resolution of abnormalities after antitreponemal therapy. Thus, while no firm conclusions can be drawn about the
invasion of the CNS and placenta by the non-pallidum treponemes, it may be erroneous to accept unquestioningly the frequently repeated belief that these organisms fail to cause such manifestations. Yaws Also known as pian, framboesia, or bouba, yaws is a chronic infection that is usually acquired in childhood and is caused by T. pallidum subspecies pertenue. The disease is characterized by the development of one or several primary lesions (called the "mother yaw"), followed by the appearance of multiple disseminated skin lesions. The early lesions may persist for many months, are infectious, and usually recur several times within the early years of infection. Late manifestations are destructive and can involve skin, bone, and joints. The infection is transmitted by direct contact with infectious lesions, and transmission may be enhanced by disruption of the skin by insect bites or abrasions. Children with open lesions and without covering clothing are most likely to transmit infection during play or group sleeping. After an average incubation period estimated at 3 to 4 weeks, the first lesion begins as a papule, usually on an extremity, and then enlarges (particularly during moist warm weather) to become papillomatous or "raspberry-like" (thus the name "framboesia") (Fig. 173-2). Regional lymphadenopathy develops, and the lesion usually heals within 6 months; dissemination is thought to occur during the early weeks and months of infection. A generalized secondary eruption, accompanied by generalized lymphadenopathy, appears either concurrent with or following the primary lesion, may take several forms (macular, papular, or papillomatous), and may become secondarily infected with other bacteria. Painful papillomatous lesions on the soles of the feet result in a painful crablike gait ("crab yaws"), and periostitis may result in nocturnal bone pain and polydactylitis. All early skin lesions are infectious, and cutaneous relapses are common during the first 5 years. Late yaws is recognized in ~10% of untreated patients and is manifested by gummas of the skin and long bone, hyperkeratoses of the palms and soles, osteitis and periostitis, and hydrarthrosis. The late gummatous lesions are characteristically very destructive and extensive. Destruction of the nose, maxilla, palate, and pharynx is termed gangosa and is similar to the destructive lesions seen in leprosy and leishmaniasis. Endemic Syphilis Endemic syphilis, also called bejel, siti, dichuchwa, njovera, or skerljevo, is a chronic infection caused by T. pallidum subspecies endemicum. Like other endemic treponematoses, endemic syphilis is chronic and is acquired in childhood. The early lesions are primarily localized to the mucocutaneous and mucosal surfaces, and the infection may be transmitted by direct contact or by shared drinking and eating utensils. A role for insects in transmission has been suggested but is unproved. The initial lesion often goes unrecognized, and the first noticeable lesion is usually an intraoral mucous patch or a mucocutaneous lesion resembling the condylomata lata of secondary syphilis (Fig. 173-2). This eruption may last for months or even years, and treponemes can readily be demonstrated in early lesions. Periostitis and regional lymphadenopathy are common. After a variable period of latency, late manifestations may appear, including osseous and cutaneous gummas. Destructive gummas, osteitis, and gangosa are more common in endemic syphilis than in late yaws. Gummas of the nipples develop in women who have previously had endemic syphilis and who breast-feed infants with oral lesions. Thus, it appears that the late lesion may result from repeated exposure of a sensitized host.
Pinta Pinta (also called mal del pinto, carate, azul, or purupuru) is the most benign of the treponemal infections and is caused by T. carateum. This disease has three stages that are characterized by marked changes in skin color, but it does not appear to cause destructive lesions or to involve other tissues. Transmission occurs by direct contact, usually during late childhood. The initial papule is most often located on the extremities or face and is pruritic. After one to many months of infection, numerous disseminated secondary lesions (pintides) appear. These lesions are initially red but become deeply pigmented, ultimately turning a dark slate blue. The secondary lesions are infectious and highly pruritic and may persist for years. Late pigmented lesions are called dyschromic macules and contain treponemes. Over time, most pigmented lesions show varying degrees of depigmentation, becoming brown and eventually white and giving the skin a mottled appearance. The white achromic lesions are characteristic of the late stage. DIAGNOSIS Diagnosis of the endemic treponematoses is based upon clinical manifestations and, when available, serologic testing. The same tests that are used for venereal syphilis (Chap. 172) become reactive during all treponemal infections, and there is no serologic test than can discriminate among the different infections. The nonvenereal treponemal infections should be considered in the evaluation of a reactive syphilis serology in any person who has immigrated from an endemic area. TREATMENT The recommended therapy for patients and their contacts is benzathine penicillin at a dose of 1.2 million units intramuscularly; that for children under 10 years of age is 600,000 units. This is half the dose recommended for patients and contacts with early venereal syphilis. There have been no controlled studies to show that the lower dose is effective in stopping relapse or progression to late disease. Definitive evidence of resistance to penicillin is lacking. However, because failure to heal existing lesions and frequent relapse following treatment for yaws have been described in Papua New Guinea, some health workers have suggested doubling the recommended dose of benzathine penicillin. Solely on the basis of experience with venereal syphilis, it is thought that doxycycline, tetracycline, and erythromycin (at doses appropriate for syphilis;Chap. 172) are therapeutic alternatives for patients allergic to penicillin. A Jarisch-Herxheimer reaction (Chap. 172) may follow treatment of endemic treponematoses. CONTROL The endemic treponematoses can be controlled with inexpensive therapy. However, the often-remote locations of the affected populations limit availability of medical care. Although the mass treatment programs of three decades ago were widely successful, time has shown that sustained control requires vigilance in regular screening and in the investigation of outbreaks -- luxuries that are often impossible in countries with more pressing medical needs. There is concern that, as HIV spreads throughout developing countries, it may markedly affect the manifestations and transmission of the endemic
treponematoses. ACKNOWLEDGEMENT The author gratefully acknowledges the substantial contributions of Dr. Peter Perine, the author of previous editions of this chapter, to the framework of the current chapter and to our insight into these little-studied infections. (Bibliography omitted in Palm version) Back to Table of Contents
174. LEPTOSPIROSIS - Peter Speelman Leptospirosis is an infectious disease caused by pathogenic leptospires and characterized by a broad spectrum of clinical manifestations, varying from inapparent infection to fulminant, fatal disease. In its mild form, leptospirosis may present as an influenza-like illness with headache and myalgias. Severe leptospirosis, characterized by jaundice, renal dysfunction, and hemorrhagic diathesis, is referred to as Weil's syndrome. ETIOLOGIC AGENTS Leptospires are spirochetes belonging to the order Spirochaetales and the family Leptospiraceae. Traditionally, the genus Leptospira comprised two species: the pathogenic L. interrogans and the free-living L. biflexa. Although seven species of pathogenic leptospires are now recognized on the basis of their DNA relatedness, it is more practical clinically and epidemiologically to use a classification based on serologic differences. The pathogenic leptospires are divided into serovars according to their antigenic composition. More than 200 serovars make up the 23 serogroups. Leptospires are coiled, thin, highly motile organisms with hooked ends and two periplasmic flagella that permit burrowing into tissue. These organisms are 6 to 20 um long and about 0.1 um wide; they stain poorly but can be seen microscopically by dark-field examination and after silver impregnation staining. Leptospires require special media and conditions for growth; it may take weeks for cultures to become positive. EPIDEMIOLOGY Leptospirosis is a zoonosis with a worldwide distribution that affects at least 160 mammalian species. Rodents, especially rats, are the most important reservoir, although other wild mammals, dogs, fish, and birds may also harbor these microorganisms. Leptospires establish a symbiotic relationship with their host and can persist in the renal tubules for years. Some serovars are associated with particular animals -- e.g., icterohaemorrhagiae/copenhageni with rats, grippotyphosa with voles, hardjo with cattle, canicola with dogs, and pomona with pigs. Transmission of leptospires may follow direct contact with urine, blood, or tissue from an infected animal or exposure to a contaminated environment; human-to-human transmission is rare. Since leptospires are excreted in the urine and can survive in water for many months, water is an important vehicle in their transmission. Epidemics of leptospirosis may result from exposure to flood waters contaminated by urine from infected animals, as has been reported from Nicaragua. Leptospirosis occurs most commonly in the tropics because the climate as well as the sometimes poor working and hygienic conditions favor the pathogen's survival. Humans are not commonly infected with leptospires. However, in the United States, the 40 to 120 cases reported annually to the Centers for Disease Control and Prevention certainly represent a significant underestimation of the total number. Certain occupational groups are at especially high risk; included are veterinarians, agricultural workers, sewage workers, slaughterhouse employees, and workers in the fishing
industry. Such individuals may acquire leptospirosis through direct exposure to or contact with contaminated water and soil. Leptospirosis has also been recognized in deteriorating inner cities where rat populations are expanding. One report described leptospirosis in urban residents of Baltimore who were sporadically exposed to rat urine. In western countries, recreational exposure and domestic animal contact are also prominent sources of leptospirosis. Recreational water activities, such as canoeing, windsurfing, swimming, and waterskiing, place persons at risk for leptospirosis. Sometimes the infection is acquired during travel abroad. In a recent study in the Netherlands, 14% of patients with confirmed leptospirosis had acquired the infection while traveling in tropical countries, mostly in Southeast Asia. Transmission via laboratory accidents has been reported but is rare. Occasionally, leptospirosis develops after unanticipated immersion in contaminated water (e.g., in an automobile accident). Most cases occur in men, with a peak incidence during the summer and fall in western countries and during the rainy season in the tropics. PATHOGENESIS The pathogenesis of leptospirosis is incompletely understood. Leptospires may enter the host through abrasions in the skin or through intact mucous membranes, especially the conjunctiva and the lining of the oro- and nasopharynx. Drinking of contaminated water may introduce leptospires through the mouth, throat, or esophagus. After entry of the organisms, leptospiremia develops, with subsequent spread to all organs. Multiplication takes place in blood and in tissues, and leptospires can be isolated from blood and cerebrospinal fluid (CSF) during the first 4 to 10 days of illness. It is not clear why the presence of leptospires in the CSF does not cause damage. All forms of leptospires can damage the wall of small blood vessels; this damage leads to vasculitis with leakage and extravasation of cells, including hemorrhages. The most important known pathogenic properties of leptospires are adhesion to cell surfaces and cellular toxicity. Vasculitis is responsible for the most important manifestations of the disease. Although leptospires mainly infect the kidneys and liver, any organ may be affected. In the kidney, leptospires migrate to the interstitium, renal tubules, and tubular lumen, causing interstitial nephritis and tubular necrosis. Hypovolemia due to dehydration or altered capillary permeability may contribute to the development of renal failure. In the liver, centrilobular necrosis with proliferation of Kupffer cells may be found. However, severe hepatocellular necrosis is not a feature of leptospirosis. Pulmonary involvement is the result of hemorrhage and not of inflammation. Invasion of skeletal muscle by leptospires results in swelling, vacuolation of the myofibrils, and focal necrosis. In severe leptospirosis, vasculitis may ultimately impair the microcirculation and increase capillary permeability, resulting in fluid leakage and hypovolemia. When antibodies are formed, leptospires are eliminated from all sites in the host except the eye, the proximal renal tubules, and perhaps the brain, where they may persist for weeks or months. The persistence of leptospires in the aqueous humor occasionally causes chronic or recurrent uveitis. The systemic immune response is effective in eliminating the organism but may also produce symptomatic inflammatory reactions. A rise in antibody titer coincides with the development of meningitis; this association
suggests that an immunologic mechanism is responsible. After the start of antimicrobial treatment for leptospirosis, a Jarisch-Herxheimer reaction similar to that seen in other spirochetal diseases may develop. Although frequently described in older publications, this reaction seems to be a rare event in leptospirosis and is certainly less frequent in this infection than in other spirochetal diseases. CLINICAL MANIFESTATIONS It is important to try to obtain a history of exposure to contaminated materials. Serologic evidence of past inapparent infection is found in 15 to 40% of persons who have been exposed but have not become ill. In symptomatic cases of leptospirosis, clinical manifestations vary from mild to serious or even fatal. More than 90% of symptomatic persons have the relatively mild and usually anicteric form of leptospirosis, with or without meningitis. Severe leptospirosis with profound jaundice (Weil's syndrome) develops in 5 to 10% of infected individuals. The incubation period is usually 1 to 2 weeks but ranges from 2 to 26 days. Typically, an acute leptospiremic phase is followed by an immune leptospiruric phase. The distinction between the first and second phases is not always clear, and milder cases do not always include the second phase. Anicteric Leptospirosis Leptospirosis may present as an acute influenza-like illness, with fever, chills, severe headache, nausea, vomiting, and myalgias. Muscle pain, which especially affects the calves, back, and abdomen, is an important feature of leptospiral infection. Less common features include sore throat and rash. The patient usually has an intense headache (frontal or retroorbital) and sometimes develops photophobia. Mental confusion may be evident. Pulmonary involvement, manifested in most cases by cough and chest pain and in a few cases by hemoptysis, is not uncommon. The most common finding on physical examination is fever with conjunctival suffusion. Less common findings include muscle tenderness, lymphadenopathy, pharyngeal injection, rash, hepatomegaly, and splenomegaly. The rash may be macular, maculopapular, erythematous, urticarial, or hemorrhagic. Mild jaundice may be present. Most patients become asymptomatic within 1 week. After an interval of 1 to 3 days, the illness recurs in a number of cases. The start of this second (immune) phase coincides with the development of antibodies. Symptoms are more variable than during the first (leptospiremic) phase. Usually the symptoms last for only a few days, but occasionally they persist for weeks. Often the fever is less pronounced and the myalgias are less severe than in the leptospiremic phase. An important event during the immune phase is the development of aseptic meningitis. Although no more than 15% of all patients have symptoms and signs of meningitis, many patients may haveCSFpleocytosis. Meningeal symptoms usually disappear within a few days but may persist for weeks. Similarly, pleocytosis generally disappears within 2 weeks but occasionally persists for months. Iritis, iridocyclitis, and chorioretinitis -- late complications that may persist for years -can become apparent as early as the third week but often present several months after the initial illness. One epidemic of uveitis among patients with leptospirosis has been reported.
Severe Leptospirosis (Weil's Syndrome) Weil's syndrome, the most severe form of leptospirosis, is characterized by jaundice, renal dysfunction, hemorrhagic diathesis, and high mortality. This syndrome is frequently but not exclusively associated with infection due to serovar icterohaemorrhagiae/copenhageni. The onset of illness is no different from that of less severe leptospirosis; however, after 4 to 9 days, jaundice as well as renal and vascular dysfunction generally develop. Although some degree of defervescence may be noted after the first week of illness, a biphasic disease pattern like that seen in anicteric leptospirosis is lacking. The jaundice of Weil's syndrome, which can be profound and give an orange cast to the skin, is usually not associated with severe hepatic necrosis. Death is rarely due to liver failure. Hepatomegaly and tenderness in the right upper quadrant are usually detected. Splenomegaly is found in 20% of cases. Renal failure may develop, often during the second week of illness. Hypovolemia and decreased renal perfusion contribute to the development of acute tubular necrosis with oliguria or anuria. Dialysis is sometimes required, although a fair number of cases can be managed without dialysis. Renal function may be completely regained. Pulmonary involvement occurs frequently, resulting in cough, dyspnea, chest pain, and blood-stained sputum, and sometimes in hemoptysis or even respiratory failure. Hemorrhagic manifestations are seen in Weil's syndrome: epistaxis, petechiae, purpura, and ecchymoses are found commonly, while severe gastrointestinal bleeding and adrenal or subarachnoid hemorrhage are detected rarely. Rhabdomyolysis, hemolysis, myocarditis, pericarditis, congestive heart failure, cardiogenic shock, adult respiratory distress syndrome, and multiorgan failure have all been described during severe leptospirosis. LABORATORY AND RADIOLOGIC FINDINGS The kidneys are invariably involved in leptospirosis (Fig. 174-CD1). Related findings range from urinary sediment changes (leukocytes, erythrocytes, and hyaline or granular casts) and mild proteinuria in anicteric leptospirosis to renal failure and azotemia in severe disease. The erythrocyte sedimentation rate is usually elevated. In anicteric leptospirosis, peripheral leukocyte counts range from 3000 to 26,000/uL, with a left shift; in Weil's syndrome, leukocytosis is often marked. Mild thrombocytopenia occurs in up to 50% of patients and is associated with renal failure. In contrast to patients with acute viral hepatitis, those with leptospirosis typically have elevated serum levels of bilirubin and alkaline phosphatase as well as mild increases (up to 200 U/L) in serum levels of aminotransferases. In Weil's syndrome, the prothrombin time may be prolonged but can be corrected with vitamin K. Levels of creatine phosphokinase, which are elevated in up to 50% of patients with leptospirosis during the first week of illness, may help to differentiate this infection from viral hepatitis. When a meningeal reaction develops, polymorphonuclear leukocytes predominate
initially and the number of mononuclear cells increases later. The protein concentration in theCSF may be elevated; CSF glucose levels are normal. In severe leptospirosis, pulmonary radiographic abnormalities are more common than would be expected on the basis of physical examination. These abnormalities most frequently develop 3 to 9 days after the onset of illness. The most common radiographic finding is a patchy alveolar pattern that corresponds to scattered alveolar hemorrhage. Radiographic abnormalities most often affect the lower lobes in the periphery of the lung fields. DIAGNOSIS A definite diagnosis of leptospirosis is based either on isolation of the organism from the patient or on seroconversion or a rise in antibody titer in the microscopic agglutination test (MAT). For a presumptive diagnosis of leptospirosis, an antibody titer of ³1:100 in the MAT or a positive macroscopic slide agglutination test in the presence of a compatible clinical illness is required. Antibodies generally do not reach detectable levels until the second week of illness. The antibody response can be affected by early treatment. The macroscopic slide agglutination test with killed antigen is useful for screening but is not specific. TheMAT, which uses a battery of live leptospiral strains, and the enzyme-linked immunosorbent assay (ELISA), which uses a broadly reacting antigen, are the standard serologic procedures. These tests usually are available only in specialized laboratories and are used for the determination of the antibody titer and for the tentative identification of the serovar involved (thus the importance of using antigens representative of the serovars prevalent in the particular geographic area). Since cross-reactions occur frequently, however, it is often impossible to identify the infecting serovar. Serologic testing cannot be used as the basis for a decision about whether to start treatment. In addition to theMAT and theELISA, various other tests with diagnostic value have been developed. Some tests, such as an indirect hemagglutination test, a microcapsule agglutination test, and an IgM ELISA, are commercially available. Dot-ELISA, gold immunoblot, and polymerase chain reaction techniques have been developed but are not yet used for routine diagnosis. Leptospires can be isolated from blood and/orCSFduring the first 10 days of illness and from urine for several weeks beginning at around 1 week. Cultures may become positive after 2 to 4 weeks, with a range of 1 week to 4 months. Sometimes urine cultures remain positive for months or years after the start of illness. For isolation of leptospires from body fluids or tissues, Ellinghausen-McCullough-Johnson-Harris (EMJH) medium is useful; other possibilities are Fletcher medium and Korthoff medium. Specimens can be mailed to a reference laboratory for culture, since leptospires remain viable in anticoagulated blood (heparin, EDTA, or citrate) for up to 11 days. Isolation of leptospires is important since it is the only way the infecting serovar can be correctly identified. Dark-field examination of blood or urine frequently results in misdiagnosis and should not be used.
DIFFERENTIAL DIAGNOSIS Leptospirosis should be differentiated from other febrile illnesses associated with headache and muscle pain, such as malaria, enteric fever, viral hepatitis, dengue, Hantavirus infections, and rickettsial diseases. In light of the strong similarity in epidemiology and clinical presentation between leptospirosis and Hantavirus infections and given the reported occurrence of dual infections, it is advisable to conduct serologic testing for Hantavirus in cases of suspected leptospirosis. When patients have a flulike disease with disproportionately severe myalgia or aseptic meningitis, a diagnosis of leptospirosis should be considered. TREATMENT The effectiveness of antimicrobial therapy for the mild febrile form of leptospirosis is controversial, but such treatment is indicated for more severe forms. Treatment should be initiated as early as possible; nevertheless, contrary to previous reports, treatment started after the first 4 days of illness is effective. For severe cases of leptospirosis, intravenous administration of penicillin G, amoxicillin, ampicillin, or erythromycin is recommended (Table 174-1). In milder cases, oral treatment with tetracycline, doxycycline, ampicillin, or amoxicillin should be considered. Although several other antibiotics, including newer cephalosporins, are highly active against leptospires in vitro, no clinical experience has yet been gained with these drugs. In rare cases, a Jarisch-Herxheimer reaction develops within hours after the start of antimicrobial therapy (see "Pathogenesis" above). Although so far the only effective mode of management is supportive, the role of antibodies to tumor necrosis factor in the treatment of this reaction deserves further study. A beneficial effect of the use of such antibodies for the modulation of the reaction has been demonstrated in patients with louse-borne relapsing fever. Patients with severe leptospirosis and renal failure may require dialysis. Those with Weil's syndrome may need transfusions of whole blood and/or platelets. Intensive care may be necessary. PROGNOSIS Most patients with leptospirosis recover. Mortality is highest among patients who are elderly and those who have Weil's syndrome. Leptospirosis during pregnancy is associated with high fetal mortality. Long-term follow-up of patients with renal failure and hepatic dysfunction has documented good recovery of renal and hepatic function. PREVENTION Individuals who may be exposed to leptospires through their occupations or their involvement in recreational water activities should be informed about the risks. Measures for controlling leptospirosis include avoidance of exposure to urine and tissues from infected animals, vaccination of animals, and rodent control. The animal vaccine used in a given area should contain the serovars known to be present in that area. Unfortunately, some vaccinated animals still excrete leptospires in their urine. Vaccination of humans against a specific serovar prevalent in an area has been
undertaken in some European and Asian countries and has proved effective. Chemoprophylaxis with doxycycline (200 mg once a week) has appeared to be efficacious in military personnel but is indicated only in rare instances of sustained short-term exposure. (Bibliography omitted in Palm version) Back to Table of Contents
175. RELAPSING FEVER - David T. Dennis, Grant L. Campbell DEFINITION The term relapsing fever describes two distinct borrelial disease entities: louse-borne relapsing fever (LBRF) and tick-borne relapsing fever (TBRF). Both are characterized by recurrent acute episodes of spirochetemia and fever alternating with spirochetal clearance and apyrexia. ETIOLOGY The worldwide distribution of relapsing spirochetal fevers was recognized in the early part of the twentieth century, and the causative agents were shown to be transmitted by lice and ticks. Borrelia recurrentis was identified as the cause ofLBRF; differing strains of borreliae causingTBRFwere usually named according to the species of Ornithodoros tick responsible for their transmission (Table 175-1). Sequencing of both flagellin and 16S ribosomal RNA genes reveals homogeneity among LBRF strains and considerable heterogeneity between Old World and New World TBRF strains. Relapsing-fever borreliae are gram-negative bacteria that belong to the family Spirochaetaceae. They are helical in shape and average 0.2 to 0.5 um in width and 5 to 20 um in length. They comprise an outer membrane, an intermediate peptidoglycan layer, and an inner cytoplasmic membrane, which encloses the protoplasmic cylinder. Periplasmic flagella (variable numbers have been described) are situated beneath the outer membrane. Relapsing-fever borreliae are slow-growing and microaerophilic; they grow best at 30 to 35°C. BothTBRF andLBRFspirochetes grow well in Barbour-Stoenner-Kelly (BSK II) medium. Relapsing-fever borreliae are distinguished by remarkable antigenic variability and strain heterogeneity. New Borrelia serotypes spontaneously emerge at a high rate, resulting from a unique process of DNA rearrangement within genes located on linear plasmids. These genes code for variable major proteins (VMPs) found on the spirochete's outer-membrane surface. This antigenic variation, generated by sequential expression of previously silent vmp genes for serotype-specific VMPs, allows the borreliae to escape the immune response of the host and results in the relapse phenomenon characteristic of infection with these organisms. Borrelial vmp gene expression also varies between mammalian and arthropod hosts. EPIDEMIOLOGY Louse-Borne Relapsing Fever Body lice (Pediculus humanus var. corporis) become infected with B. recurrentis by feeding on spirochetemic humans, the only reservoirs of infection. In lice, B. recurrentis spirochetes are found almost exclusively in the hemolymph; humans acquire infection when infected body lice are crushed and their fluids contaminate mucous membranes or breaks in the skin (such as abrasions caused by scratching of pruritic louse bites). Spirochetes are not transmitted directly by the bite of a louse (anterior station transmission) or by inoculation of louse feces (posterior station transmission). Lice have a life span of only a few weeks, feed at frequent intervals, and survive only a few days off the human host.
LBRFhas severely affected military and civilian populations disrupted by war and other disasters. During the Industrial Revolution, the disease was common among slum dwellers, prisoners, and others living in impoverished, overcrowded, and unhygienic conditions. In the first half of the twentieth century, during periods of war and famine, both LBRF and louse-borne typhus were epidemic in eastern Europe, the Balkans, and the former Soviet Union. LBRF has disappeared from its former global range as improvements have been made in standards of living, sanitation, and hygiene; it is now an important disease only in northeastern Africa, especially the highlands of Ethiopia, where an estimated 10,000 cases occur annually. In Ethiopia, the disease affects mostly homeless men crowded together in unhygienic circumstances, especially during the cool rainy season, when it is more difficult for them to change and wash their clothing. LBRF has repeatedly spilled out of Ethiopia into neighboring Somalia and Sudan, especially affecting displaced persons. LBRF does not pose a significant risk to tourists or other casual visitors but can be acquired from lice by persons (such as relief workers) in intimate contact with those affected as well as through accidental needle stick or mucocutaneous contact with infected blood. Tick-Borne Relapsing Fever Soft ticks (Argasidae, Ornithodoros spp.) transmitTBRF. The ticks become infected by feeding on spirochetemic hosts. Except for B. duttoni (a prominent cause of TBRF in sub-Saharan Africa), TBRF borreliae are zoonotic disease agents found naturally in rodents (rats, mice, chipmunks, and squirrels) and in lagomorphs (rabbits and hares). The spirochetes are transmitted by ticks to humans and animals via saliva and excretory fluids when the tick feeds. Infection in ticks is transmitted vertically from one stage to the next; in some species, infection is transmitted transovarially over several generations. Soft ticks are hardy and can survive for 10 years or more with only an occasional blood meal. These ticks feed painlessly, relatively quickly (for 20 to 45 min), and usually at night while hosts are sleeping. Thus patients with TBRF are often unaware of tick exposures. TBRFborreliae are widely distributed throughout the world. Human infection with these organisms is generally underrecognized and underreported. TBRF is most highly endemic in sub-Saharan Africa but is also found in countries of the Mediterranean littoral, Middle Eastern states, southern Russia, the Indian subcontinent, and China. In the United States, this disease occurs west of the Mississippi River, especially in mountainous areas, where B. hermsii is the causative agent. TBRF is reported at low frequency throughout Latin America. The disease typically occurs sporadically or in small -- often familial -- clusters. Infected soft ticks may cause repeated infections among persons living or sleeping in the same dwelling. In sub-Saharan Africa, O. moubata, the vector of B. duttoni, infests native huts and rest houses, hiding in crevices of floors and walls during the day and emerging at night to feed on sleeping inhabitants. In the United States, B. hermsii infections most often occur during spring and summer months among persons sleeping in rustic mountain cabins. Infections of humans are sometimes precipitated by the disappearance of rodents (e.g., as a result of epizootic plague) that nest in foundations, wall spaces, and attics and that serve as the usual maintenance hosts for O. hermsi ticks. Outbreaks caused by B. hermsii have taken place among persons staying in cabins along the north rim of the Grand Canyon and in the mountains of California, Idaho, and Colorado. Rodent-infested caves in southwestern states are associated with occasional cases of relapsing fever caused by
B. turicatae. PATHOGENESIS AND PATHOLOGY In humans, relapsing-fever borreliae penetrate the skin or mucous membranes, multiply in the blood, and circulate in great numbers during febrile periods. The organisms also may be found in the liver, spleen, central nervous system, bone marrow, and other tissues and may be sequestered at these sites during periods of remission. The severity of disease is positively related to spirochete density in the blood. Even though the pathophysiologic manifestations of the disease resemble responses to endotoxin, and although plasma from some patients with relapsing fever coagulates Limulus amebocyte lysates, borreliae and other spirochetes have not been shown to express a true lipopolysaccharide (endotoxin) molecule. Infection with B. recurrentis does, however, activate protein mediators of inflammation, such as Hageman factor, prekallikrein, and proteins of the complement system; furthermore, a spirochetal heat-stable pyrogenic factor stimulates mononuclear phagocytes to express increased amounts of leukocyte pyrogen and thromboplastin. The Jarisch-Herxheimer reaction in patients withLBRF is associated with a release of various cytokines into the plasma, including interleukin 6, interleukin 8, C-reactive protein, and enormous amounts of tumor necrosis factor a (TNF-a). Pretreatment of LBRF patients with antibody to TNF-asuppresses the Jarisch-Herxheimer reactions that follow penicillin treatment and reduces the plasma concentrations of certain other cytokines. Findings at autopsy of patients with relapsing fever most often include hepatosplenomegaly and variable edema and swelling of other organs, such as the brain, lungs, and kidneys. On microscopic examination, the spleen is congested and contains multiple microabscesses composed of mononuclear cells that replace the white pulp, the myocardium displays diffuse histiocytic inflammation and interstitial edema, and the liver has areas of midzonal necrosis. Petechial hemorrhages are commonly evident over the surfaces of the meninges, pleura, heart, spleen, liver, kidneys, and mesentery. Subcapsular and parenchymal hemorrhagic infarcts of the spleen, heart, liver, and brain are sometimes grossly visible. Icterus is a common finding in severe and fatal cases of relapsing fever. CLINICAL MANIFESTATIONS The clinical manifestations of LBRFand TBRFare similar. The mean incubation period is 7 days (range, 2 to 18 days), and the onset of illness is sudden, with fever, headache, shaking chills, sweats, myalgias, and arthralgias. The arthralgia of relapsing fever can be severe, involving small and large joints, but there is no evidence of arthritis. Dizziness, nausea, and vomiting are common. Sleep may be difficult and is sometimes accompanied by disturbing dreams. The patient is coherent but withdrawn, thirsty, and disinterested in food and other outside stimuli. The fever is high from the first, with a usual temperature of³40°C (³104°F); fever is most often irregular in pattern and is sometimes accompanied by delirium. Patients become progressively prostrate as the disease advances. The pulse is rapid and the patient is mildly tachypneic. Meningism may be found. The conjunctivae are often injected, and the patient usually exhibits
photophobia. The sclerae are sometimes icteric, most commonly in the later stages of illness. The mucous membranes are often dry, and the patient is usually dehydrated. Scattered petechiae develop on the trunk, extremities, and mucous membranes in one-third or more of patients with LBRF and in fewer patients with TBRF. A nonproductive cough is common, but chest sounds are usually normal; pleuritic pain and an accompanying pleuritic rub are sometimes noted. Cardiac findings are compatible with a high-output state; tachycardia and summation gallop are common. Tender enlargement of the spleen and liver frequently characterizes the acute phase of illness. Epistaxis and blood-tinged sputum are common complications, and gastrointestinal and central nervous system hemorrhage can occur. Because of this coagulopathy, oneLBRFoutbreak in southern Sudan was thought to be viral hemorrhagic fever. Other complications of variable incidence include iridocyclitis, optic neuritis, meningitis, coma, isolated cranial-nerve palsy, pneumonitis, myocarditis, and rupture of the spleen. Infection during pregnancy can result in spontaneous abortion, stillbirth, or neonatal infection. Life-threatening complications are unusual in otherwise healthy persons given supportive care, especially if the illness is diagnosed and treated early. Without treatment, symptoms intensify over a 2- to 7-day period (average, 5 days inLBRF and 3 days inTBRF), ending in a spontaneous crisis during which spirochetes disappear from the circulation. Treatment with one of the rapidly acting antibiotics, such as erythromycin, a tetracycline, or chloramphenicol, regularly precipitates a Jarisch-Herxheimer reaction within 1 to 4 h. The severity of this reaction is positively correlated with the density of spirochetes in the blood at the time of treatment. In the first phase of the crisis or reaction (the chill phase), rigors and rising fever are accompanied by an increasing metabolic rate, alveolar hyperventilation, high cardiac output, increasing peripheral vascular resistance, and decreased pulmonary arterial pressure. The body temperature commonly rises to³41°C (³105.8°F). This high fever is accompanied often by agitation and confusion and sometimes by delirium. Fever can be partially controlled by the use of a cooling blanket and ice packs and by sponging of the patient with tepid water and alcohol. The chill phase terminates after 10 to 30 min, giving way to a flush phase characterized by a fall in body temperature, drenching sweats, and sometimes (more commonly in LBRF) a potentially dangerous fall in systemic arterial pressure and rise in pulmonary arterial pressure. Although cardiac output is maintained at high levels, the effective circulating blood volume decreases as peripheral vascular resistance falls. Vital signs must be monitored carefully during this period of the reaction, which usually lasts£8 h. Clinical and electrocardiographic evidence of myocarditis and myocardial dysfunction includes a prolonged QTcinterval, a third heart sound (S3), elevated central venous pressure, arterial hypotension, and pulmonary edema. The crisis is followed by a period of exhaustion, sleep, and an uneventful recovery. Not uncommonly, in the first week of convalescence, patients experience 1 or 2 days of mild fever unassociated with detectable spirochetemia. In untreated patients, spirochetemia and symptoms may recur after a period of several days or weeks (average interval to first relapse, 9 days inLBRF and 7 days inTBRF). Only one or two relapses characteristically occur in untreated patients with LBRF, whereas as many as 10 (average, three) can occur in untreated patients with TBRF. In most cases, the illness
becomes shorter and milder and the afebrile intervals longer with each relapse. Because of the great antigenic variation among Borrelia strains, infection confers only partial immunity, and repeated infections of the same individual have been recorded. Diseases that should be considered in the differential diagnosis of relapsing fever or that may complicate relapsing fever include typhus fever, typhoid, nontyphoid salmonellosis, malaria, dengue and other arboviral illnesses, tuberculosis, leptospirosis, and viral hemorrhagic fevers. In the United States, the geographic distribution of Colorado tick fever overlaps that ofTBRF, and the two diseases have similar manifestations early in their courses. LABORATORY FINDINGS AND DIAGNOSIS The diagnosis of relapsing fever is confirmed most easily by the detection of spirochetes in blood, bone marrow aspirates, or cerebrospinal fluid. Motile spirochetes can be seen when fresh blood is examined by dark-field microscopy; and fixed organisms are clearly visible in Wright-, Giemsa-, or acridine orange-stained preparations of thin or dehemoglobinized thick smears of peripheral blood or buffy-coat preparations (Fig. 175-1). Organisms are found in blood taken during periods of fever preceding the crisis; smears from³70% of patients withLBRF and from fewer patients withTBRF are positive. In reference laboratories, relapsing-fever spirochetes are cultured from blood by the inoculation of BSK II medium or by the intraperitoneal inoculation of immature laboratory mice. The detection of agglutinins against Proteus OX-K (Weil-Felix reaction) in convalescent-phase serum supports the diagnosis. Serum antibodies to Borrelia can be detected by enzyme immunoassays, but these tests are unstandardized and subject to insensitivity due to antigenic variations among strains. Serologic cross-reactions occur with other spirochetes, including B. burgdorferi (the agent of Lyme disease) and Treponema pallidum. Other laboratory findings in relapsing fever are generally nonspecific. The leukocyte count is normal or moderately elevated, with an unremarkable cell differential. Serum bilirubin levels are generally only slightly elevated. Thrombocytopenia (mean platelet count, about 50,000/uL) is evident in patients withLBRFduring the acute phase of the illness; platelet counts rebound during early convalescence. Prothrombin and partial thromboplastin times are moderately prolonged during acute illness, as are standardized bleeding times. Fibrinogen concentrations in the blood are normal, and fibrinolysis is mild or absent. Results of the Rumpel-Leede tourniquet test are negative, despite the presence of petechiae. TREATMENT Relapsing-fever borreliae are exquisitely sensitive to antibiotics. Treatment with erythromycin, a tetracycline, chloramphenicol, or penicillin produces rapid clearance of spirochetes and a remission of symptoms (Table 175-2). Although a single dose of erythromycin, a tetracycline, or chloramphenicol is highly effective in the treatment ofLBRF, less is known about the efficacy of single-dose treatment ofTBRF. Empirical treatment of TBRF for 7 days is therefore recommended to reduce the risk of persisting or relapsing borreliosis. For children20% of adults. There is no convincing evidence that M. fermentans causes human disease; although it had been implicated as a possible determinant of HIV-1 disease progression, more recent data do not support such a role. M. genitalium is a fastidious organism that is difficult to cultivate.PCRstudies have identified the organism more successfully. Little is known about the epidemiology of M. genitalium. ASSOCIATION WITH HUMAN DISEASE Nongonococcal Urethritis Chlamydia trachomatis is the organism most firmly implicated in the etiology of nongonococcal urethritis (NGU). There is no doubt that both U. urealyticum and M. genitalium also cause some cases of NGU. The ubiquity of ureaplasmas among men who do not have urethritis and the difficulty of identifying M. genitalium do not allow precise estimation of the proportion of cases of NGU caused by each of these mycoplasmas. U. urealyticum and M. genitalium do, however, appear to cause most of the nonchlamydial cases. Epididymitis and Prostatitis Ureaplasmas may be an occasional cause of epididymitis. M. hominis has not been implicated in this disease. Neither organism has been convincingly associated with prostatitis. Pelvic Inflammatory Disease (PID) (See also Chap. 133) M. hominis and U. urealyticum are both prominent components of the complex microbial flora of bacterial vaginosis. Since bacterial vaginosis is associated with PID, it is difficult to determine whether either organism plays an independent role in this condition. Although M. genitalium is not associated with bacterial vaginosis, preliminary studies have linked it to PID in women who are not infected with either Neisseria gonorrhoeae or C. trachomatis. Disorders of Reproduction Ureaplasmas have been considered as causes of involuntary infertility in both men and women, but there is no convincing evidence for such an association. These organisms have been associated with chorioamnionitis and late abortion. Given the close association of ureaplasmas with bacterial vaginosis, a condition that is strongly associated with chorioamnionitis and late abortion, it is difficult to define an independent role for ureaplasmas in this condition. In infants of very low birthweight, ureaplasmas have been shown to cause pneumonia and chronic lung disease. Extragenital Infections Sexually acquired reactive arthritis and Reiter's disease may be triggered by ureaplasmas, although C. trachomatis is the usual triggering agent. Patients who have hypogammaglobulinemia may develop chronic arthritis due to ureaplasmas and some other mycoplasmal species. M. hominis has been identified in patients with postthoracotomy sternal wound infection and in rare instances of prosthetic heart valve and prosthetic joint infection. DIAGNOSIS There is seldom any reason to examine specimens from the lower genital tract (vagina, male urethra) for mycoplasmas. The ubiquity of the organisms among healthy
individuals makes a positive result uninterpretable. The organisms should be sought only in specimens from normally sterile areas, such as joint fluid with evidence of inflammation and cultures negative for conventional microorganisms. M. hominis can replicate in many routine blood culture media without changing the appearance of the media. M. hominis forms nonhemolytic pinpoint colonies on blood agar; organisms cannot be visualized in gram-stained smears of these colonies. Neither U. urealyticum nor M. genitalium will grow in ordinary microbiologic media. Microbiologic diagnosis of genital mycoplasmal infection requires specially prepared media and is beyond the capability of all but reference and research laboratories. Nucleic acid amplification tests such asPCR have been developed and may become commercially available. TREATMENT Ureaplasmas, M. genitalium, and M. hominis are usually susceptible to tetracyclines (e.g., doxycycline). Tetracycline-resistant ureaplasmas can be treated with erythromycin, while tetracycline-resistant strains of M. hominis respond to treatment with clindamycin. As noted above, a specific microbiologic diagnosis of mycoplasmal infection is seldom made. Appropriate treatment provides antimicrobial coverage for the organisms that cause the particular syndrome. Accordingly,NGU is treated with doxycycline (100 mg orally twice a day for 7 days) or azithromycin (1.0 g as a single oral dose) to provide activity against C. trachomatis, U. urealyticum, and M. genitalium. Recommended regimens for the treatment ofPIDprovide antimicrobial activity against gonococci, chlamydiae, and anaerobes as well as genital mycoplasmas. (Bibliography omitted in Palm version) Back to Table of Contents
179. CHLAMYDIAL INFECTIONS - Walter E. Stamm The genus Chlamydia contains three species that infect humans: Chlamydia psittaci, C. trachomatis, and C. pneumoniae (formerly the TWAR agent). C. psittaci is widely distributed in nature, producing genital, conjunctival, intestinal, or respiratory infections in many mammalian and avian species. Genital infections with C. psittaci have been well characterized in several species and cause abortion and infertility. Although mammalian strains of C. psittaci are not known to infect humans, avian strains occasionally do so, causing pneumonia and the systemic illness known as psittacosis. C. pneumoniae is a fastidious chlamydial species that appears to be a common cause of upper respiratory tract infection and pneumonia, primarily in children and young adults, and is a cause of recurrent respiratory infections in older adults. Studies have also linked C. pneumoniae infection to atherosclerotic cardiovascular disease and perhaps to asthma and sarcoidosis. No animal reservoir has been identified for C. pneumoniae; it appears to be an exclusively human pathogen spread via the respiratory route through close personal contact. To date, all strains of C. pneumoniae studied have been serologically homologous. C. trachomatis is also an exclusively human pathogen and was identified as the cause of trachoma in the 1940s. Since then, C. trachomatis has been recognized as a major cause of sexually transmitted and perinatal infection. Chlamydiae are obligate intracellular bacteria that are classified in their own order (Chlamydiales). They possess both DNA and RNA, have a cell wall and ribosomes similar to those of gram-negative bacteria, and are inhibited by antibiotics such as tetracycline. A unique feature of all chlamydiae is their complex reproductive cycle. Two forms of the microorganism -- the extracellular elementary body and the intracellular reticulate body -- participate in this cycle. The elementary body is adapted for extracellular survival and is the infective form transmitted from one person to another. Elementary bodies attach to susceptible target cells (usually columnar or transitional epithelial cells) and enter the cells inside a phagosome. Within 8 h of cell entry, the elementary bodies reorganize into reticulate bodies, which are adapted to intracellular survival and multiplication. They undergo binary fission, eventually producing numerous replicates contained within the intracellular membrane-bound "inclusion body," which occupies much of the infected host cell. Chlamydial inclusions resist lysosomal fusion until late in the developmental cycle. After 24 h, the reticulate bodies condense and form elementary bodies still contained within the inclusion. The inclusion then ruptures, releasing elementary bodies from the cell to initiate infection of adjacent cells or transmission to another person. Studies with monoclonal antibodies to and nucleotide sequencing of the major outer-membrane protein have delineated at least 20 serotypes of C. trachomatis. According to the classification of Wang and Grayston, strains associated with trachoma have generally been those of the A, B, Ba, and C serovars, while serovars D through K have largely been associated with sexually transmitted and perinatally acquired infections. Serovars L1, L2, and L3produce lymphogranuloma venereum (LGV) and hemorrhagic proctocolitis. The LGV strains demonstrate unique biologic behavior in that
they are more invasive than the other serovars, produce disease in lymphatic tissue, grow readily in cell culture systems and macrophages, and are fatal when inoculated intracerebrally into mice and monkeys. Non-LGV strains of C. trachomatis characteristically produce infections involving the superficial columnar epithelium of the eye, genitalia, and respiratory tract. C. trachomatis has been reported as an infrequent cause of endocarditis, peritonitis, pleuritis, and possibly periappendicitis and may occasionally cause respiratory infections in older children and adults. Some immunosuppressed patients with pneumonia have had either serologic or cultural evidence of C. trachomatis infection, but more data are necessary to define a pathogenic role for Chlamydia in these patients. SEXUALLY TRANSMITTED AND PERINATAL INFECTIONS DUE TO C. TRACHOMATIS SPECTRUM OF C. TRACHOMATIS GENITAL INFECTIONS Genital infections caused by C. trachomatis represent the most common bacterial sexually transmitted diseases (STDs) in the United States. An estimated 4 million cases occur each year. In adults, the clinical spectrum of sexually transmitted C. trachomatis infections parallels that of gonococcal infection. Both infections have been associated with urethritis, proctitis, and conjunctivitis in both sexes; with epididymitis in men; and with mucopurulent cervicitis (MPC) (Fig. 179-CD1), acute salpingitis, bartholinitis, and the Fitz-Hugh-Curtis syndrome (perihepatitis) in women. Moreover, both types of infection can be associated with septic arthritis. In general, however, chlamydial infections produce fewer symptoms and signs than corresponding gonococcal infections at the same anatomic site; in fact, chlamydial infections are often totally asymptomatic. Increasing evidence suggests that many chlamydial infections of the genital tract, especially in women, persist for months without producing symptoms. Simultaneous infection with C. trachomatis often occurs in women with cervical gonococcal infection and in heterosexual men with gonococcal urethritis. EPIDEMIOLOGY Infections due to C. trachomatis are now reportable in the United States, and national incidence data show steadily rising numbers of reported infections, undoubtedly reflecting both increased testing and increased reporting. Most testing has focused upon women to date, and thus the reported incidence is severalfold greater in women than in men; this difference likely represents a surveillance artifact. The age of peak incidence of genital C. trachomatis infections, as of other sexually transmitted infections, is the late teens and early twenties. The prevalence of chlamydial urethral infection among young men is at least 3 to 5% for those seen in general medical settings or in urban high schools,>10% for asymptomatic soldiers undergoing routine physical examination, and 15 to 20% for heterosexual men seen inSTDclinics. In areas where chlamydial control programs have been implemented, prevalence may be markedly reduced. In short, prevalence varies widely with the population group studied and with the geographic locale. The ratio of chlamydial to gonococcal urethritis is highest for heterosexual men and for those of high socioeconomic status and is lowest
for homosexual men and indigent populations. The prevalence of cervical infection among women is approximately 5% for asymptomatic college students and prenatal patients in the United States, >10% for women seen in family planning clinics, and >20% for women seen inSTDclinics. As in men, prevalence varies substantially by geographic locale. However, substantial prevalences (~8%) of asymptomatic chlamydial infection were recently demonstrated in young female military recruits from all parts of the United States. In this country, the prevalence of C. trachomatis in the cervix of pregnant women is 5 to 10 times higher than that of Neisseria gonorrhoeae. The prevalence of genital infection with either agent is highest among individuals who are between the ages of 18 and 24, single, and non-Caucasian (e.g., black or hispanic). Recurrent chlamydial infections occur frequently in these same risk groups, often acquired from untreated sexual partners. Oral contraceptive pill use and the presence of cervical ectopy also confer an increased risk of chlamydial infection. The proportion of infections that are asymptomatic appears to be higher for C. trachomatis than for N. gonorrhoeae, and symptomatic C. trachomatis infections are clinically less severe. Mild or asymptomatic chlamydial infections of the fallopian tubes nonetheless cause ongoing tubal damage and infertility. Furthermore, because the total number of C. trachomatis infections exceeds the total number of N. gonorrhoeae infections in industrialized countries, the total morbidity caused by C. trachomatis genital infections in these countries equals or exceeds that caused by N. gonorrhoeae. The prevalence of C. trachomatis is higher than that of N. gonorrhoeae in industrialized countries, in part because measures such as treatment of sex partners and routine cultures for case detection in asymptomatic individuals have been applied much more effectively to the control of gonorrhea than to the control of C. trachomatis infection. PATHOGENESIS C. trachomatis preferentially infects the columnar epithelium of the eye and the respiratory and genital tracts. The infection induces an immune response but often persists for months or years in the absence of antimicrobial therapy. Serious sequelae often occur in association with repeated or persistent infections. The precise mechanism through which repeated infection elicits an inflammatory response that leads to tubal scarring and damage in the female upper genital tract is not yet clear. One antigen, the chlamydial 60-kDa heat-shock protein, may be involved in inducing the pathologic immune response or may elicit antibodies that cross-react with human heat-shock proteins. The recent sequencing of the chlamydial genome may soon offer further insights into the pathogenic mechanisms of C. trachomatis. CLINICAL MANIFESTATIONS Nongonococcal and Postgonococcal Urethritis Nongonococcal urethritis (NGU) is a diagnosis of exclusion that is applied to men with symptoms and/or signs of urethritis who do not have gonorrhea. Postgonococcal urethritis (PGU) refers to nongonococcal urethritis developing in men 2 to 3 weeks after treatment of gonococcal urethritis with single doses of agents such as amoxicillin or cephalosporins that lack sufficient activity against chlamydiae. Since current treatment regimens for gonorrhea also include tetracycline, doxycycline, or azithromycin for possible concomitant chlamydial infection,
both the incidence of PGU and the causative role of chlamydiae in this syndrome have declined. C. trachomatis causes 20 to 40% of cases of NGU in heterosexual men but is less commonly isolated from homosexual men with this syndrome. The cause of most of the remaining cases is uncertain; considerable evidence suggests that Ureaplasma urealyticum causes many cases of NGU, while Trichomonas vaginalis and herpes simplex virus (HSV) cause some cases. NGUis diagnosed by documentation of a leukocytic urethral exudate and by exclusion of gonorrhea by Gram's staining or culture. C. trachomatis urethritis is generally less severe than gonococcal urethritis, although in an individual patient these two forms of urethritis cannot be reliably differentiated solely on clinical grounds. Symptoms include urethral discharge (often whitish and mucoid rather than frankly purulent), dysuria, and urethral itching. Physical examination may reveal meatal erythema and tenderness and a urethral exudate that is often demonstrable only by stripping of the urethra. At least one-third of males with C. trachomatis urethral infection have no demonstrable signs or symptoms of urethritis. Use of nucleic acid amplification assays on first-void urine specimens to diagnose chlamydial infections in men has facilitated more broadly based testing for asymptomatic infection in males. As a result, asymptomatic chlamydial urethritis has been demonstrated in 5 to 10% of sexually active adolescent males screened in school-based clinics or community centers. Such patients generally have first-glass pyuria (³15 leukocytes per 400´ microscopic field in the sediment of first-void urine), a positive leukocyte esterase test, or an increased number of leukocytes on Gram-stained smear prepared from a urogenital swab inserted 1 to 2 cm into the anterior urethra. For the enumeration of leukocytes, the smear is first scanned at low power to identify areas of the slide containing the highest concentration of leukocytes. These areas are then examined under oil immersion (1000´). An average of four or more leukocytes in at least three of five 1000´ (oil-immersion) fields is indicative of urethritis and correlates with the recovery of C. trachomatis. To differentiate between true urethritis and functional symptoms among symptomatic patients or to make a presumptive diagnosis of C. trachomatis infection in a "high-risk" but asymptomatic man (e.g., male patients inSTDclinics, sex partners of women with nongonococcal salpingitis orMPC, fathers of children with inclusion conjunctivitis), the examination of an endourethral specimen for increased leukocytes is useful if specific diagnostic tests for chlamydiae are not available. Alternatively, noninvasive screening for urethritis can be accomplished by testing of a first-void urine sample for pyuria, either by microscopy or by the leukocyte esterase test. Urine can also be directly tested for chlamydiae or gonococci by DNA amplification methods, as described below. Epididymitis C. trachomatis is the foremost cause of epididymitis in sexually active heterosexual men under 35 years of age, accounting for about 70% of cases. N. gonorrhoeae causes most of the remaining cases, and some men have simultaneous infections with both pathogens, usually accompanied by asymptomatic urethritis as defined above. In homosexual men, sexually transmitted coliform infection acquired via rectal intercourse may cause epididymitis. Coliform bacteria and Pseudomonas aeruginosa, usually in association with preceding urologic instrumentation or surgery, are the most common causes of epididymitis in men over 35. Men with epididymitis typically present with unilateral scrotal pain, fever, and epididymal tenderness or swelling on examination. The illness may be mild enough to treat on an outpatient basis
with oral antibiotics or severe enough to require hospitalization and parenteral therapy. Testicular torsion should be excluded promptly by radionuclide scan, Doppler flow study, or surgical exploration in a teenager or young adult who presents with acute unilateral testicular pain without urethritis. The possibility of testicular tumor or chronic infection (e.g., tuberculosis) should be excluded when a patient with unilateral intrascrotal pain and swelling does not respond to appropriate antimicrobial therapy. Reiter's Syndrome Reiter's syndrome consists of conjunctivitis, urethritis (or cervicitis in females), arthritis, and characteristic mucocutaneous lesions (Chap. 315). C. trachomatis has been recovered from the urethra of up to 70% of men with untreated nondiarrheal Reiter's syndrome and associated urethritis. In the absence of overt urethritis, it is important to exclude subclinical urethritis in the men in whom this diagnosis is suspected. The pathogenesis of Reiter's syndrome remains obscure. However, since more than 80% of affected patients have the HLA-B27 phenotype and since other mucosal infections (with Salmonella, Shigella, or Campylobacter, for example) produce an identical syndrome, chlamydial infection is thought to initiate an aberrant and hyperactive immune response that produces inflammation at the involved target organs in these genetically predisposed individuals. Evidence of exaggerated cell-mediated and humoral immune responses to chlamydial antigens in Reiter's syndrome supports this hypothesis. The presumptive demonstration of chlamydial elementary bodies and chlamydial DNA in the joint fluid and synovial tissue of patients with Reiter's syndrome suggests that chlamydiae may actually spread from genital to joint tissues in these patients, perhaps in macrophages. Proctitis C. trachomatis strains of either the genital immunotypes D through K or theLGVimmunotypes cause proctitis in homosexual men who practice receptive anorectal intercourse. In the United States, the vast majority of cases are due to immunotypes D through K and present either as asymptomatic infection or as mild proctitis not unlike gonococcal proctitis. These infections may develop in heterosexual women as well. Patients present with mild rectal pain, mucous discharge, tenesmus, and (occasionally) bleeding. Nearly all have neutrophils in their rectal Gram's stain. Anoscopy in these non-LGV cases of chlamydial proctitis reveals mild, patchy mucosal friability and mucopurulent discharge, and the disease process is limited to the distal rectum. LGV strains produce more severe ulcerative proctitis or proctocolitis that can be confused clinically withHSVproctitis (severe rectal pain, bleeding, discharge, and tenesmus) and that histologically resembles Crohn's disease in that giant cell formation and granulomas can be seen (Chap. 287). In the United States, these cases occur almost exclusively in homosexual men. Mucopurulent Cervicitis Although many women with C. trachomatis infection of the cervix have no symptoms or signs, a careful speculum examination reveals evidence ofMPC in 30 to 50% of cases. As is discussed more fully inChap. 133, MPC is associated with yellow mucopurulent discharge from the endocervical columnar epithelium and with³20 neutrophils per 1000´ microscopic field within strands of cervical mucus on a thinly smeared, Gram-stained preparation of endocervical exudate. Other characteristic findings include edema of the zone of cervical ectopy and a propensity of the mucosa to bleed on minor trauma -- e.g., when specimens are collected with a
swab. A Pap smear shows increased numbers of neutrophils as well as a characteristic pattern of mononuclear inflammatory cells, including plasma cells, transformed lymphocytes, and histiocytes. Cervical biopsy shows a predominantly mononuclear cell infiltrate of the subepithelial stroma, often with follicular cervicitis. Pelvic Inflammatory Disease (PID) (See alsoChap. 133) C. trachomatis plays an important causative role in salpingitis. Infection with C. trachomatis has been demonstrated in laparoscopically verified salpingitis, the organism has been recovered from the fallopian tubes in the absence of other pathogens, and serologic evidence of recent C. trachomatis infection has been found in women with PID. In the United States, C. trachomatis has been identified in the fallopian tubes or endometrium of up to 50% of women with PID, and its role as an important etiologic agent in this syndrome is well accepted. PIDoccurs via ascending intraluminal spread of C. trachomatis from the lower genital tract.MPC is thus followed by endometritis, endosalpingitis, and finally pelvic peritonitis. Evidence of MPC is usually found in women with laparoscopically verified salpingitis. Similarly, endometritis, demonstrated by endometrial biopsy showing plasma cell infiltration of the endometrial epithelium, is documented in most women with laparoscopically verified chlamydial (or gonococcal) salpingitis. Chlamydial endometritis can also occur in the absence of clinical evidence of salpingitis: approximately 40 to 50% of women with MPC have plasma cell endometritis. Histologic evidence of endometritis has been correlated with an "endometritis syndrome" consisting of vaginal bleeding, lower abdominal pain, and uterine tenderness in the absence of adnexal tenderness. It is not known what proportion of women who have chlamydial endometritis without adnexal tenderness also have salpingitis. However, chlamydial salpingitis produces milder symptoms than does gonococcal salpingitis and may be associated with less marked adnexal tenderness. Mild adnexal or uterine tenderness in sexually active women with cervicitis suggests PID. Infertility associated with fallopian-tube scarring has been strongly linked to antecedent C. trachomatis infection in serologic studies. Since many infertile women with tubal scarring and antichlamydial antibody have no history ofPID, it appears that subclinical tubal infection ("silent salpingitis") may produce scarring. Studies in animals and humans with salpingitis and tubal scarring suggest the continuing presence of persistent, slowly replicating chlamydial infection in tubal tissue. Ectopic pregnancy, which occurs in more than 70,000 women in the United States annually, is also thought to be related to Chlamydia-induced tubal scarring in many cases. While the pathogenesis of Chlamydia-induced tubal scarring remains poorly understood, antibodies to the chlamydial 60-kDa heat-shock protein have been correlated with tubal infertility, ectopic pregnancy, and Fitz-Hugh-Curtis syndrome (see below). Thus this antigen may initiate an immune-mediated process that ultimately damages the fallopian tube. Host genetic susceptibility, as defined by HLA type, may also play an important role. Perihepatitis, or the Fitz-Hugh-Curtis syndrome, was originally described as a complication of gonococcalPID. However, cultural and/or serologic evidence of C. trachomatis infection is found in three-quarters of women with this syndrome. C. trachomatis has also been cultured from exudate on the hepatic capsule in
laparoscopically verified cases. This syndrome should be suspected whenever a young, sexually active woman presents with an illness resembling cholecystitis (fever and right-upper-quadrant pain of subacute or acute onset). Symptoms and signs of salpingitis may be minimal. High titers of antibodies to C. trachomatis are generally present. Urethral Syndrome in Women In the absence of infection with uropathogens such as coliforms or Staphylococcus saprophyticus, C. trachomatis is the pathogen most commonly isolated from college women with dysuria, frequency, and pyuria (Chap. 280). Chlamydia can also be isolated from the urethra of women without symptoms of urethritis, and up to 25% of femaleSTDclinic patients with chlamydial urogenital infection have cultures positive for C. trachomatis from the urethra only. C. trachomatis Infection in Pregnancy C. trachomatis in pregnancy has been associated in some studies (but not in others) with premature delivery and with postpartum endometritis. Whether these complications are in part attributable to C. trachomatis is not clear. PERINATAL INFECTIONS: INCLUSION CONJUNCTIVITIS AND PNEUMONIA Epidemiology Studies in the United States have demonstrated that 5 to 25% of pregnant women have C. trachomatis infections of the cervix. In these studies, approximately one-half to two-thirds of children exposed during birth have acquired C. trachomatis infection. Roughly half of the infected infants (or 25% of the group exposed) have developed clinical evidence of inclusion conjunctivitis. In addition to infecting the eye, C. trachomatis has been isolated frequently and persistently from the nasopharynx, rectum, and vagina of such infants, occasionally for periods exceeding 1 year in the absence of treatment. Pneumonia develops in about 10% of children infected perinatally, and otitis media may in some cases result from perinatally acquired chlamydial infection. Inclusion Conjunctivitis of the Newborn (Neonatal Chlamydial Conjunctivitis) Neonatal chlamydial conjunctivitis has an acute onset 5 to 14 days after birth and often produces a profuse mucopurulent discharge. However, it is impossible to differentiate chlamydial conjunctivitis from other forms of neonatal conjunctivitis (such as that due to N. gonorrhoeae, Haemophilus influenzae, Streptococcus pneumoniae, orHSV) on clinical grounds; instead, laboratory diagnosis is required. Inclusions within epithelial cells are often detected in Giemsa-stained conjunctival smears, but these smears are considerably less sensitive than cultures, antigen detection tests, or nucleic acid hybridization tests for chlamydiae. Gram-stained smears may show gonococci or occasional small gram-negative coccobacilli in Haemophilus conjunctivitis, but smears should be accompanied by cultures for these agents. Infant Pneumonia C. trachomatis causes a distinctive pneumonia syndrome in infants. Recent epidemiologic studies have linked chlamydial pulmonary infection in infants with increased occurrence of subacute lung disease (bronchitis, asthma, wheezing) in later childhood. LYMPHOGRANULOMA VENEREUM
Definition LGV is a sexually transmitted infection caused by C. trachomatis strains of the L1, L2, and L3serovars. In the United States, most cases are caused by L2organisms. Acute LGV in heterosexual men is characterized by a transient primary genital lesion followed by multilocular suppurative regional lymphadenopathy. Women, homosexual men, and -- in occasional instances -- heterosexual men may develop hemorrhagic proctitis with regional lymphadenitis. Acute LGV is almost always associated with systemic symptoms such as fever and leukocytosis but is rarely associated with systemic complications such as meningoencephalitis. After a latent period of years, late complications include genital elephantiasis due to lymphatic involvement; strictures; and fistulas of the penis, urethra, and rectum. Epidemiology LGV is usually sexually transmitted, but occasional transmission by nonsexual personal contact, fomites, or laboratory accidents has been documented. Laboratory work involving the creation of aerosols of LGV organisms (e.g., sonication, homogenization) must be conducted only with appropriate measures for biologic containment. The peak incidence ofLGVcorresponds to the age of greatest sexual activity: the second and third decades of life. The worldwide incidence of LGV is falling, but the disease is still endemic and a major cause of morbidity in Asia, Africa, South America, and parts of the Caribbean. In the Bahamas, an apparent outbreak of LGV has been described in association with a concurrent increase in heterosexual infection with HIV. However, only 186 cases were reported in the United States in 1995, for a rate of 0.1 case per 100,000 population. The frequency of infection following exposure is believed to be much lower than that for gonorrhea and syphilis. Early manifestations are recognized far more often in men than in women, who usually present with late complications. In the United States, where the reported male-to-female ratio of cases is 3.4:1, most cases have involved homosexually active men and persons returning from abroad (travelers, sailors, and military personnel). The main reservoir of infection, although it has not been directly demonstrated, is presumed to be asymptomatically infected individuals. Clinical Manifestations In heterosexuals, a primary genital lesion develops from 3 days to 3 weeks after exposure. It is a small, painless vesicle or nonindurated ulcer or papule located on the penis in men and on the labia, posterior vagina, or fourchette in women. The primary lesion is noticed by fewer than one-third of men withLGV and only rarely by women. It heals in a few days without scarring and, even when noticed, is usually recognized as LGV only in retrospect. LGV strains of C. trachomatis have occasionally been recovered from genital ulcers and from the urethra of men and the endocervix of women who present with inguinal adenopathy; these areas may be the primary site of infection in some cases. In women and homosexual men, primary anal or rectal infection develops after receptive anorectal intercourse. In women, rectal infection withLGV (or non-LGV) strains of C. trachomatis presumably can also arise by the contiguous spread of infected secretions along the perineum (as in rectal gonococcal infections in women) or perhaps by spread to the rectum via the pelvic lymphatics.
From the site of the primary urethral, genital, anal, or rectal infection, the organism spreads via the regional lymphatics. Penile, vulvar, or anal infection can lead to inguinal and femoral lymphadenitis. Rectal infection produces hypogastric and deep iliac lymphadenitis. Upper vaginal or cervical infection results in enlargement of the obturator and iliac nodes. The most common presenting picture in heterosexual men is the inguinal syndrome, which is characterized by painful inguinal lymphadenopathy beginning 2 to 6 weeks after presumed exposure; in rare instances, the onset comes after a few months. The inguinal adenopathy is unilateral in two-thirds of cases, and palpable enlargement of the iliac and femoral nodes is often evident on the same side as the enlarged inguinal nodes (Fig. 132-CD2). The nodes are initially discrete, but progressive periadenitis results in a matted mass of nodes that becomes fluctuant and suppurative. The overlying skin becomes fixed, inflamed, and thin and finally develops multiple draining fistulas. Extensive enlargement of chains of inguinal nodes above and below the inguinal ligament ("the sign of the groove") is not specific and, although not uncommon, is documented in only a minority of cases. On histologic examination, infected nodes are initially found to have characteristic small stellate abscesses surrounded by histiocytes. These abscesses coalesce to form large, necrotic, suppurative foci. Spontaneous healing usually takes place after several months; inguinal scars or granulomatous masses of various sizes persist for life. Massive pelvic lymphadenopathy in women or homosexual men may lead to exploratory laparotomy. As cultures and serologic tests for C. trachomatis are being used more often, increasing numbers of cases ofLGVproctitis are being recognized in homosexual men. Such patients present with anorectal pain and mucopurulent, bloody rectal discharge. Although these patients may complain of diarrhea, they are often referring not to diarrhea but rather to frequent, painful, unsuccessful attempts at defecation (tenesmus). Sigmoidoscopy reveals ulcerative proctitis or proctocolitis, with purulent exudate and mucosal bleeding. The histopathologic findings in the rectal mucosa include granulomas with giant cells, crypt abscesses, and extensive inflammation. These clinical, sigmoidoscopic, and histopathologic findings may closely resemble those of Crohn's disease of the rectum. Constitutional symptoms are common during the stage of regional lymphadenopathy and, in cases of proctitis, may include fever, chills, headache, meningismus, anorexia, myalgias, and arthralgias. These findings in the presence of lymphadenopathy are sometimes mistakenly interpreted as representing malignant lymphoma. Other systemic complications are infrequent but include arthritis with sterile effusion, aseptic meningitis, meningoencephalitis, conjunctivitis, hepatitis, and erythema nodosum. Chlamydiae have been recovered from the cerebrospinal fluid and in one case were isolated from the blood of a patient with severe constitutional symptoms -- a result indicating the dissemination of infection. Laboratory-acquired infections suspected of being due to the inhalation of aerosols have been associated with mediastinal lymphadenitis, pneumonitis, and pleural effusion. Complications of untreated anorectal infection include perirectal abscess; fistula in ano; and rectovaginal, rectovesical, and ischiorectal fistulas. Secondary bacterial infection
probably contributes to these complications. Rectal stricture is a late complication of anorectal infection and usually develops 2 to 6 cm from the anal orifice -- i.e., at a site within reach on digital rectal examination. Proximal extension of the stricture for several centimeters may lead to a mistaken clinical and radiographic diagnosis of carcinoma. A small percentage of cases ofLGV in men present as chronic progressive infiltrative, ulcerative, or fistular lesions of the penis, urethra, or scrotum. Associated lymphatic obstruction may produce elephantiasis. When urethral stricture occurs, it usually involves the posterior urethra and causes incontinence or difficulty with urination. APPROACH TO THE DIAGNOSIS AND TREATMENT OF C. TRACHOMATIS GENITAL INFECTIONS Four types of laboratory procedure are available to confirm C. trachomatis infection: direct microscopic examination of tissue scrapings for typical intracytoplasmic inclusions or elementary bodies; isolation of the organism in cell culture; detection of chlamydial antigens or nucleic acid by immunologic or hybridization methods; and detection of antibody in serum or in local secretions. Except in conjunctivitis, direct microscopic examination of Giemsa-stained cell scrapings for typical inclusions has an unacceptably low degree of sensitivity, and false-positive interpretations by inexperienced observers are common. Even for conjunctivitis, this approach has been replaced by direct fluorescent antibody staining of conjunctival smears to identify chlamydial elementary bodies with specific monoclonal antibodies (see below). Cell culture techniques for isolation of C. trachomatis are available in most large medical centers but not in other clinical settings. In addition to limited availability, other disadvantages of cell culture include its low and variable level of sensitivity (60 to 80%), its requirement for rigorous transport conditions, and its high cost and technically demanding nature. Therefore, nonculture alternatives involving antigen detection or nucleic acid hybridization have been developed. In the direct immunofluorescent antibody (DFA) slide test, potentially infected genital or ocular secretions are smeared onto a slide, fixed, and stained with fluorescein-conjugated monoclonal antibody specific for chlamydial antigens. The observation of fluorescing elementary bodies confirms the diagnosis. Compared with culture, this test is 70 to 85% sensitive, and it is quite specific when used for confirmation of urethral, cervical, or ocular infection in high-risk patients with suspected C. trachomatis infection. The sensitivity and specificity of the test depend directly upon the skill of the microscopist. The apparently lower sensitivity of the test in low-risk populations, along with its relatively labor-intensive nature, limits its value as a screening tool. Enzyme-linked immunosorbent assay (ELISA) techniques for the detection of chlamydial antigens provide another alternative to culture. The reported sensitivity and specificity of these tests for genital infections (as compared with culture) have been 60 to 80% and 97 to 99%, respectively, in high-risk populations. Sensitivities have generally been higher in cervical infection and lower in urethritis among males. Like theDFAslide test, the ELISA is less sensitive and less specific in low-prevalence populations and largely asymptomatic patients. ELISAs are better suited to screening
than is DFA because large numbers of specimens can easily be processed. Assays with nucleic acid probes have also been developed for chlamydial diagnosis. One such test uses DNA-RNA hybridization and appears to be approximately equal to the bestELISAs in terms of sensitivity and specificity. Nucleic acid probes have also been developed for use in amplification assays such as ligase chain reaction and polymerase chain reaction (PCR). These tests are now the most sensitive chlamydial diagnostic methods available, being the first nonculture assays actually to surpass culture itself in sensitivity. In addition, the ability of these tests to detect chlamydial genes in urine with a high degree of sensitivity and specificity allows their use with urine specimens rather than with conventional urethral and cervical swabs for the first time. The use of urine specimens is particularly appealing for public-health chlamydial screening programs because of the ease of sample collection, even in community-based settings. Serologic tests are of limited usefulness in the diagnosis of chlamydial oculogenital infections. The complement fixation test with heat-stable, genus-specific antigen has been used with some success to diagnoseLGV but is insensitive in infections due to non-LGV strains of C. trachomatis. The microimmunofluorescence (micro-IF) test with C. trachomatis antigens is more sensitive but is generally available only in research laboratories. The test measures antibodies by serovar specificity and by immunoglobulin class (IgM, IgG, IgA, secretory IgA) in both serum and local secretions. Serologic diagnosis by themicro-IFtest may be useful in infant pneumonia (in which high-titer IgM antibody and/or fourfold rises in titer are often demonstrated), in chlamydial salpingitis (especially Fitz-Hugh-Curtis syndrome), and in LGV. In all of these more invasive syndromes, high antibody levels are present. Table 179-1 summarizes the diagnostic tests of choice for patients with suspected C. trachomatis infection. With few exceptions, the most suitable method for diagnosis is demonstration of the agent by either cell culture or one of the newer nonculture techniques. Selection of the most appropriate of these tests often depends upon local availability and expertise. However, it is clear that, in most settings and for most purposes, sensitivity and specificity will be greatest with nucleic acid amplification techniques. For patients to whom medicolegal considerations may apply (victims of sexual or child abuse), cultures or nucleic acid amplification methods should always be used. Since C. trachomatis is an intracellular pathogen, adequate specimens for chlamydial diagnostic testing must include epithelial cells. Cultures or nonculture tests of pus are less often positive. In urethritis, a thin-shafted urogenital swab should be inserted at least 2 cm into the urethra to obtain an appropriate specimen. Although cultures of urine for chlamydiae are less sensitive than urethral cultures, studies suggest that nucleic acid amplification testing of a first-void urine specimen from men is a more sensitive and less painful diagnostic alternative to the more invasive urethral swab-based tests, culture, or antigen detection tests. The first 30 mL of voided urine should be collected for testing. When a cervical sample is collected, the external os should first be cleaned of debris and purulent material; a plastic-shafted swab should then be inserted into the cervix, rotated slowly several times, and withdrawn. For the diagnosis of urogenital (cervical or urethral) infections in women, testing of a first-void urine specimen by nucleic acid amplification methods is at least as sensitive as testing of a cervical swab. When conjunctival specimens are sought, the epithelium should be
swabbed to remove cells rather than just purulent material. All specimens for chlamydial culture should be placed immediately into transport medium and then either refrigerated (if they will reach the laboratory within 12 to 18 h) or frozen at-70°C (if longer storage is anticipated). A major advantage of the nonculture diagnostic techniques is their less rigid transport requirements; neither refrigeration nor rapid transport is needed. From a public health viewpoint, the most effective use of chlamydial diagnostic testing has not been unequivocally established and varies with the clinical population, local resources, and laboratory expertise. Since chlamydial diagnostic testing has become more widely available and is now more sensitive and specific than in the past, its use for specific diagnosis in patients with suspected chlamydial syndromes (such asMPC,NGU, andPID) and their partners should be promoted. High priority should be given to the screening of asymptomatic high-risk women who would not otherwise receive treatment for presumptive chlamydial infection, especially those seen in high-risk settings (e.g.,STDclinics or abortion clinics) and those with a high-risk profile (e.g., sexually active and £21 years of age, new sex partner within the preceding 2 months, or more than one current sex partner). Similar screening programs should be used to detect and treat asymptomatic urethritis in high-risk adolescent males. Where implemented, screening programs of this type have been associated with reductions in the prevalence of chlamydial infection and of its complications, such as PID. ANTIMICROBIAL SUSCEPTIBILITY In laboratory tests that evaluate the growth of chlamydiae in cell cultures, the tetracyclines, erythromycin, rifampin, certain fluoroquinolones (especially ofloxacin), and the macrolide azithromycin are all highly active against these organisms. Sulfonamides and clindamycin are also active against C. trachomatis, but to a lesser degree. Penicillin and ampicillin suppress chlamydial multiplication but do not eradicate the organism in vitro. The cephalosporins appear to be relatively ineffective against C. trachomatis. Streptomycin, gentamicin, neomycin, kanamycin, vancomycin, ristocetin, spectinomycin, and nystatin are not effective at concentrations inhibitory for most bacteria and fungi. There does not appear to be much strain-to-strain variation in susceptibility to antibiotics, and no clinically significant antimicrobial resistance in chlamydiae has been described. Thus antimicrobial susceptibility testing is not needed in the routine management of patients with chlamydial infection, even recurrent infection. TREATMENT Until the introduction of azithromycin, chlamydial infections could not be eradicated by single-dose or short-term antimicrobial regimens. In most uncomplicated infections in adults, 7 days of treatment with doxycycline or tetracycline have to be given for genital infections, but a 2-week course of therapy is recommended for complicated chlamydial infections (e.g.,PID, epididymitis) and at least a 3-week course forLGV. Failure of treatment of genital infections with a tetracycline usually indicates poor compliance or reinfection rather than the involvement of a drug-resistant strain. Therapy for C. trachomatis urethritis is more efficacious than therapy for nonchlamydialNGU. C. trachomatis is eradicated from the urethra in nearly all cases by treatment with tetracycline hydrochloride (500 mg qid for 7 days) or doxycycline (100 mg
by mouth bid for 7 days). Eradication of C. trachomatis from the cervix by tetracycline, doxycycline, and erythromycin, with doses and durations similar to those specified above for urethritis, has been demonstrated. Erythromycin base (500 mg qid for 10 to 14 days) is the regimen of choice for pregnant women with C. trachomatis infection. Amoxicillin (500 mg tid for 10 days) has also been used successfully in pregnant women. Tetracycline hydrochloride (500 mg qid) or doxycycline (100 mg bid) for 14 days produces clinical and microbiologic cure of epididymitis andPIDassociated with C. trachomatis infection, but in this situation a tetracycline should always be used together with a drug that is highly effective against gonorrhea. Azithromycin is highly active against C. trachomatis, exhibits prolonged bioavailability, is concentrated intracellularly, and has offered the prospect of single-dose therapy for chlamydial infection for the first time. In comparative trials, a 1-g single dose of azithromycin has been as effective as 7 days of doxycycline for uncomplicated chlamydial infection. Azithromycin causes fewer adverse gastrointestinal reactions than do older macrolides such as erythromycin. The single-dose regimen of azithromycin has great appeal for the treatment of patients with uncomplicated chlamydial infection (especially those without symptoms and those with a likelihood of poor compliance) and of sexual partners of infected patients. These advantages must be weighed against the considerably greater cost of azithromycin than of doxycycline. Whenever possible, the single 1-g dose should be given as directly observed therapy. Although not approved by the U.S. Food and Drug Administration, the 1-g single-dose regimen of azithromycin appears to be safe and effective in the treatment of pregnant women. Of the newer fluoroquinolones, ofloxacin (300 mg by mouth bid for 7 days) has been shown to be as effective as doxycycline for the treatment of chlamydial infection and appears to be safe and well tolerated. It cannot be used in pregnancy. Treatment of Sex Partners The continued high prevalence of chlamydial infections in most parts of the United States is due primarily to the failure to diagnose -- and therefore treat -- patients with symptomatic or asymptomatic infection and their sex partners. C. trachomatis urethral or cervical infection has been well documented in a high proportion of the sex partners of patients withNGU, epididymitis, Reiter's syndrome, salpingitis, or endocervicitis. If possible, confirmatory laboratory tests for Chlamydia should be undertaken in these individuals, but even those without evidence of clinical disease who have recently been exposed to proven or possible chlamydial infection (e.g., NGU) should be offered therapy. Treatment of Neonates and Infants In neonates with conjunctivitis or infants with pneumonia, erythromycin ethylsuccinate or estolate can be given orally in a dose of 50 mg/kg per day, preferably in four divided doses, for 2 weeks. Careful attention must be given to compliance with therapy -- a frequent problem. Relapses of eye infection are common following treatment with topical erythromycin or tetracycline ophthalmic ointment and may also occur after oral erythromycin therapy. Thus follow-up cultures should be performed after treatment. Both parents should be examined for C. trachomatis infection and, if diagnostic testing is not readily available, should be treated with doxycycline or azithromycin.
PREVENTION Efforts to develop a vaccine for chlamydial infection have not yet been successful. Early diagnosis and treatment shorten the duration of infectiousness of the carrier and therefore constitute primary prevention of chlamydial infection. By the early 1990s, one of the 10 regions of the United States (Region X, the Pacific Northwest) had formally undertaken a chlamydial control program involving widespread screening of women attending family planning clinics. Approximately 500,000 tests per year were conducted at 150 such clinics throughout the region in women meeting the criteria for high risk. Within 5 years, the prevalence of chlamydial infection had been reduced by>30% in this population. While other regions of the United States have now initiated similar programs, many family planning andSTDclinics still do not offer chlamydial testing. The availability of highly sensitive and specific diagnostic tests that can be done with urine specimens and of single-dose therapy makes it feasible to mount an effective chlamydial control program nationwide, with screening of high-risk persons both in traditional health care settings and in novel community- and school-based settings. TRACHOMA AND ADULT INCLUSION CONJUNCTIVITIS DEFINITION Trachoma is a chronic conjunctivitis associated with infection by C. trachomatis serovar A, B, Ba, or C. It has been responsible for an estimated 20 million cases of blindness throughout the world and remains an important cause of preventable blindness. Inclusion conjunctivitis is an acute ocular infection caused by sexually transmitted C. trachomatis strains (usually serovars D through K) in adults exposed to infected genital secretions and in their newborn offspring. EPIDEMIOLOGY Epidemiologically, two types of eye disease are caused by C. trachomatis. In trachoma-endemic areas where the classic eye disease is seen, transmission is from eye to eye via hands, flies, towels, and other fomites and usually involves serovar A, B, Ba, or C. In nonendemic areas, organisms of serovars D through K can be transmitted from the genital tract to the eye, usually causing only the inclusion conjunctivitis syndrome, occasionally with keratitis. Rarely, the eye disease acquired in this way progresses, with the development of pannus and scars similar to those seen in endemic trachoma. These cases may be referred to as paratrachoma to differentiate them epidemiologically from eye-to-eye-transmitted endemic trachoma. The worldwide incidence and severity of trachoma have decreased dramatically during the past 35 years, mainly as a result of improving hygienic and economic conditions. Endemic trachoma is still the major cause of preventable blindness in northern Africa, sub-Saharan Africa, the Middle East, and parts of Asia. The endemic disease is transmitted primarily through close personal contact, particularly among young children in rural communities with limited water supplies. In endemic areas, trachoma is associated with repeated exposure and reinfection, but the infection can also become chronic and persistent. In the United States a mild form of endemic trachoma still occurs
in Mexican Americans as well as in immigrants from areas where trachoma is endemic. Acute relapse of old trachoma occasionally follows treatment with cortisone eye ointment or develops in very old persons who were exposed in their youth. CLINICAL MANIFESTATIONS Both endemic trachoma and adult inclusion conjunctivitis present initially as a conjunctivitis characterized by small lymphoid follicles in the conjunctiva. In regions with hyperendemic classic blinding trachoma, the disease usually starts insidiously before the age of 2 years. Reinfection is common and probably contributes to the pathogenesis of trachoma. Studies usingPCRtechniques indicate that chlamydial DNA is often present in the ocular secretions of patients with trachoma, even in the absence of positive cultures. Thus persistent infection may be more common than was previously thought. The cornea becomes involved, with inflammatory leukocytic infiltrations and superficial vascularization (pannus formation). As the inflammation continues, conjunctival scarring eventually distorts the eyelids, causing them to turn inward so that the inturned lashes constantly abrade the eyeball (trichiasis and entropion); eventually the corneal epithelium is abraded and may ulcerate, with subsequent corneal scarring and blindness. Destruction of the conjunctival goblet cells, lacrimal ducts, and lacrimal gland may produce a "dry-eye" syndrome, with resultant corneal opacity due to drying (xerosis) or secondary bacterial corneal ulcers. Communities with blinding trachoma often experience seasonal epidemics of conjunctivitis due to H. influenzae that contribute to the intensity of the inflammatory process. In such areas the active infectious process usually resolves spontaneously in affected persons between 10 and 15 years of age, but the conjunctival scars continue to shrink, producing trichiasis and entropion and subsequent corneal scarring in adults. In areas with milder and less prevalent disease, the process may be much slower, with active disease continuing into adulthood; blindness is rare in these cases. Eye infection with genital C. trachomatis strains in sexually active young adults presents as the acute onset of unilateral follicular conjunctivitis and preauricular lymphadenopathy similar to that seen in acute adenovirus or herpesvirus conjunctivitis. If untreated, the disease may persist for 6 weeks to 2 years. It is frequently associated with corneal inflammation in the form of discrete opacities ("infiltrates"), punctate epithelial erosions, and minor degrees of superficial corneal vascularization. Very rarely, conjunctival scarring and eyelid distortion occur, particularly in patients treated for many months with topical glucocorticoids. Recurrent eye infections develop most often in patients whose sexual consorts are not treated with antimicrobials. DIAGNOSIS The clinical diagnosis of classic trachoma can be made if two of the following signs are present: 1. Lymphoid follicles on the upper tarsal conjunctiva 2. Typical conjunctival scarring
3. Vascular pannus 4. Limbal follicles or their sequelae, Herbert's pits The clinical diagnosis of endemic trachoma should be confirmed by laboratory tests in children with more marked degrees of inflammation. Intracytoplasmic chlamydial inclusions are found in 10 to 60% of Giemsa-stained conjunctival smears in such populations, but isolation in cell cultures, newer antigen detection testing, or chlamydialPCR is more sensitive. Follicular conjunctivitis in adult Europeans or Americans living in trachomatous regions is rarely due to trachoma. Sporadic cases of adult inclusion conjunctivitis must be differentiated from keratoconjunctivitis due to adenovirus orHSV and from bacterial conjunctivitis during the first 15 days after onset; later, they must be distinguished from other forms of chronic follicular conjunctivitis. Demonstration of chlamydiae by Giemsa- or immunofluorescent-stained smears, by isolation in cell cultures, or by newer nonculture tests constitutes definitive evidence of infection. Genital examination and tests for genital chlamydial infection are indicated. Serum antibody does not constitute evidence of chlamydial eye infection since many sexually active adults have acquired serum antibody from genital infection. TREATMENT Public health control programs for endemic trachoma have consisted of the mass application of tetracycline or erythromycin ointment to the eyes of all children in affected communities for 21 to 60 days or on an intermittent schedule. These programs also include surgical correction of inturned eyelids by a mobile surgical team that visits each locale. Single-dose azithromycin therapy is now being evaluated as an alternative method of mass antibiotic treatment for trachoma in young children and pregnant women. Adult inclusion conjunctivitis responds well to treatment with full doses of systemic tetracycline or erythromycin for 3 weeks. Treatment of all sexual consorts of the patient simultaneously is also necessary to prevent ocular reinfection and to avoid genital disease due to chlamydial infection. Topical antibiotic treatment is not required for patients who receive systemic antibiotics. PREVENTION Efforts to develop a trachoma vaccine have not yet been successful. General hygienic measures associated with improved living standards are effective in the elimination of endemic trachoma. An adequate water supply for personal cleanliness may be a key factor. In some areas the reduction of numbers of flies in the household is important. PSITTACOSIS DEFINITION
Psittacosis is primarily an infectious disease of birds and mammals that is caused by C. psittaci. Transmission of infection from birds to humans results in a febrile illness characterized by pneumonitis and systemic manifestations. Inapparent infections or mild influenza-like illnesses may also occur. The term ornithosis is sometimes applied to infections contracted from birds other than parrots or parakeets, but psittacosis is the preferred generic term for all forms of the disease. EPIDEMIOLOGY Almost any avian species can harbor C. psittaci. Psittacine birds (parrots, parakeets, budgerigars) are most commonly infected, but human cases have been traced to contact with pigeons, ducks, turkeys, chickens, and many other birds. Psittacosis may be considered an occupational disease of pet-shop owners, poultry workers, pigeon fanciers, taxidermists, veterinarians, and zoo attendants. During the past 20 years, there has been an increase in incidence, with cases and outbreaks occurring primarily among employees of poultry-processing plants. It is suspected that many cases go undiagnosed and unreported. The disease appears to be especially common in England, where budgerigars are popular household pets and where restrictions on the importation of these birds have been eased. The agent is present in nasal secretions, excreta, tissues, and feathers of infected birds. Although the disease can be fatal, infected birds frequently show only minor evidence of illness, such as ruffled feathers, lethargy, and anorexia. Asymptomatic avian carriers are common, and complete recovery may be followed by continued shedding of the organism for many months. Psittacosis is almost always transmitted to humans by the respiratory route. On rare occasions the disease may be acquired from the bite of a pet bird. Prolonged contact is not essential for transmission of the disease; a few minutes spent in an environment previously occupied by an infected bird has resulted in human infection. In one outbreak, gardening rather than direct exposure to birds was associated with infection. A psittacosis-like agent has been transmitted among hospital personnel, with severe and sometimes fatal infections. There is evidence that these "human" strains are more virulent than avian organisms. There is no record of infection acquired by the ingestion of poultry products. PATHOGENESIS The psittacosis agent gains entrance to the body through the upper part of the respiratory tract, spreads via the bloodstream, and eventually localizes in the pulmonary alveoli and in the reticuloendothelial cells of the spleen and liver. Invasion of the lung probably takes place by way of the bloodstream rather than by direct extension from the upper air passages. A lymphocytic inflammatory response occurs on both the interstitial and the respiratory surfaces of the alveoli as well as in the perivascular spaces. The alveolar walls and interstitial tissues of the lung are thickened, edematous, necrotic, and occasionally hemorrhagic. Histologic examination of the affected areas reveals alveolar spaces filled with fluid, erythrocytes, and lymphocytes. The picture is not pathognomonic of psittacosis unless macrophages containing characteristic cytoplasmic inclusion bodies (Levinthal-Coles-Lillie bodies) can be identified. The respiratory
epithelium of the bronchi and bronchioles usually remains intact. CLINICAL MANIFESTATIONS The clinical manifestations and course of psittacosis are extremely variable. After an incubation period of 7 to 14 days or longer, the disease may start abruptly with shaking chills and fever, with temperatures ranging as high as 40.5°C (105°F); however, the onset is often gradual, with fever increasing over a 3- to 4-day period. Headache is almost always a prominent symptom; it is usually diffuse and excruciating and is often the patient's chief complaint. Many patients present with a dry hacking cough that is usually nonproductive, but small amounts of mucoid or bloody sputum may be raised as the disease progresses. Cough may begin early in the course of the disease or as late as 5 days after the onset of fever. Chest pain, pleurisy with effusion, or a friction rub may all occur but are rare. Pericarditis and myocarditis have been reported. Most patients have a normal or slightly increased respiratory rate; marked dyspnea with cyanosis occurs only in severe psittacosis with extensive pulmonary involvement. In psittacosis, as in mycoplasmal pneumonias, the physical signs of pneumonitis tend to be less prominent than symptoms and x-ray findings would suggest. The initial examination may reveal fine sibilant rales, or clinical evidence of pneumonia may be completely lacking. Rales usually become audible and more numerous as the illness progresses. Signs of frank pulmonary consolidation are usually absent. Symptoms of upper respiratory tract infection are not prominent, although mild sore throat, pharyngitis, and cervical adenopathy are often documented; on occasion, the last may be the only manifestation of illness. Epistaxis is encountered early in the course of nearly one-fourth of cases. Photophobia is also a common complaint. Patients often report generalized myalgia, and spasm and stiffness of the muscles of the back and neck may lead to an erroneous diagnosis of meningitis. Lethargy, mental depression, agitation, insomnia, and disorientation have been prominent features of the illness in some epidemics but not in others; delirium and stupor develop near the end of the first week in severe cases. Occasional patients are comatose when first seen, and the diagnosis of psittacosis may be elusive in these cases. Gastrointestinal manifestations such as abdominal pain, nausea, vomiting, or diarrhea are noted in some cases; constipation and abdominal distention sometimes occur as late complications. Icterus, the result of severe hepatic involvement, is a rare and ominous finding. A faint macular rash (Horder's spots) resembling the rose spots of typhoid fever has been described. Patients without cough or other clinical evidence of respiratory involvement present with fever of unknown origin (Chap. 125). The pulse rate is slow in relation to the fever. When splenomegaly is noted in a patient with acute pneumonitis, psittacosis should be considered; the reported incidence of splenomegaly in this disease ranges from 10 to 70%. Nontender hepatic enlargement also occurs, but jaundice is rare. Thrombophlebitis is not unusual during convalescence; indeed, pulmonary infarction is sometimes a late complication and may be fatal. In untreated cases of psittacosis, sustained or mildly remittent fever persists for 10 days
to 3 weeks or occasionally for as long as 3 months. Over this period, the respiratory manifestations gradually abate. Psittacosis contracted from parrots or parakeets is more likely to be a severe, prolonged illness than infection acquired from pigeons or barnyard fowl. Relapses occur but are rare. Occasional patients develop endocarditis, and C. psittaci infection should be considered in cases of culture-negative endocarditis. Secondary bacterial infections are uncommon. Immunity to reinfection is probably permanent. LABORATORY FINDINGS The chest x-ray in psittacosis is nonspecific and may show pneumonic lesions that are usually patchy in appearance but can be hazy, diffuse, homogeneous, lobar, atelectatic, wedge-shaped, nodular, or miliary. The white blood cell count is normal or moderately decreased in the acute phase of the disease but may rise in convalescence. The erythrocyte sedimentation rate frequently is not elevated. Transient proteinuria is common. The cerebrospinal fluid sometimes contains a few mononuclear cells but is otherwise normal. Despite hepatomegaly, the results of liver function tests are generally normal or mildly elevated. The diagnosis can be confirmed only by isolation of the causative microorganism or by serologic studies. The agent is present in the blood during the acute phase of the disease and in the bronchial secretions for weeks or sometimes years after infection, but it is difficult to isolate. Further, the organism is hazardous to work with in the laboratory, and most clinical laboratories do not offer culture for C. psittaci. Thus psittacosis is most readily diagnosed by the demonstration of a rising titer of complement fixation antibody in the serum of a patient with a compatible clinical syndrome. Both an acute-phase and a convalescent-phase specimen should always be tested. C. trachomatis, C. psittaci, and C. pneumoniae all share a genus-specific "group" antigen, which is the basis of the complement fixation test. Thus acute infections with C. trachomatis or C. pneumoniae can also produce titer rises in this test. However, these three species have different major outer-membrane proteins that are the principal antigens in themicro-IFtest. If there is doubt as to the interpretation of the complement fixation test, the micro-IF test can be used to differentiate among these antigens. The prompt initiation of treatment with tetracycline has been shown to delay an antibody rise in convalescence for several weeks or months. DIFFERENTIAL DIAGNOSIS A history of exposure to birds may be the only clinical basis for differentiating psittacosis from a variety of infectious and noninfectious febrile disorders. The list of pulmonary diseases that may be confused with psittacosis includes Mycoplasma pneumonia, C. pneumoniae pneumonia, legionellosis, viral pneumonia, Q fever, coccidioidomycosis, tuberculosis, enterovirus infection, carcinoma of the lung with bronchial obstruction, and common bacterial pneumonias. In the early stages, before pneumonitis appears, psittacosis may be mistaken for influenza, typhoid fever, miliary tuberculosis, or infectious mononucleosis. TREATMENT
The tetracyclines are consistently effective in the treatment of psittacosis. Defervescence and alleviation of symptoms usually take place within 24 to 48 h after the institution of therapy with 2 g daily in four divided doses. To avoid relapse, treatment should probably be continued for at least 7 to 14 days after defervescence. In severe cases, hospitalization and pulmonary intensive care may be indicated. Sulfonamides are not active against C. psittaci. Erythromycin can be used in patients allergic to or intolerant of tetracyclines. C. PNEUMONIAE INFECTIONS A third chlamydial species that causes disease in humans, C. pneumoniae, has been described in the past two decades. C. pneumoniae can be distinguished from the other two species on the basis of DNA hybridization and elementary body morphology. Although C. pneumoniae can be grown in a variety of cell cultures, it is considerably more difficult to culture than other chlamydiae, especially from clinical specimens. HL cells appear to be the most effective cell line for isolation of C. pneumoniae. Knowledge of the epidemiology of C. pneumoniae infections has been derived primarily from serologic studies. Infections begin to occur in late childhood, achieve peak incidence in young adults, but continue throughout adult life. Seroprevalence in the many adult populations that have been tested throughout the world exceeds 40% -- a figure suggesting that C. pneumoniae infections are ubiquitous. Secondary episodes (reinfections) appear to occur commonly in older adults throughout life. C. pneumoniae also produces epidemics of pneumonia and respiratory illness, especially in close residential quarters such as military barracks. The incidence of infections outside of epidemics remains poorly defined. Transmission appears to be from person to person, probably primarily in schools and family units. Little is known about the pathogenesis of C. pneumoniae infection. The infection begins in the upper respiratory tract and in many persons is a long-lived asymptomatic condition of the upper respiratory mucosal surfaces. However, in at least some individuals, the organism is transported to distant sites -- perhaps within macrophages -since evidence exists for replication within arteries and synovial membranes of joints. A C. pneumoniae outer-membrane protein may induce host immune responses whose cross-reaction with human proteins results in an autoimmune reaction. The clinical spectrum of C. pneumoniae infection includes acute pharyngitis, sinusitis, bronchitis, and pneumonitis, primarily in young adults. The clinical manifestations of primary infection appear to be more severe and prolonged than those of reinfection. The pneumonitis resembles that of M. pneumoniae pneumonia in that leukocytosis is frequently lacking and patients often have prominent antecedent upper respiratory tract symptoms, fever, nonproductive cough, a mild to moderate degree of illness, minimal findings on chest auscultation, and small segmental infiltrates on chest x-ray. In elderly patients, pneumonia due to C. pneumoniae can be especially severe and may necessitate hospitalization and respiratory support. Epidemiologic studies have demonstrated an association between serologic evidence of C. pneumoniae infection and atherosclerotic disease of the coronary and other arteries. In addition, C. pneumoniae has been identified in atherosclerotic plaques by electron
microscopy, DNA hybridization, and immunocytochemistry. Recently, the organism has been recovered in culture from atheromatous plaque -- a result indicating the presence of viable replicating bacteria in vessels. Evidence from animal models supports the hypothesis that C. pneumoniae infection of the upper respiratory tract is followed by recovery of the organism from atheromatous lesions in the aorta and that the infection accelerates the process of atherosclerosis, especially in hypercholesterolemic animals. Antimicrobial treatment of the infected animals reverses the increased risk of atherosclerosis. In humans, two small trials in patients with unstable angina or recent myocardial infarction also suggested that antibiotics reduce subsequent untoward cardiac events. Larger trials have been initiated to determine more definitively whether antibiotics affect the risk of atherosclerosis. Diagnosis of C. pneumoniae infection is currently difficult because cell culture techniques are not available for routine clinical use and nonculture tests using antigen detection methods or DNA probes have not been developed for commercial use. Acuteand convalescent-phase sera can be tested for chlamydial complement fixation antibody to make a retrospective diagnosis. However, this test does not distinguish C. pneumoniae infection from infection due to C. trachomatis or C. psittaci. Although controlled treatment trials have not been conducted, C. pneumoniae is inhibited in vitro by erythromycin and tetracycline. Recommended therapy consists of 2 g per day of either agent for 10 to 14 days. Other macrolides, such as azithromycin, and some fluoroquinolones, such as levofloxacin, also appear to be effective. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 11 -VIRAL DISEASES 180. MEDICAL VIROLOGY - Fred Wang, Elliott Kieff DEFINING A VIRUS Viruses consist of a nucleic acid surrounded by one or more proteins. Some viruses also have an outer-membrane envelope. Viruses differ from other replicating organisms in that they do not have ribosomes or enzymes for high-energy phosphate generation or for protein, carbohydrate, or lipid metabolism. Viruses are obligate intracellular parasites -- that is, they require cells in order to replicate. Typically, viral nucleic acids encode proteins necessary for replicating and packaging the nucleic acids into new viral particles. Viruses differ from viroids, prions, and virusoids. Virusoids are nucleic acids that depend on helper viruses to package the nucleic acids into virus-like particles. Viroids are simply molecules of naked, cyclical, mostly double-stranded, small RNAs and appear to be restricted to plants, in which they spread from cell to cell and are replicated by cellular RNA polymerase II. Prions (Chap. 375) are protein molecules that can spread from cell to cell and effect changes in the structure of their normal counterparts (cellular proteins). Prions have been implicated in neurodegenerative conditions such as Creutzfeldt-Jakob disease, kuru, and Gerstmann-Straussler disease. Prions have also been implicated in neurodegeneration associated with human infection with bovine spongiform encephalopathy ("mad cow disease"). VIRAL STRUCTURE Viruses have from a few to 200 genes. These genes may be embodied in a single-strand or double-strand DNA genome or in a single-strand sense, a single-strand or segmented antisense, or a double-strand segmented RNA genome. Sense-strand RNA genomes can be translated directly into protein. Sense and antisense genomes are also referred to as positive-strand and negative-strand genomes, respectively. The viral nucleic acid is usually associated with one or more virus-encoded nucleoproteins in the core of the viral particle. The viral nucleic acid is almost always enclosed in a protein shell called a capsid. Because of the limited genetic complexity of viruses, their capsids are usually composed of multimers of identical capsomers. Capsomers are in turn composed of one or a few proteins. Capsids have icosahedral or helical symmetry. Icosahedral structures approximate spheres but have two-, three-, and fivefold axes of symmetry, while helical structures have only a twofold axis of symmetry. The entire structural unit of nucleic acid, nucleoprotein(s), and capsid is called a nucleocapsid. Many human viruses have a simple nucleocapsid structure. For these viruses, the outer surface of the capsid mediates contact with uninfected cells. Other viruses are more complex and have an outer envelope that is derived from membranes of the infected cell. The piece of infected cell membrane that becomes the viral envelope has usually been modified during infection by the insertion of virus-encoded glycoproteins. These glycoproteins usually mediate contact of enveloped viruses with uninfected cells. Enveloped viruses frequently have matrix or tegument proteins that fill the space between the nucleocapsid and the envelope. In general, enveloped viruses are sensitive to solvents and nonionic detergents that can disrupt the envelope, while viruses that
consist only of nucleocapsids are usually more resistant. The schematic diagram of a large and complex herpesvirus shown in Fig. 180-1 illustrates the components of a complicated DNA virus. Prototypical pathogenic human viruses are listed inTable 180-1. The relative sizes and structures of typical pathogenic human viruses are shown inFig. 180-2. TAXONOMY OF PATHOGENIC HUMAN VIRUSES As is apparent from Table 180-1 andFig. 180-2, the classification of viruses into orders and families is based on nucleic acid composition, nucleocapsid size and symmetry, and envelopment status. Viruses of a single family have similar types of genomes and are morphologically similar in electron micrographs. Further subclassification into genus is dependent on similarities in epidemiology and biologic effects and on the degree of colinear nucleic acid sequence homology. In general, each human virus has a common name related to its pathologic effects or the circumstances of its discovery and a formal species name assigned by the International Committee on Taxonomy of Viruses. The latter designation consists of the name of the host followed by the family or genus of the virus and a number. This dual terminology has created a confusing situation in which viruses are referred to and referenced by either name -- e.g., varicella-zoster virus (VZV) or human herpesvirus (HHV) 3. VIRAL INFECTION IN VITRO STAGES OF INFECTION At the cellular level, viral infection proceeds in stages. Viral Interactions at the Cell Surface First, virus adsorbs to a receptor on the cell surface. Adsorption is the consequence of a molecular interaction of a viral surface protein with a molecule on the cell's plasma membrane. For example, a poliovirus capsid protein binds to a cell plasma-membrane protein of the immunoglobulin superfamily type; a rhinovirus capsid protein binds to intracellular adhesion molecule 1; an echovirus capsid protein binds to an integrin; the influenza A virus envelope hemagglutinin protein binds to sialic acid; the HIV envelope glycoprotein binds to CD4 and then engages one of several chemokine receptors that function as coreceptors for the virus; the herpes simplex virus (HSV) envelope glycoproteins bind to heparan sulfate on cell surfaces and then engage one of several immunoglobulin superfamily or tumor necrosis factor (TNF) receptors; and an Epstein-Barr virus (EBV) glycoprotein binds to the B lymphocyte complement receptor CD21. Adsorption characteristically proceeds almost as well at 4°C as at 37°C, and adsorbed virus can still be neutralized by antibody. Adsorption frequently initiates changes in virion surface proteins that result in destabilization and preparation for the next stage of entry into the cell. After adsorption, viruses penetrate through or fuse with the cell membrane, lose their sensitivity to neutralizing antibody, and become uncoated as they enter the cytoplasm. For all viruses, penetration and uncoating result in viral nucleocapsid or nucleoprotein entry into the cytoplasm. Penetration and uncoating as well as subsequent steps in viral replication depend on the cell's energy metabolism and on biochemical changes in the cell's plasma membrane and cytoskeleton. Therefore, penetration proceeds slowly at
temperatures48 h is the basis for the recommended dosing regimen of 5 mg/kg twice a week for the initial 2 weeks and then 5 mg/kg once a week. The major toxic effect of cidofovir is proximal renal tubular injury, as manifested by elevated serum creatinine levels and proteinuria. The risk of nephrotoxicity can be reduced by vigorous saline hydration and by concomitant oral administration of probenecid. Neutropenia, rashes, and gastrointestinal tolerance may also occur. Intravenous cidofovir has been approved for the treatment ofCMVretinitis in AIDS patients who are intolerant of ganciclovir or foscarnet or in whom those drugs have failed. In a controlled study, a maintenance dosage of 5 mg/kg a week administered to AIDS patients reduced the progression of CMV retinitis from that seen at 3 mg/kg.
Intravenous cidofovir has been reported anecdotally to be effective therapy for acyclovir-resistant mucocutaneousHSVinfections. Likewise, topically administered cidofovir is reportedly beneficial against these infections in HIV patients; it is also being studied for the treatment of anogenital warts. Intravenous cidofovir is being evaluated as therapy for progressive multifocal leukoencephalopathy and for Kaposi's sarcoma. An ophthalmic formulation is being studied as treatment for adenoviral keratoconjunctivitis. Intravitreal cidofovir has been used to treat CMV retinitis but has been associated with significant toxicity. FOMIVIRSEN Fomivirsen is the first antisense oligonucleotide approved by the U.S. Food and Drug Administration (FDA) for therapy in humans. This phosphorothioate oligonucleotide, 21 nucleotides in length, inhibitsCMVreplication through interaction with CMV messenger RNA. Fomivirsen is complementary to messenger transcripts of the major immediate early region 2 (IE2) of CMV, which codes for proteins regulating viral gene expression. In addition to its antisense mechanism of action, fomivirsen may exert activity against CMV through inhibition of viral adsorption to cells as well as direct inhibition of viral replication. Because of its different mechanism of action, fomivirsen is active against CMV isolates that are resistant to nucleoside or nucleotide analogues, such as ganciclovir, foscarnet, or cidofovir. Fomivirsen has been approved for intravitreal administration in the treatment ofCMVretinitis in AIDS patients who have failed to respond to other treatments or cannot tolerate them. Injections of 330 mg every 2 weeks have resulted in significant reductions in the rate of progression of CMV retinitis. The major toxicity is ocular inflammation, including vitritis and iritis, which usually responds to topically administered glucocorticoids. GANCICLOVIR An analogue of acyclovir, ganciclovir is active againstHSV andVZV and is markedly more active than acyclovir againstCMV. Ganciclovir triphosphate inhibits CMV DNA polymerase and can be incorporated into CMV DNA, whose elongation it eventually terminates. In HSV- and VZV-infected cells, ganciclovir is phosphorylated by virus-encoded thymidine kinases; in CMV-infected cells, it is phosphorylated by a viral kinase encoded by the UL97 gene. Ganciclovir triphosphate is present in tenfold higher concentrations in CMV-infected cells than in uninfected cells. Ganciclovir is approved for the treatment of CMV retinitis in immunosuppressed patients and for the prevention of CMV disease in transplant recipients. It is widely used for the treatment of other CMV-associated syndromes, including pneumonia, esophagogastrointestinal infections, hepatitis, and "wasting" illness. Ganciclovir is available for intravenous or oral administration. Because its oral bioavailability is low (5 to 9%), relatively large doses (1 g tid) must be administered by this route. Oral bioavailability is enhanced if the drug is administered with food, as recommended. The serum half-life of ganciclovir is 3.5 h after intravenous administration and 4.5 h after oral administration. The drug is excreted primarily by the kidney in unmetabolized form, and its dosage should be reduced in cases of renal
failure. The most commonly employed dosage for initial therapy -- 5 mg/kg intravenously every 12 h for 14 to 21 days -- is followed by a maintenance dose of 5 mg/kg intravenously per day or 5 times per week, possibly for as long as immunosuppression persists. Oral ganciclovir is approved as an alternative to the intravenous preparation in maintenance therapy forCMVretinitis, where it appears to be somewhat less effective although more convenient than intravenous therapy. Intraocular ganciclovir, given by either intravitreal injection or intraocular implantation, has also been used to treat CMV retinitis. Ganciclovir is effective as prophylaxis againstCMV-associated disease in organ and bone marrow transplant recipients. Oral ganciclovir administered prophylactically to AIDS patients with CD4+ counts of100-fold and decreased the subsequent rate of development of illness. Pleconaril treatment of adults with enteroviral meningitis decreased the overall duration of illness and headache and reduced the use of analgesics from that by placebo recipients. Clinical studies of pleconaril in other enterovirus-induced diseases are in progress. INTERFERONS Interferons are cytokines that exhibit a broad spectrum of antiviral activities as well as immunomodulating and antiproliferative properties. TheIFNs are not available for oral administration but must be given intramuscularly, subcutaneously, or intravenously. Early studies with human leukocyte IFN demonstrated an effect in the prophylaxis of experimentally induced rhinovirus infections in humans and in the treatment ofVZVinfections in immunosuppressed patients. DNA recombinant technology has made available highly purified a,b, and g IFNs that have been evaluated in a variety of viral infections. Results of such trials have confirmed the effectiveness of intranasally administered IFN in the prophylaxis of rhinovirus infections, although its use has been associated with nasal mucosal irritation. Studies have also demonstrated a beneficial
effect of intralesionally or systemically administered IFNs on genital warts. The effect of systemic administration consists primarily of a reduction in the size of lesions, and this mode of therapy may be useful in individuals who have numerous warts that cannot easily be treated by individual intralesional injection. However, lesions frequently recur after intralesional or systemic IFN therapy is discontinued. Interferons have undergone extensive study in the treatment of chronicHBVinfection. The administration ofIFN-a2b (5 million units daily for 16 weeks) to patients with stable chronic HBV infection resulted in loss of markers of HBV replication, such asHBeAgand HBV DNA, in 33 to 37% of cases; 10 to 20% of patients also became negative for hepatitis B surface antigen. In >80% of patients who lose HBeAg and HBV DNA markers, serum aminotransferases return to normal levels, and both short- and long-term improvements in liver histopathology have been described. Predictors of a favorable response to therapy include low pretherapy levels of HBV DNA, high pretherapy serum levels of alanine aminotransferase (ALT), a short duration of chronic HBV infection, and active liver histopathology. Poor responses are seen in immunosuppressed patients, including those with HIV infection. Adverse effects of the above dose of IFN are common and include fever, chills, myalgia, fatigue, neurotoxicity (primarily manifested as somnolence and confusion), and leukopenia. Approximately 25% of patients receiving a daily dose of 5 million units require dose reduction, but fewer than 5% require discontinuation of therapy. SeveralIFNpreparations, including a2a,a2b, alfacon-1, and am1 (lymphoblastoid), have been studied as therapy for chronicHCVinfections. A variety of regimens have been employed, of which the most common is IFN-a2b or -a2a at 3 million units three times per week for 12 months. A complete biochemical response, defined as a return to normal serumALTvalues at the end of treatment, has been documented in ~54% of patients. In addition, liver biopsies have shown decreases in lobular and periportal inflammation. However, relapse has occurred in approximately half of all cases upon discontinuation of therapy, so that sustained responses were documented in 28% of cases. The addition of oral ribavirin to IFN-a2b -- either as initial therapy or after failure of interferon therapy alone -- resulted in significantly higher rates of sustained response (40 to 50%) than were obtained with monotherapy. Prognostic factors for a favorable response include an age of60% subsequently identify episodes of symptomatic reactivation. Most important, these asymptomatic seropositive persons with reactivation shed virus on mucosal surfaces as frequently as those with symptomatic disease. The large reservoir of unidentified carriers of HSV-2 and the frequent asymptomatic reactivation of virus from the genital tract have fostered the continued spread of genital herpes throughout the world. HSV-2 infection is an independent risk factor for the acquisition and transmission of infection with HIV type 1. Among coinfected persons, HIV-1 virions can be shed from herpetic
lesions of the genital region. This shedding may facilitate the spread of HIV through sexual contact. HSVinfections occur throughout the year. The incubation period ranges from 1 to 26 days (median, 6 to 8 days). Transmission can result from contact with persons with active ulcerative lesions or with persons without clinical manifestations of infection who are shedding HSV or on whose mucosal surfaces the virus is replicating. Studies using the polymerase chain reaction (PCR) have shown that HSV reactivation on mucosal surfaces is much more frequent than previously recognized. Among immunocompetent adults, HSV-2 can be isolated from the genital tract on 2 to 3% of days, and HSV DNA can be detected on 20 to 30% of days. Corresponding figures for HSV-1 in oral secretions are similar. Shedding rates are highest during the initial years of acquisition and may be as high as 30 to 50% of days during this period. Immunosuppressed patients shed HSV on mucosal sites at even higher frequency (20 to 50% of days). Daily antiviral chemotherapy can markedly reduce shedding rates. These data indicate that potential exposure to HSV from sexual or other close contact (kissing, sharing of glasses or silverware) is more common than has been thought. These shedding-rate data are consistent with the high seroprevalence of HSV infections worldwide. CLINICAL SPECTRUM HSVhas been isolated from nearly all visceral or mucocutaneous sites. The clinical manifestations and course of HSV infection depend on the anatomic site involved, the age and immune status of the host, and the antigenic type of the virus. Primary HSV infections (i.e., first infections with either HSV-1 or HSV-2 in which the host lacks HSV antibodies in acute-phase serum) are frequently accompanied by systemic signs and symptoms, involve both mucosal and extramucosal sites, and have a longer duration of symptoms, a longer duration of virus isolation from lesions, and a higher rate of complications than recurrent episodes of disease. Both viral subtypes can cause genital and oral-facial infections, and the infections caused by the two subtypes are clinically indistinguishable. However, the frequency of reactivation of infection is influenced by anatomic site and virus type. Genital HSV-2 infection is twice as likely to reactivate and recurs 8 to 10 times more frequently than genital HSV-1 infection. Conversely, oral-labial HSV-1 infection recurs more frequently than oral-labial HSV-2 infection. Asymptomatic shedding rates follow the same pattern. Oral-Facial Infections Gingivostomatitis (Fig. 182-CD1) and pharyngitis are the most frequent clinical manifestations of first-episodeHSV-1 infection, while recurrent herpes labialis (Fig. 182-CD2) is the most frequent clinical manifestation of reactivation HSV infection. HSV pharyngitis and gingivostomatitis usually result from primary infection and are most commonly seen in children and young adults. Clinical symptoms and signs, which include fever, malaise, myalgias, inability to eat, irritability, and cervical adenopathy, may last from 3 to 14 days. Lesions may involve the hard and soft palate, gingiva, tongue, lip, and facial area. HSV-1 or HSV-2 infection of the pharynx usually results in exudative or ulcerative lesions of the posterior pharynx and/or tonsillar pillars. Lesions of the tongue, buccal mucosa, or gingiva may occur later in the course in one-third of cases. Fever lasting from 2 to 7 days and cervical adenopathy are common. It can be difficult to differentiate HSV pharyngitis clinically from bacterial pharyngitis, Mycoplasma pneumoniae infections, and pharyngeal ulcerations of noninfectious
etiologies (e.g., Stevens-Johnson syndrome). No substantial evidence suggests that reactivation oral-labial HSV infection is associated with symptomatic recurrent pharyngitis. Reactivation of HSVfrom the trigeminal ganglia may be associated with asymptomatic virus excretion in the saliva, development of intraoral mucosal ulcerations, or herpetic ulcerations on the vermilion border of the lip or external facial skin. About 50 to 70% of seropositive patients undergoing trigeminal nerve root decompression and 10 to 15% of those undergoing dental extraction develop oral-labial HSV infection a median of 3 days after these procedures. In immunosuppressed patients, infection may extend into mucosal and deep cutaneous layers (Fig. 182-CD3). Friability, necrosis, bleeding, severe pain, and inability to eat or drink may result. The lesions ofHSVmucositis are clinically similar to mucosal lesions caused by cytotoxic drug therapy, trauma, or fungal or bacterial infections. Persistent ulcerative HSV infections are among the most common infections in patients with AIDS. HSV and Candida infections often occur concurrently. Systemic antiviral therapy speeds the rate of healing and relieves the pain of mucosal HSV infections in immunosuppressed patients. The frequency of HSV reactivation during the early phases of transplantation or induction chemotherapy is high (50 to 90%), and prophylactic systemic antivirals such as intravenous acyclovir or penciclovir are used to reduce reactivation rates. Patients with atopic eczema also may develop severe oral-facial HSV infections (eczema herpeticum), which may rapidly come to involve extensive areas of skin and occasionally disseminate to visceral organs. Extensive eczema herpeticum has resolved promptly with the administration of intravenous acyclovir. Erythema multiforme (EM) also may be associated with HSV infections (Plate IIE-67); some evidence suggests that HSV infection is the precipitating event in ~75% of cases of cutaneous EM. HSV antigen has been demonstrated both in circulatory immune complexes and in skin lesion biopsy samples from these patients. Patients with severe HSV-associated EM are candidates for chronic suppressive oral antiviral therapy. HSV-1 has been implicated in the etiology of Bell's palsy (flaccid paralysis of the mandibular portion of the facial nerve). Whether antiviral chemotherapy can alter the clinical course and complications of this infection is unclear. Genital Infections First-episode primary genital herpes is characterized by fever, headache, malaise, and myalgias. Pain, itching, dysuria, vaginal and urethral discharge, and tender inguinal lymphadenopathy are the predominant local symptoms. Widely spaced bilateral lesions of the external genitalia are characteristic. Lesions may be present in varying stages, including vesicles, pustules, or painful erythematous ulcers. The cervix and urethra are involved in>80% of women with first-episode infections. First episodes of genital herpes in patients who have had priorHSV-1 infection are associated with less frequent systemic symptoms and faster healing than primary genital herpes. The clinical courses of acute first-episode genital herpes among patients with HSV-1 and HSV-2 infections are similar. However, the recurrence rates of genital disease (Figs. 182-CD4,182-CD5,182-CD6) differ with the viral subtype: the 12-month recurrence rates among patients with first-episode HSV-2 and HSV-1 infections are ~90% and ~55%, respectively (median number of recurrences, 4 and 50 years of age. Subtype 1 virus causes >95% of cases of HSV encephalitis. The pathogenesis ofHSVencephalitis varies. In children and young adults, primary HSV infection may result in encephalitis; presumably, exogenously acquired virus enters theCNS by neurotropic spread from the periphery via the olfactory bulb. However, most adults with HSV encephalitis have clinical or serologic evidence of mucocutaneous HSV-1 infection before the onset of the CNS symptoms. In ~25% of the cases examined, the HSV-1 strains from the oropharynx and brain tissue of the same patient differ; thus some cases may result from reinfection with another strain of HSV-1 that reaches the CNS. Two theories have been proposed to explain the development of actively replicating HSV in localized areas of the CNS in persons whose ganglionic and CNS isolates are similar. Reactivation of latent HSV-1 infection in trigeminal or autonomic nerve roots may be associated with extension of virus into the CNS via nerves innervating the middle cranial fossa. HSV DNA has been demonstrated by DNA hybridization in brain tissue obtained at autopsy -- even from healthy adults. Thus, reactivation of long-standing latent CNS infection may be another mechanism for the development of HSV encephalitis. The clinical hallmark ofHSVencephalitis has been the acute onset of fever and focal neurologic (especially temporal-lobe) symptoms. Clinical differentiation of HSV encephalitis from other viral encephalitides, focal infections, or noninfectious processes is difficult. The most sensitive noninvasive method for early diagnosis of HSV encephalitis is the demonstration of HSV DNA in cerebrospinal fluid (CSF) byPCR. Although titers of CSF and serum antibodies to HSV increase in most cases of HSV encephalitis, they rarely do so earlier than 10 days into the illness and therefore, while useful retrospectively, are generally not helpful in establishing an early clinical diagnosis. Demonstration of HSV antigen, HSV DNA, or HSV replication in brain tissue obtained by biopsy is highly sensitive and has a low complication rate; examination of such tissue also provides the best opportunity to identify alternative, potentially treatable causes of encephalitis. Antiviral chemotherapy reduces the rate of death from HSV encephalitis. Intravenous acyclovir is more effective than vidarabine. Even with therapy, however, neurologic sequelae are frequent, especially in persons over 35 years of age. Most authorities recommend the administration of intravenous acyclovir to patients with presumed HSV encephalitis until the diagnosis is confirmed or an alternative diagnosis is made. HSVhas been isolated from theCSF of 0.5 to 3% of patients presenting to the hospital with aseptic meningitis. HSV meningitis, which is usually seen in association with primary genital HSV infection, is an acute, self-limited disease manifested by headache, fever, and mild photophobia and lasting from 2 to 7 days. Lymphocytic pleocytosis in the
CSF is characteristic. Neurologic sequelae of HSV meningitis are rare. HSV is the most commonly identified cause of recurrent lymphocytic meningitis (Mollaret's meningitis). Demonstration of HSV antibodies in CSF or persistence of HSV DNA in CSF can establish the diagnosis. Daily administration of antiviral therapy aimed at reducing the likelihood of clinical HSV reactivation has been successful in such cases. Autonomic nervous system dysfunction, especially of the sacral region, has been reported in association with bothHSV and varicella-zoster virus infections. Numbness, tingling of the buttocks or perineal areas, urinary retention, constipation,CSFpleocytosis, and (in males) impotence may occur. Symptoms appear to resolve slowly over days to weeks. Occasionally, hypesthesia and/or weakness of the lower extremities may persist for many months. Rarely, transverse myelitis manifested by a rapidly progressive symmetric paralysis of the lower extremities or a Guillain-Barre syndrome may follow HSV infection. Similarly, peripheral nervous system involvement (Bell's palsy) or cranial polyneuritis also may be related to reactivation of HSV-1 infection. Transitory hypesthesia of the area of skin innervated by the trigeminal nerve and vestibular system dysfunction as measured by electronystagmography are the predominant signs of disease. Studies to determine whether antiviral chemotherapy may abort these signs or reduce their frequency and severity are unavailable. Visceral Infections HSVinfection of visceral organs usually results from viremia, and multiple-organ involvement is common. Occasionally, however, the clinical manifestations of HSV infection involve only the esophagus, lung, or liver. HSV esophagitis may result from direct extension of oral-pharyngeal HSV infection into the esophagus or may occur de novo by reactivation and spread of HSV to the esophageal mucosa via the vagus nerve. The predominant symptoms of HSV esophagitis are odynophagia, dysphagia, substernal pain, and weight loss. There are multiple oval ulcerations on an erythematous base with or without a patchy white pseudomembrane. The distal esophagus is most commonly involved. With extensive disease, diffuse friability may spread to the entire esophagus. Neither endoscopic nor barium examination can differentiate HSV esophagitis from Candida esophagitis or from esophageal ulcerations due to thermal injury, radiation, or corrosives. Endoscopically obtained secretions for cytologic examination and culture provide the most useful material for diagnosis. Systemic antiviral chemotherapy usually reduces symptoms and heals esophageal ulcerations. HSVpneumonitis is uncommon except in severely immunosuppressed patients and may result from extension of herpetic tracheobronchitis into lung parenchyma. Focal necrotizing pneumonitis usually ensues. Hematogenous dissemination of virus from sites of oral or genital mucocutaneous disease also may occur and produce bilateral interstitial pneumonitis. Bacterial, fungal, and parasitic pathogens are commonly present in HSV pneumonitis. The mortality rate from untreated HSV pneumonia in immunosuppressed patients is high (>80%). HSV has also been isolated from the lower respiratory tract of persons with adult respiratory distress syndrome (ARDS). However, the relationship between the isolation of HSV and the pathogenesis of ARDS is unclear. HSVis an uncommon cause of hepatitis in immunocompetent patients. HSV infection of the liver is associated with fever, abrupt elevations of bilirubin and serum aminotransferase levels, and leukopenia (800) of relatively uncommon (1 cm in diameter and surrounded by hairline arcuate shadows indicate the presence of bullae and are highly specific for emphysema. With hyperinflation, the chest becomes vertically elongated with low flattened diaphragms. The heart shadow is also vertical and narrow. The retrosternal airspace is increased on the lateral view, and the sternal-diaphragmatic angle exceeds 90°. In the presence of pulmonary hypertension, the pulmonary arteries become enlarged and taper rapidly. The right heart border may become prominent and impinge on the retrosternal airspace. The presence of "dirty lung fields" may reflect the presence of bronchiolitis. Computed tomography has greater sensitivity and specificity for emphysema than the plain film but is rarely necessary except for the diagnosis of bronchiectasis and evaluation of bullous disease. Nonhomogeneous distribution of emphysema is thought by some to be an indicator of suitability for lung volume reduction surgery (LVRS). PULMONARY FUNCTION TESTING (See alsoChap. 250)
Because of the imprecision of clinical findings, objective evaluation of the presence, severity, and reversibility of airflow obstruction is essential in the diagnostic evaluation ofCOPD. A normalFEV1essentially excludes the diagnosis. The spirogram in COPD shows decreased volume changes with time and a failure to reach a plateau after 3 to 5 s. Continued airflow may be evident for 10 s or more on forced exhalation. The flow-volume curve shows diminished expiratory flow at all lung volumes. Expiratory flow is concave to the volume axis. When flow is plotted against absolute lung volume, the entire curve is shifted to higher volumes, reflecting hyperinflation. Serial spirometry is important in assessing the rate of decline of FEV1. Reversibility is assessed by spirometry before and after administration of an inhaled bronchodilator, most often a short-actingb 2-adrenergic agonist. Testing should be performed when the patient is clinically stable. Short-acting bronchodilators should be withheld for 6 h, long-acting dilators for 12 h, and theophylline for 24 h prior to testing. A significant response is an increase of at least 12% and 200 mL in eitherFEV1or forced vital capacity (FVC). Postbronchodilator FEV1 is useful for prognostication. Although only one-third ofCOPDpatients show a significant response to an inhaled bronchodilator in the pulmonary function laboratory on any one day, two-thirds will show a significant response when tested with different bronchodilators on several different occasions. The degree of bronchodilator response at any one testing session does not predict the degree of clinical benefit to the patient. Therefore, bronchodilators are given irrespective of the acute response obtained in the pulmonary function laboratory. The American Thoracic Society recommends staging COPD by FEV1. Stage I, mild disease, is defined as FEV1³ 50% predicted; stage II, moderate disease, 35 to 49% predicted; and stage III, severe disease, 45 IU/L, gamma interferon> 140 pg/mL, or positive PCR for tuberculous DNA). Alternatively, the diagnosis can be established by culture of the pleural fluid, needle biopsy of the pleura, or thoracoscopy. The recommended treatment of pleural and pulmonary tuberculosis is identical (Chap. 169). Effusion Secondary to Viral Infection Viral infections are probably responsible for a sizable percentage of undiagnosed exudative pleural effusions. In many series, no diagnosis is established for approximately 20% of exudative effusions, and these effusions resolve spontaneously with no long-term residua. The importance of these effusions is that one should not be too aggressive in trying to establish a diagnosis for the undiagnosed effusion, particularly if the patient is improving clinically. AIDS Pleural effusions are uncommon in such patients. The most common cause is Kaposi's sarcoma, followed by parapneumonic effusion. Other common causes are tuberculosis, cryptococcosis, and lymphoma. Pleural effusions are very uncommon with Pneumocystis carinii infection. Chylothorax A chylothorax occurs when the thoracic duct is disrupted and chyle accumulates in the pleural space. The most common cause of chylothorax is trauma, but it also may result from tumors in the mediastinum. Patients with chylothorax present with dyspnea, and a large pleural effusion is present on the chest radiograph. Thoracentesis reveals milky fluid, and biochemical analysis reveals a triglyceride level that exceeds 110 mg/dL. Patients with chylothorax and no obvious trauma should have a lymphangiogram and a mediastinal computed tomographic (CT) scan to assess the mediastinum for lymph nodes. The treatment of choice for most chylothoraces is implantation of a pleuroperitoneal shunt. Patients with chylothoraces should not undergo prolonged tube thoracostomy with chest tube drainage because this will lead to malnutrition and immunologic incompetence. Hemothorax When a diagnostic thoracentesis reveals bloody pleural fluid, a hematocrit should be obtained on the pleural fluid. If the hematocrit is>50% that of the peripheral blood, the patient has a hemothorax. Most hemothoraces are the result of trauma; other causes include rupture of a blood vessel or tumor. Most patients with hemothorax should be treated with tube thoracostomy, which allows continuous quantification of bleeding. If the bleeding emanates from a laceration of the pleura, apposition of the two pleural surfaces is likely to stop the bleeding. If the pleural hemorrhage exceeds 200 mL/h, consideration should be given to thoracotomy. Miscellaneous Causes of Pleural Effusion There are many other causes of pleural effusion (Table 262-1). Key features of some of these conditions are as follows: If the pleural fluid amylase level is elevated, the diagnosis of esophageal rupture or pancreatic
disease is likely. If the patient is febrile, has predominantly polymorphonuclear cells in the pleural fluid, and has no pulmonary parenchymal abnormalities, an intraabdominal abscess should be considered. The diagnosis of an asbestos pleural effusion is one of exclusion. Benign ovarian tumors can produce ascites and a pleural effusion (Meigs' syndrome), as can the ovarian hyperstimulation syndrome. Several drugs can cause pleural effusion; the associated fluid is usually eosinophilic. Pleural effusions commonly occur following coronary artery bypass surgery. Effusions occurring within the first weeks are typically left-sided and bloody, with large numbers of eosinophils, and respond to one or two therapeutic thoracenteses. Effusions occurring after the first few weeks are typically left-sided and clear yellow, with predominantly small lymphocytes, and tend to recur. Other medical manipulations that induce pleural effusions include abdominal surgery, endoscopic variceal sclerotherapy, radiation therapy, liver or lung transplantation, or the intravascular insertion of central lines. PNEUMOTHORAX Pneumothorax is the presence of gas in the pleural space. A spontaneous pneumothorax is one that occurs without antecedent trauma to the thorax. A primary spontaneous pneumothorax occurs in the absence of underlying lung disease, while a secondary spontaneous pneumothorax occurs in its presence. A traumatic pneumothorax results from penetrating or nonpenetrating chest injuries. A tension pneumothorax is a pneumothorax in which the pressure in the pleural space is positive throughout the respiratory cycle. Primary Spontaneous Pneumothorax Primary spontaneous pneumothoraces are usually due to rupture of apical pleural blebs, small cystic spaces that lie within or immediately under the visceral pleura. Primary spontaneous pneumothoraces occur almost exclusively in smokers, which suggests that these patients have subclinical lung disease. Approximately one-half of patients with an initial primary spontaneous pneumothorax will have a recurrence. The initial recommended treatment for primary spontaneous pneumothorax is simple aspiration. If the lung does not expand with aspiration, or if the patient has a recurrent pneumothorax, thoracoscopy with stapling of blebs and pleural abrasion is indicated. Thoracoscopy or thoracotomy with pleural abrasion is almost 100% successful in preventing recurrences. Secondary Spontaneous Pneumothorax Most secondary spontaneous pneumothoraces are due to chronic obstructive pulmonary disease, but pneumothoraces have been reported with virtually every lung disease. Pneumothorax in patients with lung disease is more life-threatening than it is in normal individuals because of the lack of pulmonary reserve in these patients. Nearly all patients with secondary spontaneous pneumothorax should be treated with tube thoracostomy and the instillation of a sclerosing agent such as doxycycline or talc. Patients with secondary spontaneous pneumothoraces who have a persistent air leak, an unexpanded lung after 3 days of tube thoracostomy, or a recurrent pneumothorax should be subjected to thoracoscopy with bleb resection and pleural abrasion. Traumatic Pneumothorax Traumatic pneumothoraces can result from both penetrating and nonpenetrating chest trauma. Traumatic pneumothoraces should be treated with tube thoracostomy unless they are very small. If a hemopneumothorax is present, one
chest tube should be placed in the superior part of the hemithorax to evacuate the air, and another should be placed in the inferior part of the hemithorax to remove the blood. Iatrogenic pneumothorax is a type of traumatic pneumothorax which is becoming more common. The leading causes are transthoracic needle aspiration, thoracentesis, and the insertion of central intravenous catheters. The treatment differs according to the degree of distress and can be observation, supplemental oxygen, aspiration, or tube thoracostomy. Tension Pneumothorax This condition usually occurs during mechanical ventilation or resuscitative efforts. The positive pleural pressure is life-threatening both because ventilation is severely compromised and because the positive pressure is transmitted to the mediastinum, which results in decreased venous return to the heart and reduced cardiac output. Difficulty in ventilation during resuscitation or high peak inspiratory pressures during mechanical ventilation strongly suggest the diagnosis. The diagnosis is made by the finding of an enlarged hemithorax with no breath sounds and shift of the mediastinum to the contralateral side. Tension pneumothorax must be treated as a medical emergency. If the tension in the pleural space is not relieved, the patient is likely to die from inadequate cardiac output or marked hypoxemia. A large-bore needle should be inserted into the pleural space through the second anterior intercostal space. If large amounts of gas escape from the needle after insertion, the diagnosis is confirmed. The needle should be left in place until a thoracostomy tube can be inserted. DISORDERS OF THE MEDIASTINUM The mediastinum is the region between the pleural sacs. It is separated into three compartments. The anterior mediastinum extends from the sternum anteriorly to the pericardium and brachiocephalic vessels posteriorly. It contains the thymus gland; the anterior mediastinal lymph nodes; and the internal mammary arteries and veins. The middle mediastinum lies between the anterior and posterior mediastina and contains the heart; the ascending and transverse arches of the aorta; the venae cavae; the brachiocephalic arteries and veins; the phrenic nerves; the trachea, main bronchi, and their contiguous lymph nodes; and the pulmonary arteries and veins. The posterior mediastinum is bounded by the pericardium and trachea anteriorly and the vertebral column posteriorly. It contains the descending thoracic aorta; esophagus; thoracic duct; azygos and hemiazygos veins; and the posterior group of mediastinal lymph nodes. MEDIASTINAL MASSES The first step in evaluating a mediastinal mass lesion is to place it in one of the three mediastinal compartments, since each has different characteristic lesions. The most common lesions in the anterior mediastinum are thymomas, lymphomas, teratomatous neoplasms, and thyroid masses. The most common masses in the middle mediastinum are vascular masses, lymph node enlargement from metastases or granulomatous disease, and pleuropericardial and bronchogenic cysts. In the posterior mediastinum, neurogenic tumors, meningoceles, meningomyeloceles, gastroenteric cysts, and esophageal diverticula are commonly found.
CTscanning is the most valuable imaging technique for evaluating mediastinal masses and is the only imaging technique that should be done in most instances. Barium studies of the gastrointestinal tract are indicated in many patients with posterior mediastinal lesions, since hernias, diverticula, and achalasia are readily diagnosed in this manner. An131I nuclear medicine scan can efficiently establish the diagnosis of intrathoracic goiter. A definite diagnosis can be obtained with mediastinoscopy or anterior mediastinotomy in many patients with masses in the anterior or middle mediastinal compartments. A diagnosis can be established without thoracotomy via percutaneous fine-needle aspiration biopsy of mediastinal masses in any of the mediastinal compartments. In many cases the diagnosis can be established and the mediastinal mass removed with video-assisted thoracoscopy. ACUTE MEDIASTINITIS Most cases of acute mediastinitis are either due to esophageal perforation or occur after median sternotomy for cardiac surgery. Patients with esophageal rupture are acutely ill with chest pain and dyspnea due to the mediastinal infection. The esophageal rupture can occur spontaneously or as a complication of esophagoscopy or the insertion of a Blakemore tube. Appropriate treatment is exploration of the mediastinum with primary repair of the esophageal tear and drainage of the pleural space and the mediastinum. The incidence of mediastinitis following median sternotomy is 0.4 to 5.0%. Patients most commonly present with wound drainage. Other presentations include sepsis or a widened mediastinum. The diagnosis is usually established with mediastinal needle aspiration. Treatment includes immediate drainage, debridement, and parenteral antibiotic therapy, but the mortality still exceeds 20%. CHRONIC MEDIASTINITIS The spectrum of chronic mediastinitis ranges from granulomatous inflammation of the lymph nodes in the mediastinum to fibrosing mediastinitis. Most cases are due to tuberculosis or histoplasmosis, but sarcoidosis, silicosis, and other fungal diseases are at times causative. Patients with granulomatous mediastinitis are usually asymptomatic. Those with fibrosing mediastinitis usually have signs of compression of some mediastinal structure such as the superior vena cava or large airways, phrenic or recurrent laryngeal nerve paralysis, or obstruction of the pulmonary artery or proximal pulmonary veins. Other than antituberculous therapy for tuberculous mediastinitis, no medical or surgical therapy has been demonstrated to be effective for mediastinal fibrosis. PNEUMOMEDIASTINUM In this condition, there is gas in the interstices of the mediastinum. The three main causes are: (1) alveolar rupture with dissection of air into the mediastinum; (2) perforation or rupture of the esophagus, trachea, or main bronchi; and (3) dissection of air from the neck or the abdomen into the mediastinum. Typically, there is severe substernal chest pain with or without radiation into the neck and arms. The physical
examination usually reveals subcutaneous emphysema in the suprasternal notch and Hamman's sign, which is a crunching or clicking noise synchronous with the heartbeat and best heard in the left lateral decubitus position. The diagnosis is confirmed with the chest radiograph. Usually no treatment is required, but the mediastinal air will be absorbed faster if the patient inspires high concentrations of oxygen. If mediastinal structures are compressed, the compression can be relieved with needle aspiration. DISORDERS OF THE DIAPHRAGM DIAPHRAGMATIC PARALYSIS The presence of bilateral diaphragmatic paralysis almost always causes severe morbidity in adults. The most common causes include high spinal cord injury, thoracic trauma (including cardiac surgery), multiple sclerosis, anterior horn disease, and muscular dystrophy. Most patients with severe diaphragmatic weakness present with hypercapnic respiratory failure, frequently complicated by cor pulmonale and right ventricular failure, atelectasis, and pneumonia. The degree of diaphragmatic weakness is best quantitated by measuring transdiaphragmatic pressures. The treatment of choice is assisted ventilation for all or part of each day. This is best accomplished without tracheostomy using nasal intermittent positive airway pressure. If the nerve to the diaphragm is intact, diaphragmatic pacing may be a viable alternative. If the paralysis occurs during open heart surgery, recovery frequently occurs, but it may take 6 months or more. Unilateral paralysis of the diaphragm is much more common than is bilateral paralysis. The most common cause is nerve invasion from malignancy, usually a bronchogenic carcinoma. If the patient does not have malignancy, then usually no cause for the paralysis is found. The diagnosis is suggested by finding an elevated hemidiaphragm on the chest roentgenogram. Confirmation is best established with the "sniff test." When a patient is observed with fluoroscopy while sniffing, the paralyzed diaphragm will move paradoxically upward due to the negative intrathoracic pressure. Patients with a unilateral paralyzed diaphragm are usually asymptomatic. Their vital capacity and total lung capacity are each reduced about 25%. If a patient has a mediastinal mass in conjunction with the diaphragmatic paralysis, further workup should be done. However, if the patient is asymptomatic with a normal chest radiograph, no invasive procedures are warranted. (Bibliography omitted in Palm version) Back to Table of Contents
263. DISORDERS OF VENTILATION - Eliot A. Phillipson HYPOVENTILATION DEFINITION AND ETIOLOGY Alveolar hypoventilation exists by definition when arterial PCO2(PaCO2) increases above the normal range of 37 to 43 mmHg, but in clinically important hypoventilation syndromes PaCO2 is generally in the range of 50 to 80 mmHg. Hypoventilation disorders can be acute or chronic. The acute disorders, which represent life-threatening emergencies, are discussed in Chap. 265; thischapterdeals with chronic hypoventilation syndromes. Chronic hypoventilation can result from numerous disease entities (Table 263-1), but in all cases the underlying mechanism involves a defect in either the metabolic respiratory control system, the respiratory neuromuscular system, or the ventilatory apparatus. Disorders associated with impaired respiratory drive, defects in the respiratory neuromuscular system, some chest wall disorders such as obesity, and upper airway obstruction produce an increase in PaCO2, despite normal lungs, because of a reduction in overall minute volume of ventilation and hence in alveolar ventilation. In contrast, most disorders of the chest wall and disorders of the lower airways and lungs may produce an increase in PaCO2, despite a normal or even increased minute volume of ventilation, because of severe ventilation-perfusion mismatching that results in net alveolar hypoventilation. Several hypoventilation syndromes involve combined disturbances in two elements of the respiratory system. For example, patients with chronic obstructive pulmonary disease may hypoventilate not simply because of impaired ventilatory mechanics but also because of a reduced central respiratory drive, which can be inherent or secondary to a coexisting metabolic alkalosis (related to diuretic and steroid therapy). PHYSIOLOGIC AND CLINICAL FEATURES Regardless of cause, the hallmark of all alveolar hypoventilation syndromes is an increase in alveolar PCO2(PACO2) and therefore in PaCO2(Fig. 263-1). The resulting respiratory acidosis eventually leads to a compensatory increase in plasma HCO3-concentration and a decrease in Cl- concentration. The increase in PACO2produces an obligatory decrease in PAO2, resulting in hypoxemia. If severe, the hypoxemia manifests clinically as cyanosis and can stimulate erythropoiesis and induce secondary polycythemia. The combination of chronic hypoxemia and hypercapnia may also induce pulmonary vasoconstriction, leading eventually to pulmonary hypertension, right ventricular hypertrophy, and congestive heart failure. The disturbances in arterial blood gases are typically magnified during sleep because of a further reduction in central respiratory drive. The resulting increased nocturnal hypercapnia may cause cerebral vasodilation leading to morning headache; sleep quality may also be severely impaired, resulting in morning fatigue, daytime somnolence, mental confusion, and intellectual impairment. Other clinical features associated with hypoventilation syndromes are related to the specific underlying disease (Table 263-1).
DIAGNOSIS Investigation of the patient with chronic hypoventilation involves several laboratory tests that will usually localize the disorder to either the metabolic respiratory control system, the neuromuscular system, or the ventilatory apparatus (Fig. 263-2). Defects in the control system impair responses to chemical stimuli, including ventilatory, occlusion pressure, and diaphragmatic electromyographic (EMG) responses. During sleep, hypoventilation is usually more marked, and central apneas and hypopneas are common. However, because the behavioral respiratory control system (which is anatomically distinct from the metabolic control system), the neuromuscular system, and the ventilatory apparatus are intact, such patients can usually hyperventilate voluntarily, generate normal inspiratory and expiratory muscle pressures (PImax, PEmax, respectively) against an occluded airway, generate normal lung volumes and flow rates on routine spirometry, and have normal respiratory system resistance and compliance and a normal alveolar-arterial PO2 [(A-a)PO2] difference. Patients with defects in the respiratory neuromuscular system also have impaired responses to chemical stimuli but in addition are unable to hyperventilate voluntarily or to generate normal static respiratory muscle pressures, lung volumes, and flow rates. However, at least in the early stages of the disease, the resistance and compliance of the respiratory system and the alveolar-arterial oxygen difference are normal. In contrast to patients with disorders of the respiratory control or neuromuscular systems, patients with disorders of the chest wall, lungs, and airways typically demonstrate abnormalities of respiratory system resistance and compliance and have a widened (A-a)PO2 . Because of the impaired mechanics of breathing, routine spirometric tests are abnormal, as is the ventilatory response to chemical stimuli. However, because the neuromuscular system is intact, tests that are independent of resistance and compliance are usually normal, including tests of respiratory muscle strength and of respiratory control that do not involve airflow. TREATMENT The management of chronic hypoventilation must be individualized to the patient's particular disorder, circumstances, and needs and should include measures directed toward the underlying disease. Coexistent metabolic alkalosis should be corrected, including elevations of HCO3-that are inappropriately high for the degree of chronic hypercapnia. Administration of supplemental oxygen is effective in attenuating hypoxemia, polycythemia, and pulmonary hypertension, but can aggravate CO2retention and the associated neurologic symptoms. For this reason, supplemental oxygen must be prescribed judiciously and the results monitored carefully. Pharmacologic agents that stimulate respiration (particularly progesterone) are of benefit in some patients, but generally, results are disappointing. Most patients with chronic hypoventilation related to impairment of respiratory drive or neuromuscular disease eventually require mechanical ventilatory assistance for effective management. When hypoventilation is severe, treatment may be required on a 24-h basis, but in most patients ventilatory assistance only during sleep produces dramatic clinical improvement and lowering of daytime PaCO2. In patients with reduced respiratory drive but intact respiratory lower motor neurons, phrenic nerves, and
respiratory muscles, diaphragmatic pacing through an implanted phrenic electrode can be very effective. However, for patients with defects in the respiratory nerves and muscles, electrophrenic pacing is contraindicated. Such patients can usually be managed effectively with either intermittent negative-pressure ventilation in a cuirass or intermittent positive-pressure ventilation delivered through a tracheostomy or nose mask. For patients who require ventilatory assistance only during sleep, positive-pressure ventilation through a nose mask is the preferred method because it obviates a tracheostomy and avoids the problem of upper airway occlusion that can arise in a negative-pressure ventilator. Hypoventilation related to restrictive disorders of the chest wall (Table 263-1) can also be managed effectively with nocturnal intermittent positive-pressure ventilation through a nose mask or tracheostomy. HYPOVENTILATION SYNDROMES PRIMARY ALVEOLAR HYPOVENTILATION Primary alveolar hypoventilation (PAH) is a disorder of unknown cause characterized by chronic hypercapnia and hypoxemia in the absence of identifiable neuromuscular disease or mechanical ventilatory impairment. The disorder is thought to arise from a defect in the metabolic respiratory control system, but few neuropathologic studies have been reported in such patients. Recent studies in animals suggest an important role for genetic factors in the pathogenesis of hypoventilation. Isolated PAH is relatively rare, and although it occurs in all age groups, the majority of reported cases have been in males aged 20 to 50 years. The disorder typically develops insidiously and often first comes to attention when severe respiratory depression follows administration of standard doses of sedatives or anesthetics. As the degree of hypoventilation increases, patients typically develop lethargy, fatigue, daytime somnolence, disturbed sleep, and morning headaches; eventually cyanosis, polycythemia, pulmonary hypertension, and congestive heart failure occur (Fig. 263-1). Despite severe arterial blood gas derangements, dyspnea is uncommon, presumably because of impaired chemoreception and ventilatory drive. If left untreated, PAH is usually progressive over a period of months to years and ultimately fatal. The key diagnostic finding inPAH is a chronic respiratory acidosis in the absence of respiratory muscle weakness or impaired ventilatory mechanics (Fig. 263-2). Because patients can hyperventilate voluntarily and reduce PaCO2 to normal or even hypocapnic levels, hypercapnia may not be demonstrable in a single arterial blood sample, but the presence of an elevated plasma HCO3-level should draw attention to the underlying chronic disturbance. Despite normal ventilatory mechanics and respiratory muscle strength, ventilatory responses to chemical stimuli are reduced or absent (Fig. 263-2), and breath-holding time may be markedly prolonged without any sensation of dyspnea. Patients withPAHmaintain rhythmic respiration when awake, although the level of ventilation is below normal. However, during sleep, when breathing is critically dependent on the metabolic control system, there is typically a further deterioration in ventilation with frequent episodes of central hypopnea or apnea. PAHmust be distinguished from other central hypoventilationsyndromes that are secondary to underlying neurologic disease of the brainstem or chemoreceptors (Table
263-1). This distinction requires a careful neurologic investigation for evidence of brainstem or autonomic disturbances. Unrecognized respiratory neuromuscular disorders, particularly those that produce diaphragmatic weakness, are often misdiagnosed as PAH. However, such disorders can usually be suspected on clinical grounds (see below) and can be confirmed by the finding of reduced voluntary hyperventilation, as well as PImaxand PEmax. Some patients withPAHrespond favorably to respiratory stimulant medications and to supplemental oxygen. However, the majority eventually require mechanical ventilatory assistance. Excellent long-term benefits can be achieved with diaphragmatic pacing by electrophrenic stimulation or with negative- or positive-pressure mechanical ventilation. The administration of such treatment only during sleep is sufficient in most patients. RESPIRATORY NEUROMUSCULAR DISORDERS Several primary disorders of the spinal cord, peripheral respiratory nerves, and respiratory muscles produce a chronic hypoventilation syndrome (Table 263-1). Hypoventilation usually develops gradually over a period of months to years and often first comes to attention when a relatively trivial increase in mechanical ventilatory load (such as mild airways obstruction) produces severe respiratory failure. In some of the disorders (such as motor neuron disease, myasthenia gravis, and muscular dystrophy), involvement of the respiratory nerves or muscles is usually a later feature of a more widespread disease. In other disorders, respiratory involvement can be an early or even isolated feature, and hence the underlying problem is often not suspected. Included in this category are the postpolio syndrome (a form of chronic respiratory insufficiency that develops 20 to 30 years following recovery from poliomyelitis), the myopathy associated with adult acid maltase deficiency, and idiopathic diaphragmatic paralysis. Generally, respiratory neuromuscular disorders do not result in chronic hypoventilation unless there is significant weakness of the diaphragm. Distinguishing features of bilateral diaphragmatic weakness include orthopnea, paradoxical movement of the abdomen in the supine posture, and paradoxical diaphragmatic movement under fluoroscopy. However, the absence of these features does not exclude diaphragmatic weakness. Important laboratory features are a rapid deterioration of ventilation during a maximum voluntary ventilation maneuver and reduced PImaxand PEmax(Fig. 263-2). More sophisticated investigations reveal reduced or absent transdiaphragmatic pressures, calculated from simultaneous measurement of esophageal and gastric pressures; reduced diaphragmaticEMGresponses (recorded from an esophageal electrode) to transcutaneous phrenic nerve stimulation; and marked hypopnea and arterial oxygen desaturation during rapid eye movement sleep, when there is normally a physiologic inhibition of all nondiaphragmatic respiratory muscles and breathing becomes critically dependent on diaphragmatic activity. The management of chronic alveolar hypoventilation due to respiratory neuromuscular disease involves treatment of the underlying disorder, where feasible, and mechanical ventilatory assistance as described for the primary alveolar hypoventilation syndrome. However, electrophrenic diaphragmatic pacing is contraindicated in these disorders, except for high cervical spinal cord lesions in which the phrenic lower motor neurons and nerves are intact.
OBESITY-HYPOVENTILATION SYNDROME Massive obesity represents a mechanical load to the respiratory system because the added weight on the rib cage and abdomen serves to reduce the compliance of the chest wall. As a result, the functional residual capacity (i.e., end-expiratory lung volume) is reduced, particularly in the recumbent posture. An important consequence of breathing at a low lung volume is that some airways, particularly those in the lung bases, may be closed throughout part or even all of each tidal breath, resulting in underventilation of the lung bases and widening of the (A-a)P O2. Nevertheless, in the majority of obese individuals, central respiratory drive is increased sufficiently to maintain a normal PaCO2. However, a small proportion of obese patients develop chronic hypercapnia, hypoxemia, and eventually polycythemia, pulmonary hypertension, and right-sided heart failure. Recent studies in mice demonstrate that genetically obese mice lacking circulating leptin also develop chronic hypoventilation that can be reversed by leptin infusions. Those patients who also develop daytime somnolence have been designated as having the Pickwickian syndrome (Chap. 27). In many such patients, obstructive sleep apnea is a prominent feature, and even in those patients without sleep apnea, sleep-induced hypoventilation is an important element of the disorder and contributes to its progression. Most patients demonstrate a decrease in central respiratory drive, which may be inherent or acquired, and many have mild to moderate degrees of airflow obstruction, usually related to smoking. Based on these considerations, several therapeutic measures can be of considerable benefit, including weight loss, cessation of smoking, elimination of obstructive sleep apnea, and enhancement of respiratory drive by medications such as progesterone. HYPERVENTILATION AND ITS SYNDROMES DEFINITION AND ETIOLOGY Alveolar hyperventilation exists when PaCO2decreases below the normal range of 37 to 43 mmHg. Hyperventilation is not synonymous with hyperpnea, which refers to an increased minute volume of ventilation without reference to PaCO2. Although hyperventilation is frequently associated with dyspnea, patients who are hyperventilating do not necessarily complain of shortness of breath; and conversely, patients with dyspnea need not be hyperventilating. Numerous disease entities can be associated with alveolar hyperventilation (Table 263-2), but in all cases the underlying mechanism involves an increase in respiratory drive that is mediated through either the behavioral or the metabolic respiratory control systems (Fig. 263-3). Thus hypoxemia drives ventilation by stimulating the peripheral chemoreceptors, and several pulmonary disorders and congestive heart failure drive ventilation by stimulating afferent vagal receptors in the lungs and airways. Low cardiac output and hypotension stimulate the peripheral chemoreceptors and inhibit the baroreceptors, both of which increase ventilation. Metabolic acidosis, a potent respiratory stimulant, excites both the peripheral and central chemoreceptors and increases the sensitivity of the peripheral chemoreceptors to coexistent hypoxemia. Hepatic failure can also produce hyperventilation, presumably as a result of metabolic stimuli acting on the peripheral and central chemoreceptors.
Several neurologic and psychological disorders are thought to drive ventilation through the behavioral respiratory control system. Included in this category are psychogenic or anxiety hyperventilation and severe cerebrovascular insufficiency, which may interfere with the inhibitory influence normally exerted by cortical structures on the brainstem respiratory neurons. Rarely, disorders of the midbrain and hypothalamus induce hyperventilation, and it is conceivable that fever and sepsis also cause hyperventilation through effects on these structures. Several drugs cause hyperventilation by stimulating the central or peripheral chemoreceptors or by direct action on the brainstem respiratory neurons. Chronic hyperventilation is a normal feature of pregnancy and results from the effects of progesterone and other hormones acting on the respiratory neurons. PHYSIOLOGIC AND CLINICAL FEATURES Because hyperventilation is associated with increased respiratory drive, muscle effort, and minute volume of ventilation, the most frequent symptom associated with hyperventilation is dyspnea. However, there is considerable discrepancy between the degree of hyperventilation, as measured by PaCO2, and the degree of associated dyspnea. From a physiologic standpoint, hyperventilation is beneficial in patients who are hypoxemic, because the alveolar hypocapnia is associated with an increase in alveolar and arterial PO2. Conversely, hyperventilation can also be detrimental. In particular, the alkalemia associated with hypocapnia may produce neurologic symptoms, including dizziness, visual impairment, syncope, and seizure activity (secondary to cerebral vasoconstriction); parasthesia, carpopedal spasm, and tetany (secondary to decreased free serum calcium); and muscle weakness (secondary to hypophosphatemia). Severe alkalemia can also induce cardiac arrhythmias and evidence of myocardial ischemia. Patients with a primary respiratory alkalosis are also prone to periodic breathing and central sleep apnea (Chap. 264). DIAGNOSIS In most patients with a hyperventilation syndrome, the cause is readily apparent on the basis of history, physical examination, and knowledge of coexisting medical disorders (Table 263-2). In patients in whom the cause is not clinically apparent, investigation begins with arterial blood gas analysis, which establishes the presence of alveolar hyperventilation (decreased PaCO2) and its severity. Equally important is the arterial pH, which generally allows the disorder to be classified as either a primary respiratory alkalosis (elevated pH) or a primary metabolic acidosis (decreased pH). Also of importance is the PaO2 and calculation of the (A -a)PO2 , since a widened alveolar-arterial oxygen difference suggests a pulmonary disorder as the underlying cause. The finding of a reduced plasma HCO3-level establishes the chronic nature of the disorder and points toward an organic cause. Measurements of ventilation and arterial or transcutaneous PCO2during sleep are very useful in suspected psychogenic hyperventilation, since such patients do not maintain the hyperventilation during sleep. The disorders that most frequently give rise to unexplained hyperventilation are pulmonary vascular disease (particularly chronic or recurrent thromboembolism) and psychogenic or anxiety hyperventilation. Hyperventilation due to pulmonary vascular disease is associated with exertional dyspnea, a widened (A -a)P O2and maintenance of
hyperventilation during exercise. In contrast, patients with psychogenic hyperventilation typically complain of dyspnea at rest and not during mild exercise and of the need to sigh frequently. They are also more likely to complain of dizziness, sweating, palpitations, and paresthesia. During mild to moderate exercise, their hyperventilation tends to disappear and (A-a)PO2 is normal, but heart rate and cardiac output may be increased relative to metabolic rate. TREATMENT Alveolar hyperventilation is usually of relatively minor clinical consequence and therefore is generally managed by appropriate treatment of the underlying cause. In the few patients in whom alkalemia is thought to be inducing significant cerebral vasoconstriction, paresthesia, tetany, or cardiac disturbances, inhalation of a low concentration of CO2 can be very beneficial. For patients with disabling psychogenic hyperventilation, careful explanation of the basis of their symptoms can be reassuring and is often sufficient. Others have benefited fromb-adrenergic antagonists or an exercise program. Specific treatment for anxiety may also be indicated. (Bibliography omitted in Palm version) Back to Table of Contents
264. SLEEP APNEA - Eliot A. Phillipson DEFINITION AND CLASSIFICATION Sleep apnea is defined as an intermittent cessation of airflow at the nose and mouth during sleep. By convention, apneas of at least 10 s duration have been considered important, but in most patients the apneas are 20 to 30 s in duration and may be as long as 2 to 3 min. Sleep apnea syndrome refers to a clinical disorder that arises from recurrent apneas during sleep. The clinical importance of sleep apnea arises from the fact that it is one of the leading causes of excessive daytime sleepiness. Indeed, epidemiologic studies have established a prevalence of clinically important sleep apnea of at least 2% in middle-aged women and 4% in middle-aged men. Sleep apneas can be central or obstructive in type. In central sleep apnea (CSA) the neural drive to all the respiratory muscles is transiently abolished. In contrast, in obstructive sleep apnea (OSA) airflow ceases despite continuing respiratory drive because of occlusion of the oropharyngeal airway. OBSTRUCTIVE SLEEP APNEA Pathogenesis The definitive event inOSA is occlusion of the upper airway usually at the level of the oropharynx. The resulting apnea leads to progressive asphyxia until there is a brief arousal from sleep, whereupon airway patency is restored and airflow resumes. The patient then returns to sleep, and the sequence of events is repeated, often up to 400 to 500 times per night, resulting in marked fragmentation of sleep. The immediate factor leading to collapse of the upper airway inOSA is the generation of a critical subatmospheric pressure during inspiration that exceeds the ability of the airway dilator and abductor muscles to maintain airway stability. During wakefulness, upper airway muscle activity is greater than normal in patients with OSA, presumably to compensate for airway narrowing (see below) and a high upper airway resistance. Sleep plays a permissive but crucial role by reducing the activity of the muscles and their protective reflex response to subatmospheric airway pressures. Alcohol is frequently an important cofactor because of its selective depressant influence on the upper airway muscles and on the arousal response that terminates each apnea. In most patients the patency of the airway is also compromised structurally and therefore predisposed to occlusion. In a minority of patients the structural compromise is due to obvious anatomic disturbances, such as adenotonsillar hypertrophy, retrognathia, and macroglossia. However, in the majority of patients the structural defect is simply a subtle reduction in airway size that can often be appreciated clinically as "pharyngeal crowding" and that can usually be demonstrated by imaging and acoustic reflection techniques. Obesity frequently contributes to the reduction in size of the upper airways, either by increasing fat deposition in the soft tissues of the pharynx or by compressing the pharynx by superficial fat masses in the neck. More sophisticated studies also demonstrate a high airway compliance -- i.e., the airway is "floppy" and therefore prone to collapse. Pathophysiologic and Clinical Features The narrowing of the upper airways during sleep, which predisposes to OSA, inevitably results in snoring. In most patients, snoring
antedates the development of obstructive events by many years. However, the majority of snoring individuals do not have an OSA disorder, nor is there definitive evidence that snoring per se is associated with long-term health risks. Hence, in the absence of other symptoms, snoring alone does not warrant an investigation for OSA but does call for preventive counselling, particularly with regard to weight gain and alcohol consumption. The recurrent episodes of nocturnal asphyxia and of arousal from sleep that characterizeOSA lead to a series of secondary physiologic events, which in turn give rise in some patients to the clinical complications of the syndrome (Fig. 264-1). The most common manifestations are neuropsychiatric and behavioral disturbances that are thought to arise from the fragmentation of sleep and loss of slow-wave sleep induced by the recurrent arousal responses. Nocturnal cerebral hypoxia may also play an important role. The most pervasive manifestation is excessive daytime sleepiness. Initially, daytime sleepiness manifests under passive conditions, such as reading or watching television; but as the disorder progresses, sleepiness encroaches into all daily activities and can become disabling and dangerous. Several studies have demonstrated two to seven times more motor vehicle accidents in patients with OSA compared with other drivers. Other related symptoms include intellectual impairment, memory loss, and personality disturbances. The other major manifestations ofOSA are cardiorespiratory in nature and are thought to arise from the recurrent episodes of nocturnal asphyxia and of negative intrathoracic pressure, which increases left ventricular afterload (Fig. 264-1). Many patients demonstrate a cyclical slowing of the heart during the apneas to 30 to 50 beats per minute, followed by a tachycardia of 90 to 120 beats per minute during the ventilatory phase. A small number of patients develop severe bradycardia or dangerous tachyarrhythmias, leading to the notion that OSA may result in sudden death during sleep, but firm corroborative data are lacking. Unlike in healthy subjects, in patients with OSA systemic blood pressure fails to decrease during sleep. In fact, blood pressure typically rises abruptly at the termination of each obstructive event as a result of sympathetic nervous activation and reflex vasoconstriction. Furthermore, over 50% of patients with OSA have systemic hypertension. Several epidemiologic studies have implicated OSA as a risk factor for the development of systemic hypertension, and recent studies in an animal model demonstrate directly that OSA can cause sustained increases in daytime blood pressure. Emerging data also suggest that OSA can precipitate myocardial ischemia in patients with coronary artery disease and can adversely affect left ventricular function, both acutely and chronically, in patients with congestive heart failure. This complication is probably due to the combined effects of increased left ventricular afterload during each obstructive event, secondary to increased negative intrathoracic pressure (Fig. 264-1), recurrent nocturnal hypoxemia, and chronically elevated sympathoadrenal activity. Treatment of OSA in such patients often results in dramatic improvement in left ventricular function and in clinical cardiac status. Finally, up to 20% of patients with OSA develop mild pulmonary hypertension (in the absence of intrinsic lung disease), and a small proportion (75% of ARDS patients. However, the turning of these critically ill patients from the supine to the prone position is not without potential complications, such as unplanned extubation and removal of central venous catheters. The term inverse ratio ventilation is defined when the inspiratory (I) time exceeds the expiratory (E) time (i.e., > one-half of the respiratory cycle; I:E ratio > 1:1). This mode of ventilation is able to maintain a higher mean airway pressure, a major determinant of oxygenation, with lower peak airway pressures than conventional ventilation. However, due to the decrease in expiratory time, inverse ratio ventilation is potentially associated with dynamic hyperinflation and increases in end-expiratory pressure. Finally, partial liquid ventilation with perfluorocarbon, a radiopaque, inert, colorless liquid that carries a large quantity of O2, and CO2, has been studied in patients with severe ARDS. When perfluorocarbon is administered into the trachea of intubated patients, patients can be safely and adequately oxygenated and ventilated with routine mechanical ventilation. Intravascular Volume Management Although pulmonary edema inARDSpatients is a consequence of increased permeability of the alveolar-capillary membrane, elevations in the intravascular hydrostatic pressure may also contribute to the accumulation of alveolar liquid and result in worsening oxygenation. Therefore, the optimal fluid management for patients with ARDS requires a balancing between liquid restriction, which may cause hypotension and decreased perfusion to vital organs, and liquid administration, which may increase oxygen requirements. Small decrements in the intravascular volume with diuretic use produce significant decreases in extravascular lung water. Caution must be exercised in reducing intravascular volume, since vigorous diuresis, especially in the setting ofPEEP, may reduce cardiac output and perfusion of critical organs. Ideally, the lowest intravascular hydrostatic pressure that also achieves an adequate cardiac output should be maintained. The placement of a pulmonary arterial catheter may be helpful in monitoring cardiac output and pulmonary arterial occlusion pressure (a measure of intravascular volume) in order to optimize the fluid management of patients with ARDS. However, the placement of a pulmonary arterial catheter and the clinical decisions based upon information derived from the catheter do not appear to improve and may actually worsen the outcome of general intensive care unit patients. Therefore the role of the pulmonary arterial catheter for ARDS patients is presently unclear. Pharmacologic Therapies Due to their anti-inflammatory properties, glucocorticoids
have been used in patients withARDS, but when administered in high doses (30 mg/kg intravenously every 6 h for a total of four doses), they are not beneficial in the early course of the disease. In contrast, one small randomized study reported an improvement in mortality when glucocorticoids were given after 7 days of unresolving ARDS. In this study, active surveillance for infection was required before enrollment, and glucocorticoids were administered for up to 32 days. Future recommendations regarding the use of these drugs for ARDS patients will be based upon the results of an ongoing multicenter study. Patients withARDShave both quantitative and qualitative abnormalities in surfactant, rendering surfactant-replacement therapy an attractive therapeutic modality. In one large randomized study of sepsis-induced ARDS, the administration of synthetic surfactant in an aerosolized form had no significant effect on outcome. Due to concerns with the efficacy of the delivery technique and the lack of essential surfactant-associated proteins in this particular replacement therapy, further studies of different surfactant preparations and modes of administration are presently ongoing. When inhaled, nitric oxide vasodilates the pulmonary vasculature adjacent to well-ventilated alveoli, thereby improving ventilation-perfusion mismatching. Because of its subsequent inactivation by hemoglobin, nitric oxide produces a selective pulmonary vasodilation without systemic hemodynamic effects. Though inhaled nitric oxide appears to improve oxygenation initially, it is presently unknown whether this therapy will reduce mortality rates inARDSpatients. PROGNOSIS Since the initial descriptions ofARDS, mortality rates have ranged from 50 to 70%, although they may now be declining with optimal therapy. Mortality rates are higher in patients over 65 years of age, in those with an at-risk diagnosis of sepsis, and when associated with dysfunction of other organ systems. The cause of death for patients with ARDS has been traditionally divided into early causes (within 72 h) and late causes (after 3 days). Most early deaths are attributed to the original presenting illness or injury. Secondary infection and sepsis, persistent respiratory failure, and multiple-organ dysfunction are the most common causes of death in those ARDS patients who live at least 3 days. In survivors ofARDS, abnormalities in pulmonary function normally improve considerably by 3 months and reach maximum levels of correction by 6 months after extubation. Although pulmonary function markedly recovers in many survivors, over 50% of these patients will continue to have abnormalities, including restrictive impairment or decreased diffusing capacity. Patients with severe ARDS, characterized by extreme hypoxemia and a longer duration of illness, usually have more pulmonary dysfunction than individuals with mild ARDS. Survivors of ARDS also have significant reductions in their quality of life, specifically in regard to physical functioning when compared to other previously critically ill patients. (Bibliography omitted in Palm version)
Back to Table of Contents
266. MECHANICAL VENTILATORY SUPPORT - Edward P. Ingenito, Jeffrey M. Drazen Ventilators are specially designed pumps that can support the ventilatory function of the respiratory system and improve oxygenation through application of high oxygen content gas and positive pressure. They are a mainstay of physiologic supportive care and are used to stabilize patients with respiratory failure as the underlying disease process is definitively treated. INDICATIONS FOR MECHANICAL VENTILATION Respiratory failure is the primary indication for initiation of mechanical ventilation. There are two basic types of respiratory failure. Hypoxemic respiratory failure most commonly results from pulmonary conditions such as severe pneumonia, pulmonary edema, pulmonary hemorrhage, and respiratory distress syndrome causing ventilation-perfusion (/) mismatch and shunt. Hypoxemic respiratory failure is present when arterial O2saturation (SaO2)< 90% is observed despite an inspired O2fraction (FIO2)> 0.6. The goal of ventilator treatment in this setting is to provide adequate SaO2through a combination of supplemental O2 and specific patterns of ventilation that enhance oxygenation. Hypercarbic respiratory failure results from disease states causing either a decrease in minute ventilation or an increase in physiologic dead space such that, despite adequate total minute ventilation, alveolar ventilation is inadequate to meet metabolic demands. Common clinical conditions associated with hypercarbic respiratory failure include neuromuscular diseases, such as myasthenia gravis, ascending polyradiculopathy, and myopathies, as well as diseases that cause respiratory muscle fatigue due to increased workload, such as asthma, chronic obstructive pulmonary disease, and restrictive lung disease. Acute hypercarbic respiratory failure is characterized by arterial PCO2values of greater than 50 mmHg and an arterial pH above 7.30. Mechanical ventilation generally should be instituted in acute hypercarbic respiratory failure. In contrast, the decision to institute mechanical ventilation when components of both acute and chronic hypercarbic respiratory failure are present depends on blood gas parameters and clinical evaluation. In particular, if a patient is not in respiratory distress and is not mentally impaired by CO2accumulation, it is not mandatory to initiate mechanical ventilation while other forms of treatment are being administered. The goal of ventilator treatment in hypercarbic respiratory failure is to normalize arterial pH through changes in CO2tensions. In patients with severe obstructive or restrictive lung disease, elevation in airway pressures may limit tidal volumes to the extent that normalization of pH is not possible, a situation known as permissive hypercapnia. Hypoxemic and hypercarbic respiratory failure may coexist in a given individual; in such cases, the indications for and goals of mechanical ventilation are similar to those in these two individual entities. Accepted therapeutic applications of mechanical ventilation include controlled hyperventilation to reduce cerebral blood flow in patients with increased intracranial pressure or to improve pulmonary hemodynamics in patients with postoperative
pulmonary hypertension. Mechanical ventilation also has been used to reduce the work of breathing in patients with congestive heart failure, especially in the presence of myocardial ischemia. Ventilator support is also frequently used in conjunction with endotracheal intubation to prevent aspiration of gastric contents in otherwise unstable patients during gastric lavage for suspected drug overdose or during upper gastrointestinal endoscopy. In the critically ill patient, intubation and mechanical ventilation are indicated before essential diagnostic or therapeutic studies if it appears that respiratory failure may occur during these maneuvers. PHYSIOLOGIC ASPECTS OF MECHANICAL VENTILATION Most modern mechanical ventilators function by providing warmed and humidified gas to the airway opening in conformance with various specific volume, pressure, and time patterns. The ventilator serves as the energy source for inspiration, replacing the muscles of the diaphragm and chest wall. Expiration is passive, driven by the recoil of the lungs and chest wall; at the completion of inspiration, internal ventilator circuitry vents the airway to atmospheric pressure or a specified level of positive end-expiratory pressure (PEEP). PEEPhelps maintain alveolar patency in the presence of destabilizing factors and therefore reverses hypoxemia and atelectasis by improving/)">/matching of ventilation and perfusion. PEEP levels between 0 and 10 cmH2O are generally safe and effective; higher levels are recommended only in the management of significant refractory hypoxemia unresponsive to increments in FIO2 up to 0.6. ESTABLISHING AND MAINTAINING AN AIRWAY A cuffed endotracheal tube must be inserted to allow positive-pressure ventilators to deliver conditioned gas, at pressures above atmospheric pressure, to the lungs in a controlled fashion. If neuromuscular paralysis is to be induced during intubation, the use of agents whose mechanism of action includes depolarization at the neuromuscular junction, such as succinylcholine chloride, should be avoided in patients with renal failure, tumor lysis syndrome, crush injuries, medical conditions associated with elevated serum potassium levels, and muscular dystrophy syndromes. Opiates and benzodiazepines can have a deleterious effect on hemodynamics in patients with depressed cardiac function or low systemic vascular resistance and should be used cautiously in this setting. Morphine can promote histamine release from tissue mast cells and may worsen bronchospasm in patients with asthma; fentanyl, sufentanil, and alfentanil are acceptable alternatives to morphine. Ketamine may increase systemic arterial pressure as well as intracranial pressure and has been associated with dramatic hallucinatory responses; it should be used with caution in patients with hypertensive crisis, increased intracranial pressures, or a history of psychiatric disorders. Patients who require ventilator support for extended periods of time may be candidates for tracheostomy. Although definitive guidelines for performing a tracheostomy in the ventilated patient have not been established, in current clinical practice patients who are anticipated to require ventilator therapy for more than 3 weeks should be considered for this procedure. While it does not clearly reduce the incidence of laryngeal injury or tracheal stenosis, tracheostomy has been associated with improved patient comfort and
enhanced ability to partake in rehabilitation-oriented activities. VENTILATOR MODES This setting specifies the manner in which ventilator breaths are triggered, cycled, and limited; commonly used modes of mechanical ventilation are given inTable 266-1. The trigger, either an inspiratory effort or a time-based signal, defines what the ventilator senses to initiate an assisted cycle. Cycle refers to the factors that determine the end of inspiration. For example, in volume-cycled ventilation, inspiration ends when a specific tidal volume is delivered to the patient. Other types of cycling include pressure cycling, time cycling, and flow cycling. Limiting factors are operator-specified values, such as airway pressure, that are monitored by transducers internal to the ventilator circuit throughout the respiratory cycle; if the specified values are exceeded, inspiratory flow is immediately stopped, and the ventilator circuit is vented to atmospheric pressure or the specifiedPEEP. Assist Control Mode Ventilation (ACMV) An inspiratory cycle is initiated either by the patient's inspiratory effort or, if no patient effort is detected within a specified time window, by a timer signal within the ventilator. Every breath delivered consists of the operator-specified tidal volume. Ventilatory rate is determined either by the patient or by the operator-specified backup rate, whichever is of higher frequency (Fig. 266-1A). ACMV is the recommended mode for initiation of mechanical ventilation because it ensures a backup minute ventilation in the absence of an intact respiratory drive and allows for synchronization of the ventilator cycle with the patient's inspiratory effort. Problems can arise whenACMV is used in patients with tachypnea due to nonrespiratory or nonmetabolic factors such as anxiety, pain, or airway irritation. Respiratory alkalemia may develop and trigger myoclonus or seizures. Dynamic hyperinflation (so-called auto-PEEP) may occur if the patient's respiratory mechanics are such that inadequate time is available for complete exhalation between inspiratory cycles. Auto-PEEP can limit venous return, decrease cardiac output, and increase airway pressures, predisposing to barotrauma. ACMV is not effective for weaning patients from mechanical ventilation because it provides full ventilator assistance on each patient-initiated breath. Synchronized Intermittent Mandatory Ventilation (SIMV) The major difference between SIMV and ACMVis that in the former the patient is allowed to breathe spontaneously, i.e., without ventilator assist, between delivered ventilator breaths. However, mandatory breaths are delivered in synchrony with the patient's inspiratory efforts at a frequency determined by the operator. If the patient fails to initiate a breath, the ventilator delivers a fixed-tidal-volume breath and resets the internal timer for the next inspiratory cycle (Fig. 266-1B). SIMV differs from ACMV in that only the preset number of breaths is ventilator-assisted. SIMVallows patients with an intact respiratory drive to exercise inspiratory muscles between assisted breaths. This characteristic makes SIMV a useful mode of ventilation for both supporting and weaning intubated patients. SIMV may be difficult to use in patients with tachypnea because they may attempt to exhale during the ventilator-programmed inspiratory cycle. When this occurs, the airway pressure may
exceed the inspiratory pressure limit, the ventilator-assisted breath will be aborted, and minute volume may drop below that programmed by the operator. In this setting, if the tachypnea is in response to respiratory or metabolic acidosis, a change toACMVwill increase minute ventilation and help normalize the pH while the underlying process is further evaluated. Continuous Positive Airway Pressure (CPAP) This is not a true support-mode of ventilation, inasmuch as all ventilation occurs through the patient's spontaneous efforts. The ventilator provides fresh gas to the breathing circuit with each inspiration and charges the circuit to a constant, operator-specified pressure that can range from 0 to 20 cmH2O (Fig. 266-1C). CPAP is used to assess extubation potential in patients who have been effectively weaned and are requiring little ventilator support and in patients with intact respiratory system function who require an endotracheal tube for airway protection. Pressure-Control Ventilation (PCV) This form of ventilation is time triggered, time cycled, and pressure limited. During the inspiratory phase, a given pressure is imposed at the airway opening, and the pressure remains at this user-specified level throughout inspiration (Fig. 266-2A). Since inspiratory airway pressure is specified by the operator, tidal volume and inspiratory flow rate are dependent rather than independent variables and are not user specified. PCV is the preferred mode of ventilation for patients with documented barotrauma, because airway pressures can be limited, and for postoperative thoracic surgical patients, in whom the shear forces across a fresh suture line should be limited. When PCV is used, minute ventilation and tidal volume must be monitored; minute ventilation is altered through changes in rate or in the pressure-control value. The major practical limitation ofPCV is patient-ventilator asynchrony related to its time-cycled and time-triggered characteristics. Because PCV requires that the patient passively accept ventilator breaths, most patients require heavy sedation to be maintained on this ventilatory mode, which may be hazardous in the hemodynamically unstable patient. PCVwith the use of a prolonged inspiratory time is frequently applied to patients with severe hypoxemic respiratory failure. This approach, called inverse inspiratory-to-expiratory ratio ventilation (IRV), increases mean distending pressures without increasing peak airway pressures. It is thought to work in conjunction withPEEP to open collapsed alveoli and improve oxygenation. IRV may be associated with fewer deleterious effects than conventional volume-cycled ventilation, which requires higher peak airway pressures to achieve an equivalent reduction in shunt fraction. Pressure-Support Ventilation (PSV) This form of ventilation is patient triggered, flow cycled, and pressure limited; it is specifically designed for use in the weaning process. During PSV, the inspiratory phase is terminated when inspiratory airflow falls below a certain level; in most ventilators this flow rate cannot be adjusted by the operator. When PSV is used, patients receive ventilator assist only when the ventilator detects an inspiratory effort (Fig. 266-2B). PSV also can be used in combination withSIMV to ensure volume-cycled backup for patients whose respiratory drive is depressed either spontaneously or as a result of various therapeutic maneuvers.
PSVis well tolerated by most patients who are being weaned; PSV parameters can be set to provide fully or nearly fully ventilatory support and can be withdrawn slowly over a period of days in a systematic fashion to gradually load the respiratory muscles. Open Lung Ventilation (OLV) OLV is not a distinct mode of ventilation, but rather a strategy for applying either volume-cycled or pressure-control ventilation to patients with severe respiratory failure. In OLV, the primary objectives of ventilator support are maintenance of adequate oxygenation and avoidance of cyclic opening and closing of alveolar units by selecting a level ofPEEPthat allows the majority of units to remain inflated during tidal ventilation. Achievement of eucapnia and normal blood pH through adjustments in ventilator tidal volume and breathing frequency are of lower priority. Clinical and experimental observations indicate that high airway pressures and repeated opening and closing of alveoli can cause microstructural lung damage, propagation of lung injury through generation of inflammatory cytokines, and direct barotrauma. Current data suggest that a small tidal volume (i.e., 6 mL/kg) provides adequate ventilatory support with a lower incidence of adverse effects than more conventional tidal volumes of 10-15 mL/kg. These potential complications can have dire consequences in patients with respiratory failure. Alternatively, hypercapnia and consequent respiratory acidosis tend to be well tolerated physiologically, except in patients with significant hemodynamic compromise, ventricular dysfunction, cardiac dysrrhythmias, or increased intracranial pressure. OLV has been used most extensively in the management of patients with hypoxemic respiratory failure due to acute lung injury. Although few randomized clinical trials of OLV have been performed, available data suggest that OLV reduces the morality rate and improves gas exhange in patients with acute lung injury. Prone Positioning during Mechanical Ventilation Patients with acute respiratory distress syndrome (ARDS) experience hypoxemia as a result of intrapulmonary shunt due to regional atelectasis. Recent studies in patients with ARDS have demonstrated that collapse occurs most extensively in the dependent regions of the lung. Increasing airway pressures to counterbalance the compressive effects of the surrounding lung in these collapsed regions improves gas exchange but may result in potentially dangerous peak airway pressures. Prone positioning, in both experimental and clinical studies, reduces shunt and improves oxygenation by causing regional improvements in transpulmonary distending pressures wtihout overexpanding already patent alveoli. In clinical practice, prone positioning has been used in conjunction with both volume-cycled and pressure-control ventilation with equivalent clinical effectiveness and appears to be a useful adjunct to conventional ventilator support in patients with severe hypoxemic respiratory failure. Noninvasive Ventilation (NIV) Noninvasive ventilator support through a tight-fitting facemask or nasal mask, traditionally used for treatment of sleep apnea, has recently been used as primary ventilator support in patients with impending respiratory failure. Facemask and nasal devices for administering NIV therapy are most frequently combined withPSV or bi-level positive airway pressure ventilation, inasmuch as both of these modes are well tolerated by the conscious patient and optimize patient-ventilator synchrony. NIV has met with varying degrees of success when applied to patients with acute or chronic respiratory failure. The major limitation to its widespread application has been patient intolerance, because the tight-fitting mask required for NIV can cause
both physical and emotional discomfort in patients with dyspnea. In general, centers with experience using NIV have reported clinical success with minimal associated morbidity, whereas centers with less experience have reported more limited success. Aggressive medical therapy directed at the cause of impending respiratory failure, together with an experienced respiratory therapy and physician team, appear to be the keys to successful use of NIV in intensive care units. Extracorporeal Membrane Oxygenation (ECMO) This nonconventional mode of ventilator support employs a large surface area membrane system connected in series with the patient's circulation to exchange CO2 and O2. The lung functions primarily as a passive conduit with gas exchange occurring by diffusion across the membrane. ECMO was first examined in 1970 as an alternative to positive-pressure ventilation in the management of patients withARDS. Initial studies failed to demonstrate an improvement in survival rates among patients treated with ECMO. Although several uncontrolled trials have since suggested that ECMO does improve outcome among patients with ARDS, a 1993 study comparing survival rates of patients with ARDS treated with ECMO and those treated with conventional ventilator therapy showed no difference in mortality rates, but the morbidity rates and hospital costs were increased among ECMO-treated patients. Presently, the use of ECMO in patients with ARDS is not recommended. GUIDELINES FOR MANAGING THE VENTILATED PATIENT Most patients who are started on ventilator support receiveACMV orSIMV, because these modes ensure user-specified backup minute ventilation in the event that the patient fails to initiate respiratory efforts. Once the intubated patient has been stabilized with respect to oxygenation, definitive therapy for the underlying process responsible for respiratory failure is formulated and initiated. Subsequent modifications in ventilator therapy must be provided in parallel with changes in the patient's clinical status. As improvement in respiratory function is noted, the first priorities are to reducePEEP and supplemental O2. Once a patient can achieve adequate arterial saturation with an FIO2£ 0.5 and 5 cmH2O PEEP, attempts should be made to reduce the level of mechanical ventilatory support. Patients previously on full ventilator support should be switched to a ventilator mode that allows for weaning, such as SIMV,PSV, or SIMV combined with PSV. Ventilator therapy can then be gradually removed, as outlined in the section on weaning. Patients whose condition continues to deteriorate after ventilator support is initiated may require increased O2, PEEP, and alternative modes of ventilation such asIRV. GENERAL SUPPORT IN THE VENTILATED PATIENT Patients who are started on mechanical ventilation usually require some form of sedation and analgesia to maintain an acceptable level of comfort. Often, this regimen consists of a combination of a benzodiazepine and opiate administered intravenously. Medications commonly used for this purpose include lorazepam, midazolam, diazepam, morphine, and fentanyl. Immobilized patients in the intensive care unit on mechanical ventilator support are at increased risk for deep venous thrombosis; accepted practice consists of administering prophylaxis in the form of subcutaneous heparin and/or pneumatic compression boots.
Fractionated low molecular weight heparin has also been used for this purpose; it appears to be equally effective and is associated with a decreased incidence of heparin-associated thrombocytopenia. Prophylaxis against diffuse gastrointestinal mucosal injury is indicated for patients who have suffered a neurologic insult or those with severe respiratory failure in association withARDS. Histamine receptor antagonists (H2-receptor antagonists), antacids, and cytoprotective agents such as carafate have all been used for this purpose and appear to be effective. Recent data suggest that carafate use is associated with a reduction in the incidence of nosocomial pneumonias, since it does not cause changes in stomach pH and is less likely to permit colonization of the gastrointestinal tract by nosocomial organisms at pH levels near neutral. Nutrition support by enteral feeding through either a nasogastric or an orogastric tube should be maintained in all intubated patients whenever possible. In those patients with a normal baseline nutritional state, support should be initiated within 7 days. In malnourished patients, nutrition support should be initiated within 72 h. Delayed gastric emptying is common in critically ill patients on sedative medications but often responds to promotility agents such as cisapride or metoclopramide. Parenteral nutrition is an alternative to enteral nutrition in patients with severe gastrointestinal pathology. COMPLICATIONS OF MECHANICAL VENTILATION Endotracheal intubation and positive-pressure mechanical ventilation have direct and indirect effects on several organ systems, including the lung and upper airways, the cardiovascular system, and the gastrointestinal system. Pulmonary complications include barotrauma, nosocomial pneumonia, oxygen toxicity, tracheal stenosis, and deconditioning of respiratory muscles. Barotrauma, which occurs when high pressures (i.e., > 50 cmH2O) disrupt lung tissue, is clinically manifest by interstitial emphysema, pneumomediastinum, subcutaneous emphysema, or pneumothorax. Although the first three conditions may resolve simply through the reduction of airway pressures, clinically significant pneumothorax, as indicated by hypoxemia, decreased lung compliance, and hemodynamic compromise, requires tube thoracostomy. Patients intubated for longer than 72 h are at high risk for nosocomial pneumonia as a result of aspiration from the upper airways through small leaks around the endotracheal tube cuff; the most common organisms responsible for this condition are enteric gram-negative rods, Staphylococcus aureus, and anaerobic bacteria. Because the endotracheal tube and upper airways of patients on mechanical ventilation are commonly colonized with bacteria, the diagnosis of nosocomial pneumonia requires "protected brush" bronchoscopic sampling of airway secretions coupled with quantitative microbiologic techniques to differentiate colonization from infection. Oxygen toxicity is a potential complication when an FIO2 ³ 0.6 is required for more than 72 h. The condition can be prevented in some cases through the use ofPEEP to allow for FIO2values to go below 0.6 while primary therapy for the underlying condition is instituted. Although O2toxicity is thought to result from the effects of oxygen free radical on the lung interstitium, the therapeutic use of antioxidants such as superoxide dismutase, catalase, selenium, and vitamin E remains experimental.
Hypotension resulting from elevated intrathoracic pressures with decreased venous return is almost always responsive to intravascular volume repletion. In patients judged to have hypotension or respiratory failure on the basis of alveolar edema, hemodynamic monitoring with a pulmonary arterial catheter may be of value in optimizing O 2delivery via manipulation of intravascular volume and FIO2 andPEEPlevels. Gastrointestinal effects of positive-pressure ventilation include stress ulceration and mild to moderate cholestasis. It is common practice to provide prophylaxis with H2-receptor antagonists or sucralfate for stress-related ulcers. Mild cholestasis (i.e., total bilirubin values £4.0) attributable to the effects of increased intrathoracic pressures on portal vein pressures is common and generally self-limited. Cholestasis of a more severe degree should not be attributed to a positive-pressure ventilation response and is more likely due to a primary hepatic process. WEANING FROM MECHANICAL VENTILATION Removal of mechanical ventilator support requires that a number of criteria be met. Upper airway function must be intact for a patient to remain extubated but is difficult to assess in the intubated patient. Therefore, if a patient can breathe on his or her own through an endotracheal tube but develops stridor or recurrent aspiration once the tube is removed, upper airway dysfunction or an abnormal swallowing mechanism should be suspected and plans for achieving a stable airway developed. An intact cough during suctioning is a good indicator of a patient's ability to mobilize secretions. Respiratory drive and chest wall function are assessed by observation of respiratory rate, tidal volume, inspiratory pressure, and vital capacity. The weaning index, defined as the ratio of breathing frequency to tidal volume (breaths per minute per liter), is both sensitive and specific for predicting the likelihood of successful extubation. When this ratio is less than 105 with the patient breathing without mechanical assistance through an endotracheal tube, successful extubation is likely. An inspiratory pressure of more than -30 cmH2O and a vital capacity of greater than 10 mL/kg are considered indicators of acceptable chest wall and diaphragm function. Alveolar ventilation is generally adequate when elimination of CO2 is sufficient to maintain arterial pH in the range of 7.35 to 7.40, and anSaO2> 90% can be achieved with an FIO2< 0.5 and aPEEP£ 5 cmH2O. Although many patients may not meet all criteria for weaning, the likelihood that a patient will tolerate extubation without difficulty increases as more criteria are met. Many approaches to weaning patients from ventilator support have been advocated. T-piece andCPAPweaning are best tolerated by patients who have undergone mechanical ventilation for brief periods and require little respiratory muscle reconditioning, whereasSIMV andPSV are best for patients who have been intubated for extended periods and require gradual respiratory-muscle reconditioning. T-piece weaning involves brief spontaneous breathing trials with supplemental O2. These trials are usually initiated for 5 min/h followed by a 1-h interval of rest. T-piece trials are increased in 5- to 10-min increments until the patient can remain ventilator independent for periods of several hours. Extubation can then be attempted.CPAPweaning is similar to T-piece weaning except that trials of spontaneous breathing are conducted on the ventilator in CPAP mode.
Weaning by means ofSIMVinvolves gradually tapering the mandatory backup rate in increments of 2 to 4 breaths per minute while monitoring blood gas parameters and respiratory rates. Rates of greater than 25 breaths per minute on withdrawal of mandatory ventilator breaths generally indicate respiratory muscle fatigue and the need to combine periods of exercise with periods of rest. Exercise periods are gradually increased until a patient remains stable on SIMV at 4 breaths per minute or less without needing rest at higher SIMV rates. ACPAP or T-piece trial can then be attempted before planned extubation. PSV, as described in detail above, is used primarily for weaning from mechanical ventilation. PSV is usually initiated at a level adequate for full ventilator support (PSVmax); i.e., PSV is set slightly below the peak inspiratory pressures required by the patient during volume-cycled ventilation. The level of pressure support is then gradually withdrawn in increments of 5 cmH2O until a level is reached at which the respiratory rate increases to 25 breaths per minute. At this point, intermittent periods of higher-pressure support are alternated with periods of lower-pressure support to provide muscle reconditioning without causing diaphragmatic fatigue. Gradual withdrawal of PSV continues until the level of support is just adequate to overcome the resistance of the endotracheal tube (approximately 5 to 10 cmH2O). Support can be discontinued and the patient extubated. (Bibliography omitted in Palm version) Back to Table of Contents
267. LUNG TRANSPLANTATION - Janet R. Maurer Lung transplantation for end-stage lung disease has been a therapeutic option since the 1980s. Several transplant options are available for carefully selected patients: unilateral lung transplant, bilateral lung transplant, heart-lung transplant (Chap. 233), and living lobar transplant. The first successful type of lung transplant was heart-lung, which was performed for a variety of indications and in increasing numbers until 1989. Beginning in 1989, the numbers of both unilateral and bilateral lung transplants performed increased dramatically and, along with the increasing demand for heart donors, greatly reduced the number of donor organs available for heart-lung procedures. By the mid-1990s, unilateral and bilateral lung transplant numbers also had plateaued because of donor shortages and have remained relatively stable at approximately 1250 operations worldwide per year. Of these, 60% are unilateral lung transplants and 40% are bilateral. Heart-lung transplants have leveled off at between 100 and 150 per year. In its 1999 report, the Registry of the International Society for Heart and Lung Transplantation, in conjunction with the United Network for Organ Sharing, had recorded a cumulative total of 8997 isolated lung transplants and 2350 heart-lung transplants. INDICATIONS Emphysema, either smoking-induced or secondary toa1-antitrypsin deficiency, has been the single largest indication for lung transplantation. This diagnosis accounts for about 55% of unilateral lung transplants, 29% of bilateral, and 6% of heart-lung transplants. Other major indications for unilateral lung transplants include idiopathic pulmonary fibrosis (21%) and primary pulmonary hypertension (5%). Patients with cystic fibrosis comprise the largest group of bilateral lung recipients, approximately 34% of the total, followed by patients with emphysema, patients with primary pulmonary hypertension (10%), and those with idiopathic pulmonary fibrosis (7.5%). The major diagnoses among heart-lung recipients are primary and secondary pulmonary hypertension (54%) and cystic fibrosis (16%). With the widespread use of unilateral and bilateral lung transplants, the indications for heart-lung transplant have become very circumscribed, so that now most candidates for this type of transplant have either concomitant left ventricular disease and end-stage lung disease or irreparable congenital heart disease with Eisenmenger's syndrome. Patients receiving living lobar donations have been either children or young adults, and most have suffered from cystic fibrosis. In most of these operations, a lower lobe is donated from each of two adults, who are often, but not always, related to the recipient. Living donation is performed in a limited number of lung transplant programs, and the donor morbidity rate has been acceptable. RECIPIENT SELECTION Because donor lungs are the scarcest of the common solid organs transplanted, patients with end-stage lung disease undergo extensive evaluation to select the best potential candidates. In 1998, this process was further standardized with the publication of the International Guidelines for the Selection of Lung Transplant Candidates. Approximate age limits of 65 years for unilateral lung, 60 years for bilateral lung, and 55 years for heart-lung transplants were set. Chronic medical conditions that can be adequately controlled and have not resulted in end-organ damage, e.g., systemic hypertension, are acceptable in lung transplant candidates. However, in a case where a
chronic illness is often associated with nonpulmonary organ damage, e.g., diabetes mellitus, a careful assessment of target organ function is necessary. Absolute contraindications to lung transplantation include dysfunction of major organs (other than lung), infection with HIV, active malignancy within 2 years with the exception of basal cell and squamous cell skin cancer, hepatitis B antigen positivity, and hepatitis C with biopsy-proven histologic evidence of liver disease. Conditions that represent relative contraindications include symptomatic osteoporosis; severe musculoskeletal disease affecting the thorax; high-dose corticosteroid use; weight less than 70% or greater than 130% of ideal body weight; alcohol, cigarette, or narcotic abuse/addiction within 6 months before evaluation; psychosocial problems, including noncompliance, that cannot be adequately resolved through pharmacologic treatment or counseling; requirement for invasive ventilation; and colonization with fungi or atypical mycobacteria. Colonization is of particular concern when a unilateral lung transplant is being considered. Disease-specific guidelines (Table 267-1) are chosen to identify candidates who are within the transplant "window" -- that is, patients who are ill enough to fit within the category of "end-stage" and have progressive disease, yet are able to survive the pre-transplant waiting and perioperative time periods. In the past few years, increasing experience with large numbers of patients with end-stage disease has made it much easier to estimate life expectancies; however, patients with diagnoses of emphysema and Eisenmenger's type pulmonary hypertension remain problematic in this regard because posttransplant statistical analysis does not show a clear survival benefit for recipients within the first 2 years. In these types of patient, selection usually includes consideration of quality-of-life issues as well as survival rates. SELECTION OF TRANSPLANT PROCEDURE The only diseases that currently mandate a specific procedure are (1) irreparable congenital cardiac defects with Eisenmenger's syndrome (heart-lung transplant); (2) advanced lung disease with concomitant left ventricular dysfunction (heart-lung transplant); and (3) bronchiectatic lung disease, e.g., cystic fibrosis (bilateral lung transplant or bilobar living donor lung transplant). In essentially all other circumstances unilateral lung transplantation can be performed with acceptable early and midterm results. Bilateral lung transplantation, however, is often preferred if difficulty is anticipated in postoperative management, especially in patients with pulmonary hypertension; if significant bullous disease in present in emphysema; if a patient is very young; or if there are specific individual recipient considerations. As noted below, the long-term survival rates of bilateral lung recipients may be superior; nevertheless, transplant centers have generally chosen to maximize the donor organ resource by performing unilateral lung transplants whenever possible, rather than opting for potentially slightly increased survival periods. PROGNOSIS The 1- and 2-year survival rates for unilateral and bilateral lung transplant recipients are 67 and 62%, respectively. Longer term data show a divergence in survival rates by 5 years, with the half-life of bilateral transplant recipients (4.9 years) significantly longer than that of unilateral transplant recipients (3.6 years). Among unilateral graft recipients, patients with emphysema appear to have the best early survival rate (nearly 80% at one
year), and patients with idiopathic pulmonary fibrosis and those with pulmonary hypertension have the worst (60 to 65%). Living lobar recipients have early survival rates that are between the rates of these groups, but long-term data are not available for this population. FUNCTIONAL OUTCOMES Arterial blood gas levels improve markedly in unilateral and bilateral lung transplant recipients by 3 months posttransplant. In both groups, Pa CO2normalizes; in bilateral lung transplant recipients, PaO2also normalizes. Unilateral lung recipients may continue to have mild hypoxemia but rarely require supplemental oxygen. Pulmonary function studies usually reach their maximum values for both groups between 3 and 12 months postoperatively. Unilateral graft recipients who had a preoperative diagnosis of parenchymal lung disease attain 60 to 65% of their predicted FVC and FEV1values. The values for bilateral lung recipients often approach normal predicted values, but these patients can have mild restrictive physiology. Diffusing capacities are usually slightly decreased in all groups. Airway hyperresponsiveness without clinically relevant asthma can be demonstrated in the majority of lung transplant recipients. Exercise capacity has been the most interesting functional outcome observed in lung transplant recipients. With respect to nongraded exercise capacity, usually measured by 6- or 12-min walk studies, unilateral and bilateral graft recipients demonstrate marked and similar improvement in distances covered after transplantation. Typically, transplant recipients can walk 100 to 120 m/min within 6 months of transplant and are generally able to sustain this rate over time. On graded exercise studies, however, both groups achieve only 40 to 60% of predicted maximum values, with bilateral lung recipients usually performing slightly better than unilateral lung recipients. This exercise limitation has been extensively studied particularly in bilateral lung recipients. The limitation appears not to be cardiac or ventilatory but rather related to muscle deconditioning and abnormalities in skeletal muscle oxidative capacity. Rarely is the exercise limitation in these patients enough to impact on their normal daily activities or their quality of life. POSTTRANSPLANT MANAGEMENT ISSUES Airway Complications Technical improvements and surgical experience have greatly reduced significant anastomotic complications in lung transplant recipients. It is not uncommon to see small dehiscences of the airway in the first weeks posttransplant, but these generally heal without significant stricture. Probably fewer than 10% of patients will have stenosis severe enough to require balloon dilatation, laser resection, or a stent. When required, wire stents are most often used and are well tolerated. Late-occurring bronchomalacia, often at the anastomotic site, has also been treated with stents. Acute Rejection A three-pronged immunosuppressive approach, which is used in most lung transplant programs, includes either cyclosporine or tacrolimus, either azathioprine or mycophenolate mofetil, and prednisone. Cytolytic induction is rarely used in these patients because of the risk of infection. Most lung transplant recipients experience at least one episode of acute rejection, usually within the first 3 months, although episodes have been reported to occur up to several years after transplantation. From 10 to 15% of patients have recurrent acute or persistent acute rejection, which predisposes them to
chronic rejection. Symptoms include a general feeling of malaise, dyspnea, and sometimes cough. Findings may include low grade fever, rales, mild hypoxemia, decreasing FVC and FEV1values, increased white blood cell count, and ill-defined infiltrates with or without pleural effusion on chest x-ray. If a patient presents early in an episode of acute rejection, as most do, the findings are minimal and the chest radiogram is clear. Histologic diagnosis, which is the "gold standard," is routinely made by transbronchial biopsy, with a sensitivity of about 80% and a specificity approaching 100%. Bronchoscopy is also helpful in this setting to rule out infections that may have similar presentations. Acute rejection episodes occurring early after transplantation respond in at least 80% of patients to bolus methylprednisolone. Late episodes and recurrent or persistent episodes often require both intensification of immunosuppression and changes in immunosuppressive drugs. Up to 20% of asymptomatic patients have at least one episode of acute rejection detected by surveillance transbronchial biopsy in the first 2 years posttransplant. It is not clear whether asymptomatic rejection requires therapy, as the impact on outcome is unknown. Thus, the use of surveillance bronchoscopy and the treatment of asymptomatic rejection vary considerably from institution to institution, and there are at present no clear guidelines in this area. Bronchiolitis Obliterans Bronchiolitis obliterans is both the primary manifestation of chronic rejection and the most feared complication in lung transplant recipients. It occurs to some degree in at least 50% of survivors by 5 years posttransplant and is a factor in more than one-third of late deaths. Although it can occur as early as 2 months posttransplant, the onset is more often at least 6 months and the mean onset is from 1 to 2 years after surgery. The precipitating factors and initiating events in bronchiolitis obliterans are topics of intensive research both in transplant recipients and in several animal models. Those factors most consistently associated with the process include the numbers and severity of acute rejection episodes and episodes of cytomegalovirus (CMV) pneumonia, but not clearly CMV infection alone. Other factors with weaker associations include HLA mismatches, other viral infections, and the development of anti-HLA antibodies. Clinically, the onset of this process is often subacute, with a very gradual onset of dyspnea and fatigue or malaise, often accompanied by viral-type symptoms or dry cough. It can also be asymptomatic and detected by routine pulmonary function studies that show, initially, a decrease in the FEF25-75, often followed one to several months later by decreasing FEV1. This insidious development of small airway obstruction is often well established before it is clinically recognized, and for that reason frequent pulmonary function testing is recommended for lung transplant recipients. Chest radiograms are usually normal, but even early in the disease expiratory computed tomography (CT) scans show a mottled appearance with peripheral hyperlucency. Transbronchial biopsy is very specific but not sensitive in diagnosis, but patients usually undergo at least one bronchoscopy at the onset of disease to attempt histologic documentation and to rule out possible infections. Because of the difficulty in histologic diagnosis, a typical clinical picture in the absence of other etiology is considered sufficient to establish a diagnosis of bronchiolitis obliterans syndrome. The progression of this complication can be very rapid, with early death; but more often it is one of a gradually decreasing FEV1over months to years, which in the later stages is frequently accompanied by bronchomalacia, proximal bronchiectasis, and recurrent pseudomonal or other infections.
Effective treatment remains evasive. A few immunosuppressive protocols tried in small numbers of patients have been found to "stabilize" pulmonary function, but improved function is unusual. Likely, by the time the process is recognized in most patients, fibrotic obliteration of the airway is already present; the key to treatment may lie in identifying markers of incipient disease and much earlier intervention. Infections Infections rank second only to rejection as a cause of morbidity in lung transplant recipients and are the most common cause of mortality, accounting for one-third of all deaths in both the early and the late posttransplant periods. The transplanted lung may be uniquely vulnerable to infection because of impaired mucociliary clearance, loss of cough reflex, and other poorly defined local factors. In addition, the donor lungs are often colonized with organisms that are transmitted directly to the immunosuppressed recipient. Early series reported that at least 60% of lung recipients early in their course develop infections requiring treatment. Now the extensive use of broad antibacterial, antifungal, anti-pneumocystis, and antiviral prophylaxis, often maintained for at least 3 months postoperatively, seems to have reduced the early infective morbidity and mortality. Infections with paramyxoviral organisms, adenovirus, and influenza A have now been well documented and have an overall death rate of about 20%. The role of antiviral therapy is unclear. The most lethal infections are those with invasive fungal organisms, particularly those caused by Aspergillus species, which have been reported to colonize in 20 to 50% of recipients. Invasive disease caused by these organisms can vary from ulcerative bronchitis to localized parenchymal infiltrates to empyema to disseminated disease. Aspergillus is particularly likely to be problematic when patients require increased immunosuppression or have other complications; one study has reported an increased rate of invasive disease in the native lung of unilateral lung transplant recipients. The highest risk periods for infection are in the first few months posttransplant and late after transplant if bronchiolitis obliterans or other vital organ dysfunction, e.g., renal failure, develops. Since it may be very difficult to distinguish infection from rejection in the early posttransplant period, bronchoscopy with appropriate biopsies and cultures is often necessary to establish a diagnosis. Immunosuppressive and Medical Complications Medical complications related to immunosuppression, to the underlying diagnosis, or to aging are major causes of morbidity in long-term survivors of lung transplantation and account for up to 10% of late deaths. Current immunosuppressive regimens with cyclosporine or tacrolimus cause some nephrotoxicity in virtually all patients. Although few progress to renal failure, hypertension and hyperlipidemia are common. Neurotoxicity, including delirium, headaches, seizures, and, occasionally, strokes, has been reported in up to 20% of patients. Osteoporosis occurs in more than half the patients, and vertebral compression fractures are common. Other problems include thromboembolic disease, gastric complications (especially gastroparesis), hyperglycemia, and increased rates of malignancy. Posttransplant lymphoproliferative disorders associated with Epstein-Barr virus occur in
5 to 10% of lung transplant recipients. Nearly all occur within the first year after transplant. Recent data suggest a much higher incidence of this disease in patients who are Epstein-Barr naive and who receive an Epstein-Barr positive graft. Treatment for this disorder, reported to have an approximate 50% survival rate, is usually reduced immunosuppression and antiviral and anti-B lymphocyte drugs. Survivors often develop bronchiolitis obliterans. Recurrence of Underlying Disease Several different underlying diseases may recur in lung transplant recipients. These diseases include sarcoidosis, lymphangioleiomyomatosis, giant cell interstitial pneumonia, panbronchiolitis, eosinophilic granuloma, bronchoalveolar cell carcinoma, and desquamative interstitial pneumonia. (Bibliography omitted in Palm version) Back to Table of Contents
PART TEN -DISORDERS OF THE KIDNEY AND URINARY TRACT 268. DISTURBANCES OF RENAL FUNCTION - Robert M. Brenner, Barry M. Brenner Near constancy of the composition of the internal environment, including the volume, tonicity, and compartmental distribution of the body fluids, is essential to survival. With normal day-to-day variations in the intake of food and water, preservation of the internal environment requires the excretion of these substances in amounts that balance the quantities ingested. Although losses from intestines, lungs, and skin contribute to this excretory capacity, the greatest responsibility for solute and water excretion is borne by the kidneys. The kidneys regulate the composition and volume of the plasma water. This, in turn, determines the composition and volume of the entire extracellular fluid compartment. Through the continuous exchange of water and solutes across all cell membranes, the kidneys influence the intracellular fluid compartment as well. These functions are served by a variety of physiologic mechanisms that enable individuals to excrete excesses of water and nonmetabolized solutes contained in the diet, as well as the nonvolatile end products of nitrogen metabolism, such as urea and creatinine. Conversely, when faced with deficits of water or solute, excretion of water or specific solute(s) is curtailed via appropriate mechanisms for renal conservation, reducing the likelihood of volume or solute depletion. The purpose of thischapteris to review the excretory functions of the kidney and to examine how these functions are affected by chronic renal disease. EFFECTS OF NEPHRON LOSS ON RENAL EXCRETORY MECHANISMS The volume of urine excreted (averaging 1.5 L/d or roughly 1 mL/min) represents the sum of two large, directionally opposite processes -- namely, ultrafiltration of 180 L/d or more of plasma water (or 125 mL/min) and reabsorption of more than 99% of this filtrate by transport processes in the renal tubules. While renal blood flow accounts for about 20% of resting cardiac output, the kidneys comprise only about 1% of total body weight. This disproportionate allocation of cardiac output, greatly exceeding blood flow per gram of brain, heart or liver, is required for the process of ultrafiltration. GLOMERULAR ULTRAFILTRATION Urine production begins at the glomerulus where an ultrafiltrate of plasma is formed. The rate of glomerular ultrafiltration (glomerular filtration rate, GFR) is governed chiefly by forces favoring filtration on the one hand (hydraulic pressure in the glomerular capillaries) and forces opposing filtration on the other (the sum of hydraulic pressure in Bowman's space and colloid osmotic pressure in the glomerular capillaries). The rate of glomerular plasma flow and the total surface area of the glomerular capillaries are also determinants of GFR. Decreased GFR can therefore be expected when (1) glomerular hydraulic pressure is reduced (as in circulatory shock); (2) tubule (hence Bowman's space) hydraulic pressure is elevated, as in urinary tract obstruction; (3) plasma colloid osmotic pressure rises to high levels (hemoconcentration due to severe volume depletion, myeloma, or other dysproteinemias); (4) renal, and hence glomerular, blood flow is reduced (severe hypovolemia, cardiac failure); (5) permeability is reduced
(diffuse glomerular disease); or (6) filtration surface area is diminished, through focal or diffuse nephron loss in progressive renal failure. The glomerular capillary wall is specially adapted to allow passage of extremely large volumes of water while retaining all but the smallest solute molecules. Molecules the size of inulin (approximately 5200 mol wt) pass freely across the glomerular filtration barrier, appearing at approximately the same concentration in Bowman's space as in plasma. The passage of solutes across the glomerular barrier decreases progressively with increasing molecular size such that, as the molecular weight of albumin is approached, most of the solute is retained in the plasma. Albumin, a polyanionic molecule in plasma, is further retarded at the glomerular filtration barrier by electrostatic forces imparted by negatively charged cell-surface molecules on the epithelial foot processes that form the filtration slits and the slit diaphragms. With disruption of these structural and electrostatic barriers, as in many forms of glomerular injury (Chaps. 273 to 275), large quantities of plasma proteins gain access to the glomerular filtrate. Glomerular Adaptations to Nephron Loss With loss of nephron mass, the remaining functional (or least injured) nephrons tend to hypertrophy and take on an increased workload so that the overall loss of function is minimized. For example, a patient with a unilateral nephrectomy loses one-half of the nephron mass, resulting in a 50% reduction inGFR at the time of surgery. However, the GFR in the remaining kidney begins to increase after 1 or 2 weeks, and within several months GFR may rise to 80% of the preoperative value. This indicates that the GFR of the individual remaining nephrons has increased above normal, a state known as hyperfiltration. Increases in single-nephron GFR may be achieved by renal hemodynamic adjustments (increased glomerular plasma flow and increased glomerular capillary hydraulic pressure), which augment the forces driving ultrafiltration, and by glomerular hypertrophy, which increases the maximum surface area available for filtration. These structural adaptations are evident from the enlargement of glomeruli (and tubules) seen on histologic sections from people with single kidneys. Similar structural changes are observed in kidneys damaged by chronic disease processes; foci of hypertrophied glomeruli and tubules are interspersed with areas of atrophic or scarred parenchyma. Although direct measurements of single-nephron GFR cannot be made in humans, it is reasonable to conclude that focal nephron enlargement as occurs in chronically diseased kidneys generally signifies focally increased single-nephron GFR, and that these dynamic adaptations represent compensatory adjustments for the effects of nephron loss through disease. Glomerulotubular Balance The close integration of glomerular and tubular functions (glomerulotubular balance) seen in chronic renal failure (CRF) supports the notion that progressive nephron obliteration is the usual mode ofGFRreduction in CRF. Preservation of glomerulotubular balance until the terminal stages of CRF is fundamental to the intact-nephron hypothesis, which states that as CRF advances, kidney function is supported by a diminishing pool of functioning (or hyperfunctioning) nephrons, rather than relatively constant numbers of nephrons, each with diminishing function. This concept has important implications for the mechanisms of disease progression in CRF. A considerable amount of evidence suggests that nephrons subjected to increased excretory burdens for prolonged periods actually sustain injury as a result of these adaptations: thus the cost of these compensatory adaptations to
nephron loss may ultimately be relentless destruction of the remaining nephron pool. The magnitude of the single-nephron hyperfiltration induced by loss of 50% of the total nephron mass usually has no serious adverse clinical consequences, even when sustained over two to three decades. When more than 50% of the total nephron mass is lost, however, as in renal-sparing surgery for bilateral trauma or neoplasm or from a renal disease whose activity has abated, the remaining nephrons are forced to the limits of their compensatory capacity. While these adaptations achieve remarkable short-term success at offsetting the tendency forGFR to fall, over time, proteinuria and focal and segmental glomerulosclerosis develop, the more so where greater amounts of nephrons are lost or removed. As a result, a progressive decline in GFR ensues. Experimental study of the processes that advance glomerular injury show that the adverse long-term consequences of severe nephron deficits are invariably preceded by increases in glomerular capillary hydraulic pressure (glomerular capillary hypertension), glomerular hyperperfusion, and hypertrophy. Interventions directed against these compensatory and maladaptive responses can greatly ameliorate the subsequent development of renal failure. In particular, drugs (e.g., angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers) and other interventions (such as dietary protein restriction) that lower glomerular pressure can slow the rate of progression of experimental and human renal disease. In the absence of such interventions, more and more glomeruli cease to function through advancing glomerulosclerosis and disruption of tubule structure and function, leading eventually to total loss of GFR (i.e., end-stage renal disease). This final common pathway for chronic renal injury helps to explain the observed progressive nature of chronic renal failure resulting from many different kidney diseases. Biologic Consequences of Sustained Reductions in GFR Although nephron loss can proceed, to some extent, without equivalent loss ofGFR due to the compensatory mechanisms described above, determination of the total GFR of both kidneys remains the most reliable clinical index of overall excretory function. The effects of impaired GFR are to reduce the total rate of delivery of solute into the glomerular filtrate. When accompanied by comparably reduced rates of urinary excretion, retention and accumulation of the unexcreted solute occurs, resulting in increased concentrations of the substance in the plasma and other body fluids. Figure 268-1 depicts the major types of response to impairedGFR. The degree of reduction in total GFR is plotted on the abscissa, expressed as a percentage of normal (100%). The renal handling of most solutes normally present in glomerular filtrate conforms to one of three patterns. Curve A describes the pattern with substances such as creatinine and urea that normally depend largely on glomerular filtration for urinary excretion; i.e., secretion contributes little to overall excretion. Therefore, as illustrated, gradual reductions in GFR are accompanied by progressive increases in plasma levels of creatinine, urea, and other substances normally excreted primarily by filtration. The clinical course ofCRFusually also approximates the pattern described by curve A. Patients with CRF usually pass from a long asymptomatic period of "compensation" to a more accelerated and clinically overt terminal phase. In other words, despite chronic injury leading to destruction of more than 50% of nephrons, plasma elevations of creatinine and urea may still lie within the normal limits for these substances. With
further nephron loss and reduction inGFR, however, the limits of renal reserve are exceeded and continued accumulations of curve A-type solutes lead to abnormally elevated plasma concentrations (Fig. 268-1). Because some of these retained solutes are thought to exert "toxic" effects on all organ systems, clinical manifestations of CRF may now become apparent. Consequently, in patients with substantial reductions in nephron mass but near-normal plasma creatinine, overt uremia may be precipitated by a modest additional decline in GFR. The accumulation of curve A-type solutes with chronic loss of renal function proceeds until external balance is restored, i.e., intake and/or production rates exactly match excretion rates. In the case of creatinine, for example, assuming a constant rate of creatinine production, a 50% reduction inGFRresults in an approximate doubling of the plasma creatinine concentration. The latter restores the filtered load of creatinine (i.e., the product of GFR and plasma creatinine concentration) to normal, and the urinary excretion rate once again is equivalent to creatinine production. Since creatinine secretion contributes only slightly, elimination of the retained creatinine is not possible and the plasma concentration remains twice normal. With further loss of GFR, elevations in plasma creatinine are compounded by loss of nephron excretory function and creatinine retained as the result of earlier nephron destruction (Fig. 268-1). In practice, so long as the net rates of acquisition and production (i.e., liver function and muscle mass) remain reasonably constant, the inverse relationship between plasma concentrations of solutes such as creatinine and urea and GFR is sufficiently reliable to serve as clinical indices of GFR. However, where muscle mass is low, as with severe weight loss, unremarkable plasma levels of creatinine may belie substantial reductions in GFR. In contrast to solutes of the curve A type, plasma levels of phosphate (PO43-), urate, and potassium (K+) and hydrogen (H+) ions usually do not rise until theGFRfalls to a small percentage of normal. With progressive renal failure this pattern of response (curve B inFig. 268-1) reflects the participation of tubule transport mechanisms in the excretion of these substances. In other words, as GFR declines, the tubules facilitate greater elimination of these substances, by enhancing secretion and/or by diminishing reabsorption, so that a greater fraction of the filtered load is excreted. Plasma levels of curve B-type solutes, therefore, rise less than those of curve A because, with progressive reductions in GFR, excretion rate per nephron and therefore fractional excretion both increase. Eventually, however, with further loss of GFR, enhanced fractional excretion can no longer mitigate the reduction in net filtered load of these solutes and plasma levels rise (Fig. 268-1). For urate, PO 43- , and K+, at least, increased fractional excretion serves to maintain normal plasma levels until GFR falls to less than one-fourth of normal. Finally, for certain solutes, such as sodium chloride (NaCl), plasma concentrations remain normal throughout the course ofCRF, despite unrestricted intake of these substances (curve C inFig. 268-1). The compensatory mechanism required to achieve this represents a fundamental adaptation to chronic renal injury. To illustrate the magnitude of this adaptation, it is useful to compare the excretion of sodium (Na+) in a normal individual (GFR of 125 mL/min) with that of a patient with advanced renal failure (GFR of 2 mL/min). Both individuals consume a conventional diet containing 7 g/d of salt (120 mmol Na+). With a normal serum Na+ concentration of 140 mmol/L, external
Na+balance is achieved by excreting approximately 0.5% of the filtered load. By contrast, for external balance to be maintained in the patient with CRF, fractional excretion of Na+ must rise to 30%. In other words, to maintain external Na+balance, the same amount of Na+ must be excreted into the urine each day in the patient with CRF as in the normal individual. Given the drastic reduction in GFR in CRF, external balance can only be maintained by marked adaptations in the reabsorptive processes in surviving tubules. In this manner, a progressively larger fraction of the filtered load escapes reabsorption and appears in the final urine. In short, the rate of excretion of Na+per surviving nephron increases in inverse proportion to the composite GFR in surviving nephrons. ADAPTATIONS IN TUBULE TRANSPORT MECHANISMS IN RESPONSE TO NEPHRON LOSS Despite progressive nephron loss, many mechanisms that regulate renal solute and water balance differ only quantitatively, and not qualitatively, from those that operate normally. Thus, glomerulotubular balance is maintained. The most important of these mechanisms are considered below. TUBULAR TRANSPORT OF SODIUM CHLORIDE AND WATER Most of the filtered water and sodium salts are reabsorbed by the tubules, leaving small and variable amounts, equivalent on average to the quantities ingested, to reach the final urine. About two-thirds of the glomerular ultrafiltrate is reabsorbed in the proximal tubule with little change in the osmolality or Na+concentration of the unreabsorbed fraction (Fig. 268-2). In other words, fluid reabsorption in the proximal tubule is nearly isosmotic and is coupled to the active transport of Na +. Since chloride (Cl-) and bicarbonate (HCO3-) are the primary anions in the extracellular fluid, they constitute the main solutes that accompany Na+reabsorption in the renal tubules. In the earliest portion of the proximal tubule, bicarbonate is the principal anion that accompanies the reabsorption of Na+. This process occurs via a Na+/H+exchanger at the luminal brush border and is dependent on the activity of carbonic anhydrase. Glucose, amino acids, and other organic solutes (e.g., lactate) are also extensively reabsorbed in the proximal tubule by cotransport mechanisms that link the cellular entry of these organic molecules with Na+. The coupling of water absorption (i.e., volume) with solute absorption appears to be dependent upon three processes. First, given the remarkably high water permeability of this segment, very small transepithelial osmolality differences, i.e., luminal hypotonicity of the order of 2 to 3 mosmol/L produced by solute absorption, could drive water absorption. Second, due to preferential absorption of HCO 3-and organic solutes in the early portions of the proximal tubule, the concentrations of these substances decrease along the proximal tubule while that of chloride increases. Volume reabsorption would then occur if the diffusion of Na+ and Cl- down their respective electrochemical gradients across the proximal tubule epithelium occurred more easily than the back-diffusion of sodium bicarbonate into the lumen, creating an effective osmotic pressure gradient. Finally, lateral interstitial space hypertonicity produced by differences in the rates at which solutes are transported into the spaces or exit them by diffusion may also contribute to the coupling of water and solute reabsorption. Reabsorption of Fluid from Proximal Convoluted Tubules This is sensitive to
Starling forces, i.e., the hydraulic and colloid osmotic (or oncotic) pressures acting across the walls of the peritubular capillaries. Because the plasma proteins in glomerular capillaries are concentrated by ultrafiltration, oncotic pressure rises along the glomerular capillary network. This step-up in oncotic pressure is transmitted largely unchanged to the first branches of the peritubular capillaries via the efferent arterioles. These resistance vessels cause a substantial drop in hydraulic pressure, however, so that when the plasma reaches the peritubular capillaries, oncotic pressure greatly exceeds hydraulic pressure. The Starling forces are therefore oriented in an uptake mode, in contrast to their configuration at the glomerulus where hydraulic pressure exceeds oncotic pressure, favoring filtration. The extent to which oncotic pressure exceeds hydraulic pressure in the peritubular capillary network modulates the overall rate of fluid absorption by the peritubular capillaries. Therefore, when peritubular capillary oncotic pressure falls, or hydraulic pressure rises, uptake of fluid by these capillaries is reduced. As a result, fluid is retained in the interstitial space, tending to increase hydraulic pressure, ultimately retarding the egress of fluid from the lateral intercellular channels. Without an adequate route of drainage, fluid in the intercellular channels leaks back into the tubule lumen, thereby diminishing net fluid reabsorption from this tubule segment. The opposite occurs in states where peritubular oncotic pressure is increased (increased filtration fraction) or hydraulic pressure is decreased (enhanced efferent arteriolar tone). Under these circumstances, peritubular capillary uptake of reabsorbate is augmented, leading ultimately to enhanced net fluid reabsorption by the proximal tubule. Although physical factors appear to be the major determinants of fluid reabsorption in the proximal tubule, hormones (e.g., angiotensin II) may also modulate fluid reabsorption directly, by enhancing luminal Na+entry into proximal tubule cells via an apical Na+ /H+exchanger. The Limbs of Henle's Loop In contrast to the proximal tubule, active outward transport of Na has not been established for the thin ascending limb of Henle's loop. However, passive outward salt transport does occur, as indicated inFig. 268-2. In the next nephron segment, the medullary thick ascending limb of Henle, the concentration of NaCl is reduced as fluid traverses this segment. Here Cl- absorption occurs by an active process involving a Na+:K+ :2Cl-cotransport mechanism in the luminal membrane, with one-half of Na+absorption proceeding passively, driven by the lumen positive transepithelial voltage difference. This cotransporter is the site of action of the powerful loop diuretics and mutations give rise to Bartter's syndrome. Since the ascending limb of Henle is impermeable to water, net NaCl reabsorption generates a hypotonic tubule fluid and gives rise to the high NaCl concentration of the outer medullary interstitium (Fig. 268-2). In certain animals, arginine vasopressin (AVP; also called ADH) enhances NaCl absorption in the medullary portion of the thick ascending limb, but whether this occurs in humans is uncertain. Distal Tubule The fluid leaving the thick ascending limb of Henle is normally of low NaCl concentration, a characteristic independent of the organism's hydration or dietary status. In the distal tubule, water reabsorption is variable, depending on the state of hydration or, specifically, on the presence or absence ofAVP in plasma. In the absence of AVP, this and more distal nephron segments are impermeable to water, so that hypotonic fluid entering this segment is excreted as dilute urine. Indeed, continued salt reabsorption along the distal convoluted tubule (DCT) and connecting tubule segments, a process that can be inhibited by the thiazide classes of diuretics, results in further
dilution of the urine. In the presence of AVP, the permeability of these nephron segments to water increases. This is made possible by the insertion of proteins known as aquaporins into the luminal cell membrane of DCT cells. These proteins facilitate water movement from the low osmolality environment of the DCT lumen into the higher osmolality of the medullary interstitium, thereby contributing to the creation of a concentrated final urine. NaCl continues to be reabsorbed from the tubule lumen against moderately steep chemical and electrical gradients. The reabsorption of NaCl at the collecting tubule is enhanced by aldosterone. Collecting Tubules and Ducts The cortical collecting tubule possesses a low permeability to water in the absence ofAVP, whereas permeability increases in the presence of this hormone. The sensitivity of this segment to AVP appears to be more pronounced than that of the DCT. As with the DCT, the cortical collecting tubule is capable of active reabsorption of NaCl and its stimulation by aldosterone. The terminal segment of the distal nephron is the highly branched papillary collecting duct. Continued electrolyte transport in this segment results in the large ion concentration differences that normally exist between urine and plasma. As in the cortical collecting tubule, Na+ transport appears to be active, since reabsorption proceeds against sizeable electrochemical gradients. The rate of Na+transport in this segment depends on the load of Na+delivered from more proximal segments and is also affected by aldosterone. The permeability to water is also increased markedly in the presence ofAVP. Effects of Nephron Loss on Sodium Chloride Transport in Surviving Nephrons With progressive nephron loss, maintenance of external balance for NaCl requires that fractional salt excretion increases in concert with the decline inGFR. Several mechanisms contribute to this adaptive increase in fractional Na+excretion. With loss of functioning nephron units, peritubular capillary Starling forces are presumably altered in directions that serve to reduce proximal tubule reabsorption of NaCl and water. For example, a rise in peritubular capillary hydraulic pressure, which tends to inhibit net proximal fluid reabsorption, might be anticipated with systemic hypertension, a common feature of chronic renal failure. Similarly, reductions in peritubular capillary oncotic pressures may be anticipated due to reductions in both filtration fraction and hypoalbuminemia. Aldosterone, which normally exerts a potent influence on tubule transport, probably does not figure prominently in reducing fractional Na+excretion, since aldosterone levels are seldom reduced inCRF. Furthermore, external Na+balance is preserved in bilaterally adrenalectomized dogs on fixed replacement doses of mineralocorticoid. Yet another factor contributing to the suppression of fractional NaCl reabsorption in CRF may relate to the retention of various organic solutes asGFRdeclines. Several factors that regulate NaCl transport across tubules under resting conditions are also likely to contribute to the enhanced fractional excretion of salt in renal insufficiency. Atrial natriuretic peptides are released from the heart in response to elevated cardiac (atrial) filling pressures as seen with increased plasma volume or atrial tachyarrhythmias. These peptides affect natriuresis by reducing net Na +reabsorption through complementary actions on Na+ transport in the collecting duct and by altering
Starling forces in the adjacent vasa recta. The vascular actions of natriuretic peptides may also extend to glomerular hemodynamics, with afferent arteriolar vasodilatation contributing to increased single-nephronGFR and hence an increase in the amount of Na+filtered. Other modulators of tubule transport processes may also contribute to increased single-nephron natriuresis in the setting of reduced renal mass or nephron loss. Vasodilator prostaglandins are present at increased plasma levels in CRF, as are other inhibitors of transport, including inhibitor(s) of the Na+ ,K+-ATPase. This latter factor has not yet been fully characterized; whether its presence represents a homeostatic adaptation for maintenance of fluid balance or an unregulated accumulation of a toxin remains uncertain. Serum and urine from patients with uremia contain factors capable of experimentally inhibiting NaCl transport across frog skin, toad bladder, and rat renal tubule. Accumulation of natriuretic factors in uremia may not be without cost; the "trade-off" for maintenance of external Na+balance is the possibility of generalized abnormalities occurring in Na+transport across cell membranes, which often occur in advanced renal failure (Chap. 270). The obligatory high rate of solute excretion per surviving nephron (so-called osmotic diuresis due to urea and other retained solutes) also contributes to enhancing fractional NaCl excretion, much as occurs in normal individuals after the administration of mannitol or other nonreabsorbable solutes. Finally, certain forms ofCRF are associated with unusually large losses of salt in the urine. These salt-wasting nephropathies include chronic pyelonephritis and other tubulointerstitial diseases (Chap. 277) as well as polycystic and medullary cystic diseases. These disorders have in common greater destruction of medullary and tubulointerstitial, rather than cortical and glomerular, portions of the renal parenchyma. Preferential impairment of tubule reabsorptive function, rather than a primary reduction in glomerular filtration, may, therefore, underlie the salt-losing tendency in these disorders. Clinical derangements that alter renal handling of NaCl in CRF (including hypo- and hypervolemia, hypertension, etc.) are considered in Chap. 270. EFFECTS OF NEPHRON LOSS ON WATER REABSORPTION IN SURVIVING NEPHRONS As with NaCl, there is a progressive increase in the fractional excretion of water with advancing renal insufficiency, so that external water balance can be maintained even with a totalGFR of 5 mL/min or less. The adaptations of water handling by the diseased kidney are of importance in the defects in urinary concentration and dilution and hence the polyuria, nocturia, and tendency to develop water overload encountered inCRF(Chap. 47). To appreciate the mechanisms involved, the responses of a normal and a uremic individual maintaining external water balance need to be considered. Assuming both individuals have the same dietary and fluid intakes, total solute and volume excretion in both should be identical as well. If the obligatory solute load to be excreted by each is 600 mmol/d (600 mosmol/d) and the urine osmolality is 300 mmol/kg water (300 mosmol/kg), a urine volume of 2 L/d will be required to excrete the total solute. If the GFR in normal and uremic individuals totals 180 and 4 L/d, respectively, urinary volume excretion of 2 L/d represents excretion of slightly more than 1% of the total glomerular filtrate in the normal subject compared with 50% in the uremic
patient. Since the range of urine osmolalities that the diseased kidney can achieve [250 to 350 mmol/kg (250 to 350 mosmol/kg)] is narrower than in the normal kidney [40 to 1200 mmol/kg (40 to 1200 mosmol/kg)], the individual with normal function is able to excrete the obligatory daily solute load of 600 mmol (600 mosmol) in as little as 500 mL urine per day or as much as 15 L/d, compared with the narrower range in renal insufficiency, from about 1.7 to 2.4 L/d. InCRF, the limited capacity to concentrate the urine often correlates with other measures of impaired renal function. Isosthenuria (urine of similar osmolality to plasma) is therefore an almost universal finding when theGFRfalls below 25 mL/min. At this level of GFR and below, urine osmolality does not rise even when supraphysiologic doses ofAVP are administered, suggesting that the concentrating defect relates to impaired concentrating capacity in surviving nephrons. The associated increased fractional excretion per nephron of a variety of solutes produces an obligatory water loss (solute diuresis) at roughly isotonic proportions. Consequently, formation of a concentrated urine is prevented. Disease-induced abnormalities of the architecture of the renal medulla (loops of Henle, vasa recta), aberrations in medullary blood flow, and defective transport of NaCl in the ascending limb of Henle also contribute to this defect in urine concentration. Since patients withCRF are unable to excrete concentrated or dilute urine, they must have access to adequate, and to some extent, relatively constant amounts of water per day to ensure that they have adequate water to eliminate total daily solute loads. For this reason, restriction of fluid intake may be hazardous in patients with CRF. Likewise, impairment of diluting capacity may prevent many patients from excreting excess ingested fluid. The consequences of the abnormal patterns of water excretion, and the attendant susceptibilities to develop hypo- and hypernatremia, are considered inChaps. 49 and270. TUBULE TRANSPORT OF PHOSPHATE WITH NORMAL AND REDUCED NEPHRON MASS Under normal physiologic conditions, about 80 to 90% of phosphate is reabsorbed, mainly in the proximal tubule. Parathyroid hormone (PTH), by augmenting phosphate excretion via inhibition of this proximal reabsorptive process (Chap. 340), plays a central role in phosphate homeostasis. When dietary phosphate intake increases, a transient rise in plasma phosphate concentration is usually observed. This results in a similarly transient reduction in the plasma ionized calcium level (due largely to deposition of calcium phosphate in bone), which is sensed by a specific receptor on parathyroid cells, stimulating PTH secretion. By enhancing fractional phosphate excretion, PTH restores external phosphate balance and normophosphatemia. This enables plasma ionized calcium levels to return to normal, thereby removing the stimulus to PTH release and restoring the phosphate control system to the original steady state. With advancing renal failure and constant dietary intake of phosphate, external phosphate balance is achieved by progressive reduction in fractional phosphate reabsorption. EnhancedPTHsecretion is an important determinant of this phosphaturic response. With succeeding decrements in totalGFR, the amount of phosphate filtered by surviving glomeruli is reduced, leading to transient phosphate retention and,
therefore, a rise (albeit small) in plasma phosphate concentration. This leads to a small, reciprocal decline in plasma levels of ionized calcium and a corresponding increase in PTH secretion. Although the phosphaturic response of surviving tubules to this elevation in circulating PTH restores plasma phosphate and calcium to normal levels (at least in the "compensated" stage ofCRFdescribed by curve B in Fig. 268-1), the new steady-state conditions are only achieved at the cost of persistently elevated plasma PTH levels. With progressive reductions in GFR, the process is repeated, resulting in substantially elevated PTH levels. Alterations in Vitamin D Metabolism These alterations also contribute to elevatedPTHlevels inCRF. The kidney is normally the major site of conversion of vitamin D to its active metabolites. As discussed in Chap. 340, vitamin D, synthesized in skin or acquired in the diet, undergoes initial hydroxylation in the liver to form 25-hydroxyvitamin D [25(OH)D]. The kidney is the site of a second important conversion to 1,25-hydroxyvitamin D [1,25(OH2)D]. This active form of vitamin D acts directly on the parathyroid gland to suppress PTH secretion as well as to enhance intestinal absorption of calcium and phosphate resorption and promote resorption of these ions from bone. In addition, 1,25(OH)D2probably opposes the phosphaturic actions of PTH in the renal tubule by augmenting, rather than diminishing, phosphate reabsorption. With advancing renal disease, nephron loss reduces the renal capacity for vitamin D hydroxylation; phosphate retention also impairs this reaction. Not only are the circulating levels of 1,25(OH)D2diminished in uremia, but the receptors that mediate its action at the parathyroid gland are also diminished. These two effects remove inhibitory influences on PTH secretion, leading again to increased plasma PTH levels. Reduction in circulating 1,25(OH)D2levels, by suppressing intestinal calcium absorption, contributes to the development of the hypocalcemia and hyperparathyroidism of CRF (Chap. 270). Hyperparathyroidism in Chronic Renal Failure At least two additional processes are thought to contribute to hyperparathyroidism inCRF. One relates to resistance of bone to the calcemic effect ofPTH in uremia. This resistance necessitates a higher level of PTH to demineralize bone and maintain the plasma calcium concentration. The other derives from the finding that reductions in renal mass impair the kidneys' capacity to degrade circulating PTH. Ultimately, however, phosphate conforms more to a curve Brather than a curve C-type pattern inFig. 268-1, and phosphate retention occurs when theGFRfalls below about 25 mL/min, signifying that these latter forms of adaptation play limited roles. SincePTHexerts major effects on bone as well as renal tubules, the external balance of phosphate inCRF is achieved at the expense of elevated PTH levels, which, in turn, account for many of the bone changes of renal osteodystrophy (i.e., secondary hyperparathyroidism; Fig. 270-1). In support of this trade-off hypothesis, when dietary phosphate intake is reduced in proportion to the reduction inGFR in animals with CRF, external balance of phosphate no longer requires augmentation of fractional phosphate excretion in surviving nephrons. Accordingly, circulating levels of PTH no longer rise, and the bone changes of secondary hyperparathyroidism are diminished, if not prevented. HYDROGEN AND BICARBONATE TRANSPORT WITH NORMAL AND REDUCED RENAL MASS
As discussed inChap. 50, the pH of extracellular fluid is normally maintained within a narrow range (7.36 to 7.44) despite day-to-day fluctuations in the quantity of acids added to the extracellular fluid from dietary and metabolic sources (approximately 1 mmol H+ per kilogram of body weight per day). These acids consume buffers from both extracellular and intracellular fluid, of which HCO3-is the most important in the intracellular compartment. Such buffering minimizes changes in pH. Long-term effectiveness of the HCO3-buffer system, however, requires mechanisms for replenishment, otherwise unrelenting acquisition of nonvolatile acids from dietary and metabolic sources would ultimately exhaust buffering capacity, culminating in fatal acidosis. The kidneys normally function to prevent this eventuality by regenerating bicarbonate, thereby maintaining plasma concentrations of HCO3-. In addition, the kidneys also reclaim HCO3-in the glomerular ultrafiltrate. Reclamation of filtered HCO3-takes place largely in the proximal tubule and, under normal circumstances, is virtually complete below a critical plasma HCO3-concentration -- the threshold concentration -- which in humans is normally about 26 mmol/L, identical to the concentration of HCO3-in plasma. As a consequence, HCO3-wastage is prevented. Alternatively, when plasma HCO3-rises above this threshold, reabsorption becomes less complete, allowing escape of excess HCO3-into the final urine, which restores the plasma HCO3-towards normal levels. Despite complete reabsorption of HCO3-, metabolic acidosis would still ensue if HCO3-consumed in buffering nonvolatile acids were not constantly regenerated. The reabsorption of filtered HCO3-occurs by the following mechanism. Filtered bicarbonate combines with H+secreted from proximal tubule cells via the Na+/H+ exchange, to form carbonic acid (H2CO3). Dehydration of carbonic acid under the influence of luminal carbonic anhydrase yields H2O and CO2, which is free to diffuse from lumen to peritubular blood. In the proximal tubule cell, the OH - left behind by the H+secretion reacts with CO2, under the influence of intracellular carbonic anhydrase, forming HCO3-. This ion is transported across the contraluminal proximal tubule cell membrane, via an electrogenic Na/HCO3-cotransporter, to reenter the extracellular HCO3-pool. The net result is reclamation of a filtered bicarbonate ion. Secreted H+ is also free to react with nonbicarbonate buffers [e.g., phosphate or ammonia (NH4+)] in the tubule lumen, and hydrogen ions are excreted in these forms in the final urine. Again, the OH-left behind in the proximal tubule cell from H+ secretion reacts with CO2, forming bicarbonate -- also representing regeneration of an HCO3-ion. Hydrogen ions in the urine are bound to filtered buffers (e.g., phosphate) in amounts equivalent to the amounts of alkali required to titrate the pH of the urine up to the pH of the blood (the so-called titratable acid). It is not usually possible to excrete all the daily acid load in the form of titratable acid due to limits of urinary pH. Metabolism of glutamine by proximal tubule cells to yield ammonium (ammoniagenesis) serves as an additional mechanism for H+elimination and bicarbonate regeneration. Glutamine metabolism forms not only NH4+(i.e., NH3 plus H+) but also HCO3-, which is transported across the proximal tubule (HCO3-regeneration). The NH4+must be excreted in the urine for this process to be effective in bicarbonate regeneration. The excretion of ammonium involves secretion by proximal tubule cells (possibly by the Na+ /H+exchanger as Na+/NH4+), generation of high medullary interstitial NH4+concentration by an elaborate countercurrent multiplication/exchange system, and finally, secretion of the interstitial
NH4+by the collecting duct by a combination of H+ secretion and passive NH3diffusion. Ammoniagenesis is responsive to the acid-base needs of the individual. When faced with an acute acid burden and an increased need for HCO3-regeneration, the rate of renal ammonia synthesis increases sharply. The quantity of hydrogen ions excreted as titratable acid and NH4+is equal to the quantity of HCO3-regenerated in tubule cells and added to plasma. Under steady-state conditions, the net quantity of acid excreted into the urine (the sum of titratable acid and NH4+less HCO3-) must equal the quantity of acid gained by the extracellular fluid from all sources. Metabolic acidosis and alkalosis result when this delicate balance is perturbed, the former the result of insufficient net acid excretion, and the latter due to excessive acid excretion. Progressive loss of renal function usually causes little or no change in arterial pH, plasma bicarbonate concentration, or arterial carbon dioxide tension (PCO2) untilGFRfalls below 25% of normal. Thereafter, all three tend to decline as metabolic acidosis ensues. In general, the metabolic acidosis ofCRF is not due to overproduction of acids but is rather a reflection of nephron loss, which limits the amount of NH3 (and therefore also HCO3-) that can be generated. Although surviving nephrons appear to be capable of generating supranormal amounts of NH3per nephron, the diminished nephron population causes overall production to be reduced to an extent that is insufficient to permit adequate buffering of H+ in urine. As a result, although patients with CRF may be able to acidify their urine normally (i.e., urine pH as low as 4.5), the defect in NH3production limits daily net acid excretion to 30 to 40 mmol, or one-half to two-thirds the quantity of nonvolatile acid added to the extracellular fluid in the same time period. Metabolic acidosis resulting from this daily positive balance of H+ is seldom florid in CRF of mild to moderate severity. Relative stability of plasma bicarbonate (albeit at reduced levels of 14 to 18 mmol/L) is maintained at the expense of buffering by bone. Because it contains large reserves of alkaline salts (calcium phosphate and calcium bicarbonate), bone constitutes a major reserve of buffering capacity. Dissolution of these buffers contributes to the osteodystrophy of CRF (Fig. 270-1). Although the acidosis ofCRF is due to loss of tubule mass, it nevertheless depends to a large part on the level of GFR. When GFR is reduced to only a moderate extent (i.e., to about 50% of normal), retention of anions, principally sulfates and phosphates, is not pronounced. Therefore, as the plasma HCO3-falls owing to dysfunction or loss of tubules, retention of Cl- by the kidneys leads to a hyperchloremic acidosis. At this stage the anion gap is normal. With further reductions in GFR and progressive azotemia, however, the retention of phosphates, sulfates, and other unmeasured anions ensues and plasma Cl- falls to normal levels despite the reduction in plasma HCO3-concentration. An elevated anion gap therefore develops. TUBULE POTASSIUM TRANSPORT WITH NORMAL AND REDUCED NEPHRON MASS As with H+, the concentration of K+ in extracellular fluid is normally maintained within a relatively narrow range, 4 to 5 mmol/L. At least 95% of total-body K + is in the intracellular compartment, where the intracellular concentration is approximately 160 mmol/L. Normal individuals maintain external K+balance by excreting amounts into the
urine that equal the intake, less the relatively small losses in stool and sweat. K+ is freely filtered at the glomerulus, although the amount excreted usually represents no more than about 20% of the quantity filtered. The great bulk of the K+filtered is reabsorbed in the early portions of the nephron, about two-thirds in the proximal tubule, and an additional 20 to 25% in the loop of Henle. A K+ secretory process operates in the distal tubule and terminal nephron segments. This process is largely dependent on Na+reabsorption and the accompanying lumen-negative voltage creating an electrical gradient across the tubule wall, favoring K+secretion into the lumen of the distal tubule and collecting duct. The ability to maintain external K+balance and normal plasma K+ concentration until relatively late in the course ofCRF is a consequence primarily of a progressive increase in fractional excretion of K+. Greatly enhanced rates of K+ secretion occur in distal portions of surviving tubules. The augmented secretion rate of aldosterone contributes to enhanced tubule secretion of K+ . In addition, both the increased distal tubule flow rates in surviving nephrons, due to the osmotic diuresis, and enhanced luminal electronegativity, created by the increased presence of highly impermeable anions such as phosphate and sulfate, enhance K+secretion. Aldosterone also stimulates net entry of K+ into the lumen of the colon, a mechanism known to be enhanced in CRF.*More detailed discussions of abnormal K+ homeostasis in acute and chronic forms of renal failure are given in Chaps. 269 and 270. ACKNOWLEDGEMENT Harald S. MacKenzie was a co-author of thischapterin the 14th edition and some of the content of thatchapteris carried forward to the present edition. (Bibliography omitted in Palm version) Back to Table of Contents
269. ACUTE RENAL FAILURE - Hugh R. Brady, Barry M. Brenner Acute renal failure (ARF) is a syndrome characterized by rapid decline in glomerular filtration rate (hours to days), retention of nitrogenous waste products, and perturbation of extracellular fluid volume and electrolyte and acid-base homeostasis. ARF complicates approximately 5% of hospital admissions and up to 30% of admissions to intensive care units. Oliguria (urine output< 500 mL/d) is a frequent but not invariable clinical feature (~50%). ARF is usually asymptomatic and is diagnosed when biochemical screening of hospitalized patients reveals a recent increase in plasma urea and creatinine concentrations. It may complicate a wide range of diseases, which for purposes of diagnosis and management are conveniently divided into three categories: (1) diseases that cause renal hypoperfusion without compromising the integrity of renal parenchyma (prerenal ARF, prerenal azotemia) (~55%); (2) diseases that directly involve renal parenchyma (intrinsic renal ARF, renal azotemia) (~40%); and (3) diseases associated with urinary tract obstruction (postrenal ARF, postrenal azotemia) (~5%). Most ARF is reversible, the kidney being relatively unique among major organs in its ability to recover from almost complete loss of function. Nevertheless, ARF is associated with major in-hospital morbidity and mortality, in large part due to the serious nature of the illnesses that precipitate the ARF. ETIOLOGY AND PATHOPHYSIOLOGY PRERENALARF(PRERENAL AZOTEMIA) PrerenalARF is the most common form of ARF and represents a physiologic response to mild to moderate renal hypoperfusion. Prerenal ARF is rapidly reversible upon restoration of renal blood flow and glomerular ultrafiltration pressure. Renal parenchymal tissue is not damaged; indeed, kidneys from individuals with prerenal ARF function well when transplanted into recipients with normal cardiovascular function. More severe hypoperfusion may lead to ischemic injury of renal parenchyma and intrinsic renal ARF. Thus, prerenal ARF and intrinsic renal ARF due to ischemia are part of a spectrum of manifestations of renal hypoperfusion. As shown in Table 269-1, prerenal ARF can complicate any disease that induces hypovolemia, low cardiac output, systemic vasodilatation, or selective renal vasoconstriction. Hypovolemia leads to a fall in mean systemic arterial pressure, which is detected as reduced stretch by arterial (e.g., carotid sinus) and cardiac baroreceptors. Activated baroreceptors trigger a coordinated series of neural and humoral responses designed to restore blood volume and arterial pressure. These include activation of the sympathetic nervous system and renin-angiotensin-aldosterone system and release of arginine vasopressin (AVP; formerly called antidiuretic hormone). Norepinephrine, angiotensin II, and AVP act in concert in an attempt to preserve cardiac and cerebral perfusion by stimulating vasoconstriction in relatively "nonessential" vascular beds, such as the musculocutaneous and splanchnic circulations, by inhibiting salt loss through sweat glands, by stimulating thirst and salt appetite, and by promoting renal salt and water retention. Glomerular perfusion, ultrafiltration pressure, and filtration rate are preserved during mild hypoperfusion through several compensatory mechanisms. Stretch receptors in afferent arterioles, in response to a reduction in perfusion pressure, trigger afferent arteriolar vasodilatation through a local myogenic reflex (autoregulation).
Biosynthesis of vasodilator prostaglandins (e.g., prostaglandin F2 and prostacyclin) is also enhanced, and these compounds preferentially dilate afferent arterioles. In addition, angiotensin II induces preferential constriction of efferent arterioles. As a result, intraglomerular pressure is maintained, the fraction of plasma flowing through glomerular capillaries that is filtered is increased (filtration fraction), and glomerular filtration rate (GFR) is preserved. During states of more severe hypoperfusion, these compensatory responses are overwhelmed and GFR falls, leading to prerenalARF. Autoregulatory dilatation of afferent arterioles is maximal at mean systemic arterial blood pressures of ~80 mmHg, and hypotension below this level is associated with a precipitous decline inGFR. Lesser degrees of hypotension may provoke prerenalARF in the elderly and in patients with diseases affecting the integrity of afferent arterioles (e.g., hypertensive nephrosclerosis, diabetic vasculopathy). In addition, drugs that interfere with adaptive responses in the renal microcirculation may convert compensated renal hypoperfusion into overt prerenal ARF or trigger progression of prerenal ARF to ischemic intrinsic renal ARF (see below). Pharmacologic inhibitors of either renal prostaglandin biosynthesis [cyclooxygenase inhibitors; nonsteroidal anti-inflammatory drugs (NSAIDs)] or angiotensin-converting enzyme (ACE) activity (ACE inhibitors) are the major culprits and should be used judiciously in the setting of suspected renal hypoperfusion. NSAIDs do not compromise GFR in healthy individuals but may precipitate prerenal ARF in patients with volume depletion or in those with chronic renal insufficiency in whom GFR is maintained, in part, through prostaglandin-mediated hyperfiltration through the remaining functional nephrons. ACE inhibitors can also compromise GFR in individuals with renal hypoperfusion and should be used with special care in patients with bilateral renal artery stenosis or unilateral stenosis in a solitary functioning kidney. Glomerular perfusion and filtration may be exquisitely dependent on the actions of angiotensin II under the latter circumstances. Angiotensin II preserves glomerular filtration pressure distal to stenoses by elevating systemic arterial pressure and by triggering selective constriction of efferent arterioles. ACE inhibitors blunt these responses and precipitate ARF, usually reversible, in ~30% of these patients. Hepatorenal Syndrome This is a particularly aggressive form ofARF that frequently complicates hepatic failure due to advanced cirrhosis or other liver diseases, including malignancy, hepatic resection, and biliary obstruction. Intrarenal vasoconstriction and avid sodium retention are early sequelae of these diseases and may be detected before changes in systemic hemodynamics. Patients with advanced liver disease, portal hypertension, and ascites also have increased plasma volume but reduced "effective" arterial blood volume as a consequence of systemic vasodilatation and pooling of blood in the portal circulation. Renal failure typically develops slowly over weeks or months in parallel with deteriorating hepatic function but may accelerate dramatically following a variety of hemodynamic insults, including hemorrhage, paracentesis, and overzealous use of diuretics, vasodilators, or cyclooxygenase inhibitors. In full-blown hepatorenal syndrome, ARF progresses even after optimization of systemic hemodynamics and systemic arterial blood volume and removal of nephrotoxins, probably as a result of ongoing intrarenal vasoconstriction, hypoperfusion, and ischemia triggered by circulating factors or neural impulses originating in the failing liver. Indeed, it must be remembered that patients with liver disease may develop other forms of ARF (e.g., sepsis, nephrotoxic medications), and a diagnosis of hepatorenal syndrome should be
made only after exclusion of other possible reversible causes. INTRINSIC RENAL ARF (INTRINSIC RENAL AZOTEMIA) Intrinsic renal ARF can complicate many diverse diseases of the renal parenchyma. From a clinicopathologic viewpoint, it is useful to divide the causes of intrinsic renal ARF into (1) diseases of large renal vessels, (2) diseases of the renal microcirculation and glomeruli, (3) ischemic and nephrotoxic ARF, and (4) tubulointerstitial diseases (Table 269-1). Most intrinsic renal ARF is triggered by ischemia (ischemic ARF) or nephrotoxins (nephrotoxic ARF), insults that classically induce acute tubular necrosis (ATN). Accordingly, the terms ARF and ATN are usually used interchangeably in these settings. However, as many as 20 to 30% of patients with ischemic or nephrotoxic ARF do not have clinical (granular or tubular cell urinary casts) or morphologic evidence of tubular necrosis, underscoring the role of sublethal injury to tubular epithelium and injury to other renal cells (e.g., endothelial cells) in the pathophysiology of this syndrome. Etiology and Pathophysiology of IschemicARFPrerenal ARF and ischemic ARF are part of a spectrum of manifestations of renal hypoperfusion. Ischemic ARF differs from prerenal ARF in that the hypoperfusion induces ischemic injury to renal parenchymal cells, particularly tubular epithelium, and recovery typically takes 1 to 2 weeks after normalization of renal perfusion as it requires repair and regeneration of renal cells. In its most extreme form, ischemia leads to bilateral renal cortical necrosis and irreversible renal failure. Ischemic ARF occurs most frequently in patients undergoing major cardiovascular surgery or suffering severe trauma, hemorrhage, sepsis, and/or volume depletion (Table 269-1). Ischemic ARF can also complicate milder forms of true hypovolemia or reduced "effective" arterial blood volume if they occur in the presence of other insults (e.g., nephrotoxins or sepsis) or in patients with compromised autoregulatory defense mechanisms or preexisting renal disease. The course of ischemicARF is typically characterized by three phases: the initiation, maintenance, and recovery phases. The initiation phase (hours to days) is the initial period of renal hypoperfusion during which ischemic injury is evolving.GFRdeclines because (1) glomerular ultrafiltration pressure is reduced as a consequence of the fall in renal blood flow, (2) the flow of glomerular filtrate within tubules is obstructed by casts comprising epithelial cells and necrotic debris derived from ischemic tubule epithelium, and (3) there is backleak of glomerular filtrate through injured tubular epithelium (Fig. 269-1). Ischemic injury is most prominent in the terminal medullary portion of the proximal tubule (S3segment, pars recta) and the medullary portion of the thick ascending limb of the loop of Henle. Both segments have high rates of active (ATP-dependent) solute transport and oxygen consumption and are located in a zone of the kidney (the outer medulla) that is relatively ischemic, even under basal conditions, by virtue of the unique countercurrent arrangement of the medullary vasculature. Cellular ischemia results in a series of alterations in energetics, ion transport, and membrane integrity that ultimately lead to cell injury and, if severe, cell apoptosis or necrosis. These alterations include depletion of ATP, inhibition of active sodium transport and transport of other solutes, impairment of cell volume regulation and cell swelling, cytoskeletal disruption and loss of cell polarity, cell-cell and cell-matrix attachment, accumulation of intracellular calcium, altered phospholipid metabolism, oxygen free radical formation, and peroxidation of membrane lipids. Importantly, renal
injury can be limited by restoration of renal blood flow during this period. The initiation phase is followed by a maintenance phase (typically 1 to 2 weeks) during which renal cell injury is established,GFRstabilizes at its nadir (typically 5 to 10 mL/min), urine output is lowest, and uremic complications arise (see below). The reasons why the GFR remains low during this phase, despite correction of systemic hemodynamics, are still being defined. Putative mechanisms include persistent intrarenal vasoconstriction and medullary ischemia triggered by dysregulated release of vasoactive mediators from injured endothelial cells (e.g., decreased nitric oxide, increased endothelin-1 and platelet-activating factor), congestion of medullary blood vessels, and reperfusion injury induced by reactive oxygen species and other mediators derived from leukocytes or renal parenchymal cells (Fig. 269-1). In addition, epithelial cell injury per se may contribute to persistent intrarenal vasoconstriction by a process termed tubuloglomerular feedback. Specialized epithelial cells in the macula densa region of distal tubules detect increases in distal salt (probably chloride) delivery that occur as a consequence of impaired reabsorption by more proximal nephron segments. Macula densa cells in turn stimulate constriction of adjacent afferent arterioles by a poorly defined mechanism and further compromise glomerular perfusion and filtration, thereby contributing to a vicious cycle. A recovery phase is characterized by renal parenchymal cell, particularly tubule epithelial cell, repair and regeneration and a gradual return of GFR to or towards premorbid levels. The recovery phase may be complicated by a marked diuretic phase due to excretion of retained salt and water and other solutes, continued use of diuretics, and/or delayed recovery of epithelial cell function (solute and water reabsorption) relative to glomerular filtration (see below). Etiology and Pathophysiology of NephrotoxicARFAcute intrinsic renal ARF can complicate exposure to many structurally diverse pharmacologic agents (Table 269-1). With most nephrotoxins, the incidence of ARF is increased in the elderly and in patients with preexisting chronic renal insufficiency, true or "effective" hypovolemia, or concomitant exposure to other toxins. Intrarenal vasoconstriction is a pivotal event inARFtriggered by radiocontrast agents (contrast nephropathy) and cyclosporine. In keeping with this pathophysiology, both agents induce ARF that shares features with prerenal ARF: namely, an acute fall in renal blood flow andGFR, a relatively benign urine sediment, and a low fractional excretion of sodium (see below). Severe cases may show clinical or pathologic evidence ofATN. Contrast nephropathy classically presents as an acute (onset within 24 to 48 h) but reversible (peak 3 to 5 days, resolution within 1 week) rise in blood urea nitrogen and creatinine and is most common in individuals with preexisting chronic renal insufficiency, diabetes mellitus, congestive heart failure, hypovolemia, or multiple myeloma. The syndrome appears to be dose-related, and its incidence is only slightly reduced in high-risk individuals by use of more expensive low osmolality, nonionic contrast agents. Endothelin-1, a potent vasoconstrictor peptide released from endothelial cells, is an important mediator of intrarenal vasoconstriction and mesangial cell contraction in this setting. Endothelin-1 has also been implicated as an important mediator of cyclosporine-induced ARF. Direct toxicity to tubule epithelial cells and/or intratubular obstruction are major pathophysiologic events inARFinduced by many antibiotics and anticancer drugs.
Frequent offenders are the antimicrobial agents, such as acyclovir, foscarnet, aminoglycosides, amphotericin B, and pentamidine, and chemotherapeutic agents, such as cisplatin and ifosfamide. ARF complicates 10 to 30% of courses of aminoglycoside antibiotics, even in the presence of therapeutic levels. Amphotericin B causes dose-related ARF through intrarenal vasoconstriction and direct toxicity to proximal tubule epithelium. Cisplatin, like the aminoglycosides, is accumulated by proximal tubule cells and typically provokes ARF after 7 to 10 days of exposure by inducing mitochondrial injury, inhibition of ATPase activity and solute transport, free radical-mediated injury to cell membranes, apoptosis and/or necrosis. The most common endogenous nephrotoxins are calcium, myoglobin, hemoglobin, urate, oxalate, and myeloma light chains. Hypercalcemia can compromiseGFR, predominantly by inducing intrarenal vasoconstriction. Calcium phosphate deposition within the kidney may also contribute. Both rhabdomyolysis and hemolysis can induceARF, particularly in hypovolemic or acidotic individuals. Myoglobinuric ARF complicates approximately 30% of cases of rhabdomyolysis. Common causes of the latter include traumatic crush injury, acute muscle ischemia, seizures, excessive exercise, heat stroke or malignant hyperthermia, intoxications (e.g., alcohol, cocaine), and infectious or metabolic disorders. ARF due to hemolysis is relatively rare and is observed following massive blood transfusion reactions. It has been postulated that myoglobin and hemoglobin or other compounds released from muscle or red blood cells cause ARF via toxic effects on tubule epithelial cells or by inducing intratubular cast formation. Hypovolemia or acidosis may contribute to the pathogenesis of ARF in this setting by promoting intratubular cast formation. In addition, both hemoglobin and myoglobin are potent inhibitors of nitric oxide bioactivity and may trigger intrarenal vasoconstriction and ischemia in patients with borderline renal hypoperfusion. The formation of intratubular casts containing filtered immunoglobulin light chains and other proteins, including Tamm-Horsfall protein produced by thick ascending limb cells, is the major trigger for ARF in patients with multiple myeloma (myeloma cast nephropathy). Light chains may also be directly toxic to tubule epithelial cells. Intratubular obstruction may also be an important cause of ARF in patients with severe hyperuricosuria or hyperoxaluria. Acute uric acid nephropathy typically complicates treatment of lymphoproliferative or myeloproliferative disorders but occasionally occurs in other forms of primary or secondary hyperuricemia if the urine is concentrated. Pathology of Ischemic and Nephrotoxic ARF The classic pathologic features of ischemic ARF are patchy and focal necrosis of tubule epithelium with detachment from its basement membrane and occlusion of tubule lumens with casts composed of intact or degenerating epithelial cells, cellular debris, Tamm-Horsfall mucoprotein, and pigments. Leukocyte accumulation is frequently observed in vasa recta; however, the morphology of the glomeruli and renal vasculature is characteristically normal. Necrosis is most severe in the straight portion (pars recta) of proximal tubules but may also affect the medullary thick ascending limb of the loop of Henle. In nephrotoxicARF, morphologic changes tend to be most prominent in both the convoluted and straight portions of proximal tubules. Tubule cell necrosis is less pronounced than in ischemic ARF. Other Causes of Intrinsic Renal ARF Patients with advanced atherosclerosis can
develop ARF after manipulation of the aorta or renal arteries at surgery or angiography, following trauma, or, rarely, spontaneously due to embolization of cholesterol crystals to the renal vasculature (atheroembolic ARF). Cholesterol crystals lodge in small- and medium-sized arteries and incite a giant cell and fibrotic reaction in the vessel wall with narrowing or obstruction of the vessel lumen. Atheroembolic ARF is frequently irreversible. A myriad of structurally diverse pharmacologic agents induce ARF by triggering allergic interstitial nephritis, a disease characterized by infiltration of the tubulointerstitium by granulocytes (typically but not invariably eosinophils), macrophages, and/or lymphocytes and by interstitial edema. The most common offenders are antibiotics (e.g., penicillins, cephalosporins, trimethoprim, sulfonamides, rifampicin) and NSAIDs (Table 269-1). POSTRENALARF(See also Chap. 281) Urinary tract obstruction accounts for fewer than 5% of cases of ARF. Since one kidney has sufficient clearance capacity to excrete nitrogenous waste products, ARF from obstruction requires either obstruction to urine flow between the external urethral meatus and bladder neck, bilateral ureteric obstruction, or unilateral ureteric obstruction in a patient with one functioning kidney or preexisting chronic renal insufficiency. Bladder neck obstruction represents the most common cause of postrenal ARF and is usually due to prostatic disease (e.g., hypertrophy, neoplasia, or infection), neurogenic bladder, or therapy with anticholinergic drugs. Less common causes of acute lower urinary tract obstruction include blood clots, calculi, and urethritis with spasm. Ureteric obstruction may result from intraluminal obstruction (e.g., calculi, blood clots, sloughed renal papillae), infiltration of the ureteric wall (e.g., neoplasia), or external compression (e.g., retroperitoneal fibrosis, neoplasia or abscess, inadvertent surgical ligature). During the early stages of obstruction (hours to days), continued glomerular filtration leads to increased intraluminal pressure upstream to the site of obstruction. As a result there are gradual distention of proximal ureter, renal pelvis, and calyces and a fall inGFR. Acute obstruction is initially associated with modest increase in renal blood flow, but arteriolar vasoconstriction soon supervenes, leading to a further decline in glomerular filtration. CLINICAL FEATURES AND DIFFERENTIAL DIAGNOSIS Patients presenting with renal failure should be assessed initially to determine if the decline inGFR is acute or chronic. An acute process is easily established if a review of laboratory records reveals a recent rise in blood urea and creatinine levels, but previous measurements are not always available. Findings that suggest chronic renal failure (Chap. 270) include anemia, neuropathy, and radiologic evidence of renal osteodystrophy or small scarred kidneys. However, it should be noted that anemia may also complicateARF (see below), and renal size may be normal or increased in several chronic renal diseases (e.g., diabetic nephropathy, amyloidosis, polycystic kidney disease). Once a diagnosis of ARF has been established, several issues should be addressed promptly: (1) the identification of the cause of ARF, (2) the elimination of the triggering insult (e.g., nephrotoxin) and/or institution of disease-specific therapies, and (3) the prevention and management of uremic complications. CLINICAL ASSESSMENT
Clinical clues to prerenalARF are symptoms of thirst and orthostatic dizziness and physical evidence of orthostatic hypotension and tachycardia, reduced jugular venous pressure, decreased skin turgor, dry mucous membranes, and reduced axillary sweating. Case records should be reviewed for documentation of a progressive fall in urine output and body weight and treatment with NSAIDs orACEinhibitors. Careful clinical examination may reveal stigmata of chronic liver disease and portal hypertension, advanced cardiac failure, sepsis, or other causes of reduced "effective" arterial blood volume (Table 269-1). Intrinsic renalARF due to ischemia is likely following severe renal hypoperfusion complicating hypovolemic or septic shock or following major surgery. The likelihood of ischemic ARF is increased further if ARF persists despite normalization of systemic hemodynamics. Diagnosis of nephrotoxic ARF requires careful review of the clinical data and pharmacy, nursing, and radiology records for evidence of recent exposure to nephrotoxic medications or radiocontrast agents or to endogenous toxins (e.g., myoglobin, hemoglobin, uric acid, myeloma protein, or elevated levels of serum calcium). Although ischemic and nephrotoxicARFaccount for more than 90% of cases of intrinsic renal ARF, other renal parenchymal diseases must be considered (Table 269-2). Flank pain may be a prominent symptom following occlusion of a renal artery or vein and with other parenchymal diseases distending the renal capsule (e.g., severe glomerulonephritis or pyelonephritis). Subcutaneous nodules, livido reticularis, bright orange retinal arteriolar plaques, and digital ischemia, despite palpable pedal pulses, are clues to atheroembolization. ARF in association with oliguria, edema, hypertension, and an "active" urine sediment (nephritic syndrome) suggests acute glomerulonephritis or vasculitis. Malignant hypertension is a likely cause of ARF in patients with severe hypertension and evidence of hypertensive injury to other organs (e.g., left ventricular hypertrophy and failure, hypertensive retinopathy and papilledema, neurologic dysfunction). Fever, arthralgias, and a pruritic erythematous rash following exposure to a new drug suggest allergic interstitial nephritis, although systemic features of hypersensitivity are frequently absent. PostrenalARFpresents with suprapubic and flank pain due to distention of the bladder and of the renal collecting system and capsule, respectively. Colicky flank pain radiating to the groin suggests acute ureteric obstruction. Prostatic disease is likely if there is a history of nocturia, frequency, and hesitancy and enlargement or induration of the prostate on rectal examination. Neurogenic bladder should be suspected in patients receiving anticholinergic medications or with physical evidence of autonomic dysfunction. Definitive diagnosis of postrenal ARF hinges on judicious use of radiologic investigations and rapid improvement in renal function following relief of obstruction. URINALYSIS Anuria suggests complete urinary tract obstruction but may complicate severe cases of prerenal or intrinsic renal ARF. Wide fluctuations in urine output raise the possibility of intermittent obstruction, whereas patients with partial urinary tract obstruction can present with polyuria due to impairment of urine concentrating mechanisms.
In prerenalARF, the sediment is characteristically acellular and contains transparent hyaline casts ("bland," "benign," "inactive" urine sediment). Hyaline casts are formed in concentrated urine from normal constitutents of urine -- principally Tamm-Horsfall protein, which is secreted by epithelial cells of the loop of Henle. Postrenal ARF may also present with an inactive sediment, although hematuria and pyuria are common in patients with intraluminal obstruction or prostatic disease. Pigmented "muddy brown" granular casts and casts containing tubule epithelial cells are characteristic ofATN and suggest ischemic or nephrotoxic ARF. They are usually found in association with microscopic hematuria and mild "tubular" proteinuria (5% of urine leukocytes) is a common finding (~90%) in antibiotic-induced allergic interstitial nephritis when studied using Hansel's stain; however, lymphocytes may predominate in allergic interstitial nephritis induced by NSAIDs. Eosinophiluria is also a feature of atheroembolic ARF. Occasional uric acid crystals (pleomorphic in shape) are common in the concentrated urine of prerenal ARF but suggest acute urate nephropathy if seen in abundance. Oxalate (envelope-shaped) and hippurate (needle-shaped) crystals raise the possibility of ethylene glycol ingestion and toxicity. Increased urine protein excretion, but 1 g/d suggests injury to the glomerular ultrafiltration barrier ("glomerular proteinuria") or excretion of myeloma light chains. The latter are not detected by conventional dipsticks (which detect albumin) and must be sought by other means (e.g., sulfosalicylic acid test, immunoelectrophoresis). Heavy proteinuria is also a frequent finding (~80%) in patients who develop combined allergic interstitial nephritis and minimal change glomerulopathy when treated with NSAIDs. A similar syndrome can be triggered by ampicillin, rifampicin, or interferon a. Hemoglobinuria or myoglobinuria should be suspected if urine is strongly positive for heme by dipstick, but contains few red cells, and if the supernatant of centrifuged urine is positive for free heme. Bilirubinuria may provide a clue to the presence of hepatorenal syndrome. RENAL FAILURE INDICES Analysis of urine and blood biochemistry is particularly useful for distinguishing prerenalARF from ischemic or nephrotoxic intrinsic renal ARF (Table 269-3). The fractional excretion of sodium (FENa) is most useful in this regard. The FE Narelates sodium clearance to creatinine clearance. Sodium is reabsorbed avidly from glomerular filtrate in patients with prerenal ARF, in an attempt to restore intravascular volume, but not in patients with ischemic or nephrotoxic intrinsic ARF, as a result of tubular epithelial cell injury. In contrast, creatinine is not reabsorbed in either setting. Consequently, patients with prerenal ARF typically have a FENa of1.0% in prerenalARF if patients are receiving diuretics or have bicarbonaturia (accompanied by sodium to maintain electroneutrality), preexisting chronic renal failure complicated by salt wasting, or adrenal insufficiency. In contrast, the FENa is15 mg/dL)] in association with hyperkalemia, hyperphosphatemia, and increased circulating levels of intracellular enzymes such as lactate dehydrogenase may indicate acute urate nephropathy and tumor lysis syndrome following cancer chemotherapy. A wide serum anion and osmolal gap (measured serum osmolality minus the serum osmolality calculated from serum sodium, glucose, and urea concentrations) indicate the presence of an unusual anion or osmole in the circulation and are clues to diagnosis of ethylene glycol or methanol ingestion. Severe anemia in the absence of hemorrhage raises the possibility of hemolysis, multiple myeloma, or thrombotic microangiopathy. Systemic eosinophilia suggests allergic interstitial nephritis but is also a feature of atheroembolic disease and polyangiitis nodosa. RADIOLOGIC FINDINGS Imaging of the urinary tract by ultrasonography is useful to exclude postrenalARF. Computed tomography and magnetic resonance imaging are alternative imaging modalities. Whereas pelvicalyceal dilatation is usual with urinary tract obstruction (98% sensitivity), dilatation may be absent immediately following obstruction or in patients with ureteric encasement (e.g., retroperitoneal fibrosis, neoplasia). Retrograde or anterograde pyelography are more definitive investigations in complex cases and
provide precise localization of the site of obstruction. A plain film of the abdomen, with tomography if necessary, is a valuable initial screening technique in patients with suspected nephrolithiasis. Doppler ultrasonography and magnetic resonance flow imaging appear promising for assessment of patency of renal arteries and veins in patients with suspected vascular obstruction; however, contrast angiography is usually required for definitive diagnosis. RENAL BIOPSY Biopsy is reserved for patients in whom prerenal and postrenalARF have been excluded and the cause of intrinsic renal ARF is unclear. Renal biopsy is particularly useful when clinical assessment and laboratory investigations suggest diagnoses other than ischemic or nephrotoxic injury that may respond to disease-specific therapy. Examples include glomerulonephritis, vasculitis, hemolytic-uremic syndrome, thrombotic thrombocytopenic purpura, and allergic interstitial nephritis. COMPLICATIONS ARFimpairs renal excretion of sodium, potassium, and water and perturbs divalent cation homeostasis and urinary acidification mechanisms. As a result, ARF is frequently complicated by intravascular volume overload, hyponatremia, hyperkalemia, hyperphosphatemia, hypocalcemia, hypermagnesemia, and metabolic acidosis. In addition, patients are unable to excrete nitrogenous waste products and are prone to develop the uremic syndrome (Chap. 270). The speed of development and the severity of these complications reflect the degree of renal impairment and catabolic state of the patient. Expansion of extracellular fluid volume is an inevitable consequence of diminished salt and water excretion in oliguric or anuric individuals. Whereas milder forms are characterized by weight gain, bibasilar lung rales, raised jugular venous pressure, and dependent edema, continued volume expansion may precipitate life-threatening pulmonary edema. Hypervolemia may be particularly problematic in patients receiving multiple intravenous medications and enteral or parenteral nutrition. Excessive administration of free water either through ingestion and nasogastric administration or as hypotonic saline or isotonic dextrose solutions (dextrose being metabolized) can induce hypoosmolality and hyponatremia, which, if severe, lead to cerebral edema and neurologic abnormalities, including seizures. Hyperkalemia is a frequent complication ofARF. Serum potassium typically rises by 0.5 mmol/L per day in oliguric and anuric patients due to impaired excretion of ingested or infused potassium and potassium released from injured tissue. Coexistent metabolic acidosis may exacerbate hyperkalemia by promoting potassium efflux from cells. Hyperkalemia may be particularly severe, even at the time of diagnosis, in patients with rhabdomyolysis, hemolysis, and tumor lysis syndrome. Mild hyperkalemia (24,000 mmol of Na+ are filtered per day. An overwhelming fraction of this Na+ load is reabsorbed by the tubules, leaving only a small fraction (usually90% of patients with anti-GBM nephritis by specific immunoassay. If immunoassays are not available, circulating anti-GBM antibodies can be detected in 60 to 80% of patients by indirect immunofluorescence, i.e., by incubating the patient's serum with stored sections of normal human kidneys. Complement levels are normal. About 20% of patients have low titers ofANCA, usually a perinuclear ANCA (Chap. 275), the pathophysiologic significance of which is unclear. Occasional patients have a positive cytoplasmic ANCA, which may signal the presence of coexistent extraglomerular renal vasculitis. Patients with lung involvement frequently have microcytic, hypochromic, iron-deficiency anemia from alveolar hemorrhage, and abnormal bilateral hilar and basilar interstitial shadowing on chest x-ray that may be difficult to distinguish from pulmonary edema or infection. The diffusion capacity for carbon dioxide is a useful tool for distinguishing among the latter diagnoses, being increased in patients with lung hemorrhage due to uptake of carbon monoxide by alveolar blood, and reduced in patients with infection or pulmonary edema. Renal biopsy is the gold standard for diagnosis of anti-GBMnephritis. The typical morphologic pattern on light microscopy is diffuse proliferative glomerulonephritis, with focal necrotizing lesions and crescents in >50% of glomeruli (crescentic glomerulonephritis). Immunofluorescence microscopy reveals linear ribbon-like deposition of IgG along the GBM (Fig. 274-2). C3 is present in the same distribution in 70% of patients. Prominent IgG deposition along the tubule basement membrane and tubulointerstitial inflammation is found occasionally. Electron microscopy reveals nonspecific inflammatory changes without immune deposits. Typical features on lung biopsy include alveolar hemorrhage, disruption of alveolar septa, hemosiderin-laden macrophages, and linear staining of IgG along the alveolar capillary basement membrane. It should be noted that Goodpasture's syndrome is not the only cause of the pulmonary-renal syndrome (i.e., renal failure and lung hemorrhage). Other important causes of this clinical complex include severe cardiac failure complicated by pulmonary edema (often blood-tinged) and prerenal azotemia; renal failure from any cause complicated by hypervolemia and pulmonary edema; immune complex-mediated vasculitides such as systemic lupus erythematosus (SLE), Henoch-Schonlein purpura, and cryoglobulinemia; pauci-immune vasculitides such as Wegener's granulomatosis and polyarteritis nodosa; infections such as Legionnaire's disease; and renal vein thrombosis with pulmonary embolism. In general, these disorders can be differentiated
by astute analysis of the clinical, serologic, and histopathologic findings. TREATMENT Prior to the introduction of immunosuppressive therapy, greater than 80% of patients with anti-GBMnephritis developedESRDwithin 1 year, and many patients died from pulmonary hemorrhage or complications of uremia. With early and aggressive use of plasmapheresis, glucocorticoids, cyclophosphamide, and azathioprine, renal and patient survival have improved dramatically. In general, emergency plasmapheresis is performed daily or on alternate days until anti-GBM antibodies are not detected in the circulation (usually 1 to 2 weeks). Prednisone (1 mg/kg per day) is started simultaneously, in combination with either cyclophosphamide (2 to 3 mg/kg per day) or azathioprine (1 to 2 mg/kg per day) to suppress new synthesis of anti-GBM antibodies. The speed of initiation of therapy is a critical determinant of outcome. One-year renal survival approaches 90% if treatment is started before serum creatinine exceeds 442 umol/L (5 mg/dL) and falls to about 10% if renal failure is more advanced. Patients who require dialysis at presentation rarely recover renal function. Serial anti-GBM titers are monitored to gauge response to therapy. Relapses are not unusual and are often heralded by rising antibody titers. In patients with ESRD, renal transplantation is a viable treatment option. Recurrence of anti-GBM nephritis in the allograft is extremely unusual provided that anti-GBM antibody titers have been consistently negative for 2 to 3 months prior to transplantation. However, in occasional patients with Alport's syndrome, when the allograft presents normal GBM components to the immune system of the recipient for the first time, anti-GBM nephritis can occur de novo in renal allografts. Pauci-Immune Glomerulonephritis The major pauci-immune glomerulonephritides are idiopathic renal-limited crescentic glomerulonephritis, microscopic polyarteritis nodosa, and Wegener's granulomatosis (Fig. 274-1).RPGN is a more common clinical presentation than acute nephritic syndrome, and the usual pathology is necrotizing glomerulonephritis with crescents affecting>50% of glomeruli (crescentic glomerulonephritis). The marked overlap of clinical features and glomerular histopathology, and the presence of circulatingANCA in most patients, suggest that these entities are a spectrum of a single disease. Here, we focus on idiopathic renal-limited crescentic glomerulonephritis.*The ANCA-associated glomerulopathies with extrarenal features, namely Wegener's granulomatosis, Churg-Strauss syndrome, and microscopic polyarteritis nodosa, are discussed in Chap. 275. Idiopathic Renal-Limited Crescentic Glomerulonephritis This is more common in middle-aged and older patients and shows a slight male preponderance. Patients typically present withRPGN, nephritic syndrome being rare.ANCA, usually a perinuclear ANCA IgG with specificity for myeloperoxidase (Chap. 275), are detected in 70 to 90% of patients (Fig. 274-1). The erythrocyte sedimentation rate and C-reactive protein levels may be elevated; however, C3 levels are typically normal, and circulating immune complexes, cryoglobulins, and anti-GBMantibodies are not detected. Most patients have crescents on light microscopy, often associated with necrotizing glomerulonephritis. Immune deposits are scanty or absent. Immunofluorescence microscopy reveals abundant fibrin deposits within crescents (Fig. 274-2). Most cases are treated aggressively with glucocorticoids, with or without cyclophosphamide or azathioprine (Chap. 275).
NEPHROTIC SYNDROME GENERAL FEATURES AND COMPLICATIONS The nephrotic syndrome is a clinical complex characterized by a number of renal and extrarenal features, the most prominent of which are proteinuria of>3.5 g per 1.73 m2 per 24 h (in practice, >3.0 to 3.5 g per 24 h), hypoalbuminemia, edema, hyperlipidemia, lipiduria, and hypercoagulability. It should be stressed that the key component is proteinuria, which results from altered permeability of the glomerular filtration barrier for protein, namely theGBM and the podocytes and their slit diaphragms. The other components of the nephrotic syndrome and the ensuing metabolic complications are all secondary to urine protein loss and can occur with lesser degrees of proteinuria or may be absent even in patients with massive proteinuria. In general, the greater the proteinuria, the lower the serum albumin level. Hypoalbuminemia is compounded further by increased renal catabolism and inadequate, albeit usually increased, hepatic synthesis of albumin. The pathophysiology of edema formation in nephrotic syndrome is poorly understood. The underfilling hypothesis postulates that hypoalbuminemia results in decreased intravascular oncotic pressure, leading to leakage of extracellular fluid from blood to the interstitium. Intravascular volume falls, thereby stimulating activation of the renin-angiotensin-aldosterone axis and the sympathetic nervous system and release of vasopressin (antidiuretic hormone), and suppressing atrial natriuretic peptide release. These neural and hormonal responses promote renal salt and water retention, thereby restoring intravascular volume and triggering further leakage of fluid to the interstitium. This hypothesis does not, however, explain the occurrence of edema in many patients in whom plasma volume is expanded and the renin-angiotensin-aldosterone axis is suppressed. The latter finding suggests that primary renal salt and water retention may also contribute to edema formation in some cases. Hyperlipidemia is believed to be a consequence of increased hepatic lipoprotein synthesis that is triggered by reduced oncotic pressure and may be compounded by increased urinary loss of proteins that regulate lipid homeostasis. Low-density lipoproteins and cholesterol are increased in the majority of patients, whereas very low density lipoproteins and triglycerides tend to rise in patients with severe disease. Although not proven conclusively, hyperlipidemia may accelerate atherosclerosis and progression of renal disease. Hypercoagulability is probably multifactorial in origin and is caused, at least in part, by increased urinary loss of antithrombin III, altered levels and/or activity of proteins C and S, hyperfibrinogenemia due to increased hepatic synthesis, impaired fibrinolysis, and increased platelet aggregability. As a consequence of these perturbations, patients can develop spontaneous peripheral arterial or venous thrombosis, renal vein thrombosis, and pulmonary embolism. Clinical features that suggest acute renal vein thrombosis include sudden onset of flank or abdominal pain, gross hematuria, a left-sided varicocele (the left testicular vein drains into the renal vein), increased proteinuria, and an acute decline inGFR. Chronic renal vein thrombosis is usually asymptomatic. Renal vein thrombosis is particularly common (up to 40%) in patients with nephrotic syndrome
due to membranous glomerulopathy, membranoproliferative glomerulonephritis, and amyloidosis. Other metabolic complications of nephrotic syndrome include protein malnutrition and iron-resistant microcytic hypochromic anemia due to transferrin loss. Hypocalcemia and secondary hyperparathyroidism can occur as a consequence of vitamin D deficiency due to enhanced urinary excretion of cholecalciferol-binding protein, whereas loss of thyroxine-binding globulin can result in depressed thyroxine levels. An increased susceptibility to infection may reflect low levels of IgG that result from urinary loss and increased catabolism. In addition, patients are prone to unpredictable changes in the pharmacokinetics of therapeutic agents that are normally bound to plasma proteins. ETIOLOGY AND DIFFERENTIAL DIAGNOSIS Proteinuria>150 mg per 24 h is abnormal and can result from a number of mechanisms. Glomerular proteinuria results from leakage of plasma proteins through a perturbed glomerular filtration barrier; tubular proteinuria results from failure of tubular reabsorption of low-molecular-weight plasma proteins that are normally filtered and then reabsorbed and metabolized by tubular epithelium; overflow proteinuria results from filtration of proteins, usually immunoglobulin light chains, that are present in excess in the circulation. Tubular proteinuria virtually never exceeds 2 g per 24 h and thus, by definition, never causes nephrotic syndrome. Overflow proteinuria should be suspected in patients with clinical or laboratory evidence of multiple myeloma or other lymphoproliferative malignancy. Suspicion is heightened when there is a discrepancy between proteinuria detected by dipsticks, which are sensitive to albumin but not light chains, and the sulfosalicylic acid precipitation method, which detects both. Nephrotic syndrome can complicate any disease that perturbs the negative electrostatic charge or architecture of theGBM and the podocytes and their slit diaphragms. Six entities account for greater than 90% of cases of nephrotic syndrome in adults: minimal change disease (MCD), focal and segmental glomerulosclerosis (FSGS), membranous glomerulopathy,MPGN, diabetic nephropathy, and amyloidosis. Diabetic nephropathy and amyloidosis, being manifestations of systemic diseases, are discussed inChap. 275. Renal biopsy is a valuable tool in adults with nephrotic syndrome for establishing a definitive diagnosis, guiding therapy, and estimating prognosis. Renal biopsy is not required in the majority of children with nephrotic syndrome as most cases are due to MCD and respond to empiric treatment with glucocorticoids. MINIMAL CHANGE DISEASE This glomerulopathy accounts for about 80% of nephrotic syndrome in children of younger than 16 years and 20% in adults (Table 274-1). The peak incidence is between 6 and 8 years. Patients typically present with nephrotic syndrome and benign urinary sediment. Microscopic hematuria is present in 20 to 30%. Hypertension and renal insufficiency are very rare. MCD(also called nil disease, lipoid nephrosis, or foot process disease) is so named because glomerular size and architecture are normal by light microscopy. Immunofluorescence studies are typically negative for immunoglobulin and C3. Mild
mesangial hypercellularity and sparse deposits of C3 and IgM may be detected. Occasionally, mesangial proliferation is associated with scanty IgA deposits, similar to those found in IgA nephropathy. However, the natural history of this variant and response to therapy resemble classic MCD. Electron microscopy reveals characteristic diffuse effacement of the foot processes of visceral epithelial cells (Fig. 274-3). This morphologic finding is referred to as foot process fusion in the older literature. The etiology ofMCD is unknown and the vast majority of cases are idiopathic (Table 274-1). MCD occasionally develops after upper respiratory tract infection, immunizations, and atopic attacks. Patients with atopy and MCD have an increased incidence of HLA-B12, suggesting a genetic predisposition. MCD, often in association with interstitial nephritis, is a rare side effect of nonsteroidal anti-inflammatory drugs (NSAIDs), rifampin, and interferon-a. The occasional association with lymphoproliferative malignancies (such as Hodgkin's lymphoma), the tendency for idiopathic MCD to remit during intercurrent viral infection such as measles, and the good response of idiopathic forms to immunosuppressive agents (see below) suggest an immune etiology. In children, the urine contains albumin principally and minimal amounts of higher molecular weight proteins such as IgG anda 2-macroglobulin. This selective proteinuria in conjunction with foot process effacement suggests injury to podocytes and loss of the fixed negative charge in the glomerular filtration barrier for protein. Proteinuria is typically nonselective in adults, suggesting more extensive perturbation of membrane permeability. TREATMENT MCDis highly steroid-responsive and carries an excellent prognosis. Spontaneous remission occurs in 30 to 40% of childhood cases but is less common in adults. Approximately 90% of children and 50% of adults enter remission following 8 weeks of high-dose oral glucocorticoids. In a typical regimen using prednisone, children receive 60 mg/m2 of body surface area daily for 4 weeks, followed by 40 mg/m2 on alternate days for an additional 4 weeks; adults receive 1 to 1.5 mg/kg body weight per day for 4 weeks, followed by 1 mg/kg per day on alternate days for 4 weeks. Up to 90% of adults enter remission if therapy is extended for 20 to 24 weeks. Nephrotic syndrome relapses in over 50% of cases following withdrawal of glucocorticoids. Alkylating agents are reserved for the small number of patients who fail to achieve lasting remission. These include patients who relapse during or shortly after withdrawal of steroids (steroid-dependent) and those who relapse more than three times per year (frequently relapsing). In these settings, cyclophosphamide (2 to 3 mg/kg per day) or chlorambucil (0.1 to 0.2 mg/kg per day) is started after steroid-induced remission and continued for 8 to 12 weeks. Cytotoxic agents may also induce remission in occasional steroid-resistant cases. These benefits must be balanced against the risk of infertility, cystitis, alopecia, infection, and secondary malignancies, particularly in children and young adults. Azathioprine has not been proven to be a useful adjunct to steroid therapy. Cyclosporine induces remission in 60 to 80% of patients; it is an alternative to cytotoxic agents and an option in patients who are resistant to cytotoxic agents. Unfortunately, relapse is usual when cyclosporine is withdrawn, and long-term therapy carries the risk of nephrotoxicity and other side effects. Long-term renal and patient survival is excellent in MCD.
FOCAL AND SEGMENTAL GLOMERULOSCLEROSIS WITH HYALINOSIS The pathognomonic morphologic lesion inFSGS is sclerosis with hyalinosis involving portions (segmental) of fewer than 50% (focal) of glomeruli on a tissue section. The incidence of idiopathic (primary) FSGS has increased over the past two decades so that it now accounts for about one-third of cases of nephrotic syndrome in adults and as many as one-half of cases of nephrotic syndrome in blacks. FSGS can complicate a number of systemic diseases and sustained glomerular capillary hypertension following nephron loss from any cause (Table 274-2 andChap. 273). IdiopathicFSGStypically presents as nephrotic syndrome (~66%) or subnephrotic proteinuria (~33%) in association with hypertension, mild renal insufficiency, and an abnormal urine sediment that contains red blood cells and leukocytes. Proteinuria is nonselective in most cases. Light microscopy of renal biopsy tissue revealsFSGSwith entrapment of amorphous hyaline material, a process that shows a predilection for juxtamedullary glomeruli. The sclerotic scars contain areas of glomerular capillary collapse and hyaline material composed of collagen types I, III, and IV. Adhesions occur between areas of capillary collapse and Bowman's capsule. Immunofluorescence studies are usually negative. Electron microscopy reveals evidence of damage to visceral epithelial cells, including swelling and detachment of podocytes from theGBM, effacement of foot processes, transition to foam cells, and overt cell degeneration and necrosis. The etiology of primaryFSGS is unclear (Table 274-2). There is evidence that a circulating nonimmunoglobulin permeability factor triggers FSGS in at least a subgroup of patients. The latter individuals tend to be young and prone to develop early recurrence of FSGS following renal transplantation. Plasmapharesis has been employed with variable success to control the nephrotic syndrome in this group. The overlap of clinical and morphologic features betweenMCD and FSGS has prompted some authorities to speculate that they are a spectrum of morphologic manifestations of a single pathogenetic process. FSGS is a potential long-term consequence of nephron loss from any cause. It can complicate congenital renal diseases such as congenital oligomeganephronia, in which both kidneys have a reduced complement of nephrons, and congenital unilateral agenesis. In addition, FSGS may develop following acquired loss of nephrons from extensive surgical ablation of renal mass; reflux nephropathy; glomerulonephritis; interstitial nephritis; sickle cell disease; and the combined effects of ischemia, cyclosporine nephrotoxicity, and rejection on renal allograft function (Table 274-2). It appears that>50% of nephrons must be lost for development of secondary FSGS. TREATMENT In contrast toMCD, spontaneous remission of primaryFSGS is rare and renal prognosis is relatively poor. Proteinuria remits in only 20 to 40% of patients treated with glucocorticoids for 8 weeks. Uncontrolled studies suggest that up to 70% respond when steroid therapy is prolonged for 16 to 24 weeks. Cyclophosphamide and cyclosporine, when used at doses described above for MCD, induce partial or complete remission in 50 to 60% of steroid-responsive patients but are generally ineffective in steroid-resistant
cases. Poor prognostic factors at presentation include hypertension, abnormal renal function, black race, and persistent heavy proteinuria. Renal transplantation is complicated by recurrence of FSGS in the allograft in about 50% of cases and graft loss in about 10%. Factors associated with an increased risk of recurrence include a short time interval between the onset of the FSGS andESRD, young age at onset, and possibly the presence of mesangial hypercellularity on renal biopsy. MEMBRANOUS GLOMERULOPATHY This lesion is a leading cause of idiopathic nephrotic syndrome in adults (30 to 40%) and a rare cause in children (80%) present with nephrotic syndrome, proteinuria usually being nonselective. Microscopic hematuria is present in up to 50% of cases, but red blood cells casts, macroscopic hematuria, and leukocytes are extremely rare. Hypertension is documented in only 10 to 30% of patients at the outset but is common later in patients with progressive renal failure. Serologic tests such as antinuclear antibody,ANCA, anti-GBM antibody, cryoglobulin titers, and complement levels are normal in the idiopathic form. Light microscopy of renal biopsy sections reveals diffuse thickening of theGBMwithout evidence of inflammation or cellular proliferation. Silver staining demonstrates characteristic spikes along the GBM, which represent projections of new basement membrane engulfing subepithelial immune deposits. Immunofluorescence reveals granular deposition of IgG, C3, and the terminal components of complement (C5b-9) along the glomerular capillary wall. Electron-microscopic appearances vary depending on the stage of disease. The earliest finding is the presence of subepithelial immune deposits (Fig. 274-3). As these deposits enlarge, spikes of new basement membrane extend out between the immune deposits and begin to engulf them. With time, the deposits are completely surrounded and incorporated into the basement membrane. The pathogenesis of idiopathic human membranous glomerulopathy is incompletely understood. The presence of electron-dense immune deposits that contain IgG and C3 suggest an immune process. About one-third of adult membranous nephropathy occurs in association with systemic diseases such asSLE, infections such as hepatitis B, malignancy, and drug therapy with gold and penicillamine (Table 274-3). Nephrotic syndrome remits spontaneously and completely in up to 40% of patients with membranous glomerulopathy. The natural history of another 30 to 40% is characterized by repeated relapses and remissions. The final 10 to 20% suffer a slow progressive decline inGFR that typically culminates inESRDafter 10 to 15 years. Presenting features that predict a poor prognosis include male gender, older age, hypertension, severe proteinuria and hyperlipidemia, and impaired renal function. Controlled trials of glucocorticoids have failed to show consistent improvement in proteinuria or renal protection. Cyclophosphamide, chlorambucil, and cyclosporine have each been shown to reduce proteinuria and/or slow the decline in GFR in patients with progressive disease in small or uncontrolled studies. These observations need to be confirmed in
controlled prospective studies. Transplantation is a successful treatment option for patients who reach ESRD. MEMBRANOPROLIFERATIVE GLOMERULONEPHRITIS This morphologic entity, also known as mesangiocapillary glomerulonephritis, is characterized by thickening of theGBM and proliferative changes on light microscopy (Table 274-4). Two major types are identified; both are characterized by a diffuse increase in mesangial cellularity and matrix, and by thickening and reduplication of the GBM such that the lobular pattern of the glomerular tuft is exaggerated. The hallmark of type IMPGN is the presence of subendothelial and mesangial deposits on electron microscopy that contain C3 and IgG or IgM; rarely, IgA deposits are demonstrated by immunofluorescence microscopy (Fig. 274-3). The hallmark of type II MPGN (dense deposit disease) is the presence of electron-dense deposits within the GBM and other renal basement membranes (shown by electron microscopy) that stain for C3, but little or no immunoglobulin. Most patients with type IMPGNpresent with heavy proteinuria or nephrotic syndrome, active urinary sediment, and normal or mildly impairedGFR. C3 levels are usually depressed, and C1q and C4 levels are borderline or low. Type I MPGN is an immune-complex glomerulonephritis and can be associated with a variety of chronic infections (e.g., bacterial endocarditis, HIV, hepatitis B and C), systemic immune-complex diseases (e.g.,SLE, cryoglobulinemia), and malignancies (e.g., leukemias, lymphomas). Type I MPGN is a relatively benign disease, and 70 to 85% of patients survive without clinically significant impairment of GFR. There is no proven therapy for patients with progressive disease beyond eradicating the underlying infection, malignancy, or systemic disease, when possible. The incidence of type I MPGN appears to be falling, possibly because the overall incidence of hepatitis C infection has fallen dramatically in western society over the past decade. Type IIMPGN can also present with proteinuria and nephrotic syndrome; however, some patients present with nephritic syndrome,RPGN, or recurrent macroscopic hematuria. Type II MPGN is an autoimmune disease in which patients have an IgG autoantibody, termed C3 nephritic factor, that binds to C3 convertase, the enzyme that metabolizes C3, and renders it resistant to inactivation (Chap. 273). Type II MPGN runs a variable course; theGFRremains stable in some patients and declines gradually toESRDover 5 to 10 years in others. There is no effective therapy for this disease. FIBRILLARY-IMMUNOTACTOID GLOMERULOPATHY This emerging clinicopathologic entity accounts for 1% of diagnoses in most large renal biopsy series. Virtually all patients present with proteinuria, and>50% have nephrotic syndrome. The majority of patients also have hematuria, hypertension, and renal insufficiency. The light-microscopic appearances vary from mesangial expansion and basement membrane thickening withPAS-positive material to proliferative and crescentic glomerulonephritis. On electron microscopy, this PAS-positive material is observed to be composed of randomly arranged (fibrillary glomerulopathy) or organized bundles (immunotactoid glomerulopathy) of microfibrils and microtubules, the compositon of which has yet to be defined. The etiology of fibrillary-immunotactoid
glomerulopathy remains to be determined. Patients with the immunotactoid variant have an increased incidence of lymphoproliferative malignancy. There is no proven therapy for fibrillary-immunotactoid glomerulopathy, and many patients progress to ESRDover 1 to 10 years. Transplantation appears to be a viable option in the latter setting. MESANGIAL PROLIFERATIVE GLOMERULONEPHRITIS In 5 to 10% of patients with idiopathic nephrotic syndrome, renal biopsy reveals a diffuse increase in glomerular cellularity, predominantly due to proliferation of mesangial and endothelial cells, and infiltration by monocytes. Findings on immunofluorescence microscopy vary and include deposits of IgA, IgG, IgM, and/or complement, or absence of immune reactants. It is likely that this morphologic entity is, in fact, a heterogeneous group of diseases that includes atypical forms ofMCD andFSGS and milder or resolving forms of the immune-complex and pauci-immune glomerulopathies described above under nephritic syndrome andRPGN. In keeping with the heterogeneity of this diagnosis, the prognosis is variable. In general, persistent nephrotic-range proteinuria signals a poor prognosis, with many patients progressing toESRDover 10 to 20 years despite immunosuppressive therapy. TREATMENT Nephrotic Syndrome and Complications The treatment of nephrotic syndrome involves (1) specific treatment of the underlying morphologic entity and, when possible, causative disease (see above); (2) general measures to control proteinuria if remission is not achieved through immunosuppressive therapy and other specific measures; and (3) general measures to control nephrotic complications. General measures may be warranted to control proteinuria in nephrotic syndrome if patients do not respond to immunosuppressive therapy and other specific measures and suffer progressive renal failure or severe nephrotic complications. Nonspecific measures that may reduce proteinuria include angiotensin-converting enzyme (ACE) inhibitors, andNSAIDs. The first of these measures aim to reduce proteinuria and slow the rate of progression of renal failure by lowering intraglomerular pressure and preventing the development of hemodynamically mediated focal segmental glomerulosclerosis. There is conclusive evidence that ACE inhibitors are renoprotective in human diabetic nephropathy (Chap. 275) and that ACE inhibitors slow the development of secondaryFSGS in experimental animals. Their role in the treatment of nephrotic syndrome in other settings is unproven. NSAIDs also reduce proteinuria in some patients with nephrotic syndrome, probably by altering glomerular hemodynamics andGBMpermeability characteristics. This potential benefit must be balanced against the risk of inducing acute renal failure, hyperkalemia, salt and water retention, and other side effects. Complications of nephrotic syndrome that may require treatment include edema, hyperlipidemia, thromboembolism, malnutrition, and vitamin D deficiency. Edema should be managed cautiously by moderate salt restriction, usually 1 to 2 g/day, and the judicious use of loop diuretics. It is unwise to remove >1.0 kg of edema per day as more aggressive diuresis may precipitate intravascular volume depletion and prerenal azotemia. Administration of salt-poor albumin is not recommended as most is excreted
within 24 to 48 h. Whereas many nephrologists advocate lowering low-density lipoproteins and cholesterol levels with lipid-lowering drugs to prevent accelerated atherosclerosis and slow the rate of decline ofGFR, the value of such interventions in this setting has not been conclusively shown. Anticoagulation is indicated for patients with deep venous thrombosis, arterial thrombosis, and pulmonary embolism. Patients may be relatively resistant to heparin as a consequence of antithrombin III deficiency. Renal vein and vena caval angiography are probably indicated only when embolization occurs on anticoagulation and insertion of a caval filter is contemplated. There is no consensus regarding the optimal diet for patients with nephrotic syndrome. High-protein diets to prevent protein malnutrition are now in disfavor, since protein supplements have little, if any, effect on serum albumin levels and may hasten the progression of renal disease by increasing urinary protein excretion. The potential value of dietary protein restriction for reducing proteinuria must be balanced against the risk of contributing to malnutrition. Vitamin D supplementation is advisable in patients with clinical or biochemical evidence of vitamin D deficiency. ASYMPTOMATIC ABNORMALITIES OF THE URINARY SEDIMENT HEMATURIA Most asymptomatic glomerular hematuria is due to IgA nephropathy (Berger's disease) or thin basement membrane (TBM) disease (benign hematuria). A rarer but more ominous cause of isolated hematuria is Alport's syndrome. The latter is the most common form of hereditary nephritis, is usually transmitted as an X-linked dominant trait, and is associated with sensorineural deafness, ophthalmologic abnormalities, and progressive renal insufficiency (Chap. 275). TBM disease is sometimes familial but, in contrast to Alport's syndrome, is usually a benign disorder. Asymptomatic hematuria may also be the presenting feature of indolent forms of most other primary and secondary proliferative glomerulopathies (Fig. 274-1). Glomerular hematuria must be distinguished from a variety of renal parenchymal and extrarenal causes of hematuria. It is particularly important to exclude malignancy of the kidney or urinary tract, particularly in older male patients (Chap. 94). Other potential diagnoses include vascular, cystic, and tubulointerstitial diseases; papillary necrosis; hypercalciuria and hyperuricosuria; benign prostatic hypertrophy; and renal calculi. Important clues to the presence of glomerular hematuria are the presence of urinary red blood cell casts, dysmorphic urinary red blood cells, proteinuria of greater than 2.0 g per 24 h, and clinical or serologic evidence of nephritic syndrome,RPGN, or a compatible systemic disease. IgA Nephropathy (Berger's Disease) IgA nephropathy is the most common glomerulopathy worldwide and accounts for 10 to 40% of glomerulonephritis in most series (Table 274-5). The disease is particularly common in southern Europe and Asia and appears to be more common in blacks than whites. Familial clustering has been reported but is rare. No consistent HLA association has emerged, although HLA-B35 appears to be more common in French patients. Most cases are idiopathic. The renal and serologic abnormalities in IgA nephropathy and Henoch-Schonlein purpura (Chap. 275) are indistinguishable, and most authorities consider these to be a spectrum of a single disease. Less commonly, IgA nephropathy is found in association with systemic diseases, including chronic liver disease, Crohn's disease, gastrointestinal adenocarcinoma, chronic obstructive bronchiolitis, idiopathic interstitial pneumonia,
dermatitis herpetiformis, mycosis fungoides, leprosy, ankylosing spondylitis, relapsing polychondritis, and Sjogren's syndrome. In many of these conditions, IgA is deposited in the glomerulus without inducing inflammation, and this may be a clinically insignificant consequence of perturbed IgA homeostasis. Patients with IgA nephropathy typically present with gross hematuria, often 24 to 48 h after a pharyngeal or gastrointestinal infection, vaccination, or strenuous exercise. Other cases are diagnosed upon detection of microscopic hematuria during routine physical examinations. Hypertension (20 to 30%) and nephrotic syndrome (~10%) are unusual at presentation. Light microscopy of renal biopsy specimens typically shows mesangial expansion by increased matrix and cells. Diffuse proliferation, cellular crescents, interstitial inflammation, and areas of glomerulosclerosis may be evident in severe cases. The diagnostic finding, for which the disease is named, is mesangial deposition of IgA, detected by immunofluorescence microscopy. C3 is usually detected in the area of immune deposits, and IgG is observed in 50% of cases. Electron microscopy reveals electron-dense deposits in the mesangium and, in severe cases, these extend into the paramesangial subendothelial space. The pathogenesis of IgA nephropathy is incompletely understood. TREATMENT There is no proven therapy for IgA nephropathy. A recent, relatively large randomized controlled trial suggested a benefit of fish oils in patients with progressive disease and heavy proteinuria; however, this experience has not been universal. Some authorities advocate a trial of high-dose glucocorticoids with or without cytotoxic agents in patients with severe nephrotic syndrome and those with nephritic syndrome orRPGN and evidence of active inflammation on renal biopsy. IgA nephropathy typically smolders for decades, with patients often suffering exacerbations of hematuria and renal impairment during intercurrent infections. As many as 20 to 50% of patients developESRDwithin 20 years. Clinical predictors of a poor prognosis include older age, male sex, hypertension, nephrotic-range proteinuria, and renal insufficiency at presentation. Histologic features that predict an aggressive course include diffuse severe disease, extracapillary proliferation (crescents), extension of immune attack into the paramesangial subendothelial space, glomerulosclerosis, interstitial fibrosis, and arteriolar hyalinosis. Thin Basement Membrane Disease (Benign Hematuria) This disorder can be heredofamilial or sporadic and is as common as IgA nephropathy in some series of asymptomatic hematuria. When familial, it is usually inherited as an autosomal dominant trait and is due to a defect in the gene encoding the a4 chain of type IV collagen.TBMdisease typically manifests in childhood as persistent hematuria. Intermittent hematuria and exacerbation of hematuria during upper respiratory tract infections have also been reported. The kidney is normal on light and immunofluorescence microscopy. TheGBM is thin (usually $15 billion per year in the United States. GASTRIC PHYSIOLOGY Despite the constant attack on the gastroduodenal mucosa by a host of noxious agents (acid, pepsin, bile acids, pancreatic enzymes, drugs, and bacteria), integrity is maintained by an intricate system that provides mucosal defense and repair. Gastric Anatomy The gastric epithelial lining consists of rugae that contain microscopic gastric pits, each branching into four or five gastric glands made up of highly specialized epithelial cells. The makeup of gastric glands varies with their anatomic location. Glands within the gastric cardia comprise50% over the past 30 years. The reason for the reduction in the frequency of DUs is likely related to the decreasing frequency of Helicobacter pylori. Before the discovery of H. pylori, the natural history of DUs was typified by frequent recurrences after initial therapy. Eradication of H. pylori has greatly reduced these recurrence rates. Gastric Ulcers GUs tend to occur later in life than duodenal lesions, with a peak incidence reported in the sixth decade. More than half of GUs occur in males and are less common thanDUs, perhaps due to the higher likelihood of GUs being silent and presenting only after a complication develops. Autopsy studies suggest a similar incidence of DUs and GUs. Pathology Duodenal Ulcers DUsoccur most often in the first portion of duodenum (>95%), with ~90% located within 3 cm of the pylorus. They are usually £1 cm in diameter but can occasionally reach 3 to 6 cm (giant ulcer). Ulcers are sharply demarcated, with depth at
times reaching the muscularis propria. The base of the ulcer often consists of a zone of eosinophilic necrosis with surrounding fibrosis. Malignant duodenal ulcers are extremely rare. Gastric Ulcers In contrast toDUs,GUs can represent a malignancy. Benign GUs are most often found distal to the junction between the antrum and the acid secretory mucosa. This junction is variable, but in general the antral mucosa extends about two thirds of the distance of the lesser curvature and one third the way up the greater curvature. Benign GUs are quite rare in the gastric fundus and are histologically similar to DUs. Benign GUs associated with H. pylori are associated with antral gastritis. In contrast,NSAID-related GUs are not accompanied by chronic active gastritis but may instead have evidence of a chemical gastropathy. Pathophysiology It is now clear that H. pylori andNSAID-induced injury account for the majority ofDUs. Gastric acid contributes to mucosal injury but does not play a primary role. Duodenal Ulcers Many acid secretory abnormalities have been described inDUpatients (Table 285-1). Of these, average basal and nocturnal gastric acid secretion appear to be increased in DU patients as compared to control; however, the level of overlap between DU patients and control subjects is substantial. The reason for this altered secretory process is unclear, but H. pylori infection may contribute to this finding. Accelerated gastric emptying of liquids has been noted in some DU patients but is not consistently observed; its role in DU formation, if any, is unclear. Bicarbonate secretion is significantly decreased in the duodenal bulb of patients with an active DU as compared to control subjects. H. pylori infection may also play a role in this process. Gastric Ulcer As inDUs, the majority ofGUs can be attributed to either H. pylori orNSAID-induced mucosal damage. GUs that occur in the prepyloric area or those in the body associated with a DU or a duodenal scar are similar in pathogenesis to DUs. Gastric acid output (basal and stimulated) tends to be normal or decreased in GU patients. When GUs develop in the presence of minimal acid levels, impairment of mucosal defense factors may be present. Abnormalities in resting and stimulated pyloric sphincter pressure with a concomitant increase in duodenal gastric reflux have been implicated in someGUpatients. Although bile acids, lysolecithin, and pancreatic enzymes may injure gastric mucosa, a definite role for these in GU pathogenesis has not been established. Delayed gastric emptying of solids has been described in GU patients but has not been reported consistently. The observation that patients who have undergone disruption of the normal pyloric barrier (pyloroplasty, gastroenterostomy) often have superficial gastritis without frank ulceration decreases enthusiasm for duodenal gastric reflux as an explanation for GU pathogenesis. H. pylori and acid peptic disorders Gastric infection with the bacterium H. pylori accounts for the majority ofPUD. This organism also plays a role in the development of gastric mucosal-associated lymphoid tissue (MALT) lymphoma and gastric adenocarcinoma. Although the entire genome of H. pylori has been sequenced, it is still not clear how this organism, which is in the stomach, causes ulceration in the
duodenum, or whether its eradication will lead to a decrease in gastric cancer. THE BACTERIUM The bacterium, initially named Campylobacter pyloridis, is a gram-negative microaerophilic rod found most commonly in the deeper portions of the mucous gel coating the gastric mucosa or between the mucous layer and the gastric epithelium. It may attach to gastric epithelium but under normal circumstances does not appear to invade cells. It is strategically designed to live within the aggressive environment of the stomach. It is S-shaped (~0.5´ 3 um in size) and contains multiple sheathed flagella. Initially, H. pylori resides in the antrum but, over time, migrates towards the more proximal segments of the stomach. The organism is capable of transforming into a coccoid form, which represents a dormant state that may facilitate survival in adverse conditions. The bacterium expresses a host of factors that contribute to its ability to colonize the gastric mucosa and produce mucosal injury. Several of the key bacterial factors include urease (converting urea to NH3 and water, thus alkalinizing the surrounding acidic environment), catalase, lipase, adhesins, platelet-activating factor, cytotoxin-associated gene protein (Cag A), pic B (induces cytokines), and vacuolating cytotoxin (Vac A). Multiple strains of H. pylori exist and are characterized by their ability to express several of these factors (Cag A, Vac A, etc.). It is possible that the different diseases related to H. pylori infection can be attributed to different strains of the organism with distinct pathogenic features. EPIDEMIOLOGY The prevalence of H. pylori varies throughout the world and depends to a great extent on the overall standard of living in the region. In developing parts of the world, 80% of the population may be infected by the age of 20. In contrast, in the United States, this organism is rare in childhood. The overall prevalence of H. pylori in the United States is ~30%, with individuals born before 1950 having a higher rate of infection than those born later. About 10% of Americans50% when compared to 30 years ago. PATHOPHYSIOLOGY H. pylori infection is virtually always associated with a chronic active gastritis, but only 10 to 15% of infected individuals develop frank peptic ulceration. The basis for this difference is unknown. Initial studies suggested that>90% of all DUs were associated with H. pylori, but H. pylori is present in only 30 to 60% of individuals withDU and 70% of patients withGU. The pathophysiology of ulcers not associated with H. pylori orNSAIDingestion [or the rare Zollinger-Ellison syndrome (ZES)] is unclear. The particular end result of H. pylori infection (gastritis,PUD, gastricMALTlymphoma, gastric cancer) is determined by a complex interplay between bacterial and host factors
(Fig. 285-6). 1. Bacterial factors: H. pylori is able to facilitate gastric residence, induce mucosal injury, and avoid host defense. Different strains of H. pylori produce different virulence factors. A specific region of the bacterial genome, the pathogenicity island, encodes the virulence factors Cag A and pic B. Vac A also contributes to pathogenicity, though it is not encoded within the pathogenicity island. These virulence factors, in conjunction with additional bacterial constituents, can cause mucosal damage. Urease, which allows the bacteria to reside in the acidic stomach, generates NH3, which can damage epithelial cells. The bacteria produce surface factors that are chemotactic for neutrophils and monocytes, which in turn contribute to epithelial cell injury (see below). H. pylori makes proteases and phospholipases that break down the glycoprotein lipid complex of the mucous gel, thus reducing the efficacy of this first line of mucosal defense. H. pylori expresses adhesins, which facilitate attachment of the bacteria to gastric epithelial cells. Although lipopolysaccharide (LPS) of gram-negative bacteria often plays an important role in the infection, H. pylori LPS has low immunologic activity compared to that of other organisms. It may promote a smoldering chronic inflammation. 2. Host factors: The host responds to H. pylori infection by mounting an inflammatory response, which contributes to gastric epithelial cell damage without providing immunity against infection. The neutrophil response is strong both in acute and chronic infection. In addition, T lymphocytes and plasma cells are components of the chronic inflammatory infiltrate, supporting the involvement of antigen-specific cellular and humoral responses. A number of cytokines are released from both epithelial and immune modulatory cells in response to H. pylori infection including the proinflammatory cytokines tumor necrosis factor (TNF)a, interleukin (IL)1a/b, IL-6, interferon (IFN)g, and granulocyte-macrophage colony stimulating factor. Several chemokines such as IL-8 and growth-regulated oncogene (GRO)a, involved in neutrophil recruitment/activation, and RANTES, which recruits mononuclear cells, have been observed in H. pylori-infected mucosa. The reason for H. pylori-mediated duodenal ulceration remains unclear. One potential explanation is that gastric metaplasia in the duodenum ofDUpatients permits H. pylori to bind to it and produce local injury secondary to the host response. Another hypothesis is that H. pylori antral infection could lead to increased acid production, increased duodenal acid, and mucosal injury. Basal and stimulated [meal, gastrin-releasing peptide (GRP)] gastrin release are increased in H. pylori-infected individuals, and somatostatin-secreting D cells may be decreased. H. pylori infection might induce increased acid secretion through both direct and indirect actions of H. pylori and proinflammatory cytokines (IL-8,TNF, and IL-1) on G, D, and parietal cells (Fig. 285-7). H. pylori infection has also been associated with decreased duodenal mucosal bicarbonate production. Data supporting and contradicting each of these interesting theories have been demonstrated. Thus, the mechanism by which H pylori infection of the stomach leads to duodenal ulceration remains to be established. NSAIDs-induced disease EPIDEMIOLOGY NSAIDsrepresent one of the most commonly used medications in the United States. More than 30 billion over-the-counter tablets and 70 million prescriptions
are sold yearly in the United States alone. The spectrum of NSAID-induced morbidity ranges from nausea and dyspepsia (prevalence reported as high as 50 to 60%) to a serious gastrointestinal complication such as frank peptic ulceration complicated by bleeding or perforation in as many as 3 to 4% of users per year. About 20,000 patients die each year from serious gastrointestinal complications from NSAIDs. Unfortunately, dyspeptic symptoms do not correlate with NSAID-induced pathology. Over 80% of patients with serious NSAID-related complications did not have preceding dyspepsia. In view of the lack of warning signs, it is important to identify patients who are at increased risk for morbidity and mortality related to NSAID usage. A summary of established and possible risk factors is presented inTable 285-3. PATHOPHYSIOLOGY Prostaglandins play a critical role in maintaining gastroduodenal mucosal integrity and repair. It therefore follows that interruption of prostaglandin synthesis can impair mucosal defense and repair, thus facilitating mucosal injury via a systemic mechanism. A summary of the pathogenetic pathways by which systemically administeredNSAIDsmay lead to mucosal injury is shown inFig. 285-8. Injury to the mucosa also occurs as a result of the topical encounter withNSAIDs. Aspirin and many NSAIDs are weak acids that remain in a nonionized lipophilic form when found within the acid environment of the stomach. Under these conditions, NSAIDs migrate across lipid membranes of epithelial cells, leading to cell injury once trapped intracellularly in an ionized form. Topical NSAIDs can also alter the surface mucous layer, permitting back diffusion of H+ and pepsin, leading to further epithelial cell damage. Miscellaneous Pathogenetic Factors in Acid Peptic Disease Cigarette smoking has been implicated in the pathogenesis ofPUD. Not only have smokers been found to have ulcers more frequently than do nonsmokers, but smoking appears to decrease healing rates, impair response to therapy, and increase ulcer-related complications such as perforation. The mechanism responsible for increased ulcer diathesis in smokers is unknown. Theories have included altered gastric emptying, decreased proximal duodenal bicarbonate production, and cigarette-induced generation of noxious mucosal free radicals. Acid secretion is not abnormal in smokers. Despite these interesting theories, a unifying mechanism for cigarette-induced peptic ulcer diathesis has not been established. Genetic predisposition has also been considered to play a role in ulcer development. First-degree relatives ofDUpatients are three times as likely to develop an ulcer; however, the potential role of H. pylori infection in contacts is a major consideration. Increased frequency of blood group O and of the nonsecretor status have also been implicated as genetic risk factors for peptic diathesis. However, H. pylori preferentially binds to group O antigens. Therefore, the role of genetic predisposition in commonPUD has not been established. Psychological stress has been thought to contribute toPUD, but studies examining the role of psychological factors in its pathogenesis have generated conflicting results. Although PUD is associated with certain personality traits (neuroticism), these same traits are also present in individuals with nonulcer dyspepsia (NUD) and other functional and organic disorders. Although more work in this area is needed, no typical PUD
personality has been found. Diet has also been thought to play a role in peptic diseases. Certain foods can cause dyspepsia, but no convincing studies indicate an association between ulcer formation and a specific diet. This is also true for beverages containing alcohol and caffeine. Specific chronic disorders have been associated withPUD(Table 285-4). Multiple factors play a role in the pathogenesis ofPUD. The two predominant causes are H. pylori infection andNSAIDingestion. PUD not related to H. pylori or NSAIDs may be increasing. Independent of the inciting or injurious agent, peptic ulcers develop as a result of an imbalance between mucosal protection/repair and aggressive factors. Gastric acid plays an essential role in mucosal injury. CLINICAL FEATURES History Abdominal pain is common to many gastrointestinal disorders, includingDU andGU, but has a poor predictive value for the presence of either DU or GU. Up to 10% of patients withNSAID-induced mucosal disease can present with a complication (bleeding, perforation, and obstruction) without antecedent symptoms. Despite this poor correlation, a careful history and physical examination are essential components of the approach to a patient suspected of having peptic ulcers. Epigastric pain described as a burning or gnawing discomfort can be present in bothDU andGU. The discomfort is also described as an ill-defined, aching sensation or as hunger pain. The typical pain pattern in DU occurs 90 min to 3 h after a meal and is frequently relieved by antacids or food. Pain that awakes the patient from sleep (between midnight and 3 A.M.) is the most discriminating symptom, with two-thirds of DU patients describing this complaint. Unfortunately, this symptom is also present in one-third of patients withNUD. The pain pattern in GU patients may be different from that in DU patients, where discomfort may actually be precipitated by food. Nausea and weight loss occur more commonly in GU patients. In the United States, endoscopy detects ulcers in60 years old. The higher incidence in the elderly is likely due to the increased use ofNSAIDsin this group. As many as 20% of patients with ulcer-related hemorrhage bleed without any preceding warning signs or symptoms. Perforation The second most common ulcer-related complication is perforation, being reported in as many as 6 to 7% ofPUDpatients. As in the case of bleeding, the incidence of perforation in the elderly appears to be increasing secondary to increased use ofNSAIDs. Penetration is a form of perforation in which the ulcer bed tunnels into an adjacent organ.DUs tend to penetrate posteriorly into the pancreas, leading to pancreatitis, whereasGUs tend to penetrate into the left hepatic lobe. Gastrocolic fistulas associated with GUs have also been described. Gastric Outlet Obstruction Gastric outlet obstruction is the least common ulcer-related complication, occurring in 1 to 2% of patients. A patient may have relative obstruction secondary to ulcer-related inflammation and edema in the peripyloric region. This process often resolves with ulcer healing. A fixed, mechanical obstruction secondary to scar formation in the peripyloric areas is also possible. The latter requires endoscopic (balloon dilation) or surgical intervention. Signs and symptoms relative to mechanical obstruction may develop insidiously. New onset of early satiety, nausea, vomiting, increase of postprandial abdominal pain, and weight loss should make gastric outlet obstruction a possible diagnosis. Differential Diagnosis The list of gastrointestinal and nongastrointestinal disorders that can mimic ulceration of the stomach or duodenum is quite extensive. The most commonly encountered diagnosis among patients seen for upper abdominal discomfort isNUD. NUD, also known as functional dyspepsia or essential dyspepsia, refers to a group of heterogeneous disorders typified by upper abdominal pain without the presence of an ulcer. Dyspepsia has been reported to occur in up to 30% of the U.S. population. Up to 60% of patients seeking medical care for dyspepsia have a negative diagnostic evaluation. The etiology of NUD is not established, and the potential role of H. pylori in NUD remains controversial. Several additional disease processes that may present with "ulcer-like" symptoms include proximal gastrointestinal tumors, gastroesophageal reflux, vascular disease, pancreaticobiliary disease (biliary colic, chronic pancreatitis), and gastroduodenal Crohn's disease.
Diagnostic Evaluation In view of the poor predictive value of abdominal pain for the presence of a gastroduodenal ulcer and the multiple disease processes that can mimic this disease, the clinician is often confronted with having to establish the presence of an ulcer. Documentation of an ulcer requires either a radiographic (barium study) or an endoscopic procedure. Barium studies of the proximal gastrointestinal tract are still commonly used as a first test for documenting an ulcer. The sensitivity of older single-contrast barium meals for detecting aDU is as high as 80%, with a double-contrast study providing detection rates as high as 90%. Sensitivity for detection is decreased in small ulcers (3 cm in size or those associated with a mass are more often malignant. Unfortunately, up to 8% of GUs that appear to be benign by radiographic appearance are malignant by endoscopy or surgery. Radiographic studies that show a GU must be followed by endoscopy and biopsy. Endoscopy provides the most sensitive and specific approach for examining the upper gastrointestinal tract. In addition to permitting direct visualization of the mucosa, endoscopy facilitates photographic documentation of a mucosal defect and tissue biopsy to rule out malignancy (GU) or H. pylori. Endoscopic examination is particularly helpful in identifying lesions too small to detect by radiographic examination, for evaluation of atypical radiographic abnormalities, or to determine if an ulcer is a source of blood loss. Although the methods for diagnosing H. pylori are outlined in Chap. 154, a brief summary will be included here (Table 285-5). PyloriTek, a biopsy urease test, has a sensitivity and specificity of >90 to 95%. In the interest of making a diagnosis of H. pylori without the need for performing endoscopy, several noninvasive methods for detecting this organism have been developed. Three types of studies routinely used include serologic testing, the13C- or14C-urea breath test, and the fecal H. pylori antigen test. Occasionally, specialized testing such as serum gastrin and gastric acid analysis or sham feeding may be needed in individuals with complicated or refractoryPUD (see "Zollinger-Ellison Syndrome," below). Screening for aspirin orNSAIDS(blood or urine) may also be necessary in refractory, H. pylori-negative PUD patients. TREATMENT Before the discovery of H. pylori, the therapy ofPUDdisease was centered on the old dictum by Schwartz of "no acid, no ulcer." Although acid secretion is still important in the pathogenesis of PUD, eradication of H. pylori and therapy/prevention ofNSAID-induced disease is the mainstay. A summary of commonly used drugs for treatment of acid peptic disorders is shown in Table 285-6. Acid Neutralizing/Inhibitory Drugs
Antacids Before we understood the important role of histamine in stimulating parietal cell activity, neutralization of secreted acid with antacids constituted the main form of therapy for peptic ulcers. They are now rarely, if ever, used as the primary therapeutic agent but instead are often used by patients for symptomatic relief of dyspepsia. The most commonly used agents are mixtures of aluminum hydroxide and magnesium hydroxide. Aluminum hydroxide can produce constipation and phosphate depletion; magnesium hydroxide may cause loose stools. Many of the commonly used antacids (e.g., Maalox, Mylanta) have a combination of both aluminum and magnesium hydroxide in order to avoid these side effects. The magnesium-containing preparation should not be used in chronic renal failure patients because of possible hypermagnesemia, and aluminum may cause chronic neurotoxicity in these patients. Calcium carbonate and sodium bicarbonate are potent antacids with varying levels of potential problems. The long-term use of calcium carbonate (converts to calcium chloride in the stomach) can lead to milk-alkali syndrome (hypercalcemia, hyperphosphatemia with possible renal calcinosis and progression to renal insufficiency). Sodium bicarbonate may induce systemic alkalosis. H2Receptor antagonists Four of these agents are presently available (cimetidine, ranitidine, famotidine, and nizatidine), and their structures share homology with histamine (Fig. 285-9). Although each has different potency, all will significantly inhibit basal and stimulated acid secretion to comparable levels when used at therapeutic doses. Moreover, similar ulcer-healing rates are achieved with each drug when used at the correct dosage. Presently, this class of drug is often used for treatment of active ulcers (4 to 6 weeks) in combination with antibiotics directed at eradicating H. pylori (see below). Cimetidine was the first H2receptor antagonist used for the treatment of acid peptic disorders. The initial recommended dosing profile for cimetidine was 300 mg four times per day. Subsequent studies have documented the efficacy of using 800 mg at bedtime for treatment of active ulcer, with healing rates approaching 80% at 4 weeks. Cimetidine may have weak antiandrogenic side effects resulting in reversible gynecomastia and impotence, primarily in patients receiving high doses for prolonged periods of time (months to years, as inZES). In view of cimetidine's ability to inhibit cytochrome P450, careful monitoring of drugs such as warfarin, phenytoin, and theophylline is indicated with long-term usage. Other rare reversible adverse effects reported with cimetidine include confusion and elevated levels of serum aminotransferases, creatinine, and serum prolactin. Ranitidine, famotidine, and nizatidine are more potent H2receptor antagonists than cimetidine. Each can be used once a day at bedtime. Comparable nighttime dosing regimens are ranitidine, 300 mg, famotidine, 40 mg, and nizatidine, 300 mg. Additional rare, reversible systemic toxicities reported with H2receptor antagonists include pancytopenia, neutropenia, anemia, and thrombocytopenia, with a prevalence rate varying from 0.01 to 0.2%. Cimetidine and rantidine (to a lesser extent) can bind to hepatic cytochrome P450, whereas the newer agents, famotidine and nizatidine, do not. Proton pump (H +,K+-ATPase) inhibitors Omeprazole, lansoprazole, and the newest additions, rabeprazole and pantoprazole, are substituted benzimidazole derivatives that
covalently bind and irreversibly inhibit H+,K+ -ATPase. These are the most potent acid inhibitory agents available. Omeprazole and lansoprazole are the proton pump inhibitors (PPIs) that have been used for the longest time. Both are acid labile and are administered as enteric-coated granules in a sustained-release capsule that dissolves within the small intestine at a pH of 6. These agents are lipophilic compounds; upon entering the parietal cell, they are protonated and trapped within the acid environment of the tubulovesicular and canalicular system. These agents potently inhibit all phases of gastric acid secretion. Onset of action is rapid, with a maximum acid inhibitory effect between 2 and 6 h after administration and duration of inhibition lasting up to 72 to 96 h. With repeated daily dosing, progressive acid inhibitory effects are observed, with basal and secretagogue-stimulated acid production being inhibited by >95% after 1 week of therapy. The half-life of PPIs is approximately 18 h, thus it can take between 2 and 5 days for gastric acid secretion to return to normal levels once these drugs have been discontinued. Because the pumps need to be activated for these agents to be effective, their efficacy is maximized if they are administered before a meal (e.g., in the morning before breakfast). Standard dosing for omeprazole and lansoprazole is 20 mg and 30 mg once per day, respectively. Mild to moderate hypergastrinemia has been observed in patients taking these drugs. Carcinoid tumors developed in some animals given the drugs preclinically; however, extensive experience has failed to demonstrate gastric carcinoid tumor development in humans. Serum gastrin levels return to normal levels within 1 to 2 weeks after drug cessation. As with any agent that leads to significant hypochlorhydria, PPIs may interfere with absorption of drugs such as ketoconazole, ampicillin, iron, and digoxin. Hepatic cytochrome P450 can be inhibited by these agents, but the overall clinical significance of this observation is not definitely established. Caution should be taken when using warfarin, diazepam, and phenytoin concomitantly with PPIs. Cytoprotective Agents Sucralfate Sucralfate is a complex sucrose salt in which the hydroxyl groups have been substituted by aluminum hydroxide and sulfate. This compound is insoluble in water and becomes a viscous paste within the stomach and duodenum, binding primarily to sites of active ulceration. Sucralfate may act by several mechanisms. In the gastric environment, aluminum hydroxide dissociates, leaving the polar sulfate anion, which can bind to positively charged tissue proteins found within the ulcer bed, and providing a physicochemical barrier impeding further tissue injury by acid and pepsin. Sucralfate may also induce a trophic effect by binding growth factors such asEGF, enhance prostaglandin synthesis, stimulate mucous and bicarbonate secretion, and enhance mucosal defense and repair. Toxicity from this drug is rare, with constipation being the most common one reported (2 to 3%). It should be avoided in patients with chronic renal insufficiency to prevent aluminum-induced neurotoxicity. Hypophosphatemia and gastric bezoar formation have also been rarely reported. Standard dosing of sucralfate is 1 g four times per day. Bismuth-containing preparations Sir William Osler considered bismuth-containing compounds the drug of choice for treatingPUD. The resurgence in the use of these agents is due to their effect against H. pylori. Colloidal bismuth subcitrate (CBS) and bismuth subsalicylate (BSS, Pepto-Bismol) are the most widely used preparations. The mechanism by which these agents induce ulcer healing is unclear. Potential
mechanisms include ulcer coating; prevention of further pepsin/HCl-induced damage; binding of pepsin; and stimulation of prostaglandins, bicarbonate, and mucous secretion. Adverse effects with short-term usage are rare with bismuth compounds. Long-term usage with high doses, especially with the avidly absorbed CBS, may lead to neurotoxicity. These compounds are commonly used as one of the agents in an anti-H. pylori regimen (see below). Prostaglandin analogues In view of their central role in maintaining mucosal integrity and repair, stable prostaglandin analogues were developed for the treatment ofPUD. The prostaglandin E1derivative misoprostal is the only agent of this class approved by the U.S. Food and Drug Administration for clinical use in the prevention ofNSAID-induced gastroduodenal mucosal injury (see below). The mechanism by which this rapidly absorbed drug provides its therapeutic effect is through enhancement of mucosal defense and repair. Prostaglandin analogues enhance mucous bicarbonate secretion, stimulate mucosal blood flow, and decrease mucosal cell turnover. The most common toxicity noted with this drug is diarrhea (10 to 30% incidence). Other major toxicities include uterine bleeding and contractions; misoprostal is contraindicated in women who may be pregnant, and women of childbearing age must be made clearly aware of this potential drug toxicity. The standard therapeutic dose is 200 ug four times per day. Miscellaneous drugs A number of drugs aimed at treating acid peptic disorders have been developed over the years. In view of their limited utilization in the United States, if any, they will only be listed briefly. Anticholinergics, designed to inhibit activation of the muscarinic receptor in parietal cells, met with limited success due to their relatively weak acid-inhibiting effect and significant side effects (dry eyes, dry mouth, urinary retention). Tricyclic antidepressants have been suggested by some, but again the toxicity of these agents in comparison to the safe, effective drugs already described, precludes their utility. Finally, the licorice extract carbenoxolone has aldosterone-like side effects with fluid retention and hypokalemia, making it an undesirable therapeutic option. Therapy of H. pylori Extensive effort has been placed into determining who of the many individuals with H. pylori infection should be treated. The common conclusion arrived at by multiple consensus conferences (National Institutes of Health Consensus Development, American Digestive Health Foundation International Update Conference, European Maastricht Consensus, and Asia Pacific Consensus Conference) is that H. pylori should be eradicated in patients with documentedPUD. This holds true independent of time of presentation (first episode or not), severity of symptoms, presence of confounding factors such as ingestion ofNSAIDs, or whether the ulcer is in remission. Some have advocated treating patients with a history of documented PUD who are found to be H. pylori-positive by serology or breath testing. Over half of patients with gastricMALTlymphoma experience complete remission of the tumor in response to H. pylori eradication. Treating patients withNUD or to prevent gastric cancer remains controversial. Multiple drugs have been evaluated in the therapy of H. pylori. No single agent is effective in eradicating the organism. Combination therapy for 14 days provides the greatest efficacy. A short-time course administration (7 to 10 days), although attractive,
has not proven as successful as the 14-day regimens. The agents used with the greatest frequency include amoxicillin, metronidazole, tetracycline, clarithromycin, and bismuth compounds. The physician's goal in treatingPUD is to provide relief of symptoms (pain or dyspepsia), promote ulcer healing, and ultimately prevent ulcer recurrence and complications. The greatest impact of understanding the role of H. pylori in peptic disease has been the ability to prevent recurrence of what was often a recurring disease. Documented eradication of H. pylori in patients with PUD is associated with a dramatic decrease in ulcer recurrence to 4% (as compared to 59%) inGUpatients and 6% (compared to 67%) inDUpatients. Eradication of the organism may lead to diminished recurrent ulcer bleeding. The impact of its eradication on ulcer perforation is unclear. Suggested treatment regimens for H. pylori are outlined in Table 285-7. Choice of a particular regimen will be influenced by several factors including efficacy, patient tolerance, existing antibiotic resistance, and cost of the drugs. The aim for initial eradication rates should be 85 to 90%. Dual therapy [PPIplus amoxicillin, PPI plus clarithromycin, ranitidine bismuth citrate (Tritec) plus clarithromycin] are not recommended in view of studies demonstrating eradication rates of 400 pg/mL observed in ~80% of gastrinoma patients. The lower accuracy, coupled with it being a more cumbersome study with greater potential for adverse effects, makes calcium infusion less useful and therefore rarely, if ever, utilized. Rarely, one may observe increasedBAO and hypergastrinemia in a patient who in the past has been categorized as having G cell hyperplasia or hyperfunction. This set of findings may have been due to H. pylori. The standard meal test was devised to assist in making the diagnosis of G cell-related hyperactivity, by observing a dramatic increase in gastrin after a meal (>200%). This test is not useful in differentiating between G cell hyperfunction and ZES. Tumor Localization Once the biochemical diagnosis of gastrinoma has been confirmed, the tumor must be located. Multiple imaging studies have been utilized in an
effort to enhance tumor localization (Table 285-12). The broad range of sensitivity is due to the variable success rates achieved by the different investigative groups. Endoscopic ultrasound (EUS) permits imaging of the pancreas with a high degree of resolution (75%. Up to 50% of patients have metastatic disease at diagnosis. Success in controlling gastric acid hypersecretion has shifted the emphasis of therapy towards providing a surgical cure. Detecting the primary tumor and excluding metastatic disease are critical in view of this paradigm shift. Once a biochemical diagnosis has been confirmed, the patient should first undergo an abdominal computed tomographic scan, magnetic resonance imaging, or octreoscan (depending on availability) to exclude metastatic disease. Once metastatic disease has been excluded, an experienced endocrine surgeon may opt for exploratory laparotomy with intraoperative ultrasound or transillumination. In other centers, careful examination of the peripancreatic area withEUS, accompanied by endoscopic exploration of the duodenum for primary tumors, will be performed before surgery. Selective arterial secretin injection (SASI) may be a useful adjuvant for localizing tumors in a subset of patients. TREATMENT Treatment of functional endocrine tumors is directed at ameliorating the signs and symptoms related to hormone overproduction, curative resection of the neoplasm, and attempts to control tumor growth in metastatic disease. PPIsare the treatment of choice and have decreased the need for total gastrectomy. Initial doses of omeprazole or lansoprazole should be in the range of 60 mg/d. Dosing can be adjusted to achieve aBAO500 pg/mL) in
patients with pernicious anemia.ECL cell hyperplasia with frank development of gastric carcinoid tumors may result from gastrin trophic effects. The role of gastrin in carcinoid development is confirmed by the observation that antrectomy leads to regression of these lesions. Hypergastrinemia and achlorhydria may also be seen in non-pernicious anemia-associated type A gastritis. Type B gastritis Type B, or antral-predominant, gastritis is the more common form of chronic gastritis. H. pylori infection is the cause of this entity. Although described as "antral-predominant," this is likely a misnomer in view of studies documenting the progression of the inflammatory process towards the body and fundus of infected individuals. The conversion to a pan-gastritis is time-dependent -- estimated to require 15 to 20 years. This form of gastritis increases with age, being present in up to 100% of people over age 70. Histology improves after H. pylori eradication. The number of H. pylori organisms decreases dramatically with progression to gastric atrophy, and the degree of inflammation correlates with the level of these organisms. Early on, with antral-predominant findings, the quantity of H. pylori is highest and a dense chronic inflammatory infiltrate of the lamina propria is noted accompanied by epithelial cell infiltration with polymorphonuclear leukocytes (Fig. 285-12). Multifocal atrophic gastritis, gastric atrophy with subsequent metaplasia, has been observed in chronic H. pylori-induced gastritis. This may ultimately lead to development of gastric adenocarcinoma (Fig. 285-13;Chap. 90). H. pylori infection is now considered an independent risk factor for gastric cancer. Worldwide epidemiologic studies have documented a higher incidence of H. pylori infection in patients with adenocarcinoma of the stomach as compared to control subjects. Seropositivity for H. pylori is associated with a three- to sixfold increased risk of gastric cancer. This risk may be as high as ninefold after adjusting for the inaccuracy of serologic testing in the elderly. The mechanism by which H. pylori infection leads to cancer is unknown. However, eradication of H. pylori as a general preventative measure for gastric cancer is not recommended. Infection with H. pylori is also associated with development of a low grade B cell lymphoma, gastricMALTlymphoma (Chap. 112). The chronic T cell stimulation caused by the infection leads to production of cytokines that promote the B cell tumor. Tumor growth remains dependent upon the presence of H. pylori in that its eradication is often associated with complete regression of the tumor. The tumor may take more than a year to regress after treating the infection. Such patients should be followed byEUSevery 2 to 3 months. If the tumor is stable or decreasing in size, no other therapy is necessary. If the tumor grows, it may have become a high-grade B cell lymphoma. When the tumor becomes a high-grade aggressive lymphoma histologically, it loses responsiveness to H. pylori eradication. TREATMENT Treatment in chronic gastritis is aimed at the sequelae and not the underlying inflammation. Patients with pernicious anemia will require parenteral vitamin B12supplementation on a long-term basis. Eradication of H. pylori is not routinely recommended unlessPUD or a low-gradeMALTlymphoma is present.
Miscellaneous Forms of Gastritis Lymphocytic gastritis is characterized histologically by intense infiltration of the surface epithelium with lymphocytes. The infiltrative process is primarily in the body of the stomach and consists of mature T cells and plasmacytes. The etiology of this form of chronic gastritis is unknown. It has been described in patients with celiac sprue, but whether there is a common factor associating these two entities is unknown. No specific symptoms suggest lymphocytic gastritis. A subgroup of patients has thickened folds noted on endoscopy. These folds are often capped by small nodules that contain a central depression or erosion; this form of the disease is called varioliform gastritis. H. pylori probably plays no significant role in lymphocytic gastritis. Therapy with glucocorticoids or sodium cromoglycate has obtained unclear results. Marked eosinophilic infiltration involving any layer of the stomach (mucosa, muscularis propria, and serosa) is characteristic of eosinophilic gastritis. Affected individuals will often have circulating eosinophilia with clinical manifestation of systemic allergy. Involvement may range from isolated gastric disease to diffuse eosinophilic gastroenteritis. Antral involvement predominates, with prominent edematous folds being observed on endoscopy. These prominent antral folds can lead to outlet obstruction. Patients can present with epigastric discomfort, nausea, and vomiting. Treatment with glucocorticoids has been successful. Several systemic disorders may be associated with granulomatous gastritis. Gastric involvement has been observed in Crohn's disease. Involvement may range from granulomatous infiltrates noted only on gastric biopsies to frank ulceration and stricture formation. Gastric Crohn's disease usually occurs in the presence of small-intestinal disease. Several rare infectious processes can lead to granulomatous gastritis, including histoplasmosis, candidiasis, syphilis, and tuberculosis. Other unusual causes of this form of gastritis include sarcoidosis, idiopathic granulomatous gastritis, and eosinophilic granulomas involving the stomach. Establishing the specific etiologic agent in this form of gastritis can be difficult, at times requiring repeat endoscopy with biopsy and cytology. Occasionally, a surgically obtained full-thickness biopsy of the stomach may be required to exclude malignancy. MENETRIER'S DISEASE Menetrier's disease is a rare entity characterized by large, tortuous gastric mucosal folds. The differential diagnosis of large gastric folds includesZES, malignancy, infectious etiologies (CMV, histoplasmosis, syphilis), and infiltrative disorders such as sarcoidosis. The mucosal folds in Menetrier's disease are often most prominent in the body and fundus. Histologically, massive foveolar hyperplasia (hyperplasia of surface and glandular mucous cells) is noted, which replaces most of the chief and parietal cells. This hyperplasia produces the prominent folds observed. The pits of the gastric glands elongate and may become extremely tortuous. Although the lamina propria may contain a mild chronic inflammatory infiltrate, Menetrier's disease is not considered a form of gastritis. The etiology of this unusual clinical picture is unknown. Overexpression of growth factors such asTGF-amay be involved in the process. Epigastric pain at times accompanied by nausea, vomiting, anorexia, and weight loss are signs and symptoms in patients with Menetrier's disease. Occult gastrointestinal
bleeding may occur, but overt bleeding is unusual and, when present, is due to superficial mucosal erosions. Between 20 and 100% of patients (depending on time of presentation) develop a protein-losing gastropathy accompanied by hypoalbuminemia and edema. Gastric acid secretion is usually reduced or absent because of the replacement of parietal cells. Large gastric folds are readily detectable by either radiographic (barium meal) or endoscopic methods. Endoscopy with deep mucosal biopsy (and cytology) is required to establish the diagnosis and exclude the other entities that may present in a similar manner. A nondiagnostic biopsy may lead to a surgically obtained full-thickness biopsy to exclude malignancy. TREATMENT Medical therapy with anticholinergic agents, prostaglandins,PPIs, prednisone, and H2receptor antagonists has obtained varying results. Anticholinergics decrease protein loss. A high-protein diet should be recommended to replace protein loss in patients with hypoalbuminemia. Ulcers should be treated with a standard approach. Severe disease with persistent and substantial protein loss may require total gastrectomy. Subtotal gastrectomy is performed by some; it may be associated with higher morbidity and mortality secondary to the difficulty in obtaining a patent and long-lasting anastomosis between normal and hyperplastic tissues. ACKNOWLEDGEMENT The author acknowledges the contribution of material to this chapter by Dr. Lawrence Friedman and Dr. Walter Peterson from their chapter on this subject in the 14th edition and is grateful to Pamela Glazer for typing this manuscript. (Bibliography omitted in Palm version) Back to Table of Contents
286. DISORDERS OF ABSORPTION - Henry J. Binder Disorders of absorption represent a broad spectrum of conditions with multiple etiologies and varied clinical manifestations. Almost all of these clinical problems are associated with diminished intestinal absorption of one or more dietary nutrients and are often referred to as the malabsorption syndrome. This latter term is not ideal as it represents a pathophysiologic state, does not provide an etiologic explanation for the underlying problem, and should not be considered an adequate final diagnosis. The only clinical situations in which absorption is increased are hemochromatosis and Wilson's disease, in which there is increased absorption of iron and copper, respectively. Most, but not all, of these clinical conditions are associated with steatorrhea, an increase in stool fat excretion of >6% of dietary fat intake. Some disorders of absorption are not associated with steatorrhea: Primary lactase deficiency, which represents a congenital absence of the small intestinal brush border disaccharidase enzyme lactase, is only associated with lactose "malabsorption," and pernicious anemia is associated with a marked decrease in intestinal absorption of cobalamin (vitamin B12) due to an absence of gastric parietal cell intrinsic factor required for cobalamin absorption. Disorders of absorption must be included in the differential diagnosis of diarrhea (Chap. 42) for several reasons. First, diarrhea is frequently associated with and/or is a consequence of the diminished absorption of one or more dietary nutrients. The diarrhea may be secondary either to the intestinal process that is responsible for the steatorrhea or to steatorrhea per se. Thus, celiac sprue (see below) is associated with both extensive morphologic changes in the small intestinal mucosa and reduced absorption of several dietary nutrients; in contrast, the diarrhea of steatorrhea is the result of the effect of nonabsorbed dietary fatty acids on intestinal, usually colonic, ion transport. For example, oleic acid and ricinoleic acid (a bacterially hydroxylated fatty acid that is also the active ingredient in castor oil, a widely used laxative) induce active colonic Cl secretion, most likely secondary to increasing intracellular Ca. In addition, diarrhea per se may result in mild steatorrhea (24 h) fasting on stool output can be very effective in suggesting that a dietary nutrient is responsible for the individual's diarrhea. A secretory diarrhea associated with enterotoxin-induced traveler's diarrhea would not be affected by prolonged fasting, as enterotoxin-induced stimulation of intestinal fluid and electrolyte secretion is not altered by eating. In contrast, diarrhea secondary to lactose malabsorption in primary lactase deficiency would undoubtedly cease during a prolonged fast. Thus, a substantial decrease in stool output while fasting during a quantitative stool collection of at least 24 h is presumptive evidence that the diarrhea is related to malabsorption of a dietary nutrient. The persistence of stool output while fasting indicates the likelihood that the diarrhea is secretory and that the cause of diarrhea is not due to a dietary nutrient. Either a luminal (e.g., E. coli enterotoxin) or circulating (e.g., vasoactive intestinal peptide) secretogogue could be responsible for the patient's diarrhea persisting unaltered during a prolonged fast. The observed effects of fasting can be compared and correlated with stool electrolyte and osmolality determinations. Measurement of stool electrolytes and osmolality requires the comparison of stool Na+ and K+concentrations determined in liquid stool to the stool osmolality to determine the presence or absence of a so-called stool osmotic gap. The following formula is used:
The cation concentrations are doubled to estimate stool anion concentrations. The presence of a significant osmotic gap suggests the presence in stool water of a substance(s) other than Na/K/anions that presumably is responsible for the patient's diarrhea. Originally, stool osmolality was measured, but it is almost invariably greater than the required 290 to 300 mosmol/kg H2O, reflecting bacterial degradation of nonabsorbed carbohydrate either immediately before defecation or in the stool jar while awaiting chemical analysis, even when the stool is refrigerated. As a result, the stool osmolality should be assumed to be 300 mosmol/kg H2O. When the calculated difference is >50, an osmotic gap is present, suggesting that the diarrhea is due to a nonabsorbed dietary nutrient, e.g., a fatty acid and/or carbohydrate. When this difference is3 weeks) (Chap. 42). The ready availability of endoscopic equipment to examine the stomach and duodenum has led to their almost uniform use as the preferred method to obtain histologic material of proximal small-intestinal mucosa. The primary indications for a small-intestinal biopsy are (1) evaluation of a patient either with documented or suspected steatorrhea or with chronic diarrhea, and (2) diffuse or focal abnormalities of the small intestine defined on a small-intestinal series. Lesions seen on small-bowel biopsy can be classified into three different categories (Table 286-7): (1) diffuse, specific; (2) patchy, specific; and (3) diffuse, nonspecific.
1. Diffuse, specific lesions. There are relatively few diseases associated with altered nutrient absorption that have specific histopathologic abnormalities on small-intestinal mucosal biopsy, and they are uncommon. Whipple's disease is characterized by the presence of periodic acid-Schiff (PAS)-positive macrophages in the lamina propria, while the bacilli that are also present may require electron-microscopic examination for identification (Fig. 286-4). Abetalipoproteinemia is characterized by a normal mucosal appearance except for the presence of mucosal absorptive cells that contain lipid postprandially and disappear following a prolonged period of either fat-free intake or fasting. Immune globulin deficiency is associated with a variety of histopathologic findings on small-intestinal mucosal biopsy. The characteristic feature is the absence or substantial reduction in the number of plasma cells in the lamina propria; the mucosal architecture may be either perfectly normal or flat, i.e., villus atrophy. As patients with immune globulin deficiency are often infected with Giardia lamblia, Giardia trophozoites may also be seen in the biopsy. 2. Patchy, specific lesions. Several diseases are associated with abnormal small-intestinal mucosal biopsies, but the characteristic features that are present have a patchy distribution. As a result, biopsies obtained randomly or in the absence of abnormalities visualized endoscopically may not reveal these diagnostic features. Intestinal lymphoma can at times be diagnosed on mucosal biopsy by the identification of malignant lymphoma cells in the lamina propria and submucosa (Chap. 112). The presence of dilated lymphatics in the submucosa and sometimes in the lamina propria indicates the presence of lymphangiectasia associated with hypoproteinemia secondary to protein loss into the intestine. Eosinophilic gastroenteritis represents a heterogeneous group of disorders with a spectrum of presentations and symptoms with a eosinophilic infiltrate of the lamina propria, with or without peripheral eosinophilia. The patchy nature of the infiltrate as well as its presence in the submucosa often leads to an absence of histopathologic findings on mucosal biopsy. As the involvement of the duodenum in Crohn's disease is also submucosal and not necessarily continuous, mucosal biopsies are not the most direct approach to the diagnosis of duodenal Crohn's disease (Chap. 287). Amyloid deposition can be identified by Congo Red stain in some patients with amyloidosis involving the duodenum (Chap. 319). Several microorganisms can be identified on small-intestinal biopsies, establishing a correct diagnosis. Many of these microorganisms are associated with diarrhea that occurs in immunodeficient individuals, especially those with HIV infection, and include Cryptosporidium, Isospora belli, cytomegalovirus, Mycobacterium avium intracellulare, and G. lamblia. 3. Diffuse, nonspecific lesions. Celiac sprue presents with a characteristic mucosal appearance on duodenal/proximal jejunal mucosal biopsy that is not diagnostic of the disease. The diagnosis of celiac sprue is established by clinical, histologic, and immunologic response to a gluten-free diet. Tropical sprue is associated with histopathologic findings similar to those of celiac sprue after a tropical or subtropical exposure but does not respond to gluten restriction; most often symptoms improve with antibiotics and folate administration. Patients with steatorrhea require assessment of pancreatic exocrine function, which is often abnormal in chronic pancreatitis. No test assesses pancreatic exocrine function
well. Endoscopic approaches provide excellent assessment of pancreatic duct anatomy but do not assess exocrine function (Chap. 303). One noninvasive study (bentiromide test) of pancreatic exocrine function is based on the feeding of a tripeptide containing p-aminobenzoic acid (PABA). Following splitting of PABA by pancreatic proteases, PABA is liberated, absorbed, and excreted in urine. Reduced proteolysis results in reduced urinary excretion of PABA. This test is neither sensitive nor specific. Table 286-8 summarizes the results of the D-xylose test, Schilling test, and small-intestinal mucosal biopsy in patients with five different causes of steatorrhea. SPECIFIC DISEASE ENTITIES CELIAC SPRUE Celiac sprue is a not uncommon cause of malabsorption of one or more nutrients in Caucasians, especially those of European descent. Celiac sprue has had several other names including nontropical sprue, celiac disease (in children), adult celiac disease, and gluten-sensitive enteropathy. The etiology of celiac sprue is not known, but environmental, genetic, and immunologic factors are important. Celiac sprue has protean manifestations, almost all of which are secondary to nutrient malabsorption, and a varied natural history, with the onset of symptoms occurring at ages ranging from the first year of life through the eighth decade. The hallmark of celiac sprue is the presence of an abnormal small-intestinal biopsy (Fig. 286-4) and the response of both symptoms, evidence of malabsorption and the histopathologic changes on the small-intestinal biopsy, to the elimination of gluten from the diet. The histopathologic changes have a proximal to distal intestinal distribution of severity, which probably reflects the exposure of the intestinal mucosa to varied amounts of dietary gluten; the degree of symptoms is related to the extent of these histopathologic changes. The symptoms of celiac sprue may appear with the introduction of cereals in an infant's diet, although there is frequently a spontaneous remission during the second decade of life that may be either permanent or followed by the reappearance of symptoms over several years. Alternatively, the symptoms of celiac sprue may first become evident at almost any age throughout adulthood. In many patients, frequent spontaneous remissions and exacerbations occur. The symptoms range from significant malabsorption of multiple nutrients with diarrhea, steatorrhea, weight loss, and the consequences of nutrient depletion (i.e., anemia and metabolic bone disease) to the absence of any gastrointestinal symptoms but with evidence of the depletion of a single nutrient (e.g., folate deficiency, osteomalacia, edema from protein loss). Asymptomatic relatives of patients with celiac sprue have been identified as having this disease either by small-intestinal biopsy or by serologic studies (e.g., antiendomysial antibodies). Etiology The etiology of celiac sprue is not known, but environmental, genetic, and immunologic factors all appear to contribute to the disease. One environmental factor is the clear association of the disease with gliadin, a component of gluten that is present in wheat, barley, rice, and, in smaller amounts, oats.
In addition to the role of gluten restriction in treatment, the instillation of gluten into both normal-appearing rectum and distal ileum of patients with celiac sprue results within hours in morphologic changes. An immunologic component to etiology is suspected for three reasons. First, serum antibodies, IgA antigliadin and antiendomysial antibodies, are present, but it is also not known whether such antibodies are primary or secondary to the tissue damage. The antiendomysial antibody has 90 to 95% sensitivity and 90 to 95% specificity, and the antigen recognized by the antiendomysial antibody test is tissue transglutaminase. The relationship of this autoantibody to pathogenetic mechanism(s) responsible for celiac sprue remains to be established. Nonetheless, this antibody will undoubtedly prove extremely useful in establishing the true prevalence of celiac sprue in the general population and may provide important clues to its etiology. Second, treatment with prednisolone for 4 weeks of a patient with celiac sprue who continues to eat gluten will induce a remission and convert the "flat" abnormal duodenal biopsy to a more normal appearing one. Third, gliadin peptides may interact with gliadin-specific T cells that may either mediate tissue injury or induce the release of one or more cytokines that are responsible for the tissue injury. Genetic factor(s) also appear to be involved in celiac sprue. The incidence of celiac sprue varies widely in different population groups (high in Caucasians, low in blacks and orientals) and is 10% in first-degree relatives of celiac sprue patients. Furthermore, about 95% of patients with celiac sprue express the HLA-DQ2 allele, though only a minority of all persons expressing DQ2 have celiac sprue. Diagnosis A small-intestinal biopsy is required to establish a diagnosis of celiac sprue (Fig. 286-4). A biopsy should be performed in patients with symptoms and laboratory findings suggestive of nutrient malabsorption and/or deficiency. Since the presentation of celiac sprue is often subtle, without overt evidence of malabsorption or nutrient deficiency, it is important to have a relatively low threshold to perform a biopsy. It is more prudent to perform a biopsy than to obtain another test of intestinal absorption, which can never completely exclude or establish this diagnosis. The diagnosis of celiac sprue requires the presence of characteristic histopathologic changes on small-intestinal biopsy together with a prompt clinical and histopathologic response following the institution of a gluten-free diet. If serologic studies have detected the presence of IgA antiendomysial antibodies, they too should disappear after a gluten-free diet is started. The changes seen on duodenal/jejunal biopsy are restricted to the mucosa and include: (1) absence or reduced height of villi, resulting in a "flat" appearance; (2) increased loss of villus cells in association with increased crypt cell proliferation resulting in crypt hyperplasia and loss of villus structure, with consequent villus, but not mucosal, atrophy; (3) cuboidal appearance and nuclei that are no longer oriented basally in surface epithelial cells and increased intraepithelial lymphocytes; and (4) increased lymphocytes and plasma cells in the lamina propria. Although these histopathologic features are characteristic of celiac sprue, they are not diagnostic because a similar appearance can be seen in tropical sprue, eosinophilic enteritis, and milk-protein intolerance in children and occasionally in lymphoma, bacterial overgrowth, Crohn's disease, and gastrinoma with acid hypersecretion. However, the presence of a characteristic histopathologic appearance that reverts to normal following the initiation of
a gluten-free diet establishes the diagnosis of celiac sprue. Readministration of gluten with or without an additional small-intestinal biopsy is not necessary. Failure to Respond to Gluten Restriction The most common cause of persistent symptoms in a patient who fulfills all the criteria of the diagnosis of celiac sprue is continued intake of gluten. Gluten is ubiquitous, and significant effort must be made to exclude all gluten from the diet. Use of rice in place of wheat flour is very helpful, and several support groups provide important aid to patients with celiac sprue and to their families. About 90% of patients who have the characteristic findings of celiac sprue will respond to complete dietary gluten restriction. The remainder represent a heterogeneous group (whose condition is often called refractory sprue) that includes some patients who (1) respond to restriction of other dietary protein, e.g., soy; (2) respond to glucocorticoids; (3) are "temporary," i.e., the clinical and morphologic findings disappear after several months or years; or (4) fail to respond to all measures and have a fatal outcome, with or without documented complications of celiac sprue. Mechanism of Diarrhea The diarrhea in celiac sprue has several pathogenetic mechanisms. Diarrhea may be secondary to (1) steatorrhea, which is primarily a result of the changes in jejunal mucosal function; (2) secondary lactase deficiency, a consequence of changes in jejunal brush border enzymatic function; (3) bile acid malabsorption resulting in bile acid-induced fluid secretion in the colon, in cases with more extensive disease involving the ileum; and (4) endogenous fluid secretion resulting from the crypt hyperplasia. Patients with more severe involvement with celiac sprue may obtain temporary improvement with dietary lactose and fat restriction while awaiting the full effects of total gluten restriction, which represents primary therapy. Associated Diseases Celiac sprue is associated with dermatitis herpetiformis (DH), though the association has not been explained. Patients with DH have characteristic papulovesicular lesions that respond to dapsone. Almost all patients with DH have histopathologic changes in the small intestine consistent with celiac sprue, although usually much milder and less diffuse in distribution. Most patients with DH have mild, or no, gastrointestinal symptoms. In contrast, relatively few patients with celiac sprue have DH. Celiac sprue is also associated with insulin-dependent diabetes mellitus and IgA globulin deficiency. The clinical importance of the former association is that although severe watery diarrhea without evidence of malabsorption is most often seen in patients with "diabetic diarrhea" (Chap. 333), a small-intestinal biopsy must at times be considered to exclude this association. Complications The most important complication of celiac sprue is the development of a malignancy. An increased incidence of both gastrointestinal and nongastrointestinal neoplasms as well as intestinal lymphoma exists in patients with celiac sprue. For unexplained reasons the occurrence of lymphoma in patients with celiac sprue is higher in Ireland and the United Kingdom than in the United States. The possibility of lymphoma must be considered whenever a patient with celiac sprue previously doing well on a gluten-free diet is no longer responsive to gluten restriction or a patient who presents with clinical and histopathologic features consistent with celiac sprue does not respond to a gluten-free diet. Other complications of celiac sprue include the
development of intestinal ulceration independent of lymphoma and so-called refractory sprue (see above) and collagenous sprue. In collagenous sprue, a layer of collagen-like material is present beneath the basement membrane; these patients generally do not respond to a gluten-free diet and often have a poor prognosis. TROPICAL SPRUE Tropical sprue is a poorly understood syndrome that affects both expatriates and natives in certain but not all tropical areas and is manifested by chronic diarrhea, steatorrhea, weight loss, and nutritional deficiencies, including those of both folate and cobalamin. This disease affects 5 to 10% of the population in some tropical areas. Chronic diarrhea in a tropical environment is most often caused by infectious agents including G. lamblia, Yersinia enterocolitica, C. difficile, Cryptosporidium parvum, and Cyclospora cayetanensis, among other organisms. Tropical sprue should not be entertained as a possible diagnosis until the presence of cysts and trophozoites has been excluded in three stool samples.*Chronic infections of the gastrointestinal tract and diarrhea in patients with or without AIDS are discussed in Chaps. 309 and 131. The small-intestinal mucosa in individuals living in tropical areas is not identical to that of individuals who reside in temperate climates. Biopsies reveal a mild alteration of villus architecture with a modest increase in mononuclear cells in the lamina propria, which on occasion can be as severe as that seen in celiac sprue. These changes are observed both in native residents and in expatriates living in tropical regions, are usually associated with mild decreases in absorptive function, but revert to "normal" when an individual moves or returns to a temperate area. Some have suggested that the changes seen in tropical enteropathy and in tropical sprue represent different ends of the spectrum of a single entity, but convincing evidence to support this concept is lacking. Etiology The etiology of tropical sprue is not known, though because tropical sprue responds to antibiotics, the consensus is that tropical sprue may be caused by one or more infectious agents. Nonetheless, multiple uncertainties regarding the etiology and pathogenesis of tropical sprue exist. First, its occurrence is not evenly distributed in all tropical areas; rather, it is found in specific locations including South India, the Philippines, and several Caribbean islands (e.g., Puerto Rico, Haiti) but is rarely observed in Africa, Jamaica, or Southeast Asia. Second, an occasional individual will not develop symptoms of tropical sprue until long after having left an endemic area. This is the reason why the original term for celiac sprue was nontropical sprue to distinguish it from tropical sprue. Third, multiple microorganisms have been identified on jejunal aspirate with relatively little consistency among studies. Klebsiella pneumoniae, Enterobacter cloacae, or E. coli have been implicated in some studies of tropical sprue, while other investigations have favored a role for a toxin produced by one or more of these bacteria. Fourth, the incidence of tropical sprue appears to have decreased substantially during the past decade. One speculation for this reduced occurrence of tropical sprue is the wider use of antibiotics in acute diarrhea especially in travelers to tropical areas from temperate countries. Fifth, the role of folic acid deficiency in the pathogenesis of tropical sprue requires clarification. Folic acid is absorbed exclusively in the duodenum and proximal jejunum, and most patients with tropical sprue have
evidence of folate malabsorption and depletion. Although folate deficiency can cause changes in small-intestinal mucosa that are corrected by folate replacement, the several earlier studies reporting that tropical sprue could be cured by folic acid did not provide an explanation for the "insult" that was initially responsible for folate malabsorption. The clinical pattern of tropical sprue varies in different areas of the world (e.g., India vs. Puerto Rico). Not infrequently, individuals in India initially will report the occurrence of an acute enteritis before the development of steatorrhea and malabsorption. In contrast, in Puerto Rico, a most insidious onset of symptoms and a more dramatic response to antibiotics is seen than in some other locations. Tropical sprue in different areas of the world may not be the same disease; there may be similar clinical entities but with different etiologies. Diagnosis The diagnosis of tropical sprue is best made by the presence of an abnormal small-intestinal mucosal biopsy in an individual with chronic diarrhea and evidence of malabsorption who is either residing or has recently lived in a tropic country. The small-intestinal biopsy in tropical sprue does not have pathognomonic features but resembles, and can often be indistinguishable from, that seen in celiac sprue (Fig. 286-4). The biopsy in tropical sprue will have less villus architectural alteration and more mononuclear cell infiltrate in the lamina propria. In contrast to celiac sprue, the histopathologic features of tropical sprue are present with a similar degree of severity throughout the small intestine, and a gluten-free diet does not result in either clinical or histopathologic improvement in tropical sprue. TREATMENT Broad-spectrum antibiotics and folic acid are most often curative, especially if the patient leaves the tropical area and does not return. Tetracycline should be used for up to 6 months and may be associated with improvement within 1 to 2 weeks. Folic acid alone will induce a hematologic remission as well as improvement in appetite, weight gain, and some morphologic changes in small intestinal biopsy. Because of the presence of marked folate deficiency, folic acid is most often given together with antibiotics. SHORT BOWEL SYNDROME This is a descriptive term for the myriad clinical problems that often occur following resection of varying lengths of small intestine. The factors that determine both the type and degree of symptoms include: (1) the specific segment (jejunum vs. ileum) resected, (2) the length of the resected segment, (3) the integrity of the ileocecal valve, (4) whether any large intestine has also been removed, (5) residual disease in the remaining small and/or large intestine (e.g., Crohn's disease, mesenteric artery disease), and (6) the degree of adaptation in the remaining intestine. Short bowel syndrome can occur at any age from neonates through the elderly. Three different situations in adults demand intestinal resections: (1) mesenteric vascular disease including both atherosclerosis, thrombotic phenomena, and vasculitidies; (2) primary mucosal and submucosal disease, e.g., Crohn's disease; and (3) operations without preexisting small intestinal disease, such as trauma and jejunoileal bypass for
obesity. Following resection of the small intestine, the residual intestine undergoes adaptation of both structure and function that may last for up to 6 to 12 months. Adaptation requires the continued intake of dietary nutrients and calories to stimulate it via direct contact with ileal mucosa, the release of one or more intestinal hormones, and pancreatic and biliary secretions. Thus, enteral nutrition and calorie administration must be maintained, especially in the early postoperative period, even if an extensive intestinal resection requiring total parenteral nutrition (TPN) was required. The subsequent ability of such patients to absorb nutrients will not be known for several months until after adaptation is completed. Multiple factors besides the absence of intestinal mucosa (required for both lipid and fluid and electrolyte absorption) contribute to the diarrhea and steatorrhea in these patients. Removal of the ileum and especially the ileocecal valve is often associated with more severe diarrhea than jejunal resection. Without part or all of the ileum, diarrhea can be caused by an increase in bile acids entering the colon, leading to their stimulation of colonic fluid and electrolyte secretion. Absence of the ileocecal valve is also associated with a decrease in intestinal transit time and bacterial overgrowth from the colon. Lactose intolerance as a result of the removal of lactase-containing mucosa as well as gastric hypersecretion will also contribute to the diarrhea. In addition to diarrhea and/or steatorrhea, a range of nonintestinal symptoms are also observed in some patients. A significant increase in renal calcium oxalate calculi is observed in patients with a small-intestinal resection with an intact colon and is due to an increase in oxalate absorption by the large intestine, with subsequent hyperoxaluria. Since oxalate is high in relatively few foods (e.g., spinach, rhubarb, tea), dietary restrictions alone are not adequate treatment. Cholestyramine, an anion-binding resin, and calcium have proved useful in reducing the hyperoxaluria. Similarly, an increase in cholesterol gall stones is seen that is related to a decrease in the bile acid pool size, which results in the generation of cholesterol supersaturation in gall bladder bile. Gastric hypersecretion of acid occurs in many patients following large resections of the small intestine. The etiology is unclear but may be related to either reduced hormonal inhibition of acid secretion or increased gastrin levels due to reduced small-intestinal catabolism of circulating gastrin. The resulting gastric acid secretion may be an important factor contributing to the diarrhea and steatorrhea. A reduced pH in the duodenum can inactivate pancreatic lipase and/or precipitate duodenal bile acids, thereby increasing steatorrhea, and an increase in gastric secretion can create a volume overload relative to the reduced small-intestinal absorptive capacity. Inhibition of gastric acid secretion with either proton pump inhibitors or H2receptor antagonists can help in reducing the diarrhea and steatorrhea. TREATMENT Treatment of short bowel syndrome depends on the severity of symptoms and whether the individual is able to maintain caloric and electrolyte balance with oral intake alone. Initial treatment includes judicious use of opiates to reduce stool output and to establish an effective diet. An initial diet should be low-fat, high-carbohydrate to minimize the diarrhea from fatty acid stimulation of colonic fluid secretion. BothMCT (see above), a
low-lactose diet, and various fiber-containing diets should also be tried. In the absence of an ileocecal valve, the possibility of bacterial overgrowth must be considered and treated. If gastric acid hypersecretion is contributing to the diarrhea and steatorrhea, a proton pump inhibitor may be helpful. Usually none of these therapeutic approaches will provide an instant solution but will reduce disabling diarrhea. The patient's vitamin and mineral status must also be monitored, and replacement therapy initiated, if indicated. Fat-soluble vitamins, folate, cobalamin, calcium, iron, magnesium, and zinc are the most critical factors to monitor on a regular basis. If these approaches are not successful, homeTPNrepresents an established therapy that can be maintained for many years. Intestinal transplantation is beginning to become established as a possible approach for individuals with extensive intestinal resection who cannot be maintained without TPN. BACTERIAL OVERGROWTH SYNDROME Bacterial overgrowth syndrome comprises a group of disorders with diarrhea, steatorrhea, and macrocytic anemia whose common feature is the proliferation of colon-type bacteria within the small intestine. This bacterial proliferation is due to stasis caused by impaired peristalsis (i.e., functional stasis), changes in intestinal anatomy (i.e., anatomic stasis), or direct communication between the small and large intestine. These conditions have also been referred to as stagnant bowel syndrome or blind loop syndrome. Pathogenesis The manifestations of bacterial overgrowth syndromes are a direct consequence of the presence of increased amounts of a colonic-type bacterial flora, such as E. coli or Bacteroides, in the small intestine. Macrocytic anemia is due to cobalamin, not folate, deficiency. Most bacteria require cobalamin for growth, and increasing concentrations of bacteria use up the relatively small amounts of dietary cobalamin. Steatorrhea is due to impaired micelle formation as a consequence of a reduced intraduodenal concentration of bile acids and the presence of unconjugated bile acids. Certain bacteria, e.g., Bacteroides, deconjugate conjugated bile acids to unconjugated bile acids. In the presence of bacterial overgrowth, unconjugated bile acids will be absorbed more rapidly than conjugated bile acids, and, as a result, the intraduodenal concentration of bile acids will be reduced. In addition, theCMC of unconjugated bile acids is higher than that of conjugated bile acids, resulting in a decrease in micelle formation. Diarrhea is due, at least in part, to the steatorrhea, when it is present. However, some patients manifest diarrhea without steatorrhea, and it is assumed that the colonic-type bacteria in these patients are producing one or more bacterial enterotoxins that are responsible for fluid secretion and diarrhea. Etiology The etiology of these different disorders is bacterial proliferation in the small intestinal lumen secondary to either anatomic or functional stasis or to a communication between the relatively sterile small intestine and the colon with its high levels of aerobic and anaerobic bacteria. Several examples of anatomic stasis have been identified: (1) one or more diverticula (both duodenal and jejunal) (Figure 286-3C); (2) fistulas and strictures related to Crohn's disease (Figure 286-3D); (3) a proximal duodenal afferent loop following a subtotal gastrectomy and gastrojejunostomy; (4) a bypass of the intestine, e.g., jejunoileal bypass for obesity; and (5) dilatation at the site of a previous
intestinal anastomosis. The common feature of all of these anatomic derangements is the presence of a segment(s) of intestine that is out of continuity of propagated peristalsis, resulting in stasis and bacterial proliferation. Bacterial overgrowth syndromes can also occur in the absence of an anatomic blind loop when functional stasis is present. The best example of impaired peristalsis and bacterial overgrowth in the absence of a blind loop is scleroderma, where motility abnormalities exist in both the esophagus and small intestine (Chap. 313). Functional stasis and bacterial overgrowth can also occur in association with diabetes mellitus and in the small intestine when a direct connection exists between the small and large intestine, including an ileocolonic resection, or occasionally following an enterocolic anastomosis that permits entry of bacteria into the small intestine as a result of bypassing the ileocecal valve. Diagnosis The diagnosis may be suspected from the combination of a low serum cobalamin level and an elevated serum folate level as enteric bacteria frequently produce folate compounds that will be absorbed in the duodenum. Ideally, the diagnosis of the bacterial overgrowth syndrome is the demonstration of increased levels of aerobic and/or anaerobic colonic-type bacteria in a jejunal aspirate obtained by intubation. This specialized test is rarely available, and bacterial overgrowth is best established by a Schilling test (Table 286-6), which should be abnormal following the administration of58Co-labeled cobalamin, with or without the administration of intrinsic factor. Following the administration of tetracycline for 5 days, the Schilling test will become normal, confirming the diagnosis of bacterial overgrowth. TREATMENT Primary treatment should be directed, if at all possible, to the surgical correction of an anatomic blind loop. In the absence of functional stasis, it is important to define the anatomic relationships responsible for stasis and bacterial overgrowth. For example, bacterial overgrowth secondary to strictures, one or more diverticula, or a proximal afferent loop can potentially be cured by surgical correction of the anatomic state. In contrast, the functional stasis of scleroderma or certain anatomic stasis states (e.g., multiple jejunal diverticula), cannot be corrected surgically, and these conditions should be treated with broad-spectrum antibiotics. Tetracycline used to be the initial treatment of choice but with increasing resistance, other antibiotics such as metronidazole, amoxicillin/clavulinic acid (Augmentin), and cephalosporin have been employed. The antibiotic should be given for approximately 3 weeks or until symptoms remit. Since the natural history of these conditions is chronic, antibiotics should not be given continuously, and symptoms usually remit within 2 to 3 weeks of initial antibiotic therapy. Therapy need not be repeated until symptoms recur. In the presence of frequent recurrences several treatment strategies exist, but the use of antibiotics for 1 week per month whether or not symptoms are present is often most effective. Unfortunately, therapy for bacterial overgrowth syndrome is largely empirical, with an absence of clinical trials on which to base decisions regarding the antibiotic to be used, the duration of treatment, and/or the best approach for treating recurrences. Bacterial overgrowth may also occur as a component of another chronic disease, e.g., Crohn's disease, radiation enteritis, or short bowel syndrome. Treatment of the bacterial overgrowth in these settings will not cure the underlying problem but may be very important in ameliorating a subset of clinical problems that are related to bacterial
overgrowth. WHIPPLE'S DISEASE Whipple's disease is a chronic multisystem disease associated with diarrhea, steatorrhea, weight loss, arthralgia, and central nervous system and cardiac problems that is caused by the bacteria Tropheryma whippelii. Until the identification of T. whippelii by polymerase chain reaction during the past decade, the hallmark of Whipple's disease had been the presence ofPAS-positive macrophages in the small intestine and other organs with evidence of disease. Long before the establishment of T. whippelii as the causative agent of Whipple's disease, gram-positive bacilli had been identified both within and outside of macrophages. Etiology Whipple's disease is caused by a small gram-positive bacillus, T. whippelii. The bacillus, an actinobacterium, has low virulence but high infectivity, and relatively minimal symptoms are observed compared to the extent of the bacilli in multiple tissues. Clinical Presentation The onset of Whipple's disease is insidious and is characterized by diarrhea, steatorrhea, abdominal pain, weight loss, migratory large-joint arthropathy, and fever as well as ophthalmologic and central nervous system symptoms. The development of dementia is a relatively late symptom and is an extremely poor prognostic sign, especially in patients who relapse following the induction of a remission with antibiotics. For unexplained reasons, the disease occurs primarily in middle-aged (50-year-old) Caucasian men. The steatorrhea in these patients is generally believed secondary to both small-intestinal mucosal injury and lymphatic obstruction secondary to the increased number ofPAS-positive macrophages in the lamina propria of the small intestine. Diagnosis The diagnosis of Whipple's disease is suggested by a multisystem disease in a 50-year-old Caucasian male with diarrhea and steatorrhea. Obtaining tissue biopsies from the small intestine and/or other organs that may be involved (e.g., liver, lymph nodes, heart, eyes, central nervous system, or synovial membranes), based on the patient's symptoms, is the primary approach to establish the diagnosis of Whipple's disease. The presence ofPAS-positive macrophages containing the characteristic small (0.25´ 1 to 2 um) bacilli is suggestive of this diagnosis. However, Whipple's disease can be confused with the PAS-positive macrophages containing M. avian complex, which may be a cause of diarrhea in AIDS. The presence of the T. whippelii bacillus outside of macrophages is a more important indicator of active disease than their presence within the macrophages. T. whippelii has now been successfully grown in culture. TREATMENT The treatment for Whipple's disease is prolonged use of antibiotics. At the present time the drug of choice is double-strength trimethoprim/sulfamethoxazole for approximately 1 year.PAS-positive macrophages can persist following successful treatment, and the presence of bacilli outside of macrophages is indicative of persistent infection or an early sign of recurrence. Recurrence of disease activity, especially with dementia, is an extremely poor prognostic sign and requires an antibiotic that crosses the blood-brain barrier. If trimethoprim/sulfamethoxazole is not tolerated, chloramphenicol is an
appropriate second choice. PROTEIN-LOSING ENTEROPATHY Protein-losing enteropathy is not a specific disease but rather describes a group of gastrointestinal and nongastrointestinal disorders with hypoproteinemia and edema in the absence of either proteinuria or defects in protein synthesis, e.g., chronic liver disease. These diseases are characterized by excess protein loss into the gastrointestinal tract. Normally, about 10% of the total protein catabolism occurs via the gastrointestinal tract. Evidence of increased protein loss into the gastrointestinal tract has been established in>65 different diseases, which can be classified into three primary groups: (1) mucosal ulceration such that the protein loss primarily represents exudation across damaged mucosa, e.g., ulcerative colitis, gastrointestinal carcinomas, peptic ulcer; (2) nonulcerated mucosa but with evidence of mucosal damage so that the protein loss represents loss across epithelia with altered permeability, e.g., celiac sprue and Menetrier's disease in the small intestine and stomach, respectively; (3) lymphatic dysfunction, either representing primary lymphatic disease or secondary to partial lymphatic obstruction that may occur as a result of enlarged lymph nodes or cardiac disease. Diagnosis The diagnosis of protein-losing enteropathy is suggested by the presence of peripheral edema and low serum albumin and globulin levels in the absence of renal and hepatic disease. It is extremely rare for an individual with protein-losing enteropathy to have selective loss of only albumin or only globulins. Therefore, marked reduction of serum albumin with normal serum globulins should not initiate an evaluation for protein-losing enteropathy but should suggest the presence of renal and/or hepatic disease. Likewise, reduced serum globulins with normal serum albumin levels is more likely a result of reduced globulin synthesis rather than enhanced globulin loss into the intestine. Documentation of an increase in protein loss into the gastrointestinal tract has been established by the administration of one of several radiolabeled proteins and its quantitation in stool during a 24- or 48-h period. Unfortunately, none of these radiolabeled proteins are available for routine clinical use.a 1-Antitrypsin, a protein that represents approximately 4% of total serum proteins and is resistant to proteolysis, can be used to document enhanced rates of serum protein loss into the intestinal tract but cannot be used to assess gastric protein loss due to its degradation in an acid milieu.a1-Antitrypsin clearance is measured by determining stool volume and both stool and plasmaa1-antitrypsin concentrations. In addition to the loss of protein via abnormal and distended lymphatics, peripheral lymphocytes may also be lost via lymphatics, resulting in a relative lymphopenia. Thus, the presence of lymphopenia in a patient with hypoproteinemia supports the presence of increased loss of protein into the gastrointestinal tract. Patients with increased protein loss into the gastrointestinal tract from lymphatic obstruction often have steatorrhea and diarrhea. The steatorrhea is a result of altered lymphatic flow as lipid-containing chylomicrons exit from intestinal epithelial cells via intestinal lymphatics (Table 286-4;Fig. 286-4). In the absence of mechanical or anatomic lymphatic obstruction, instrinsic intestinal lymphatic dysfunction, with or without lymphatic dysfunction in the peripheral extremities, has been named intestinal lymphangiectasia. Similarly, about 50% of individuals with intrinsic peripheral lymphatic
disease (Milroy's disease) will also have intestinal lymphangiectasia and hypoproteinemia. Other than steatorrhea and enhanced protein loss into the gastrointestinal tract, all other aspects of intestinal absorptive function are normal in intestinal lymphangiectasia. Other Causes Patients who appear to have idiopathic protein-losing enteropathy without any evidence of gastrointestinal disease should be examined for cardiac disease and especially right-sided valvular disease and chronic pericarditis (Chap. 236). On occasion, hypoproteinemia can be the only presentation for these two types of heart disease. Menetrier's disease (also called hypertrophic gastropathy) is an uncommon entity that involves the body and fundus of the stomach and is characterized by large gastric folds, reduced gastric acid secretion, and, at times, enhanced protein loss into the stomach. TREATMENT As excess protein loss into the gastrointestinal tract is most often secondary to a specific disease, treatment should be directed primarily to the underlying disease process and not to the hypoproteinemia. For example, if significant hypoproteinemia with resulting peripheral edema is present secondary to either celiac sprue or ulceratic colitis, a gluten-free diet or mesalamine, respectively, would be the initial therapy. When enhanced protein loss is secondary to lymphatic obstruction, it is critical to establish the nature of this obstruction. Identification of mesenteric nodes or lymphoma may be possible by imaging studies. Similarly, it is important to exclude cardiac disease as a cause of protein-losing enteropathy either by echosonography or, on occasion, by a right-heart catheterization. The increased protein loss that occurs in intestinal lymphangiectasia is a result of distended lymphatics associated with lipid malabsorption. Treatment of the hypoproteinemia is accomplished by a low-fat diet and the administration ofMCTs(Table 286-3), which do not exit from the intestinal epithelial cells via lymphatics but are delivered to the body via the portal vein. SUMMARY A pathophysiologic classification of the many conditions that can produce malabsorption is given in Table 286-9. A summary of the pathophysiology of the various clinical manifestations of malabsorption is given inTable 286-10. (Bibliography omitted in Palm version) Back to Table of Contents
287. INFLAMMATORY BOWEL DISEASE - Sonia Friedman, Richard S. Blumberg Inflammatory bowel disease (IBD) is an idiopathic and chronic intestinal inflammation. Ulcerative colitis (UC) and Crohn's disease (CD) are the two major types of IBD. EPIDEMIOLOGY The incidence ofIBDvaries within different geographic areas. Northern countries, such as the United States, United Kingdom, Norway, and Sweden, have the highest rates. The incidence rates ofUC andCD in the United States are about 11 per 100,000 and 7 per 100,000, respectively (Table 287-1). Countries in southern Europe, South Africa, and Australia have lower incidence rates: 2 to 6.3 per 100,000 for UC, and 0.9 to 3.1 per 100,000 for CD. In Asia and South America, IBD is rare; incidence rates of UC and CD are 0.5 and 0.08 per 100,000, respectively. The highest mortality in IBD patients is during the first years of disease and in long duration disease due to the risk of colon cancer. In a Swedish population study, the standardized mortality ratios for CD and UC were 1.51 and 1.37, respectively. The peak age of onset ofUC andCD is between 15 and 30 years. A second peak occurs between the ages of 60 and 80. The male to female ratio for UC is 1:1 and for CD is 1.1 to 1.8:1. A two- to fourfold increased frequency of UC and CD in Jewish populations has been described in the United States, Europe, and South Africa. Furthermore, disease frequency differs within the Jewish populations. The prevalence ofIBD in Ashkenazi Jews is about twice that of Israeli-born, Sephardic, or Oriental Jews. The prevalence decreases progressively in non-Jewish Caucasian, African-American, Hispanic, and Asian populations. Urban areas have a higher prevalence of IBD than rural areas and high socioeconomic classes have a higher prevalence than lower socioeconomic classes. The effects of cigarette smoking are different inUC andCD. The risk of UC in smokers is 40% that of nonsmokers. Additionally, former smokers have a 1.7-fold increased risk for UC than people who have never smoked. In contrast, smoking is associated with a twofold increased risk of CD. Oral contraceptives are also linked to CD; the relative risk of CD for oral contraceptive users is about 1.9. Appendectomy appears to be protective against UC but further studies are needed. IBDruns in families. If a patient has IBD, the lifetime risk that a first-degree relative will be affected is ~10%. If two parents have IBD, each child has a 36% chance of being affected. In twin studies, 67% of monozygotic twins are concordant forCD and 20% are concordant forUC, whereas 8% of dizygotic twins are concordant for CD and none are concordant for UC. There is also concordance for anatomic site and clinical type of CD within families. Additional evidence for genetic predisposition toIBDcomes from its association with certain genetic syndromes.UC andCD are both associated with Turner's syndrome, and Hermansky-Pudlak syndrome is associated with a granulomatous colitis. Glycogen storage disease type 1b can present with Crohn's-like lesions of the large and small bowel. Other immunodeficiency disorders, such as hypogammaglobulinemia, selective IgA deficiency, and hereditary angioedema, also exhibit an increased association with
IBD. ETIOLOGY AND PATHOGENESIS AlthoughIBD has been described as a clinical entity for over 100 years, its etiology and pathogenesis have not been definitively elaborated. Various studies have led to a consensus hypothesis that in genetically predisposed individuals, both exogenous factors (e.g., infectious agents, normal lumenal flora) and host factors (e.g., intestinal epithelial cell barrier function, vascular supply, neuronal activity) together cause a chronic state of dysregulated mucosal immune function that is further modified by specific environmental factors (e.g., smoking). Although it is possible that the chronic activation of the mucosal immune system may represent an appropriate response to a chronic unidentified infectious agent, a search for such an agent has thus far been unrewarding. As such, IBD must currently be considered an inappropriate response to either the endogenous microbial flora within the intestine, with or without some component of autoimmunity. Importantly, the normal intestine contains a significant concentration of immune cells in a chronic state of so-called physiologic inflammation, in which the gut is poised for, but actively restrained from, full immunologic responses. During the course of infections in the normal host, full activation of the gut-associated lymphoid tissue occurs but is rapidly superceded by downregulation of the immune response and tissue repair. In IBD this process is not regulated normally. GENETIC CONSIDERATIONS IBDis a polygenic disorder that gives rise to multiple clinical subgroups withinUC andCD. Genome-wide searches have shown that potential disease-associated loci are present on chromosomes 16, 12, 7, 3, and 1, although the specific gene associations are undefined. HLA alleles may play a role. UC patients disproportionately express DR2-related alleles, whereas in CD an increased use of the DR5 DQ1 haplotype or the DRB*0301 allele has been described. In UC patients with pancolitis undergoing total proctocolectomy, 14.3% versus 3.2% of non-IBD controls express the HLA DRB1*0103 allele. This allele is associated with extensive disease and extraintestinal manifestations such as mouth ulcers, arthritis, and uveitis. Other associations with immunoregulatory genes include the intercellular adhesion molecule R241 allele in UC and CD and the interleukin (IL) 1 receptor antagonist allele 2 in UC patients that is associated with total colonic inflammation. Although not proven at the genetic level, patients with IBD and their first-degree relatives may exhibit diminished intestinal epithelial cell barrier function. DEFECTIVE IMMUNE REGULATION IN IBD The normal state of the mucosal immune system is one of inhibited immune responses to lumenal contents due to oral tolerance that occurs in the normal individual. When soluble antigens are administered orally rather than subcutaneously or intramuscularly, antigen-specific non-responsiveness is induced. Multiple mechanisms are involved in the induction of oral tolerance and include deletion or anergy of antigen-reactive T cells or activation of CD4+ T cells that suppress gut inflammation through secretion of inhibitory cytokines (IL-10, TGF-b). Oral tolerance may be responsible for the lack of immune responsiveness to dietary antigens and the commensal flora in the intestinal
lumen. In IBD this tightly regulated state of suppression of inflammation is altered, leading to uncontrolled inflammation. The mechanisms that maintain this regulated state of immune suppression are unknown. Gene knockout (-/-) or transgenic (Tg) mouse models of colitis have revealed that deleting specific cytokines (e.g.,IL-2, IL-10, TGF-b) or their receptors, deleting molecules associated with T-cell antigen recognition (e.g., T-cell antigen receptors, MHC class II), or interfering with intestinal epithelial cell barrier function (e.g., blocking N-cadherin, deleting multidrug resistance gene 1a or trefoil factor) leads to colitis. Thus, a variety of specific alterations can lead to unregulated autoimmunity directed at the colon in mice. In bothUC andCD, activated CD4+ T cells present in the lamina propria and peripheral blood secrete inflammatory cytokines. Some directly activate other inflammatory cells (macrophages and B cells) and others act indirectly to recruit other lymphocytes, inflammatory leukocytes, and mononuclear cells from the peripheral vasculature into the gut through interactions between homing receptors on leukocytes (e.g.,a4b7 integrin) and addressins on vascular endothelium (e.g., MadCAM1). CD4+ T cells can be subdivided into two major categories both of which may be associated with colitis in animal models and humans: TH1 cells (IFN-g, TNF) and TH2 cells (IL-4, IL-5, IL-13). TH1 cells appear to induce transmural granulomatous inflammation that resembles CD, and TH2 cells appear to induce superficial mucosal inflammation more characteristic of UC. The TH1 cytokine pathway is initiated by IL-12, a key cytokine in the pathogenesis of experimental models of mucosal inflammation. Thus, use of antibodies to block proinflammatory cytokines (e.g., anti-TNF-a, anti-IL-12) or molecules associated with leukocyte recruitment (e.g., anti-a4b7) or use of cytokines that inhibit inflammation (e.g., IL-10) or promote intestinal barrier function (e.g., IL-11) may be beneficial to humans with colitis. THE INFLAMMATORY CASCADE INIBD Once initiated in IBD, the immune inflammatory response is perpetuated as a consequence of T-cell activation. A sequential cascade of inflammatory mediators acts to extend the response; each step is a potential target for therapy. Inflammatory cytokines, such asIL-1, IL-6, and tumor necrosis factor (TNF) have diverse effects on tissue. They promote fibrogenesis, collagen production, activation of tissue metalloproteinases, and the production of other inflammatory mediators; they also activate the coagulation cascade in local blood vessels (e.g., increased production of von Willebrand's factor). These cytokines are normally produced in response to infection, but are usually turned off or inhibited at the appropriate time to limit tissue damage. In IBD their activity is not regulated, resulting in an imbalance between the proinflammatory and anti-inflammatory mediators. Therapies such as the 5-ASA compounds are potent inhibitors of these inflammatory mediators through inhibition of transcription factors such as NF-kB that regulate their expression. EXOGENOUS FACTORS IBDmay have an as yet undefined infectious etiology. Three specific agents have received the greatest attention, Mycobacterium paratuberculosis, Paramyxovirus, and
Helicobacter species. The immune response to a specific organism could be expressed differently, depending upon the individual's genetic background. Although M. paratuberculosis had initially been identified inCDpatients, further studies have not confirmed a disease association. In addition, antimycobacterial agents have not been effective in treating CD. A role for the measles virus or paramyxoviruses in the development of CD has been suggested based on an increase in the incidence of CD in England that paralleled use of the measles vaccine. However, studies in the United States have not substantiated this finding. In an animal model of IBD, Helicobacter hepaticus has been implicated as a trigger for the inflammatory response; evidence in people is lacking. Multiple pathogens (e.g., Salmonella, Shigella sp., Campylobacter sp.) may initiateIBD by triggering an inflammatory response that the mucosal immune system may fail to control. However, in an IBD patient the normal flora is likely perceived as if it were a pathogen. Anaerobic organisms, particularly Bacteroides species, may be responsible for the induction of inflammation. Such a notion is supported by the response in patients withCD to agents that alter the intestinal flora, such as metronidazole, ciprofloxacin, and elemental diets. CD also responds to fecal diversion, demonstrating the ability of lumenal contents to exacerbate disease. On the other hand, other bacterial organisms, so-called probiotics such as Lactobacillus sp., downregulate inflammation in animal models and humans. Psychosocial factors can contribute to clinical exacerbation of symptoms. Major life events such as illness or death in the family, divorce or separation, interpersonal conflict, or other major loss, are associated with an increase inIBDsymptoms such as pain, bowel dysfunction, and bleeding. Acute daily stress can exacerbate bowel symptoms even after controlling for major life events. When the sickness impact profile, a measurement of overall psychological and physical functioning is used, IBD patients have functional impairment greater than that of a health maintenance organization population but less than that of patients with chronic back pain or amyotrophic lateral sclerosis. IBD patients have been hypothesized to have a characteristic personality that renders them susceptible to emotional stresses. However, emotional dysfunction could also be the result of chronic illness and should be considered when treating these patients. PATHOLOGY ULCERATIVE COLITIS: MACROSCOPIC FEATURES UCis a mucosal disease that usually involves the rectum and extends proximally to involve all or part of the colon. Approximately 40 to 50% of patients have disease limited to the rectum and rectosigmoid, 30 to 40% have disease extending beyond the sigmoid but not involving the whole colon, and 20% have a total colitis. Proximal spread occurs in continuity without areas of uninvolved mucosa. When the whole colon is involved, the inflammation extends 1 to 2 cm into the terminal ileum in 10 to 20% of patients. This is called backwash ileitis and has little clinical significance. Although variations in macroscopic activity may suggest skip areas, biopsies from normal-appearing mucosa are usually abnormal. Thus, it is important to obtain multiple biopsies from apparently uninvolved mucosa, whether proximal or distal, during endoscopy.
With mild inflammation, the mucosa is erythematous and has a fine granular surface that looks like sandpaper. In more severe disease, the mucosa is hemorrhagic, edematous, and ulcerated (Fig. 287-1). In long-standing disease, inflammatory polyps (pseudopolyps) may be present as a result of epithelial regeneration (Fig. 287-2). The mucosa may appear normal in remission but in patients with many years of disease it appears atrophic and featureless and the entire colon becomes narrowed and foreshortened (Fig. 287-3). Patients with fulminant disease can develop a toxic colitis or a toxic megacolon where the bowel wall becomes very thin and the mucosa is severely ulcerated, which may lead to perforation. ULCERATIVE COLITIS: MICROSCOPIC FEATURES Histologic findings correlate well with the endoscopic appearance and clinical course ofUC. The process is limited to the mucosa and superficial submucosa with deeper layers unaffected except in fulminant disease (Fig. 287-4). In UC, two major histologic features are indicative of chronicity and help distinguish it from infectious or acute self-limited colitis. First, the crypt architecture of the colon is distorted; crypts may be bifid and reduced in number, often with a gap between the crypt bases and the muscularis mucosae. Second, some patients have basal plasma cells and multiple basal lymphoid aggregates. Mucosal vascular congestion with edema and focal hemorrhage, and an inflammatory cell infiltrate of neutrophils, lymphocytes, plasma cells, and macrophages may be present. The neutrophils invade the epithelium, usually in the crypts, and give rise to cryptitis and, ultimately, to crypt abscesses (Fig. 287-5). The cryptitis is associated with mucus discharge from goblet cells and increased epithelial cell turnover. Histologically, this results in goblet cell depletion. Other chronic changes that are sometimes seen are neuronal hypertrophy and fibromuscular hyperplasia of the muscularis mucosae. CROHN'S DISEASE: MACROSCOPIC FEATURES CDcan affect any part of the gastrointestinal tract from the mouth to the anus. Some 30 to 40% of patients have small bowel disease alone, 40 to 55% have disease involving both the small and large intestines, and 15 to 25% have colitis alone. In the 75% of patients with small intestinal disease, the terminal ileum is involved in 90%. UnlikeUC, which almost always involves the rectum, the rectum is often spared in CD. CD is segmental, with skip areas in the midst of diseased intestine (Fig. 287-6). Perirectal fistulas, fissures, abscesses, and anal stenosis are present in one-third of patients with CD, particularly those with colonic involvement. CD may also involve the liver and the pancreas. UnlikeUC,CD is a transmural process (Fig. 287-7). Endoscopically, aphthous or small superficial ulcerations characterize mild disease; in more active disease, stellate ulcerations fuse longitudinally and transversely to demarcate islands of mucosa that frequently are histologically normal. This "cobblestone" appearance is characteristic of CD, both endoscopically and by barium radiography. As in UC, pseudopolyps can form in CD. Active CD is characterized by focal inflammation and formation of fistula tracts, which
resolve by fibrosis and stricturing of the bowel. The bowel wall thickens and becomes narrowed and fibrotic, leading to chronic, recurrent bowel obstructions. Projections of thickened mesentery encase the bowel ("creeping fat") and serosal and mesenteric inflammation promote adhesions and fistula formation. CROHN'S DISEASE: MICROSCOPIC FEATURES The earliest lesions are aphthoid ulcerations and focal crypt abscesses with loose aggregations of macrophages, which form noncaseating granulomas in all layers of the bowel wall from mucosa to serosa (Fig. 287-8). Granulomas can be seen in lymph nodes, mesentery, peritoneum, liver, and pancreas. Although granulomas are a pathognomonic feature ofCD, only half of cases reveal granulomas on surgical or endoscopic biopsy specimens. Other histologic features of CD include submucosal or subserosal lymphoid aggregates, particularly away from areas of ulceration, gross and microscopic skip areas, and transmural inflammation that is accompanied by fissures that penetrate deeply into the bowel wall and sometimes form fistulous tracts or local abscesses. CLINICAL PRESENTATION ULCERATIVE COLITIS Signs and Symptoms The major symptoms ofUC are diarrhea, rectal bleeding, tenesmus, passage of mucus, and crampy abdominal pain. The severity of symptoms correlates with the extent of disease. Although UC can present acutely, symptoms usually have been present for weeks to months. Occasionally, diarrhea and bleeding are so intermittent and mild that the patient does not seek medical attention. Patients with proctitis usually pass fresh blood or blood-stained mucus, either mixed with stool or streaked onto the surface of a normal or hard stool. They also have tenesmus, or urgency with a feeling of incomplete evacuation. They rarely have abdominal pain. With proctitis or proctosigmoiditis, proximal transit slows, which may account for the constipation that is commonly seen in patients with distal disease. When the disease extends beyond the rectum, blood is usually mixed with stool, or grossly bloody diarrhea may be noted. Colonic motility is altered by inflammation with rapid transit through the inflamed intestine. When the disease is severe, patients pass a liquid stool containing blood, pus, and fecal matter. Diarrhea is often nocturnal and/or postprandial. Although severe pain is not a prominent symptom, some patients with active disease may experience vague lower abdominal discomfort or mild central abdominal cramping. Severe cramping and abdominal pain can occur in association with severe attacks of the disease. Other symptoms in moderate to severe disease include anorexia, nausea, vomiting, fever, and weight loss. Physical signs of proctitis include a tender anal canal and blood on rectal exam. With more extensive disease, patients have tenderness to palpation directly over the colon. Patients with a toxic colitis have severe pain and bleeding, and those with megacolon have hepatic tympany. Both may have signs of peritonitis if a perforation has occurred. The classification of disease activity is shown inTable 287-2.
Laboratory, Endoscopic, and Radiographic Features Active disease can be associated with a rise in acute phase reactants (C-reactive protein, orosomucoid levels), platelet count, erythrocyte sedimentation rate (ESR) and a decrease in hemoglobin. In severely ill patients, the serum albumin level will fall rather quickly. Leukocytosis may be present but is not a specific indicator of disease activity. Proctitis or proctosigmoiditis rarely causes a rise in C-reactive protein. Diagnosis relies upon the patient's history; clinical symptoms, negative stool examination for bacteria, Clostridium difficile toxin, and ova and parasites; sigmoidoscopic appearance; and histology of rectal or colonic biopsy specimens. Sigmoidoscopy is used to assess disease activity and is often performed before treatment. Histologic features change more slowly than clinical features but can also be used to grade disease activity (Table 287-3). Patients with a severe attack ofUCshould have a plain, supine film of the abdomen. In the presence of severe disease, the margin of the colon becomes edematous and irregular (Fig. 287-9). Colonic thickening and toxic dilation can both be seen on a plain radiograph. The earliest radiologic change ofUC seen on single-contrast barium enema is a fine mucosal granularity (Fig. 287-10). With increasing severity, the mucosa becomes thickened and superficial ulcers are seen. Deep ulcerations can appear as "collar-button" ulcers, which indicate that the ulceration has penetrated the mucosa. Haustral folds may be normal in mild disease, but as activity progresses they become edematous and thickened. Loss of haustration can occur, especially in patients with long-standing disease. In addition, the colon becomes shortened and narrowed. Polyps in the colon may be postinflammatory polyps or pseudopolyps (Fig. 287-11), adenomatous polyps, or carcinoma. Computed tomography (CT) scanning is not as helpful as endoscopy and barium enema in making the diagnosis ofUC, but typical findings include mild mural thickening (2 cm, homogeneous wall density, mural thickening of small bowel, mesenteric fat stranding, perianal disease, and adenopathy. CT scanning can help identify abscesses, fistulas, and sinus tracts. Magnetic resonance imaging (MRI) may prove superior for demonstrating pelvic lesions such as ischiorectal abscesses. Complications BecauseCD is a transmural process, serosal adhesions develop that provide direct pathways for fistula formation and reduce the incidence of free perforation. Free perforation occurs in 1 to 2% of patients, usually in the ileum but occasionally in the jejunum or as a complication of toxic megacolon. The peritonitis of free perforation, especially colonic, may be fatal. Generalized peritonitis may also result from the rupture of an intraabdominal abscess. Other complications include intestinal obstruction in 40%, massive hemorrhage, malabsorption, and severe perianal disease. Serologic Markers Several serologic markers may be used to differentiate betweenCD andUC and help to predict the course of disease. Two antibodies that can be detected in the serum ofIBDpatients are perinuclear antineutrophil cytoplasmic antibody (pANCA) and anti-Saccharomyces cerevisiae antibodies (ASCA). A distinct set of antineutrophil cytoplasmic antibodies with perinuclear staining by indirect immunofluorescence is associated with UC. The antigens to which these antibodies are directed have not been identified, but they are distinct from those associated with vasculitis and may be related to histones. pANCA positivity is found in about 60 to 70% of UC patients and 5 to 10% of CD patients; 5 to 15% of first-degree relatives of UC patients are pANCA positive, whereas only 2 to 3% of the general population is pANCA positive. pANCA may also identify specific disease phenotypes. pANCA positivity is more often associated with pancolitis, early surgery, pouchitis, or inflammation of the pouch after ileal pouch-anal anastamosis (IPAA) and primary sclerosing cholangitis. pANCA in CD is associated with colonic disease that resembles UC.
ASCAantibodies recognize mannose sequences in the cell wall mannan of S. cerevisiae; 60 to 70% ofCDpatients, 10 to 15% of UC patients, and up to 5% of non-IBDcontrols are ASCA positive. The combined measurement ofpANCAand ASCA has been advocated as a valuable diagnostic approach to IBD. In one report, pANCA positivity with ASCA negativity yielded a 57% sensitivity and 97% specificity for UC, whereas pANCA negativity with ASCA positivity yielded a 49% sensitivity and 97% specificity for CD. ASCA was associated with small bowel CD. These antibody tests may help decide whether a patient with indeterminate colitis should undergo an IPAA, because patients with predominant features of CD often have a more difficult postoperative course. Anti-goblet cell autoantibodies (GABs) -- autoantibodies against two target antigens in colonic epithelial cells -- are present in 39% ofUCpatients, 30% ofCDpatients, 21% of first-degree relatives of UC patients, 19% of first-degree relatives of CD patients, and 2% of healthy controls. An anti-colon antibody is found in 36% of UC patients and 13% of CD patients and healthy controls. In addition, 31% of CD patients and 4% of UC patients have serum antibodies against pancreatic acinar cells or pancreatic autoantibodies (PABs). Antibodies to red cell membrane antigens that cross-react with enteropathogens such as Campylobacter sp. may be associated with hemolytic anemia in CD. None of these antibodies are useful in the diagnosis and management of patients withIBD. DIFFERENTIAL DIAGNOSIS OF UC AND CD UC and CD have similar features to many other diseases. In the absence of a key diagnostic test, a combination of clinical, laboratory, histopathologic, radiographic, and therapeutic observations is required (Table 287-4). Once a diagnosis ofIBD is made, distinguishing between UC and CD is impossible in 10 to 20% of cases. These are termed indeterminate colitis. INFECTIOUS DISEASE Infections of the small intestines and colon can mimicCD orUC. They may be bacterial, fungal, viral, or protozoal in origin (Table 287-5). Campylobacter colitis can mimic the endoscopic appearance of severe UC and can cause a relapse of established UC. Salmonella can cause watery or bloody diarrhea, nausea, and vomiting. Shigellosis causes watery diarrhea, abdominal pain, and fever followed by rectal tenesmus and by the passage of blood and mucus per rectum. All three are usually self-limited but 1% of patients infected with Salmonella become asymptomatic carriers. Yersinia enterocolitica infection occurs mainly in the terminal ileum and causes mucosal ulceration, neutrophil invasion, and thickening of the ileal wall. Other bacterial infections that may mimicIBDinclude C. difficile, which presents with watery diarrhea, tenesmus, nausea, and vomiting, and Escherichia coli, three categories of which can cause colitis. These are enterohemorrhagic, enteroinvasive, and enteroadherent E. coli, all of which can cause bloody diarrhea and abdominal tenderness. Diagnosis of bacterial colitis is made by sending stool specimens for bacterial culture and C. difficile toxin analysis. Gonorrhea, Chlamydia, and syphilis can also cause proctitis Gastrointestinal involvement with mycobacterial infection occurs primarily in the
immunosuppressed patient but may occur in patients with normal immunity. Distal ileal and cecal involvement predominates and patients present with symptoms of small bowel obstruction and a tender abdominal mass. The diagnosis is made most directly by colonoscopy with biopsy and culture. Mycobacterium avium intracellulare complex infection occurs in advanced stages of HIV infection and in other profoundly immunocompromised states, and usually manifests as a systemic infection with diarrhea, abdominal pain, weight loss, fever, and malabsorption. Diagnosis is established by acid-fast smear and culture of mucosal biopsies. Although most of the patients with viral colitis are immunosuppressed, cytomegalovirus (CMV) and herpes simplex proctitis may occur in immunocompetent individuals. CMV occurs most commonly in the esophagus, colon, and rectum, but may also involve the small intestine. Symptoms include abdominal pain, bloody diarrhea, fever, and weight loss. With severe disease, necrosis and perforation can occur. Diagnosis is made by identification of intranuclear inclusions in mucosal cells on biopsy. Herpes simplex infection of the gastrointestinal tract is limited to the oropharynx, anorectum, and perianal areas. Symptoms include anorectal pain, tenesmus, constipation, inguinal adenopathy, difficulty with urinary voiding, and sacral paresthesias. Diagnosis is made by rectal biopsy. HIV itself can cause diarrhea, nausea, vomiting, and anorexia. Small intestinal biopsies show partial villus atrophy; small bowel bacterial overgrowth and fat malabsorption may also be noted. Protozoan parasites include Isospora belli, which can cause a self-limited infection in healthy hosts but causes a chronic profuse, watery diarrhea and weight loss in AIDS patients. Entamoeba histolytica or related species infect about 10% of the world's population; symptoms include abdominal pain, tenesmus, frequent loose stool containing blood and mucus, and abdominal tenderness. Colonoscopy reveals focal punctate ulcers with normal intervening mucosa; diagnosis is made by biopsy or serum amebic antibodies. Fulminant amebic colitis is rare but has a mortality rate of>50%. Other parasitic infections that may mimic IBD include hookworm (Necator americanus), whipworm (Trichuris trichiura), and Strongyloides stercoralis. In severely immunocompromised patients Candida or Aspergillus can be identified in the submucosa. Disseminated histoplasmosis can involve the ileocecal area. NONINFECTIOUS DISEASE Many diseases may mimicIBD(Table 287-5). Diverticulitis can be confused withCDclinically and radiographically. Both diseases cause fever, abdominal pain, tender abdominal mass, leukocytosis, elevatedESR, partial obstruction, and fistulas. Perianal disease or ileitis on small bowel series favors the diagnosis of CD. Significant endoscopic mucosal abnormalities are more likely in CD than in diverticulitis. Endoscopic or clinical recurrence following segmental resection favors CD. Diverticular-associated colitis is similar to CD, but mucosal abnormalities are limited to the sigmoid and descending colon. Ischemic colitis is commonly confused withIBD. The ischemic process can be chronic and diffuse as inUC, or segmental as inCD. Colonic inflammation due to ischemia may resolve quickly or may persist and result in transmural scarring and stricture formation.
Ischemic bowel disease should be considered in the elderly following abdominal aortic aneurysm repair or when a patient has a hypercoagulable state or a severe cardiac or peripheral vascular disorder. Patients usually present with sudden onset of left lower quadrant pain, urgency to defecate, and the passage of bright red blood per rectum. Endoscopic examination often demonstrates a normal-appearing rectum and a sharp transition to an area of inflammation in the descending colon and splenic flexure. The effects of radiation therapy on the gastrointestinal tract can be difficult to distinguish fromIBD. Acute symptoms can occur within 1 to 2 weeks of starting radiotherapy. When the rectum and sigmoid are irradiated, patients develop bloody, mucoid diarrhea and tenesmus, as in distalUC. With small bowel involvement, diarrhea is common. Late symptoms include malabsorption and weight loss. Stricturing with obstruction and bacterial overgrowth may occur. Fistulas can penetrate the bladder, vagina, or abdominal wall. Flexible sigmoidoscopy reveals mucosal granularity, friability, numerous telangiectasias, and occasionally discrete ulcerations. Biopsy can be diagnostic. Solitary rectal ulcer syndrome is uncommon and can be confused withIBD. It occurs in mostly young females and may be caused by impaired evacuation and failure of relaxation of the puborectalis muscle. Ulceration may arise from anal sphincter overactivity, higher intrarectal pressures during defecation, and digital removal of stool. Patients complain of constipation with straining and pass blood and mucus per rectum. Other symptoms include abdominal pain, diarrhea, tenesmus, and perineal pain. The ulceration, which can be as large as 5 cm in diameter, is usually seen anteriorly or anteriorlaterally 3 to 15 cm from the anal verge. Biopsies can be diagnostic. Several types of colitis have been associated with nonsteroidal anti-inflammatory drugs (NSAID), including de novo colitis, reactivation ofIBD, and proctitis caused by use of suppositories. Most patients with NSAID-related colitis present with diarrhea and abdominal pain and complications include stricture, bleeding, obstruction, perforation, and fistulization. Withdrawal of these agents is crucial, and in cases of reactivated IBD, standard therapies are indicated. INDETERMINITE COLITIS Cases ofIBD that cannot be categorized asUC orCD are called indeterminate colitis. Long-term follow-up reduces the number of patients labeled indeterminate to about 10%. The disease course of indeterminate colitis is unclear and surgical recommendations are difficult, especially since up to 20% of pouches fail, requiring ileosotomy. A multistage ileal pouch-anal anastamosis (the initial stage consisting of a subtotal colectomy with Hartman pouch) with careful histologic evaluation of the resected specimen to exclude CD is advised. Medical therapy is similar to UC and CD; most clinicians use 5-ASA drugs, glucocorticoids, and immunomodulators as necessary. THE ATYPICAL COLITIDIES Two atypical colitides -- collagenous colitis and lymphocytic colitis -- have completely normal endoscopic appearances. Collagenous colitis has two main histologic components: increased subepithelial collagen deposition and colitis with increased intraepithelial lymphocytes. Female to male ratio is 9:1, and most patients present in the
sixth or seventh decades of life. The main symptom is chronic watery diarrhea. Treatments range from sulfasalazine and lomotil to bismuth to glucocorticoids for refractory disease. Lymphocytic colitis has features similar to collagenous colitis including age at onset and clinical presentation, but it has almost equal incidence in men and women and no subepithelial collagen deposition on pathologic section. However, intraepithelial lymphocytes are increased. Diarrhea stops in the majority of patients treated with sulfasalazine or prednisone. Diversion colitis is an inflammatory process that arises in segments of the large intestine that are excluded from the fecal stream. It usually occurs in patients with ileostomy or colostomy when a mucus fistula or a Hartman's pouch has been created. Diversion colitis is reversible by surgical reanastamosis. Clinically, patients have mucus or bloody discharge from the rectum. Erythema, granularity, friability, and, in more severe cases, ulceration can be seen on endoscopy. Histopathology shows areas of active inflammation with foci of cryptitis and crypt abscesses. Crypt architecture is normal and this differentiates it fromUC. It may be impossible to distinguish fromCD. Short-chain fatty acid enemas will help in diversion colitis, but the definitive therapy is surgical reanastamosis. EXTRAINTESTINAL MANIFESTATIONS IBDis associated with a variety of extraintestinal manifestations; up to one-third of patients have at least one. Patients with perianalCD are at higher risk for developing extraintestinal manifestations than other IBD patients. DERMATOLOGIC Erythema nodosum (EN) occurs in up to 15% ofCDpatients and 10% ofUCpatients. Attacks usually correlate with bowel activity; skin lesions develop after the onset of bowel symptoms, and patients frequently have concomitant active peripheral arthritis. The lesions of EN are hot, red, tender nodules measuring 1 to 5 cm in diameter and are found on the anterior surface of the lower legs, ankles, calves, thighs, and arms. Therapy is directed toward the underlying bowel disease. Pyoderma gangrenosum (PG) is seen in 1 to 12% ofUCpatients and less commonly inCDcolitis. Although it usually presents after the diagnosis ofIBD, PG may occur years before the onset of bowel symptoms, run a course independent of the bowel disease, respond poorly to colectomy, and even develop years after proctocolectomy. It is usually associated with severe disease. Lesions are commonly found on the dorsal surface of the feet and legs but may occur on the arms, chest, stoma, and even the face. PG usually begins as a pustule and then spreads concentrically to rapidly undermine healthy skin. Lesions then ulcerate with violaceous edges surrounded by a margin of erythema. Centrally, they contain necrotic tissue with blood and exudates. Lesions may be single or multiple and grow as large as 30 cm. They are sometimes very difficult to treat and often require intravenous antibiotics, intravenous glucocorticoids, dapsone, purinethinol, thalidomide, or intravenous cyclosporine.
Other dermatologic manifestations include pyoderma vegetans that occurs in intertriginous areas, pyostomatitis vegetans that involves the mucous membranes, and metastaticCD, a rare disorder defined by cutaneous granuloma formation. Psoriasis affects 5 to 10% of patients withIBD and is unrelated to bowel activity. Perianal skin tags are found in 75 to 80% of patients with CD, especially those with colonic involvement. Oral mucosal lesions are seen often in CD and rarely inUC and include aphthous stomatitis and "cobblestone" lesions of the buccal mucosa. RHEUMATOLOGIC Peripheral arthritis develops in 15 to 20% ofIBDpatients, is more common inCD, and worsens with exacerbations of bowel activity. It is asymmetric, polyarticular, and migratory, and most often affects large joints of the upper and lower extremities. Treatment is directed at reducing bowel inflammation. In severe UC, colectomy frequently cures the arthritis. Ankylosing spondylitis (AS) occurs in about 10% ofIBDpatients and is more common inCD thanUC. About two-thirds of IBD patients with AS test positive for the HLA-B27 antigen. The activity of AS is not related to bowel activity and does not remit with glucocorticoids or colectomy. It most often affects the spine and pelvis, producing symptoms of diffuse low-back pain, buttock pain, and morning stiffness. The course is continuous and progressive leading to permanent skeletal damage and deformity. Sacroiliitis is symmetrical, occurs equally inUC andCD, is often asymptomatic, does not correlate with bowel activity, and does not necessarily progress toAS. Other rheumatic manifestations include hypertrophic osteoarthropathy, osteoporosis and osteomalacia secondary to malabsorption of calcium and vitamin D as well as glucocorticoid therapy, pelvic/femoral osteomyelitis, and relapsing polychondritis. OCULAR The incidence of ocular complications inIBDpatients is 1 to 10%. The most common are conjunctivitis, anterior uveitis/iritis, and episcleritis. Uveitis is associated with bothUC andCDcolitis, may be found during periods of remission, and may develop in patients following bowel resection. Symptoms include ocular pain, photophobia, blurred vision, and headache. Prompt intervention, sometimes with systemic glucocorticoids, is required to prevent scarring and visual impairment. Episcleritis is a benign disorder that presents with symptoms of mild ocular burning. It occurs in 3 to 4% of IBD patients, more commonly in CD colitis, and is treated with topical glucocorticoids. HEPATOBILIARY Hepatic steatosis is detectable in about half of the abnormal liver biopsies from patients withCD andUC; patients usually present with hepatomegaly. Fatty liver usually results from a combination of chronic debilitating illness, malnutrition, and glucocorticoid therapy. Cholelithiasis is more common in CD than UC and occurs in 10 to 35% of patients with ileitis or ileal resection. Gallstone formation is caused by malabsorption of bile acids resulting in depletion of the bile salt pool and the secretion of lithogenic bile.
Primary sclerosing cholangitis (PSC) is characterized by both intrahepatic and extrahepatic bile duct inflammation and fibrosis (Fig. 287-17), frequently leading to biliary cirrhosis and hepatic failure; 1 to 5% of patients withIBD have PSC, but 50 to 75% of patients with PSC have IBD. Although it can be recognized after the diagnosis of IBD, PSC can be detected earlier or even years after proctocolectomy. Most patients have no symptoms at the time of diagnosis; when symptoms are present they consist of fatigue, jaundice, abdominal pain, fever, anorexia, and malaise. Diagnosis is made by endoscopic retrograde cholangiopancreatography (ERCP), which demonstrates multiple bile duct strictures alternating with relatively normal segments. The bile acid ursodeoxycholic acid (ursodiol) may reduce alkaline phosphatase and serum aminotransferase levels, but histologic improvement has been marginal and it has no definitive long-term benefit. Endoscopic stenting may be palliative for cholestasis secondary to bile duct obstruction. Patients with symptomatic disease develop cirrhosis and liver failure over 5 to 10 years and eventually require liver transplantation. Ten percent of PSC patients develop cholangiocarcinoma and cannot be transplanted. Pericholangitis is a subset of PSC found in about 30% of IBD patients; it is confined to small bile ducts and is usually benign. UROLOGIC The most frequent genitourinary complications are calculi, ureteral obstruction, and fistulas. The highest frequency of nephrolithiasis (10 to 20%) occurs in patients withCDfollowing small bowel resection or ileostomy. Calcium oxalate stones develop secondary to hyperoxaluria, which results from increased absorption of dietary oxalate. Normally, dietary calcium combines with luminal oxalate to form insoluble calcium oxalate, which is eliminated in the stool. In patients with ileal dysfunction, however, nonabsorbed fatty acids bind calcium and leave oxalate unbound. The unbound oxalate is then delivered to the colon, where it is readily absorbed, especially in the presence of colonic inflammation. OTHER The risk of thromboembolic disease increases whenIBDbecomes active, and patients may present with deep vein thrombosis, pulmonary embolism, cerebrovascular accidents, and arterial emboli. Factors responsible for the hypercoagulable state include reactive thrombocytosis, increased levels of fibrinopeptide A, factor V, factor VIII, fibrinogen, accelerated thromboplastin generation, antithrombin III deficiency secondary to increased gut losses or increased catabolism, and free protein S deficiency. A spectrum of vasculitidies involving small, medium, and large vessels has also been observed in IBD patients. Patients withIBD have an increased prevalence of osteoporosis secondary to vitamin D deficiency, calcium malabsorption, malnutrition, and corticosteroid use. Deficiencies of vitamin B12 and fat-soluble vitamins may occur after ileal resection or with ileal disease. More common cardiopulmonary manifestations include endocarditis, myocarditis, pleuropericarditis, and interstitial lung disease. A secondary or reactive amyloidosis can occur in patients with long-standingIBD, especially in patients withCD. Amyloid material is deposited systemically and can cause diarrhea, constipation, and renal failure. The
renal disease can be successfully treated with colchicine. Pancreatitis is a rare extra-intestinal manifestation of IBD and results from duodenal fistulas, ampullary CD, gallstones, PSC, drugs such as 6-mercaptopurine or azathioprine, autoimmune pancreatitis, and primary CD of the pancreas. TREATMENT 5-ASA Agents The mainstay of therapy for mild to moderateUC andCDcolitis is sulfasalazine and the other 5-ASA agents. Sulfasalazine was originally developed to deliver both antibacterial (sulfapyridine) and anti-inflammatory (5-aminosalicylic acid, 5-ASA) therapy into the connective tissues of joints and the colonic mucosa. The molecular structure provides a convenient delivery system to the colon by allowing the intact molecule to pass through the small intestine after only partial absorption, and to be broken down in the colon by bacterial azo reductases that cleave the azo bond linking the sulfa and 5-ASA moieties. Sulfasalazine is effective in inducing and maintaining remission in mild to moderate UC and CD ileocolitis and colitis, but its high rate of side effects limits its use. Although sulfasalazine is more effective at higher doses, at 6 or 8 g/d up to 30% of patients experience allergic reactions or intolerable side effects such as headache, anorexia, nausea, and vomiting that are attributable to the sulfapyridine moiety. Hypersensitivity reactions, independent of sulfapyridine levels, include rash, fever, hepatitis, agranulocytosis, hypersensitivity pneumonitis, pancreatitis, worsening of colitis, and reversible sperm abnormalities. Sulfasalazine can also impair folate absorption and patients should be supplemented with folic acid. Newer sulfa-free aminosalicylate preparations deliver increased amounts of the pharmacologically active ingredient of sulfasalazine (5-ASA, mesalamine) to the site of active bowel disease while limiting systemic toxicity. 5-ASA may function through inhibition of NF-kB activity. Sulfa-free aminosalicylate formulations include alternative azo-bonded carriers, 5-ASA dimers, pH-dependent tablets, and continuous-release preparations. Each has the same efficacy as sulfasalazine when equimolar concentrations are used. Olsalazine is composed of two 5-ASA radicals linked by an azo bond which is split in the colon by bacterial reduction and two 5-ASA molecules are released. Olsalazine is similar in effectiveness to sulfasalazine in treatingCD andUC, but up to 17% of patients experience non-bloody diarrhea caused by increased secretion of fluid in the small bowel. Balsalazide contains an azo bond binding mesalamine to the carrier molecule 4-amino benzoyl balanine; it is effective in the colon. Claversal is an enteric-coated form of 5-ASA that consists of mesalamine surrounded by an acrylic-based polymer resin and a cellulose coating that releases mesalamine at pH> 6.0, a level that is present from the mid-jejunum continuously to the distal colon. The most commonly used drugs besides sulfasalazine in the United States are Asacol and Pentasa. Asacol is also an enteric-coated form of mesalamine, but it has a slightly different release pattern, with 5-ASA liberated at pH> 7.0. The disintegration of Asacol is variable with complete break-up of the tablet occurring in many different parts of the gut ranging from the small intestine to the splenic flexure; it has increased gastric residence when taken with a meal. Asacol is used to induce and maintain remission inUC and in CDileitis, ileocolitis, and colitis. Appropriate doses of Asacol and the other 5-ASA compounds are shown in Table 287-6. Some 50 to 75% of patients with mild to moderate UC and CD improve when treated with 2 g/d of 5-ASA; the dose response
continues up to at least 4.8 g/d. Doses of 1.5-4 g/d maintain remission in 50 to 75% of patients with UC and CD. Pentasa is another mesalamine formulation that uses an ethylcellulose coating to allow water absorption into small beads containing the mesalamine. Water dissolves the 5-ASA, which then diffuses out of the bead into the lumen. Disintegration of the capsule occurs in the stomach. The microspheres then disperse throughout the entire gastrointestinal tract from the small intestine through the distal colon in both fasted or fed conditions. Controlled trials of Pentasa and Asacol in activeCDdemonstrate a 40 to 60% clinical improvement or remission, and meta-analyses demonstrate maintenance of CD remission with 1.5 to 3 g/d of 5-ASA in 68 to 95% of patients. Pentasa at a dose of 2 g/d is more effective than placebo in postoperative prophylaxis of CD. Topical mesalamine enemas are effective in mild-to-moderate distalUC andCD. Clinical response occurs in up to 80% of UC patients with colitis distal to the splenic flexure. Mesalamine suppositories, which are no longer available in the United States but are available in Canada, at doses of 500 mg twice a day are effective in treating proctitis. Glucocorticoids The majority of patients with moderate to severeUCbenefit from oral or parenteral glucocorticoids. Prednisone is usually started at doses of 40 to 60 mg/d for active UC that is unresponsive to 5-ASA therapy. Parenteral glucocorticoids may be administered as intravenous hydrocortisone 300 mg/d or methylprednisolone 40 to 60 mg/d. Adrenocorticotropic hormone (ACTH) is occasionally preferred for glucocorticoid-naive patients despite a risk of adrenal hemorrhage. ACTH has equivalent efficacy to intravenous hydrocortisone in both glucocorticoid-naive and -experiencedCDpatients. Topically applied glucocorticoids are also beneficial for distal colitis and may serve as an adjunct in those who have rectal involvement plus more proximal disease. Hydrocortisone enemas or foam may control active disease, although they have no proven role as maintenance therapy. These glucocorticoids are significantly absorbed from the rectum and can lead to adrenal suppression with prolonged administration. The systemic effects of standard glucocorticoid formulations have led to the development of more potent formulations that are less well absorbed and have increased first-pass metabolism. Budesonide is being used in enema form with favorable preliminary results in distalUC. Glucocorticoids are also effective for treatment of moderate-to-severeCD and induce a 60 to 70% remission rate compared to a 30% placebo response. Controlled ileal-release budesonide has been nearly equal to prednisone for ileocolonic CD at a dose of 9 mg/d. Glucocorticoids play no role in maintenance therapy in eitherUC orCD. Once clinical remission has been induced, they should be tapered according to the clinical activity, normally at a rate of no more than 5 mg per week. They can usually be tapered to 20 mg/d within 4 to 5 weeks but often take several months to be discontinued altogether. The side effects are numerous, including fluid retention, abdominal striae, fat redistribution, hyperglycemia, subcapsular cataracts, osteonecrosis, myopathy, emotional disturbances, and withdrawal symptoms. Most of these side effects, aside from osteonecrosis, are related to the dose and duration of therapy.
Antibiotics Despite numerous trials, antibiotics have no role in the treatment of active or quiescentUC. However, pouchitis, which occurs in about a third of UC patients after colectomy and ileal pouch-anal anastamosis, usually responds to treatment with metronidazole or ciprofloxacin. Metronidazole is effective in active inflammatory, fistulous, and perianalCD and may prevent recurrence after ileal resection. The most effective dose is 15 to 20 mg/kg per day in three divided doses; it is usually continued for several months. Common side effects include nausea, metallic taste, and disulfiram-like reaction. Peripheral neuropathy can occur with prolonged administration (several months) and on rare occasions is permanent despite discontinuation. Ciprofloxacin (500 mg bid) is also beneficial for inflammatory, perianal, and fistulous CD. These two antibiotics should be used as second-line drugs in active CD after 5-ASA agents and as first-line drugs in perianal and fistulous CD. Azathioprine and 6-Mercaptopurine Azathioprine and 6-mercaptopurine (6-MP) are purine analogues commonly employed in the management of glucocorticoid-dependentIBD. Azathioprine is rapidly absorbed and converted to 6-MP, which is then metabolized to the active end product, thioinosinic acid, an inhibitor of purine ribonucleotide synthesis and cell proliferation. These agents also inhibit the immune response. Efficacy is seen at 3 to 4 weeks. Compliance can be monitored by measuring the level of 6-thioguanine, an end product of 6-MP metabolism. Azathioprine (2.0 to 2.5 mg/kg per day) or 6-MP (1.0-1.5 mg/kg per day) have been employed successfully as glucocorticoid-sparing agents in up to two-thirds of UCand CDpatients previously unable to be weaned from glucocorticoids. The role of these immunomodulators as maintenance therapy in UC and CD and for treating active perianal disease and fistulas in CD appears promising. In addition, 6-MP at a dose of 50 mg/d is more effective than Pentasa or placebo for postoperative prophylaxis of CD. Although azathioprine and6-MP are usually well tolerated, pancreatitis occurs in 3 to 4% of patients, typically presents within the first few weeks of therapy, and is always completely reversible when the drug is stopped. Other side effects include nausea, fever, rash, and hepatitis. Bone marrow suppression (particularly leukopenia) is dose-related and often delayed, necessitating regular monitoring of the complete blood count. Additionally, 1 in 300 individuals lacks thiopurine methyltransferase, the enzyme responsible for drug metabolism; an additional 11% of the population are heterozygotes with intermediate enzyme activity. Both are at increased risk of toxicity because of increased accumulation of thioguanine metabolites. No increased risk of cancer has been documented inIBDpatients taking these medications long-term. Methotrexate Methotrexate (MTX) inhibits dihydrofolate reductase, resulting in impaired DNA synthesis. Additional anti-inflammatory properties may be related to decreasedIL-1 production. Intramuscular or subcutaneous MTX (25 mg per week) is effective in inducing remission and reducing glucocorticoid dosage and 15 mg per week is effective in maintaining remission in activeCD. Potential toxicities include leukopenia and hepatic fibrosis, necessitating periodic evaluation of complete blood counts and liver enzymes. The role of liver biopsy in patients on long-term MTX is uncertain. Hypersensitivity pneumonitis is a rare but serious complication of therapy. MTX should only be used
when either6-MP or azathioprine are ineffective or poorly tolerated. Cyclosporine Cyclosporine (CSA) alters the immune response by acting as a potent inhibitor of T cell-mediated responses. Although CSA acts primarily via inhibition ofIL-2 production from T helper cells, it also decreases recruitment of cytotoxic T cells and blocks other cytokines, including IL-3, IL-4, interferona, andTNF. It has a more rapid onset of action than6-MP and azathioprine. CSAis most effective given at 4 mg/kg per day IV in severeUC that is refractory to intravenous glucocorticoids, with 82% of patients responding. CSA can be an alternative to colectomy. The long-term success of oral CSA is not as dramatic, but if patients are started on6-MP or azathioprine at the time of hospital discharge, remission can be maintained. Intravenous CSA is effective in 80% of patients with refractory fistulas, but 6-MP or azathioprine must be used to maintain remission. Oral CSA alone is only effective at a higher dose (7.5 mg/kg per day) in active disease but is not effective in maintaining remission without 6-MP/azathioprine. Serum levels should be monitored and kept in the range of 200 to 400 ng/mL. CSAhas the potential for significant toxicity, and renal function should be frequently monitored. Hypertension, gingival hyperplasia, hypertrichosis, paresthesias, tremors, headaches, and electrolyte abnormalities are common side effects. Creatinine elevation calls for dose reduction or discontinuation. Seizures may also complicate therapy, especially if serum cholesterol levels are less than 120 mg/dL. Opportunistic infections, most notably Pneumocystis carinii pneumonia, have occurred with combination immunosuppressive treatment; prophylaxis should then be given. Nutritional Therapies Dietary antigens may act as stimuli of the mucosal immune response. Patients with activeCDrespond to bowel rest, along with total enteral or total parenteral nutrition (TPN). Bowel rest and TPN are as effective as glucocorticoids for inducing remission of active CD but are not as effective as maintenance therapy. Enteral nutrition in the form of elemental or peptide-based preparations are also as effective as glucocorticiods or TPN, but these diets are not palatable. Enteral diets may provide the small intestine with nutrients vital to cell growth and do not have the complications of TPN. In contrast to CD, activeUC is not effectively treated with either elemental diets or TPN. Standard medical management of UC and CD is reviewed inTable 287-7. Newer Medical Therapies Anti-tumor Necrosis Factor Antibody TNF is a key inflammatory cytokine and mediator of intestinal inflammation. The expression of TNF is increased inIBD. Infliximab is a chimeric mouse-human monoclonal antibody against TNF that is extremely effective in CD. It blocks TNF in the serum and at the cell surface and likely lyses TNF-producing macrophages and T cells through complement fixation and antibody-dependent cytotoxicity. Of activeCDpatients refractory to glucocorticoids, 6-MP, or 5-ASA, 65% will respond to intravenous infliximab (5 mg/kg); one-third will enter complete remission. Patients who experience an initial response will respond again to repeated infusions of infliximab every 8 weeks up to 44 weeks. Thus infliximab may be also be efficacious in maintaining remission. However, more trials need to be completed on remission
maintenance after infliximab therapy. Infliximab is also effective inCDpatients with refractory perianal and enterocutaneous fistulas, with a 68% response rate (50% reduction in fistula drainage) and a 50% complete remission rate. The effects of infliximab for both inflammatory and fistulous disease last 12 weeks on average but longer in some patients. The incidence of antibodies to infliximab (25% of the molecule is murine) is 13%. One side effect is a lupus-like syndrome, which is rare and reversible after stopping the drug. Anti-double-stranded DNA antibodies occur in 9% but are not associated with clinical lupus. Among more than 1000 patients treated with infliximab, four developed lymphoma: one patient withCD, two with rheumatoid arthritis, and one with AIDS. Since the risk of lymphoma is already increased in these conditions, it is unclear whether infliximab is the cause. Thus, infliximab is extremely effective in refractory inflammatory and fistulous CD, but should be used only when necessary. Results on the efficacy of infliximab in UC are mixed. Newer Immunosuppressive Agents Tacrolimus has a mechanism of action similar to cyclosporine. It has shown efficacy in children with refractoryIBD and in adults with extensive involvement of the small bowel. Mycophenolate mofetil inhibits the de novo pathway of purine synthesis in lymphocytes, disrupting the conversion of inosine monophosphate to guanosine monophosphate (GMP) by reversible inhibition of inosine monophosphate dehydrogenase. The resulting depletion of intracellular GMP suppresses the generation of cytotoxic T cells and formation of antibodies by activated B cells. Patients withCD orUC who received either 500 mg twice a day or 15 mg/kg per day in two divided doses have tolerated the drug well and have experienced benefit with reduction of glucocorticoid requirements. Thalidomide has been shown to inhibitTNFproduction by monocytes and other cells. Thalidomide is effective in glucocorticoid refractory and fistulousCD, but randomized controlled trials still need to be performed. The Anti-Inflammatory Cytokines IL-10 is an anti-inflammatory and immunosuppressive cytokine produced by subsets of T and B cells, macrophages, and monocytes. It decreases TH1 production of IL-2 and interferon g, and limits production of IL-1, IL-6, IL-8,TNF, IL-12, and granulocyte-macrophage colony-stimulating factor. IL-10 has a moderate benefit in activeCD. IL-11 is a cytokine with thrombopoietic activity and mucosal protective effects that is effective in reducing inflammation in animal models of colitis. It seems to be effective in activeCD, but more trials are needed. Surgical Therapy Ulcerative Colitis Nearly half of patients with extensive chronicUCundergo surgery within the first 10 years of their illness. The indications for surgery are listed in Table
287-7. Morbidity is about 20% in elective, 30% for urgent, and 40% for emergency proctocolectomy. The risks are primarily hemorrhage, contamination and sepsis, and neural injury. Although single-stage total proctocolectomy with ileostomy has been the operation of choice, newer operations maintain continence while surgically removing the involved rectal mucosa. TheIPAA is the most frequent continence-preserving operation performed. BecauseUC is a mucosal disease, the rectal mucosa can be dissected out and removed down to the dentate line of the anus or about 2 cm proximal to it. The ileum is fashioned into a pouch that serves as a neorectum. This ileal pouch is then sutured circumferentially to the anus in an end-to-end fashion. If performed carefully, this operation preserves the anal sphincter and maintains continence. The overall operative morbidity is 10%, with the major complication being bowel obstruction. Pouch failure necessitating conversion to permanent ileostomy occurs in 5 to 10% of patients. Some inflamed rectal mucosa is usually left behind, and thus endoscopic surveillance is necessary. Primary dysplasia of the ileal mucosa of the pouch has occurred rarely. Patients withIPAAs usually have about six to eight bowel movements a day. On validated quality of life indices, they report better performance in sports and sexual activities than ileostomy patients. The most frequent late complication of IPAA is pouchitis in about one-third of patients withUC. This syndrome consists of increased stool frequency, watery stools, cramping, urgency, nocturnal leakage of stool, arthralgias, malaise, and fever. Although it usually responds to antibiotics, in 3% of patients it is refractory and requires pouch take-down. Crohn's Disease Most patients withCDrequire at least one operation in their lifetime. The need for surgery is related to duration of disease and the site of involvement. Patients with small bowel disease have an 80% chance of requiring surgery. Those with colitis alone have a 50% chance. The indications for surgery are shown inTable 287-7. SMALL INTESTINAL DISEASE BecauseCD is chronic and recurrent with no clear surgical cure, as little intestine as possible is resected. Current surgical alternatives for treatment of obstructing CD include resection of the diseased segment and strictureplasty. Surgical resection of the diseased segment is the most frequently performed operation, and in most cases primary anastomosis can be done to restore continuity. If much of the small bowel has already been resected and the strictures are short with intervening areas of normal mucosa, strictureplasties should be done to avoid a functionally insufficient length of bowel. The strictured area of intestine is incised longitudinally and the incision sutured transversely, thus widening the narrowed area. Complications of strictureplasty include prolonged ileus, hemorrhage, fistula, abscess, leak, and restricture. Colorectal Disease A greater percentage of patients withCDcolitis require surgery for intractability, fulminant disease, and anorectal disease. Several alternatives are available, ranging from the use of a temporary loop ileostomy to resection of segments of diseased colon or even the entire colon and rectum. For patients with segmental involvement, segmental colon resection with primary anastomosis can be performed. In 20 to 25% of patients with extensive colitis, the rectum is spared sufficiently to consider rectal preservation. Most surgeons believe that anIPAA is contraindicated in CD due to
the high incidence of pouch failure. A diverting colostomy may help heal severe perianal disease or rectovaginal fistulas, but disease almost always recurs with reanastomosis. Often, these patients require a total proctocolectomy and ileostomy. INFLAMMATORY BOWEL DISEASE AND PREGNANCY When adjusted for patient age, the fertility rate inUC is probably normal. In contrast, fertility is reduced inCD in proportion to disease activity and can be restored when remission is induced. The ovaries and fallopian tubes can be affected by the inflammatory process of CD, especially on the right side because of the proximity of the terminal ileum. In addition, perirectal, perineal, rectovaginal abscesses, and fistulae can result in dyspareunia. Infertility in men can be caused by sulfasalazine but reverses when treatment is stopped. InUC, fetal outcome approximates that in the normal population. InCD, spontaneous abortions, stillbirths, and developmental defects are increased with increased disease activity, not medications. The courses of CD and UC during pregnancy mostly correlate with disease activity at the time of conception. Most CD patients can deliver vaginally, but cesarean section may be the preferred route of delivery for patients with anorectal and perirectal abscesses and fistulas to reduce the likelihood of fistulas developing or extending into the episiotomy scar. Sulfasalazine, mesalamine, and olsalazine are safe for use in pregnancy, but folate supplementation must be given with sulfasalazine. No adverse affects have been reported from sulfasalazine in nursing infants. Topical 5-ASA agents are also safe during pregnancy. Glucocorticoids are generally safe for use during pregnancy and are indicated for patients with moderate to severe disease activity. The amount of glucocorticoids received by the nursing infant is minimal. The safest antibiotics to use forCD in pregnancy are ampicillin, cephalosporin, or ciprofloxicin. Flagyl is teratogenic and tumorigenic in high doses, passes into breast milk, and should be avoided. 6-MPand azathioprine pose minimal or no risk during pregnancy, but experience is limited. If the patient cannot be weaned from the drug or has an exacerbation that requires 6-MP/azathioprine during pregnancy, she should continue the drug with informed consent. Their effects during nursing are unknown. There is little data on cyclosporine in pregnancy. In a small number of patients with severeIBDtreated with intravenous cyclosporine during pregnancy, 80% of pregnancies were successfully completed without development of renal toxicity, congenital malformations, or developmental defects. However, because of the lack of data, cyclosporine should probably be avoided unless the patient would otherwise require surgery. Methotrexate is contraindicated in pregnancy and nursing. Surgery inUCshould be performed only for emergency indications, including severe hemorrhage, perforation, and megacolon refractory to medical therapy. Total colectomy and ileostomy carries a 60% risk of postoperative spontaneous abortion. Fetal mortality is also high in CD requiring surgery. Patients with ileostomies andIPAAs tolerate pregnancy well.
INFLAMMATORY BOWEL DISEASE IN THE ELDERLY The most common presenting symptoms in the elderly are diarrhea, weight loss, and abdominal pain. CDin the elderly is mostly colonic with a distal distribution and occurs predominantly in women. Proctitis has been documented in 50% of elderly patients and the diagnosis is often delayed. Diseases that can mimic CD in the elderly are ischemic colitis, diverticular disease, irritable bowel, infectious colitides, and malignancies, including carcinoma, lymphoma, and carcinoid. The incidence of surgery is high in elderly patients, with up to 50% of patients with ileitis, ileocolitis, or extensive colitis requiring urgent or early surgery for first-time disease. In addition, surgery has a much higher morbidity than in younger patients, although the rate of postoperative recurrence is less. Most elderly patients respond as well as younger individuals to medical management. UCin the elderly is more common in men, presents usually with diarrhea and weight loss, and may have a more distal distribution than in younger patients. Most elderly patients have a favorable response to medical therapy, especially 5-ASA agents, and immunosuppressives used in conjunction with low doses of glucocorticoids. Cyclosporine has been used more frequently in the elderly, but the age-related decreases in renal clearance may affect dosing. Glucocorticoid complications such as osteoporosis and hyperglycemia are also increased in the elderly.6-MP and azathioprine are well tolerated in the elderly. Surgery also has a higher morbidity and mortality in UC, and elderly patients have a longer hospital stay than younger patients. The risk of colon cancer in UC andCDcolitis is no greater than that in the general population since the duration of disease is short and the extent of disease is often distal. CANCER IN INFLAMMATORY BOWEL DISEASE ULCERATIVE COLITIS Patients with long-standingUC are at increased risk for developing colonic epithelial dysplasia and carcinoma (Figs. 287-18,287-19, and Fig. 287-20). Several features distinguish sporadic (SCC) and colitis-associated (CAC) colon cancers. First, SCC usually arise from an adenomatous polyp; CAC typically arise from either flat dysplasia or a dysplasia-associated lesion or mass (DALM). Second, multiple synchronous colon cancers occur in 3 to 5% of SCC but in 12% of CAC. Third, the mean age of individuals with SCC is in the sixties; the mean age of those with CAC is in the thirties. Fourth, SCC exhibits a left-sided predominance, whereas CAC is distributed more uniformly throughout the colon. Fifth, mucinous and anaplastic cancers are more common in CAC than SCC. At the molecular level, p53 mutations occur much earlier and APC gene mutations much later in CAC than SCC. The risk of neoplasia in chronicUCincreases with duration and extent of disease. For patients with pancolitis, the risk of cancer rises 0.5 to 1% per year after 8 to 10 years of disease. This observed increase in cancer rates has led to the endorsement of surveillance colonoscopies with biopsies for patients with chronic UC as the standard of care. Annual or biennial colonoscopy with multiple biopsies has been advocated for patients with more than 8 to 10 years of pancolitis or 12 to 15 years of left-sided colitis and has been widely employed to screen and survey for subsequent dysplasia and
carcinoma. CROHN'S DISEASE Risk factors for developing colorectal cancer inCD are a history of colonic (or ileocolonic) involvement and long disease duration. The cancer risks in CD andUC are probably equivalent for similar extent and duration of disease. In patients with extensive colonic involvement, the overall risk is increased 18-fold and the cumulative risk is 8% at 22 years. Thus, the same endoscopic surveillance strategy used for UC is recommended for patients with chronic CD colitis. A pediatric colonoscope can be used to pass narrow strictures in CD patients and impassable strictures can be surveyed with annual barium enemas. A colon resection should be performed if there is evidence of malignancy. MANAGEMENT OF DYSPLASIA AND CANCER If high grade dysplasia (HGD) is encountered on colonoscopic surveillance, the usual treatment forUC is colectomy and forCD is either colectomy or segmental resection. If low grade dysplasia (LGD) is found, the management is controversial. Many investigators recommend immediate colectomy, but some repeat the colonoscopy in 1 to 6 months and search for recurrent dysplasia. Polyps in chronic colitis can be removed endoscopically provided that biopsies of the surrounding mucosa are free of dysplasia. IBDpatients are also at greater risk for other malignancies. Patients withCD may have an increased risk of developing non-Hodgkin's lymphoma and squamous cell carcinoma of the skin. Although CD patients have a twelvefold increased risk of developing small bowel cancer, this type of carcinoma is extremely rare. QUALITY OF LIFE IN INFLAMMATORY BOWEL DISEASE The assessment of health-related quality of life plays an important role in the evaluation and treatment ofIBDpatients. Although clinical trials have generally relied upon traditional disease activity indices such as the Crohn's Disease Activity Index (CDAI) to measure therapeutic efficacy, these measures do not reflect quality of life. The Inflammatory Bowel Disease Questionnaire (IBDQ) is a validated, disease-specific instrument that has been used to measure quality of life. It is a 32-item questionnaire that measures global function, systemic and bowel symptoms, functional and social impairment, and emotional function. When compared to the general population, IBD patients have an impaired quality of life in all six categories. The most frequent concerns ofUCpatients are having an ostomy bag, developing cancer, effects of medication, the uncertain nature of the disease, and having surgery. The most frequent concerns of CDpatients are the uncertain nature of the disease, energy level, effects of medication, having surgery, and having an ostomy bag. (Bibliography omitted in Palm version) Back to Table of Contents
288. IRRITABLE BOWEL SYNDROME - Chung Owyang Irritable bowel syndrome (IBS) is a gastrointestinal (GI) disorder characterized by altered bowel habits and abdominal pain in the absence of detectable structural abnormalities. No clear diagnostic markers exist for IBS, so all definitions of the disease are based on the clinical presentation. The Rome criteria for the diagnosis of IBS are summarized inTable 288-1. IBS is one of the most common conditions encountered in clinical practice but one of the least well understood. Until recently, many physicians did not consider IBS to be a disease at all; they viewed it as nothing more than a somatic manifestation of psychological stress. With the availability of better techniques to study colonic and GI motility and visceral sensory function, along with the development of newer concepts about the importance of the brain in regulating gut function, significant progress has been made toward a better understanding of the pathogenesis of IBS. Improved methods of treatment may result from these insights. CLINICAL FEATURES IBSis a disorder of young people, with most new cases presenting before age 45. However, some reports suggest that the elderly are troubled by IBS symptoms up to 92% as often as middle-aged persons. Indeed, many of the diagnoses of "painful diverticular disease" given the elderly patients may represent IBS. Women are diagnosed with IBS two to three times as often as men and make up 80% of the population with severe IBS. Patients with IBS may fall into two broad clinical groups. Most commonly, patients have abdominal pain associated with altered bowel habits that include constipation, diarrhea, or both. In the second group, patients have painless diarrhea. This latter group accounts for20,000/uL, are common, especially in generalized peritonitis. Plain abdominal films may reveal dilation of the large and small bowel with edema of the small-bowel wall, as evidenced by the distance between adjacent loops of gas-filled small intestine. Diagnostic paracentesis is sometimes valuable in determining the nature of the exudate as well as whether bacteria can be demonstrated or cultured. GONOCOCCAL PERITONITIS This usually involves an extension of gonococcal infection from a primary focus in the female reproductive tract. The signs of inflammation usually are limited to the pelvis, but there may be findings of a mild generalized peritonitis. Occasionally, the patient has right upper quadrant pain and tenderness caused by gonococcal perihepatitis involving the liver capsule and adjacent peritoneum (Fitz-Hugh-Curtis syndrome) (Chap. 147). STARCH PERITONITIS An acute granulomatous peritonitis can develop in some patients as a foreign-body reaction to cornstarch used to powder surgical gloves. The clinical picture is that of acute abdominal pain and fever 10 to 30 days after an abdominal operation. The diagnosis can be made by paracentesis and demonstration of starch granules in monocytes. However, most patients are reexplored because of the fear of abscess or bacterial peritonitis, with the finding of foreign-body granuloma studding the peritoneum. PSEUDOMYXOMA PERITONEI This is a rare condition resulting from rupture of a mucocele of the appendix, a mucinous ovarian cyst, or mucin-secreting intestinal or ovarian adenocarcinoma. The abdomen becomes filled with masses of jelly-like mucus. Occasional patients are cured with removal of the mucocele or the ovarian cyst and most of the myxomatous material. In other cases, however, the mucoid material recurs, leading to progressive wasting and eventual death. Colloid carcinoma arising from the stomach or colon with peritoneal implants may resemble pseudomyxoma at laparotomy. The course of this type of highly
malignant tumor is one of rapid cachexia and early death. The diagnosis usually can be made by the appearance of many highly malignant cells in the peritoneal implants. PNEUMATOSIS CYSTOIDES INTESTINALIS This is a condition in which multiple gas-filled blebs or cysts accumulate in the intestinal wall beneath the serosal surface of the bowel. The exact source of the gas has not been explained satisfactorily. In some instances, this disease is associated with specific ulceration of the intestinal mucosa, in particular peptic ulcer with outlet obstruction. Cysts in the wall of the small bowel are seen as an occasional complication of mesenteric vascular occlusion. In the large bowel, these cysts are usually benign, may be seen with a variety of other disorders, and usually disappear over time. Physical findings are not specific and the diagnosis is made either by x-ray or at laparotomy. Occasionally, the subserosal cysts may rupture, resulting in pneumoperitoneum. CHYLOUS ASCITES SeeChap. 46. MESENTERIC LIPODYSTROPHY This is a rare disorder usually affecting middle-aged women and characterized pathologically by infiltration of the mesentery with lipid-laden macrophages and fibrous tissue. These patients present with ill-defined abdominal pain and occasionally an abdominal mass. The diagnosis is made at laparotomy by demonstration of thick fibrofatty masses at the root of the mesentery with retraction and distortion of the bowel loops. FAMILIAL MEDITERRANEAN FEVER Familial Mediterranean fever (FMF, familial paroxysmal polyserositis) is an inherited disorder, characterized by recurrent episodes of fever, peritonitis, and/or pleuritis. Arthritis, skin lesions, and amyloidosis are seen in some patients. FMFoccurs predominantly in patients of non-Ashkenazi (Sephardic) Jewish, Armenian, and Arabic ancestry. However, the disease has been seen in patients of Italian, Ashkenazi Jewish, and Anglo-Saxon descent as well as others. ETIOLOGY FMFis an autosomal recessive trait characterized by mutations in the MEFV gene located on 16p. The gene encodes a 781-amino acid protein called pyrin expressed in cells of the myeloid lineage. The gene is in the RoRet family, and the product appears to function as a transcription factor based on the presence of a nuclear localization signal, a zinc finger, and a coiled-coil domain. Its expression in granulocytes is increased by proinflammatory cytokines and reduced by anti-inflammatory cytokines. Mutations associated with FMF cluster in exon 10. Different mutations appear to be associated
with distinct disease manifestations; replacement of methionine 694 by valine is common in patients who have amyloidosis as a feature of the disease. Valine 726 replacement by alanine is rarely associated with amyloidosis. Other as yet unknown genes may modify the phenotype or be responsible for FMF in non-Mediterranean populations. PATHOLOGY Despite the striking clinical manifestations during an acute attack ofFMF, no specific pathologic alterations have been found. At laparotomy there is acute peritoneal inflammation with an exudate that contains a predominance of polymorphonuclear leukocytes. A disproportionately large number of male patients develop gallbladder disease with and without cholelithiasis. Pleural and joint inflammation are also nonspecific. In the amyloidosis that accompaniesFMF, amyloid is deposited in the intima and media of the arterioles, the subendothelial region of venules, the glomeruli, and the spleen. Aside from their vessels, the heart and liver are uninvolved. MANIFESTATIONS The symptoms ofFMFoften begin between the ages of 5 and 15, although attacks sometimes commence during infancy and onset has occurred as late as age 50. The duration and frequency of attacks vary greatly in the same patient, and their occurrence follows no set pattern. The usual acute episode lasts 1 to 2 days, but some may be prolonged for 7 to 10 days. The attacks range in frequency from twice weekly to once a year, but 2 to 4 weeks is the most common interval. Spontaneous remissions lasting years have been seen. The severity and frequency of the attacks decrease with age or with development of amyloidosis. Fever Fever is a cardinal manifestation and is present during most attacks. Rarely, fever may be present without serositis. The temperature may be preceded by a chill and will peak in 12 to 24 h. Defervescence is often accompanied by diaphoresis. The fever ranges from 38.5° to 40°C but is quite variable. Abdominal Pain Abdominal pain occurs in >95% of patients and may vary in severity in the same patient. Minor premonitory discomfort may precede an acute episode by 24 to 48 h. The pain usually starts in one quadrant and then spreads to involve the whole abdomen. The initial site is usually very tender. Tenderness may remain localized with referred pain in other areas, and may radiate to the back. Diaphragmatic irritation may lead to splinting of the chest and pain in one or both shoulders. Nausea and vomiting sometimes occur. The abdomen is usually distended and may become rigid, with decreased or absent bowel sounds. On x-ray the wall of the small intestine may appear edematous, transit of barium is slowed, and fluid levels may be seen. An abdominal operation may precipitate an acute attack ofFMF, which may be confused with other postoperative complications. Chest Pain Most patients with abdominal attacks have referred chest pain at one time or another, and 75% also develop acute pleuritic pain with or without abdominal
symptoms. In 30%, the attacks of pleuritis precede the onset of abdominal attacks by varying periods of time, and a small number of patients never develop abdominal attacks. Chest pain is usually unilateral and is associated with diminished breath sounds, a friction rub, or a transient pleural effusion. Joint Pain In Israel, 75% of patients report at least one episode of acute arthritis. Arthritis can be distinct from abdominal or pleural attacks, can be acute or, rarely, chronic, and may involve one or several joints. Effusions are common, with large joints most frequently involved. Radiologic findings are nonspecific. Despite careful search, frank arthritis rarely has been seen in the United States. Some patients have a history of rheumatic fever-like illness in childhood, but in a large series of patients, including 30 from the Middle East, acute arthritis was not observed. Mild arthralgia is common during acute attacks but is nonspecific. Skin Manifestations Skin involvement occurs in one-third of patients. These lesions consist of painful, erythematous areas of swelling from 5 to 20 cm in diameter, usually located on the lower legs, the medial malleolus, or the dorsum of the foot. They may occur without abdominal or pleural pain and subside within 24 to 48 h. Other Signs and Symptoms Involvement of other serosal membranes has been reported, but pericarditis and meningitis are rare. Hematuria, splenomegaly, and small white dots called colloid bodies in the ocular fundus are findings of questionable significance. Rarely, migraine-like headaches accompany acute abdominal attacks, and some patients have become somewhat irrational or show extreme emotional lability during attacks. Whether these are primary manifestations ofFMF or secondary effects of pain and fever is unclear. Complications Depression and lack of motivation are common, and patients withFMFrequire considerable support. A striking number of patients have developed gallbladder disease. Amyloidosis has been reported in Israel, North Africa, and elsewhere in the Middle East, but its occurrence is rare in the United States. These findings are even more striking because there are probably as many knownFMFpatients in the United States as in Israel. Thus, environmental or nutritional, as well as genetic, factors may play a role in the development of amyloidosis in FMF. LABORATORY FINDINGS Polymorphonuclear leukocytosis ranging from 10,000 to 30,000 cells/uL is almost invariable during acute attacks. The erythrocyte sedimentation rate is elevated during attacks but returns to normal between attacks. Plasma fibrinogen, serum haptoglobin, ceruloplasmin, and C-reactive protein increase during the episodes. Plasma lipids are normal, and no consistent abnormalities of hepatic or renal function are seen. When amyloidosis is present, laboratory findings are typical of a nephrotic syndrome followed by renal insufficiency. DIAGNOSIS
When the typical acute attacks ofFMFoccur in an individual of appropriate ethnic background with a family history of FMF, the diagnosis is easy. When a patient is seen for the first time, a variety of other febrile illnesses must be excluded, such as acute appendicitis, pancreatitis, porphyria, cholecystitis, intestinal obstruction, and other major abdominal catastrophes. Some inherited hyperlipidemias may mimic the clinical picture ofFMF, but lipid analysis will eliminate them from consideration. The FMF patient is not immune to other diseases, and when an attack differs from the usual pattern or is more prolonged, consideration should be given to other diagnostic possibilities. The pleural form of the disease is sometimes difficult to differentiate from acute pulmonary infection or infarction, but the rapid disappearance of signs and symptoms resolves the problem. The erythema is sometimes difficult to differentiate from superficial thrombophlebitis or cellulitis. The most difficult diagnostic problem inFMF is the patient who presents with fever alone. In this situation an extensive diagnostic workup for fever of unknown origin may be required. Fortunately, such patients are rare, and all eventually develop serosal involvement. Until specific diagnostic tests for FMF are available, patients with recurrent fever but without signs of inflammation of one of the serosal membranes should not be categorized as having FMF. PROGNOSIS Despite the severity of the symptoms during some acute attacks, most patients are remarkably free of debilitation between attacks and are able to lead fairly normal lives. The greatest hazard to patients is prolonged periods of hospitalization due to erroneous diagnoses or failure to understand the disease. In the United States, the prognosis of patients withFMF does not seem to be different from that of patients with other chronic nonfatal illnesses. Death usually results from causes unrelated to the underlying disease. In the past, ~25% ofFMFpatients in Israel developed amyloidosis, and this complication usually led to death. However, the widespread use of colchicine has resulted in a dramatic decrease in the incidence of amyloidosis. TREATMENT During the past 25 years, the outlook of patients withFMF has been altered dramatically. Chronic administration of colchicine greatly reduces the number of acute attacks of FMF. It is recommended that 0.6 mg colchicine be taken by mouth three times a day. If patients develop gastrointestinal side effects with this dose, it should be reduced to 0.6 mg taken twice a day. Although an occasional patient will respond to 0.6 mg taken only once a day, this amount is less likely to be beneficial. Most FMF patients will respond favorably to colchicine prophylaxis. (Bibliography omitted in Palm version)
Back to Table of Contents
290. ACUTE INTESTINAL OBSTRUCTION - William Silen ETIOLOGY AND CLASSIFICATION Intestinal obstruction may be mechanical or nonmechanical (resulting from neuromuscular disturbances that produce either adynamic or dynamic ileus). The causes of mechanical obstruction of the lumen are conveniently divided into (1) lesions extrinsic to the intestine, e.g., adhesive bands, internal and external hernias; (2) lesions intrinsic to the wall of the intestine, e.g., diverticulitis, carcinoma, regional enteritis; and (3) obturation of the lumen, e.g., gallstone obstruction, intussusception. Clinically, however, it is most useful to consider whether the obstructive mechanism involves the small or large intestine, because the causes, symptoms, and treatments are different (see below). Adhesions and external hernias are the most common causes of obstruction of the small intestine, constituting 70 to 75% of cases of this type. Adhesions, however, almost never produce obstruction of the colon, where carcinoma, sigmoid diverticulitis, and volvulus, in that order, are the most common causes and together account for about 90% of the cases. Primary intestinal pseudoobstruction (Chap. 289) is a chronic motility disorder that frequently mimics mechanical obstruction. Unnecessary operations in such patients should be avoided. Adynamic ileus is probably the most common overall cause of obstruction. The development of this condition is mediated via the hormonal component of the sympathoadrenal system. Adynamic ileus may occur after any peritoneal insult, and its severity and duration will be dependent to some degree on the type of peritoneal injury. Hydrochloric acid, colonic contents, and pancreatic enzymes are among the most irritating substances, whereas blood and urine are less so. Adynamic ileus occurs to some degree after any abdominal operation. Retroperitoneal hematomas, particularly associated with vertebral fracture, commonly cause severe adynamic ileus, and the latter may occur with other retroperitoneal conditions, such as ureteral calculus or severe pyelonephritis. Thoracic diseases, including lower-lobe pneumonia, fractured ribs, and myocardial infarction, frequently produce adynamic ileus, as do electrolyte disturbances, particularly potassium depletion. Finally, intestinal ischemia, whether the result of vascular occlusion or intestinal distention itself, may perpetuate an adynamic ileus. Spastic ileus or dynamic ileus is very uncommon and results from extreme and prolonged contraction of the intestine. It has been observed in heavy metal poisoning, uremia, porphyria, and extensive intestinal ulcerations. PATHOPHYSIOLOGY Distention of the intestine is caused by the accumulation of gas and fluid proximal to and within the obstructed segment. Between 70 and 80% of intestinal gas consists of swallowed air, and because this is composed mainly of nitrogen, which is poorly absorbed from the intestinal lumen, removal of air by continuous gastric suction is a useful adjunct in the treatment of intestinal distention. The accumulation of fluid proximal to the obstructing mechanism results not only from ingested fluid, swallowed saliva, gastric juice, and biliary and pancreatic secretions but also from interference with normal sodium and water transport. During the first 12 to 24 h of obstruction, there is a
marked depression of flux from lumen to blood of sodium and consequently water in the distended proximal intestine. After 24 h, there is movement of sodium and water into the lumen, contributing further to the distention and fluid losses. Intraluminal pressure rises from a normal of 2 to 4 cmH2O to 8 to 10 cmH2O. During peristalsis, when simple obstruction or a "closed loop" is present, pressures reach 30 to 60 cmH2O. Closed-loop obstruction of the small intestine results when the lumen is occluded at two points by a single mechanism such as a hernial ring or adhesive band, thus producing a closed loop whose blood supply is often obstructed at the same time. Strangulation of the loop itself is thus common in association with marked distention proximal to the involved loop. A form of closed-loop obstruction is encountered when complete obstruction of the colon exists in the presence of a competent ileocecal valve (85% of individuals). Although the blood supply of the colon is not entrapped within the obstructing mechanism, distention of the cecum is extreme because of its greater diameter (Laplace's law), and impairment of the intramural blood supply is considerable with consequent gangrene of the cecal wall, usually anteriorly. Necrosis of the small intestine may occur by the same mechanism of interference with intramural blood flow when distention is extreme, but this sequence is uncommon in the small intestine. Once impairment of blood supply occurs, bacterial invasion supervenes, and peritonitis develops. The systemic effects of extreme distention include elevation of the diaphragm with restricted ventilation and subsequent atelectasis. Venous return via the inferior vena cava may also be impaired. The loss of fluids and electrolytes may be extreme and, unless replacement is prompt, leads to hemoconcentration, hypovolemia, renal insufficiency, shock, and death. Vomiting, accumulation of fluids within the lumen by the mechanisms described above, and the sequestration of fluid into the edematous intestinal wall and peritoneal cavity as a result of impairment of venous return from the intestine all contribute to massive loss of fluid and electrolytes, especially potassium. As soon as significant impedance to venous return is present, the intestine becomes severely congested, and blood begins to seep into the intestinal lumen. Blood loss may reach significant levels when long segments of intestine are involved. SYMPTOMS Mechanical small-intestinal obstruction is characterized by cramping midabdominal pain, which tends to be more severe the higher the obstruction. The pain occurs in paroxysms, and the patient is relatively comfortable in the intervals between the pains. Audible borborygmi are often noted by the patient simultaneously with the paroxysms of pain. The pain may become less severe as distention progresses, probably because motility is impaired in the edematous intestine. When strangulation is present, the pain is usually more localized and may be steady and severe without a colicky component, a fact that often causes delay in diagnosis of obstruction. Vomiting is almost invariable, and it is earlier and more profuse the higher the obstruction. The vomitus initially contains bile and mucus and remains as such if the obstruction is high in the intestine. With low ileal obstruction, the vomitus becomes feculent, i.e., orange-brown in color with a foul odor, which results from the overgrowth of bacteria proximal to the obstruction. Hiccups (singultus) are common. Obstipation and failure to pass gas by rectum are invariably present when the obstruction is complete, although some stool and gas may be passed spontaneously or after an enema shortly after onset of the complete obstruction. Diarrhea is occasionally observed in partial obstruction. Blood in the stool is
rare but does occur in cases of intussusception. Other than some minor but inconsistent differences in pain patterns noted above, the symptoms of strangulating obstructions cannot be distinguished from those of nonstrangulating obstructions. Mechanical colonic obstruction produces colicky abdominal pain similar in quality to that of small-intestinal obstruction but of much lower intensity. Complaints of pain are occasionally absent in stoic elderly patients. Vomiting occurs late, if at all, particularly if the ileocecal valve is competent. Paradoxically, feculent vomitus is very rare. A history of recent alterations in bowel habits and blood in the stool is common because carcinoma and diverticulitis are the most frequent causes. Constipation becomes progressive, and obstipation with failure to pass gas ensues. Acute symptoms may develop over a period of a week. Cecal volvulus more closely resembles obstruction of the small intestine clinically, whereas patients with sigmoid volvulus more typically have the picture of colonic obstruction in which marked distention predominates, with relatively less pain. In adynamic ileus, colicky pain is absent, and only discomfort from distention is evident. Vomiting may be frequent but is rarely profuse. It usually consists of gastric contents and bile and is almost never feculent. Complete obstipation may or may not occur. Singultus (hiccups) is common. PHYSICAL FINDINGS Abdominal distention is the hallmark of all forms of intestinal obstruction. It is least marked in cases of obstruction high in the small intestine and most marked in colonic obstruction. Early, especially in closed-loop strangulating small-bowel obstruction, distention may be barely perceptible or absent. Tenderness and rigidity are usually minimal; the temperature is rarely above 37.8°C (100°F) in nonstrangulating obstruction of the small and large intestine. Contrary to popular belief, the same is true of strangulating obstruction until very late, a fact that has often resulted in unfortunate delay in treatment. Signs and symptoms of shock also occur very late in strangulating obstruction. The appearance of shock, tenderness, rigidity, and fever often means that contamination of the peritoneum with infected intestinal content has occurred. Hernial orifices should always be carefully examined for the presence of a mass. The presence of a palpable abdominal mass usually signifies a closed-loop strangulating small-bowel obstruction because the tense fluid-filled loop is the palpable lesion. Auscultation may reveal loud, high-pitched borborygmi coincident with the colicky pain, but this finding is often absent late in strangulating or nonstrangulating obstruction. A quiet abdomen does not eliminate the possibility of obstruction, nor does it necessarily establish the diagnosis of adynamic ileus. LABORATORY AND X-RAY FINDINGS Leukocytosis, with shift to the left, usually occurs when strangulation is present, but a normal white blood cell count does not exclude strangulation. Elevation of the serum amylase level is encountered occasionally in all forms of intestinal obstruction, especially the strangulating variety. The x-ray is extremely valuable but under certain circumstances may also be
misleading. In nonstrangulating complete small-bowel obstruction, x-rays are almost completely reliable. Distention of fluid- and gas-filled loops of small intestine usually arranged in a "stepladder" pattern with air-fluid levels and an absence or paucity of colonic gas are pathognomonic (Fig. 290-1). These findings, however, are absent in slightly over half the cases of strangulating small-bowel obstruction, especially early in the disease. A general haze due to peritoneal fluid and sometimes a "coffee bean"-shaped mass are seen in strangulating obstruction. Occasionally, the films are normal, but when symptoms are consistent with obstruction of the small intestine, a normal film should suggest strangulation. In these circumstances, computed tomography may be very useful. Roentgenographic differentiation of partial mechanical small-bowel obstruction from adynamic ileus may be impossible because gas is present in both the small and large intestines; however, colonic distention is usually more prominent in adynamic ileus. A radiopaque dye given by mouth is useful in making this distinction. Colonic obstruction with a competent ileocecal valve is easily recognized because distention with gas is mainly confined to the colon. Barium enema, sigmoidoscopy, or colonoscopy, depending on the suspected site of obstruction, is usually advisable to determine the nature of the lesion, except when concomitant perforation is suspected, a rare occurrence. Sigmoidoscopy may be therapeutic in cases of sigmoid volvulus. When the ileocecal valve is incompetent, the films resemble those of partial small-bowel obstruction or adynamic ileus, and barium enema or colonoscopy is necessary to establish the correct diagnosis. Barium given by mouth is perfectly safe when obstruction is in the small intestine, since the barium sulfate does not become inspissated in this location. Barium should never be given by mouth to a patient with possible colonic obstruction until that possibility has been excluded by barium enema. TREATMENT Small-Intestinal Obstruction The overall mortality rate for obstruction of the small intestine is about 10%, even under the most optimal conditions. While the mortality rate for nonstrangulating obstruction is as low as 5 to 8%, that for strangulating obstruction has been reported to be between 20 and 75%. Well over half the deaths from small-bowel obstruction occur in those with strangulation; however, the latter constitute only one-fourth to one-third of the cases. Careful studies indicate that the clinical, laboratory, and x-ray findings are not reliable in distinguishing strangulating from nonstrangulating obstruction when obstruction is complete. Complete obstruction is suggested when passage of gas or stool per rectum has ceased and when gas is absent in the distal intestine by x-ray. Since strangulating small-bowel obstruction is always complete, operation should always be undertaken in such patients after suitable preparation. Before operation, fluid and electrolyte balance should be restored and decompression instituted by means of a nasogastric tube. Replacement of potassium is especially important because intake is nil and losses in vomitus are large. >From 6 to 8 h of preparation may be necessary. During this period, broad-spectrum antibiotics are indicated if strangulation is felt to be likely, but operation should not be delayed unless there is unequivocal clinical and roentgenographic evidence of resolution of the obstruction during the period of preparation. Attempts to pass a long tube into the small intestine usually fail while putting the patient through uncomfortable, unproductive manipulations that delay appropriate fluid replacement and decompression. There are
few, if any, indications for the use of a long intestinal tube. Procrastination of operation because of improvement in well-being of the patient during resuscitation and gastric decompression usually leads to unnecessary and hazardous delay in proper treatment. Purely nonoperative therapy is safe only in the presence of incomplete obstruction and is best utilized in patients with (1) repeated episodes of partial obstruction, (2) recent postoperative partial obstruction, and (3) partial obstruction following a recent episode of diffuse peritonitis. Colonic Obstruction The mortality rate for colonic obstruction is about 20%. As in small-bowel obstruction, nonoperative treatment is contraindicated unless the obstruction is incomplete. Occasionally, but not always, when the obstruction is incomplete, nonoperative therapy may result in sufficient decompression that a definitive operative procedure can be undertaken at a later date. This can usually be accomplished by discontinuation of all oral intake and perhaps by nasogastric suction, although attempts to decompress a completely obstructed colon by intubation are almost invariably futile. A long intestinal tube will not decompress an obstructed colon with a competent ileocecal valve. When obstruction is complete, early operation is mandatory, especially when the ileocecal valve is competent; cecal gangrene is likely if the cecal diameter exceeds 10 cm on plain abdominal film. For obstruction on the left side of the colon, the most common site, preliminary operative decompression by cecostomy or transverse colostomy followed by definitive resection of the primary lesion has been the treatment of choice. Recently, primary resection of obstructing left-sided lesions with on-table washout of the colon has been accomplished safely. For a lesion of the right or transverse colon, primary resection and anastomosis can be performed safely because distention of the ileum with consequent discrepancy in size and hazard in suture are not present. Adynamic Ileus This type of ileus usually responds to nonoperative continuous decompression and adequate treatment of the primary disease. The prognosis is usually good. Successful decompression of severe colonic ileus has been accomplished by colonoscopy, but this should be avoided if tenderness in the right lower quadrant suggests possible cecal gangrene. Neostigmine is effective in cases of colonic ileus that have not responded to other conservative treatment. Rarely, adynamic colonic distention may become so great that cecostomy is required if cecal gangrene is feared. Spastic ileus usually responds to treatment of the primary disease. (Bibliography omitted in Palm version) Back to Table of Contents
291. ACUTE APPENDICITIS - William Silen INCIDENCE AND EPIDEMIOLOGY The peak incidence of acute appendicitis is in the second and third decades of life; it is relatively rare at the extremes of age. Males and females are equally affected, except between puberty and age 25, when males predominate in a 3:2 ratio. Perforation is more common in infancy and in the aged, during which periods mortality rates are highest. The mortality rate has decreased steadily in Europe and the United States from 8.1 per 100,000 of the population in 1941 to less than 1 per 100,000 in 1970 and subsequently. The absolute incidence of the disease also decreased by about 40% between 1940 and 1960 but since then has remained unchanged. Although various factors such as changing dietary habits, altered intestinal flora, and better nutrition and intake of vitamins have been suggested to explain the reduced incidence, the exact reasons have not been elucidated. The overall incidence of appendicitis is much lower in underdeveloped countries, especially parts of Africa, and in lower socioeconomic groups. PATHOGENESIS Luminal obstruction has long been considered the pathogenetic hallmark. However, obstruction can be identified in only 30 to 40% of cases; ulceration of the mucosa is the initial event in the majority. The cause of the ulceration is unknown, although a viral etiology has been postulated. Infection with Yersinia organisms may cause the disease, since high complement fixation antibody titers have been found in up to 30% of cases of proven appendicitis. Whether the inflammatory reaction seen with ulceration is sufficient to obstruct the tiny appendiceal lumen even transiently is not clear. Obstruction, when present, is most commonly caused by a fecalith, which results from accumulation and inspissation of fecal matter around vegetable fibers. Enlarged lymphoid follicles associated with viral infections (e.g., measles), inspissated barium, worms (e.g., pinworms, Ascaris, and Taenia), and tumors (e.g., carcinoid or carcinoma) may also obstruct the lumen. Secretion of mucus distends the organ, which has a capacity of only 0.1 to 0.2 mL, and luminal pressures rise as high as 60 cmH 2O. Luminal bacteria multiply and invade the appendiceal wall as venous engorgement and subsequent arterial compromise result from the high intraluminal pressures. Finally, gangrene and perforation occur. If the process evolves slowly, adjacent organs such as the terminal ileum, cecum, and omentum may wall off the appendiceal area so that a localized abscess will develop, whereas rapid progression of vascular impairment may cause perforation with free access to the peritoneal cavity. Subsequent rupture of primary appendiceal abscesses may produce fistulas between the appendix and bladder, small intestine, sigmoid, or cecum. Occasionally, acute appendicitis may be the first manifestation of Crohn's disease. While chronic infection of the appendix with tuberculosis, amebiasis, and actinomycosis may occur, a useful clinical aphorism states that chronic appendiceal inflammation is not usually the cause of prolonged abdominal pain of weeks' or months' duration. In contrast, recurrent acute appendicitis does occur, often with complete resolution of inflammation and symptoms between attacks. Recurrent acute appendicitis may become more frequent as antibiotics are dispensed more freely and if a long
appendiceal stump is left after laparoscopic appendectomy. CLINICAL MANIFESTATIONS The history and sequence of symptoms are important diagnostic features of appendicitis. The initial symptom is almost invariably abdominal pain of the visceral type, resulting from appendiceal contractions or distention of the lumen. It is usually poorly localized in the periumbilical or epigastric region with an accompanying urge to defecate or pass flatus, neither of which relieves the distress. This visceral pain is mild, often cramping, and rarely catastrophic in nature, usually lasting 4 to 6 h, but it may not be noted by stoic individuals or by some patients during sleep. As inflammation spreads to the parietal peritoneal surfaces, the pain becomes somatic, steady, and more severe, aggravated by motion or cough, and usually located in the right lower quadrant. Anorexia is nearly universal; a hungry patient does not have acute appendicitis. Nausea and vomiting occur in 50 to 60% of cases, but vomiting is usually self-limited. The development of nausea and vomiting before the onset of pain is extremely rare. Change in bowel habit is of little diagnostic value, since any or no alteration may be observed, although the presence of diarrhea caused by an inflamed appendix in juxtaposition to the sigmoid may cause serious diagnostic difficulties. Urinary frequency and dysuria occur if the appendix lies adjacent to the bladder. The typical sequence of symptoms (poorly localized periumbilical pain followed by nausea and vomiting with subsequent shift of pain to the right lower quadrant) occurs in only 50 to 60% of patients. Physical findings vary with time after onset of the illness and according to the location of the appendix, which may be situated deep in the pelvic cul-de-sac; in the right lower quadrant in any relation to the peritoneum, cecum, and small intestine; in the right upper quadrant (especially during pregnancy); or even in the left lower quadrant. The diagnosis cannot be established unless tenderness can be elicited. While tenderness is sometimes absent in the early visceral stage of the disease, it ultimately always develops and is found in any location corresponding to the position of the appendix. Abdominal tenderness may be completely absent if a retrocecal or pelvic appendix is present, in which case the sole physical finding may be tenderness in the flank or on rectal or pelvic examination. Percussion, rebound tenderness, and referred rebound tenderness are often, but not invariably, present; they are most likely to be absent early in the illness. Flexion of the right hip and guarded movement by the patient are due to parietal peritoneal involvement. Hyperesthesia of the skin of the right lower quadrant and a positive psoas or obturator sign are often late findings and are rarely of diagnostic value. When the inflamed appendix is in close proximity to the anterior parietal peritoneum, muscular rigidity is present yet is often minimal early. The temperature is usually normal or slightly elevated [37.2 to 38°C (99 to 100.5°F)], but a temperature> 38.3°C (101°F) should suggest perforation. Tachycardia is commensurate with the elevation of the temperature. Rigidity and tenderness become more marked as the disease progresses to perforation and localized or diffuse peritonitis. Distention is rare unless severe diffuse peritonitis has developed. The disappearance of pain and tenderness just before perforation is extremely unusual. A mass may develop if localized perforation has occurred but usually will not be detectable before 3 days after onset. Earlier presence of a mass suggests carcinoma of the cecum or Crohn's disease. Perforation is rare before 24 h after onset of symptoms,
but the rate may be as high as 80% after 48 h. Diagnosis is based primarily on clinical grounds. Although moderate leukocytosis of 10,000 to 18,000 cells/uL is frequent (with a concomitant left shift), the absence of leukocytosis does not rule out acute appendicitis. Leukocytosis of >20,000 cells/uL suggests probable perforation. Anemia and blood in the stool suggest a primary diagnosis of carcinoma of the cecum, especially in elderly individuals. The urine may contain a few white or red blood cells without bacteria if the appendix lies close to the right ureter or bladder. Urinalysis is most useful in excluding genitourinary conditions that may mimic acute appendicitis. Radiographs are rarely of value except when an opaque fecalith (5% of patients) is observed in the right lower quadrant (especially in children). Consequently, abdominal films are not routinely obtained unless other conditions such as intestinal obstruction or ureteral calculus may be present. In some patients with recurrent or prolonged symptoms, a careful barium enema or computed tomography (CT) scan may reveal an extrinsic defect on the medial wall of the cecum or a calcified fecalith. The value of CT scan in acute appendicitis is being evaluated. The diagnosis may also be established by the ultrasonic demonstration of an enlarged and thick-walled appendix. Ultrasound is most useful to exclude ovarian cysts, ectopic pregnancy, or tuboovarian abscess. While the typical historic sequence and physical findings are present in 50 to 60% of cases, a wide variety of atypical patterns of disease are encountered, especially at the age extremes and during pregnancy. Infants under 2 years of age have a 70 to 80% incidence of perforation and generalized peritonitis. Any infant or child with diarrhea, vomiting, and abdominal pain is highly suspect. Fever is much more common in this age group, and abdominal distention is often the only physical finding. In the elderly, pain and tenderness are often blunted, and thus the diagnosis is frequently delayed and leads to a 30% incidence of perforation in patients over 70. Elderly patients often present initially with a slightly painful mass (a primary appendiceal abscess) or with adhesive intestinal obstruction 5 or 6 days after a previously undetected perforated appendix. Appendicitis occurs about once in every 1000 pregnancies and is the most common extrauterine condition requiring abdominal operation. The diagnosis may be missed or delayed because of the frequent occurrence of mild abdominal discomfort and nausea and vomiting during pregnancy. During the last trimester, when the mortality rate from appendicitis is highest, uterine displacement of the appendix to the right upper quadrant and laterally leads to confusion in diagnosis because pain and tenderness are similarly displaced. DIFFERENTIAL DIAGNOSIS Appendicitis can be confused with any condition that causes abdominal pain. Diagnostic accuracy is about 75 to 80% for experienced clinicians and must be based solely on the clinical criteria outlined. It is probably better to err slightly in the direction of overdiagnosis, since delay is associated with perforation and increased morbidity and mortality. In unperforated appendicitis, the mortality rate is 0.1%, little more than that associated with general anesthesia; for perforated appendicitis, overall mortality is 3%,
(15% in the elderly). In doubtful cases, 4 to 6 h of observation is always more beneficial than harmful. The most common conditions discovered at operation when acute appendicitis is erroneously diagnosed are, in order of frequency, mesenteric lymphadenitis, no organic disease, acute pelvic inflammatory disease, ruptured graafian follicle or corpus luteum cyst, and acute gastroenteritis. In addition, acute cholecystitis, perforated ulcer, acute pancreatitis, acute diverticulitis, strangulating intestinal obstruction, ureteral calculus, and pyelonephritis may present diagnostic difficulties. Differentiation of pelvic inflammatory disease from acute appendicitis on clinical grounds may be virtually impossible. Gram-negative intracellular diplococci on cervical smear are not pathognomonic unless Neisseria gonorrhoeae can be cultured. Pain on movement of the cervix is not specific and may occur in appendicitis if perforation has occurred or if the appendix lies adjacent to the uterus or adnexa. Rupture of a graafian follicle (mittelschmerz) occurs at midcycle and will spill off blood and fluid to produce pain and tenderness more diffuse and usually of a less severe degree than in appendicitis. Fever and leukocytosis are usually absent. Rupture of a corpus luteum cyst is identical clinically to rupture of a graafian follicle but develops about the time of menstruation. The presence of an adnexal mass, evidence of blood loss, and a positive pregnancy test help differentiate ruptured tubal pregnancy, but a negative pregnancy test is present when tubal abortion has occurred. Twisted ovarian cyst and endometriosis are occasionally difficult to distinguish from appendicitis. In all these female conditions, ultrasonography, laparoscopy, and occasionallyCT may be of great value. Acute mesenteric lymphadenitis is the diagnosis usually given when enlarged, slightly reddened lymph nodes at the root of the mesentery and a normal appendix are encountered at operation in a patient who usually has right lower quadrant tenderness. Whether this is a single, discrete entity is unclear, since the causative factor is not known. Some of these patients have infection with Y. pseudotuberculosis or Y. enterocolytica, in which case the diagnosis can be established by culture of the mesenteric nodes or by serologic titers (Chap. 162). The diagnosis is essentially impossible clinically, although retrospectively these patients may have a higher temperature and more diffuse pain and tenderness. Children seem to be affected more frequently than adults. Acute gastroenteritis usually causes profuse watery diarrhea, often with nausea and vomiting, but without localized findings. Between cramps, the abdomen is completely relaxed. In Salmonella gastroenteritis, the abdominal findings are similar, although the pain may be more severe and more localized, and fever and chills are common. The occurrence of similar symptoms among other members of the family may be helpful. When the diagnosis of acute pelvic appendicitis with perforation has been missed, gastroenteritis is the most common previous working diagnosis. Persistent abdominal or rectal tenderness should eliminate the diagnosis of gastroenteritis. Regional enteritis (Crohn's disease) is usually associated with a more prolonged history, often with previous exacerbations regarded as episodes of gastroenteritis unless the diagnosis has been established previously. Meckel's diverticulitis usually cannot be distinguished from acute appendicitis but is very rare. TREATMENT Cathartics and enemas should be avoided if appendicitis is under consideration, and antibiotics should not be administered when the diagnosis is in question, since they will
only mask the perforation. The treatment is early operation and appendectomy as soon as the patient can be prepared. Appendectomy is increasingly accomplished laparoscopically and may have some benefits over the open technique. Preparation for operation rarely takes more than 1 to 2 h in early appendicitis but may require 6 to 8 h in cases of severe sepsis and dehydration associated with late perforation. The only circumstance in which operation is not indicated is the presence of a palpable mass 3 to 5 days after the onset of symptoms. Should operation be undertaken at that time, a phlegmon rather than a definitive abscess will be found, and complications from its dissection are frequent. Such patients treated with broad-spectrum antibiotics, parenteral fluids, and rest usually show resolution of the mass and symptoms within 1 week. Interval appendectomy should be done safely 3 months later. Should the mass enlarge or the patient become more toxic, drainage of the abscess is necessary. The complications of subphrenic, pelvic, or other intraabdominal abscesses usually follow perforation with generalized peritonitis and can be avoided by early diagnosis of the disease. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 2 -LIVER AND BILIARY TRACT DISEASE 292. APPROACH TO THE PATIENT WITH LIVER DISEASE - Marc Ghany, Jay H. Hoofnagle In most instances, a diagnosis of liver disease can be made accurately by a careful history, physical examination, and application of a few laboratory tests. In some instances, radiologic examinations are helpful or, indeed, diagnostic. Liver biopsy is considered the "gold standard" in evaluation of liver disease but is now needed less for diagnosis than for grading and staging disease. This chapter provides an introduction to diagnosis and management of liver disease, briefly reviewing the structure and function of the liver; the major clinical manifestations of liver disease; and the use of clinical history, physical examination, laboratory tests, imaging studies, and liver biopsy. LIVER STRUCTURE AND FUNCTION The liver is the largest organ of the body, weighing 1 to 1.5 kg and representing 1.5 to 2.5% of the lean body mass. The size and shape of the liver vary and generally match the general body shape -- long and lean or squat and square. The liver is located in the right upper quadrant of the abdomen under the right lower rib cage against the diaphragm and projects for a variable extent into the left upper quadrant. The liver is held in place by ligamentous attachments to the diaphragm, peritoneum, great vessels, and upper gastrointestinal organs. It receives a dual blood supply; approximately 20% of the blood flow is oxygen-rich blood from the hepatic artery, and 80% is nutrient-rich blood from the portal vein arising from the stomach, intestines, and spleen. The majority of cells in the liver are hepatocytes, which constitute two-thirds of the mass of the liver. The remaining cell types are Kupffer cells (members of the reticuloendothelial system), stellate (Ito or fat-storing) cells, endothelial cells and blood vessels, bile ductular cells, and supporting structures. Viewed by light microscopy, the liver appears to be organized in lobules, with portal areas at the periphery and central veins in the center of each lobule. However, from a functional point of view, the liver is organized into acini, with both hepatic arterial and portal venous blood entering the acinus from the portal areas and then flowing through the sinusoids to the terminal hepatic veins. The advantage of viewing the acinus as the physiologic unit of the liver is that it helps to explain the morphologic patterns of many vascular and biliary diseases not explained by the lobular arrangement. Portal areas of the liver consist of small veins, arteries, bile ducts, and lymphatics organized in a loose stroma of supporting matrix and small amounts of collagen. Blood flowing into the portal areas is distributed through the sinusoids, passing from zone 1 to zone 3 of the acinus and draining into the terminal hepatic veins ("central veins"). The sinusoids are lined by unique endothelial cells that have prominent fenestrae of variable size, allowing the free flow of plasma but not cellular elements. The plasma is thus in direct contact with hepatocytes in the subendothelial space of Disse. Hepatocytes have distinct polarity. The basolateral side of the hepatocyte lines the space of Disse and is richly lined with microvilli; it demonstrates endocytotic and pinocytotic activity, with passive and active uptake of nutrients, proteins, and other
molecules. The apical pole of the hepatocyte forms the cannicular membranes through which bile components are secreted. The canniculi of hepatocytes form a fine network, which fuses into the bile ductular elements near the portal areas. Kupffer cells usually lie within the sinusoidal vascular space and represent the largest group of fixed macrophages in the body. The stellate cells are located in the space of Disse but are not usually prominent unless activated, when they produce collagen and matrix. Red blood cells stay in the sinusoidal space as blood flows through the lobules, but white blood cells can migrate through or around endothelial cells into the space of Disse and from there to portal areas, where they can return to the circulation through lymphatics. Hepatocytes perform numerous and vital roles in maintaining homeostasis and health. These functions include the synthesis of most essential serum proteins (albumin, carrier proteins, coagulation factors, many hormonal and growth factors), the production of bile and its carriers (bile acids, cholesterol, lecithin, phospholipids), the regulation of nutrients (glucose, glycogen, lipids, cholesterol, amino acids), and metabolism and conjugation of lipophilic compounds (bilirubin, cations, drugs) for excretion in the bile or urine. Measurement of these activities to assess liver function is complicated by the multiplicity and variability of these functions. The most commonly used liver "function" tests are measurements of serum bilirubin, albumin, and prothrombin time. The serum bilirubin level is a measure of hepatic conjugation and excretion, and the serum albumin level and prothrombin time are measures of protein synthesis. Abnormalities of bilirubin, albumin, and prothrombin time are typical of hepatic dysfunction. Frank liver failure is incompatible with life, and the functions of the liver are too complex and diverse to be subserved by a mechanical pump; dialysis membrane; or concoction of infused hormones, proteins, and growth factors. LIVER DISEASES While there are many causes of liver disease (Table 292-1), they generally present clinically in a few distinct patterns, usually classified as either hepatocellular or cholestatic (obstructive). In hepatocellular diseases (such as viral hepatitis or alcoholic liver disease), features of liver injury, inflammation, and necrosis predominate. In cholestatic diseases (such as gall stone or malignant obstruction, primary biliary cirrhosis, many drug-induced liver diseases), features of inhibition of bile flow predominate. The pattern of onset and prominence of symptoms can rapidly suggest a diagnosis, particularly if major risk factors are considered, such as the age and sex of the patient and a history of exposure or risk behaviors. Typical presenting symptoms of liver disease include jaundice, fatigue, itching, right upper quadrant pain, abdominal distention, and intestinal bleeding. At present, however, many patients are diagnosed with liver disease who have no symptoms and who have been found to have abnormalities in biochemical liver tests as a part of a routine physical examination or screening for blood donation or for insurance or employment. The wide availability of batteries of liver tests makes it relatively simple to demonstrate the presence of liver injury as well as to rule it out in someone suspected of liver disease. Evaluation of patients with liver disease should be directed at (1) establishing the etiologic diagnosis, (2) estimating the disease severity (grading), and (3) establishing
the disease stage (staging). Diagnosis should focus on the category of disease, such as hepatocellular versus cholestatic injury, as well as on the specific etiologic diagnosis. Grading refers to assessing the severity or activity of disease -- active or inactive, and mild, moderate, or severe. Staging refers to estimating the place in the course of the natural history of the disease, whether acute or chronic; early or late; precirrhotic, cirrhotic, or end-stage. The goal of this chapter is to introduce general, salient concepts in the evaluation of patients with liver disease that help lead to the diagnoses discussed in subsequent chapters. CLINICAL HISTORY The clinical history should focus on the symptoms of liver disease -- their nature, pattern of onset, and progression -- and on potential risk factors for liver disease. The symptoms of liver disease include constitutional symptoms such as fatigue, weakness, nausea, poor appetite, and malaise and the more liver-specific symptoms of jaundice, dark urine, light stools, itching, abdominal pain, and bloating. Symptoms can also suggest the presence of cirrhosis, end-stage liver disease, or complications of cirrhosis such as portal hypertension. Generally, the constellation of symptoms and their pattern of onset rather than a specific symptom points to an etiology. Fatigue is the most common and most characteristic symptom of liver disease. It is variously described as lethargy, weakness, listlessness, malaise, increased need for sleep, lack of stamina, and poor energy. The fatigue of liver disease typically arises after activity or exercise and is rarely present or severe in the morning after adequate rest (afternoon versus morning fatigue). Fatigue in liver disease is often intermittent and variable in severity from hour to hour and day to day. In some patients, it may not be clear whether fatigue is due to the liver disease or to other problems such as stress, anxiety, sleep disturbance, or a concurrent illness. Nausea occurs with more severe liver disease and may accompany fatigue or be provoked by odors of food or eating fatty foods. Vomiting can occur but is rarely persistent or prominent. Poor appetite with weight loss occurs commonly in acute liver diseases but is rare in chronic disease, except when cirrhosis is present and advanced. Diarrhea is uncommon in liver disease, except with severe jaundice, in which case lack of bile acids reaching the intestine can lead to steatorrhea. Right upper quadrant discomfort or ache ("liver pain") occurs in many liver diseases and is usually marked by tenderness over the liver area. The pain arises from stretching or irritation of Glisson's capsule, which surrounds the liver and is rich in nerve endings. Severe pain is most typical of gall bladder disease, liver abscess, and severe venoocclusive disease but is an occasional accompaniment of acute hepatitis. Itching occurs with acute liver disease, appearing early in obstructive jaundice (from biliary obstruction or drug-induced cholestasis) and somewhat later in hepatocellular disease (acute hepatitis). Itching also occurs in chronic liver diseases, typically the cholestatic forms such as primary biliary cirrhosis and sclerosing cholangitis where it is often the presenting symptom, occurring before the onset of jaundice. However, itching
can occur in any liver disease, particularly once cirrhosis is present. Jaundice is the hallmark symptom of liver disease and perhaps the most reliable marker of severity. Patients usually report darkening of the urine before they notice scleral icterus. Jaundice is rarely detectable with a bilirubin level less than 43 umol/L (2.5 mg/dL). With severe cholestasis there will also be lightening of the color of the stools and steatorrhea. Jaundice without dark urine usually indicates indirect (unconjugated) hyperbilirubinemia and is typical of hemolytic anemia and the genetic disorders of bilirubin conjugation, the common and benign form being Gilbert's syndrome and the rare and severe form being Crigler-Najjar syndrome. Gilbert's syndrome affects up to 5% of the population; the jaundice is more noticeable after fasting and with stress. Major risk factors for liver disease that should be sought in the clinical history include details of alcohol use, medications (including herbal compounds, birth control pills, and over-the-counter medications), personal habits, sexual activity, travel, exposure to jaundiced or other high-risk persons, injection drug use, recent surgery, remote or recent transfusion with blood and blood products, occupation, accidental exposure to blood or needlestick, and familial history of liver disease. For assessing the risk of viral hepatitis, a careful history of sexual activity is of particular importance and should include life-time number of sexual partners and, for men, a history of having sex with men. Sexual exposure is a common mode of spread of hepatitis B but is rare for hepatitis C. Maternal-infant transmission occurs with both hepatitis B and C. Vertical spread of hepatitis B can now be prevented by passive and active immunization of the infant at birth. Vertical spread of hepatitis C is uncommon, but there are no known means of prevention. A history of injection drug use, even in the remote past, is of great importance in assessing the risk for hepatitis B and C. Injection drug use is now the single most common risk factor for hepatitis C. Transfusion with blood or blood products is no longer an important risk factor for acute viral hepatitis. However, blood transfusions received before the introduction of sensitive enzyme immunoassays for antibody to hepatitis C virus (anti-HCV) in 1992 is an important risk factor for chronic hepatitis C. Blood transfusion before 1986, when screening for antibody to hepatitis B core antigen (anti-HBc) was introduced, is also a risk factor for hepatitis B. Travel to an underdeveloped area of the world, exposure to persons with jaundice, and exposure to young children in day-care centers are risk factors for hepatitis A. Tattooing and body piercing (for hepatitis B and C) and eating shellfish (for hepatitis A) are frequently mentioned but actually quite rate types of exposure for acquiring hepatitis. A history of alcohol intake is important in assessing the cause of liver disease and also in planning management and recommendations. In the United States, for example, at least 70% of adults drink alcohol to some degree, but significant alcohol intake is less common; in population-based surveys, only 5% have more than two drinks per day, the average drink representing 11 to 15 g alcohol. Alcohol consumption associated with an increased rate of alcoholic liver disease is probably more than two drinks (22 to 30 g) per day in women and three drinks (33 to 45 g) in men. Most patients with alcoholic cirrhosis have a much higher daily intake and have drunk excessively for 10 years or more before onset of liver disease. In assessing alcohol intake, the history should also focus upon whether alcohol abuse or dependence is present. Alcoholism is usually
defined on the behavioral patterns and consequences of alcohol intake, not on the basis of the amount of alcohol intake. Abuse is defined by a repetitive pattern of drinking alcohol that has adverse effects on social, family, occupational, or health status. Dependence is defined by alcohol-seeking behavior, despite its adverse effects. Many alcoholics demonstrate both dependence and abuse, and dependence is considered the more serious and advanced form of alcoholism. A clinically helpful approach to diagnosis of alcohol dependence and abuse is the use of the CAGE questionnaire (Table 292-2), which is recommended in all medical history taking. Family history can be helpful in assessing liver disease. Familial causes of liver disease include Wilson's disease; hemochromatosis anda 1-antitrypsin (a 1AT) deficiency; and the more uncommon inherited pediatric liver diseases of familial intrahepatic cholestasis (FIC), benign recurrent intrahepatic cholestasis (BRIC), and Alagille's syndrome. Onset of severe liver disease in childhood or adolescence with a family history of liver disease or neuropsychiatric disturbance should lead to investigation for Wilson's disease. A family history of cirrhosis, diabetes, or endocrine failure and the appearance of liver disease in adulthood should suggest hemochromatosis and lead to investigation of iron status. Patients with abnormal iron studies warrant genotyping of the HFE gene for the C282Y and H63D mutations typical of genetic hemochromatosis. A family history of emphysema should provoke investigation ofa 1AT levels and, if low, for Pi genotype. PHYSICAL EXAMINATION The physical examination rarely demonstrates evidence of liver dysfunction in a patient without symptoms or laboratory findings, nor are most signs of liver disease specific to one diagnosis. Thus, the physical examination usually complements rather than replaces the need for other diagnostic approaches. In many patients, the physical examination is normal unless the disease is acute or severe and advanced. Nevertheless, the physical examination is important in that it can be the first evidence for the presence of hepatic failure, portal hypertension, and liver decompensation. In addition, the physical examination can reveal signs that point to a specific diagnosis, either in risk factors or in associated diseases or findings. Typical physical findings in liver disease are icterus, hepatomegaly, hepatic tenderness, splenomegaly, spider angiomata, palmar erythema, and excoriations. Signs of advanced disease include muscle-wasting, ascites, edema, dilated abdominal veins, hepatic fetor, asterixis, mental confusion, stupor, and coma. Icterus is best appreciated by inspecting the sclera under natural light. In fair-skinned individuals, a yellow color of the skin may be obvious. In dark-skinned individuals, the mucous membranes below the tongue can demonstrate jaundice. Jaundice is rarely detectable if the serum bilirubin level is 1000 U/L -- occur almost exclusively in disorders associated with extensive hepatocellular injury such as (1) viral hepatitis, (2) ischemic liver injury (prolonged hypotension or acute heart failure), or (3) toxin or drug-induced liver injury. The pattern of the aminotransferase elevation can be helpful diagnostically. In most acute hepatocellular disorders, the ALTis higher than or equal to theAST. An AST:ALT ratio >2:1 is suggestive while a ratio>3:1 is highly suggestive of alcoholic liver disease. The AST in alcoholic liver disease is rarely>300 U/L and the ALT is often normal. A low level of ALT in the serum is due to an alcohol-induced deficiency of pyridoxal phosphate. The aminotransferases are usually not greatly elevated in obstructive jaundice. One notable exception occurs during the acute phase of biliary obstruction caused by the passage of a gallstone into the common bile duct. In this setting, the aminotransferases
can briefly be in the 1,000 to 2,000 U/L range. However, aminotransferase levels decrease quickly and the liver function tests rapidly evolve into one typical of cholestasis. Enzymes that Reflect Cholestasis The activities of three enzymes -- alkaline phosphatase, 5¢-nucleotidase, and gamma glutamyl transpeptidase (GGT) -- are usually elevated in cholestasis. Alkaline phosphatase and 5¢-nucleotidase are found in or near the bile canalicular membrane of hepatocytes, while GGT is located in the endoplasmic reticulum and in bile duct epithelial cells. Reflecting its more diffuse localization in the liver, GGT elevation in serum is less specific for cholestasis than are elevations of alkaline phosphatase or 5¢-nucleotidase. Some have advocated the use of GGT to identify patients with occult alcohol use. Its lack of specificity makes its use in this setting questionable. The normal serum alkaline phosphatase consists of many distinct isoenzymes found in the liver, bone, placenta, and, less commonly, small intestine. Patients over age 60 can have a mildly elevated alkaline phosphatase (1 to 11/2times normal), while individuals with blood types O and B can have an elevation of the serum alkaline phosphatase after eating a fatty meal due to the influx of intestinal alkaline phosphatase into the blood. It is also nonpathologically elevated in children and adolescents undergoing rapid bone growth because of bone alkaline phosphatase, and late in normal pregnancies due to the influx of placental alkaline phosphatase. Elevation of liver-derived alkaline phosphatase is not totally specific for cholestasis and a less than threefold elevation can be seen in almost any type of liver disease. Alkaline phosphatase elevations greater than four times normal occur primarily in patients with cholestatic liver disorders, infiltrative liver diseases such as cancer, and bone conditions characterized by rapid bone turnover (e.g., Paget's disease). In bone diseases, the elevation is due to increased amounts of the bone isoenzymes. In liver diseases, the elevation is almost always due to increased amounts of the liver isoenzyme. If an elevated serum alkaline phosphatase is the only abnormal finding in an apparently healthy person, or if the degree of elevation is higher than expected in the clinical setting, identification of the source of elevated isoenzymes is helpful (Fig. 293-1). This problem can be approached in several ways. First, and most precise, is the fractionation of the alkaline phosphatase by electrophoresis. The second approach is based on the observation that alkaline phosphatases from individual tissues differ in susceptibility to inactivation by heat. The finding of an elevated serum alkaline phosphatase level in a patient with a heat-stable fraction strongly suggests that the placenta or a tumor is the source of the elevated enzyme in serum. Susceptibility to inactivation by heat increases, respectively, for the intestinal, liver, and bone alkaline phosphatases, bone being by far the most sensitive. The third, best substantiated, and most available approach involves the measurement of serum 5¢-nucleotidase orGGT. These enzymes are rarely elevated in conditions other than liver disease. In the absence of jaundice or elevated aminotransferases, an elevated alkaline phosphatase of liver origin often, but not always, suggests early cholestasis and, less often, hepatic infiltration by tumor or granulomata. Other conditions that cause isolated elevations of the alkaline phosphatase include Hodgkin's disease, diabetes,
hyperthyroidism, congestive heart failure, and inflammatory bowel disease. The level of serum alkaline phosphatase elevation is not helpful in distinguishing between intrahepatic and extrahepatic cholestasis. There is essentially no difference among the values found in obstructive jaundice due to cancer, common duct stone, sclerosing cholangitis, or bile duct stricture. Values are similarly increased in patients with intrahepatic cholestasis due to drug-induced hepatitis, primary biliary cirrhosis, rejection of transplanted livers, and, rarely, alcohol-induced steatonecrosis. Values are also greatly elevated in hepatobiliary disorders seen in patients with AIDS (e.g., AIDS cholangiopathy due to cytomegalovirus or cryptosporidial infection and tuberculosis with hepatic involvement). TESTS THAT MEASURE BIOSYNTHETIC FUNCTION OF THE LIVER Serum Albumin Serum albumin is synthesized exclusively by hepatocytes. Serum albumin has a long half-life: 15 to 20 days, with approximately 4% degraded per day. Because of this slow turnover, the serum albumin is not a good indicator of acute or mild hepatic dysfunction; only minimal changes in the serum albumin are seen in acute liver conditions such as viral hepatitis, drug-related hepatoxicity, and obstructive jaundice. In hepatitis, albumin levels below 3 g/dL should raise the possibility of chronic liver disease. Hypoalbuminemia is more common in chronic liver disorders such as cirrhosis and usually reflects severe liver damage and decreased albumin synthesis. One exception is the patient with ascites in whom synthesis may be normal or even increased, but levels are low because of the increased volume of distribution. However, hypoalbuminemia is not specific for liver disease and may occur in protein malnutrition of any cause, as well as protein-losing enteropathies, nephrotic syndrome, and chronic infections that are associated with prolonged increases in serum interleukin-1 and/or tumor necrosis factor levels that inhibit albumin synthesis. Serum albumin should not be measured for screening in patients in whom there is no suspicion of liver disease. A general medical clinic study of consecutive patients in whom no indications were present for albumin measurement showed that while 12% of patients had abnormal test results, the finding was of clinical importance in only 0.4%. Serum Globulins Serum globulins are a group of proteins made up of gamma globulins (immunoglobulins) produced by B lymphocytes and alpha and beta globulins produced primarily in hepatocytes. Gamma globulins are increased in chronic liver disease, such as chronic hepatitis and cirrhosis. In cirrhosis, the increased serum gamma globulin concentration is due to the increased synthesis of antibodies, some of which are directed against intestinal bacteria. This occurs because the cirrhotic liver fails to clear bacterial antigens that normally reach the liver through the hepatic circulation. Increases in the concentration of specific isotypes of gamma globulins are often helpful in the recognition of certain chronic liver diseases. Diffuse polyclonal increases in IgG levels are common in autoimmune hepatitis; increases greater than 100% should alert the clinician to this possibility. Increases in the IgM levels are common in primary biliary cirrhosis, while increases in the IgA levels occur in alcoholic liver disease. Coagulation Factors With the exception of factor VIII, the blood clotting factors are made exclusively in hepatocytes. Their serum half-lives are much shorter than albumin,
ranging from 6 hours for factor VII to 5 days for fibrinogen. Because of their rapid turnover, measurement of the clotting factors is the single best acute measure of hepatic synthetic function and helpful in both the diagnosis and assessing the prognosis of acute parenchymal liver disease. Useful for this purpose is the serum prothrombin time, which collectively measures factors II, V, VII, and X. Biosynthesis of factors II, VII, IX, and X depends on vitamin K. The prothrombin time may be elevated in hepatitis and cirrhosis as well as in disorders that lead to vitamin K deficiency such as obstructive jaundice or fat malabsorption of any kind. Marked prolongation of the prothrombin time,>5 s above control and not corrected by parenteral vitamin K administration, is a poor prognostic sign in acute viral hepatitis and other acute and chronic liver diseases. OTHER DIAGNOSTIC TESTS While tests may direct the physician to a category of liver disease, additional radiologic testing and procedures are often necessary to make the proper diagnosis, as shown inFig. 293-1. The two most commonly-used ancillary tests are reviewed here. Percutaneous Liver Biopsy Percutaneous biopsy of the liver is a safe procedure that can be easily performed at the bedside with local anesthesia. Liver biopsy is of proven value in the following situations: (1) hepatocellular disease of uncertain cause; (2) prolonged hepatitis with the possibility of chronic active hepatitis; (3) unexplained hepatomegaly; (4) unexplained splenomegaly; (5) hepatic filling defects by radiologic imaging; (6) fever of unknown origin; (7) staging of malignant lymphoma. Liver biopsy is most accurate in disorders causing diffuse changes throughout the liver and is subject to sampling error in focal infiltrative disorders such as hepatic metastases. Liver biopsy should not be the initial procedure in the diagnosis of cholestasis. The biliary tree should first be assessed for signs of obstruction. Ultrasonography Ultrasonography is the first diagnostic test to use in patients whose liver tests suggest cholestasis, to look for the presence of a dilated intrahepatic or extrahepatic biliary tree, or to identify gallstones. In addition, it shows space-occupying lesions within the liver, enables the clinician to distinguish between cystic and solid masses, and helps direct percutaneous biopsies. Ultrasound with Doppler imaging can detect the patency of the portal vein, hepatic artery, and hepatic veins and determine the direction of blood flow. This is the first test ordered in patients suspected of having Budd-Chiari syndrome. USE OF LIVER TESTS As previously noted, the best way to increase the sensitivity and specificity of laboratory tests in the detection of liver disease is to employ a battery of tests that include the aminotransferases, alkaline phosphatase, bilirubin, albumin, and prothrombin time along with the judicious use of the other tests described in this chapter.Table 293-1 shows how patterns of liver tests can lead the clinician to a category of disease which will direct further evaluation. However, it is important to remember that no single set of liver tests will necessarily provide a diagnosis. It is often necessary to repeat these tests on several occasions over days to weeks for a diagnostic pattern to emerge.Figure 293-1 is an algorithm for the evaluation of chronically abnormal liver tests.
(Bibliography omitted in Palm version) Back to Table of Contents
294. BILIRUBIN METABOLISM AND THE HYPERBILIRUBINEMIAS - Paul D. Berk, Allan W. Wolkoff BILIRUBIN METABOLISM SOURCES OF BILIRUBIN Bilirubin is the end-product of the metabolic degradation of heme, the prosthetic group of hemoglobin, myoglobin, the cytochrome P450s, and various other hemoproteins. The first step in the conversion of heme to bilirubin is the stereospecific oxidative opening of the heme molecule at its a-bridge carbon by the microsomal enzyme heme oxygenase, resulting in the formation of equimolar quantities of carbon monoxide and of the green tetrapyrrole biliverdin. Biliverdin is then reduced by a second enzyme, biliverdin reductase, to bilirubin. Between 70 and 90% of bilirubin is derived from degradation of the hemoglobin of senescent or injured circulating red blood cells. The remainder has several sources, including hemoglobin produced during the process of ineffective erythropoiesis within the bone marrow and the turnover of nonhemoglobin hemoproteins in cells throughout the body. Degradation of red-cell hemoglobin occurs principally in the spleen but also throughout the rest of the peripheral reticuloendothelial system, including the Kupffer cells within the liver. Bilirubin produced in the periphery is transported to the liver within the plasma, where, due to its insolubility in aqueous solutions, it is tightly bound to albumin. The anatomy of the hepatic acinus is highly specialized to facilitate the extraction of such tightly protein-bound compounds (Fig. 294-1). Cuboidal hepatocytes within the hepatic cell plates are immediately adjacent to sinusoids on two surfaces. The endothelial cells of the sinusoids are fenestrated, allowing ready exchange of plasma between the sinusoidal blood and the extracellular space of Disse and affording direct access of the bilirubin-albumin complex to the surface of the hepatocyte, which is greatly expanded by the elaboration of microvilli. HEPATIC DISPOSITION OF BILIRUBIN Since bilirubin is a potentially toxic waste product, hepatic handling is designed to eliminate it from the body via the biliary tract. Transfer of bilirubin from blood to bile involves four distinct but interrelated steps, described below (Fig. 294-1). Hepatocellular Uptake Bilirubin most likely enters the hepatocyte both by a facilitated transport mechanism and by passive diffusion. While kinetic data suggest that facilitated transport is the predominant process and several putative bilirubin transporters have been identified, none has been cloned successfully. Cloned transporters such as organic anion transport protein (OATP) and sodium taurocholate co-transporting polypeptide (NTCP), which are responsible for the hepatocellular uptake of other organic substrates, including sulfobromophthalein and bile acids, specifically do not transport bilirubin. Therefore, the precise mechanism of bilirubin uptake remains to be determined. Intracellular Binding Having crossed the plasma membrane to enter the cell, bilirubin partitions between the lipid environment of intracellular membranes and the aqueous
cytosol, in which it is kept in solution by binding as a nonsubstrate ligand to several of the glutathione-S-transferases, formerly called ligandins. Conjugation The aqueous insolubility of bilirubin reflects a rigid, highly ordered molecular structure in which internal hydrogen bonding involving the propionic acid carboxyl groups of one dipyrrolic half of the molecule and the imino and lactam groups of the opposite half blocks solvent access to these polar residues. When the carboxyl groups are esterified by conjugation with glucuronic acid residues, the internal hydrogen bonding is disrupted, rendering the resulting mono- and diglucuronide conjugates highly soluble in aqueous solution. Bilirubin glucuronidation is catalyzed by a specific UDP-glucuronosyltransferase. The UDP-glucuronosyltransferases have been classified into gene families based on the degree of homology between the various protein isoforms. Those that conjugate bilirubin and certain other substrates have been designated the UGT1 family and have been shown to be expressed from a single gene complex by alternative splicing. This gene complex contains multiple substrate-specific first exons, designated A1, A2, . . . (Fig. 294-2), each with its own promoter and each encoding the amino-terminal end of a specific isoform, as well as four common exons (exons 2 to 5) that encode the shared carboxyl-terminal end of all of the UGT1 isoforms. The various first exons encode the specific substrate-binding sites for each isoform, while the shared exons encode common glycosylation, UDP-glucuronic acid-binding, transmembrane, and stop transfer domains. Exon A1 and the four common exons, collectively designated the UGT1A1 gene (Fig. 294-2), encode the physiologically critical enzyme bilirubin-UDP-glucuronosyltransferase (UGT1A1). A critical corollary of the organization of the UGT1 gene is that a mutation in one of the first exons will affect only a single enzyme isoform. By contrast, a mutation in exons 2 to 5 will alter all isoforms encoded by the UGT1 gene complex. Biliary Excretion Normal bile typically contains less than 5% unconjugated bilirubin, an average of 7% bilirubin monoconjugates, and 90% bilirubin diconjugates. The proportion of monoconjugates increases in the presence of an increased bilirubin load (hemolysis) or a reduced bilirubin-conjugating capacity. Bilirubin mono- and diglucuronides are excreted across the canalicular plasma membrane into the canaliculus by an ATP-dependent transport process mediated by a canalicular membrane protein called multidrug resistance-associated protein 2 (MRP2). MRP2 is a member of the MRP gene family, other members of which pump certain types of drug conjugates, as well as unmodified anticancer drugs, out of cells. It is also a member of the ATP-binding cassette (ABC) superfamily. Mutations in the rat homologue of MRP2 result in conjugated hyperbilirubinemia in several jaundiced strains that serve as models of the Dubin-Johnson syndrome. It has recently been established that the Dubin-Johnson syndrome in humans also results from mutations in MRP2 (see below). BILIRUBIN IN PLASMA Although physicians equate the direct-reacting fraction of bilirubin in plasma with conjugated bilirubin and the indirect fraction with unconjugated bilirubin, modern analytical methods document that normal plasma contains virtually no bilirubin conjugates. The 10 to 20% of bilirubin in normal plasma that gives a prompt (direct)
diazo reaction is an artifact of the kinetics of the van den Bergh reaction, which, along with various modifications, is the method most commonly used to quantitate bilirubin in clinical laboratories. Indeed, when the direct-reacting fraction is less than 15% of total bilirubin at virtually any total bilirubin concentration, the bilirubin in the sample can be considered as essentially all unconjugated. The canalicular transport mechanism for excretion of bilirubin conjugates is very sensitive to injury. Accordingly, in hepatocellular disease, as well as with either cholestasis or mechanical obstruction to the bile ducts, bilirubin conjugates within the hepatocyte, prevented from taking their normal path into the canaliculi and down the bile ducts, may reflux into the bloodstream, resulting in a mixed or, less often, a truly conjugated hyperbilirubinemia. EXTRAHEPATIC ASPECTS OF BILIRUBIN DISPOSITION Bilirubin in the Gut Following secretion into bile, conjugated bilirubin reaches the duodenum and passes down the gastrointestinal tract without reabsorption by the intestinal mucosa. Although some reaches the feces unaltered, an appreciable fraction is converted to urobilinogen and related compounds by bacterial metabolism within the ileum and colon. Urobilinogen is reabsorbed from these sites, reaches the liver via the portal circulation, and is reexcreted into bile, undergoing an enterohepatic circulation. Urobilinogen not taken up by the liver reaches the systemic circulation, from which some is cleared by the kidneys. Urinary urobilinogen excretion normally does not exceed 4 mg/d. In the presence of hemolysis, which increases the amount of bilirubin entering the gut (and hence the amount of urobilinogen formed and reabsorbed), or in the presence of hepatic disease, which decreases hepatic extraction of urobilinogen, plasma urobilinogen levels rise, as does the amount excreted in the urine. Severe cholestasis, bile duct obstruction, or administration of broad-spectrum antibiotics that eliminate the enteric flora required for the conversion of bilirubin to urobilinogens, markedly decrease formation of urobilinogen and its urinary excretion. Unconjugated bilirubin ordinarily does not reach the gut except in neonates or, by ill-defined alternative pathways, in the presence of severe unconjugated hyperbilirubinemia (e.g., Crigler-Najjar syndrome type I). In these circumstances, however, unconjugated bilirubin is readily reabsorbed from the gut lumen, amplifying the underlying hyperbilirubinemia. Renal Excretion of Bilirubin Conjugates Unconjugated bilirubin is not excreted in urine no matter how high its plasma concentration, since it is too tightly bound to albumin for effective glomerular filtration and there is no tubular mechanism for its renal secretion. By contrast, the polar bilirubin conjugates are far less tightly bound to albumin and are readily filtered at the glomerulus. Bilirubin conjugates are not secreted by the renal tubules but may be minimally reabsorbed. Since normal plasma contains virtually exclusively unconjugated bilirubin, no bilirubin normally appears in the urine. Indeed, bilirubinuria indicates the presence of conjugated bilirubin in plasma and, therefore, hepatobiliary dysfunction. CLINICAL PHYSIOLOGY The plasma concentration of unconjugated bilirubin ([Br]) is determined by the rate at which newly synthesized bilirubin enters the plasma (plasma bilirubin turnover, BrT) and
hepatic bilirubin clearance (CBr), according to the following relationship:
where k is a constant related to the different units of time employed in the conventional expression of BrTand CBr. BrT closely reflects total bilirubin production; CBr, analogous to the creatinine clearance test widely used to assess kidney function, is a measure of the rate at which bilirubin is extracted from plasma and is a true quantitative test of liver function. While not easily quantified in routine clinical settings, investigative measurements of BrT and CBr have yielded useful pathophysiologic insights into the unconjugated hyperbilirubinemias. Equation (1) indicates that the unconjugated bilirubin concentration will increase in the presence of either an increase inBrT or a reduction in hepatic CBr. This equation therefore provides a basis for classifying unconjugated hyperbilirubinemias according to pathogenesis. Furthermore, for an individual with a given value for CBr, or for a population in which CBrvaries within a narrow range, [Br] will increase as a linear function of BrT, with a slope relating increases in the plasma [Br] to increased BrT equal to k/CBr. For individuals or populations with reduced bilirubin clearance (e.g., in Gilbert's syndrome, see below), this slope will be steeper than in normal individuals. Conversely, for a given BrT, the relationship between [Br] and CBr is hyperbolic, like the relation between serum creatinine concentration and creatinine clearance. In any patient, if CBr is reduced from its baseline value, [Br] will be increased in consequence, in direct proportion to the extent of the decrease in CBr. DISORDERS OF BILIRUBIN METABOLISM LEADING TO UNCONJUGATED HYPERBILIRUBINEMIA INCREASED BILIRUBIN PRODUCTION Hemolysis Increased destruction of erythrocytes leads to increased bilirubin turnover and unconjugated hyperbilirubinemia. With normal liver function, the hyperbilirubinemia is usually modest. In particular, since the bone marrow is only capable of a sustained eightfold increase in erythrocyte production in response to a hemolytic stress, hemolysis alone cannot result in a sustained hyperbilirubinemia of more than approximately 68 umol/L (4 mg/dL). Higher values imply concomitant hepatic dysfunction. The causes of hemolysis are numerous. Besides specific hemolytic disorders, mild hemolytic processes accompany many acquired systemic diseases. When hemolysis is the only abnormality in an otherwise healthy individual, the result is a purely unconjugated hyperbilirubinemia, with the direct-reacting fraction as measured in a typical clinical laboratory being £15% of the total serum bilirubin. In the presence of systemic disease, which may include a degree of hepatic dysfunction, hemolysis may produce a component of conjugated hyperbilirubinemia in addition to an elevated unconjugated bilirubin concentration. Prolonged hemolysis may lead to the precipitation of bilirubin salts within the gall bladder or biliary tree, resulting in the formation of gallstones in which bilirubin, rather than cholesterol, is the major component. Such pigment stones may lead to acute or
chronic cholecystitis, biliary obstruction, or any other biliary tract consequence of calculous disease. Ineffective Erythropoiesis During erythroid maturation, small amounts of hemoglobin may be lost during nuclear extrusion, and a fraction of developing erythroid cells is destroyed within the marrow. These processes normally account for 10 to 15% of bilirubin produced. In various disorders, including thalassemia major, frankly megaloblastic anemias due to folate or vitamin B12deficiency, congenital erythropoietic porphyria, lead poisoning, and various congenital and acquired dyserythropoietic anemias, the fraction of total bilirubin production derived from ineffective erythropoiesis is increased, reaching as much as 70% of the total, and may be sufficient to produce modest degrees of unconjugated hyperbilirubinemia. Miscellaneous Degradation of the hemoglobin of extravascular collections of erythrocytes, such as those seen in massive tissue infarctions or large hematomas, may lead transiently to unconjugated hyperbilirubinemia. DECREASED HEPATIC BILIRUBIN CLEARANCE Decreased Hepatic Uptake As noted above, the mechanisms by which bilirubin enters hepatocytes are not fully defined but probably include both diffusion and facilitated transport. Decreased hepatic bilirubin uptake is believed to contribute to the unconjugated hyperbilirubinemia of Gilbert's syndrome (GS), although the molecular basis for this finding remains unclear (see below). Several drugs, including flavispidic acid, novobiocin, and various cholecystographic contrast agents, have been reported to inhibit bilirubin uptake. The resulting unconjugated hyperbilirubinemia resolves with cessation of the medication. Impaired Conjugation Physiologic Neonatal Jaundice Bilirubin produced by the fetus is cleared by the placenta and eliminated by the maternal liver. Consequently, bilirubin concentrations in normal neonates at birth are low. The presence of jaundice at birth is pathologic and requires investigation. Immediately after birth, the neonatal liver must assume responsibility for bilirubin clearance and excretion. However, many aspects of hepatic physiology are incompletely developed at birth. Levels of UGT1A1 are low, and alternative pathways allow passage of unconjugated bilirubin into the gut. Since the intestinal flora that converts bilirubin to urobilinogen is also undeveloped, an enterohepatic circulation of unconjugated bilirubin ensues. In consequence, most neonates develop mild unconjugated hyperbilirubinemia between days 2 and 5 after birth. Peak levels are typically less than 85 to 170 umol/L (5 to 10 mg/dL) and decline to normal adult concentrations within 2 weeks, as mechanisms required for bilirubin disposition mature. Prematurity, with more profound immaturity of hepatic function, or hemolysis, such as occurs with erythroblastosis fetalis, results in higher levels of unconjugated hyperbilirubinemia. A rapidly rising unconjugated bilirubin concentration, or absolute levels in excess of 340 umol/L (20 mg/dL), puts the infant at risk for bilirubin encephalopathy, or kernicterus, in which bilirubin crosses an immature blood-brain barrier and precipitates in the basal ganglia and other areas of the brain. The
consequences range from appreciable neurologic deficits to death. Principal treatment options include phototherapy, which converts bilirubin into photoisomers that are soluble in aqueous media and readily excretable in bile without conjugation, and exchange transfusion. The canalicular mechanisms responsible for bilirubin excretion are also immature at birth, and their maturation may, on occasion, lag behind that of UGT1A1. This may lead to transient conjugated neonatal hyperbilirubinemia, especially in infants with hemolysis. Acquired Conjugation Defects A modest reduction in bilirubin-conjugating capacity may be observed in advanced hepatitis or cirrhosis. However, in this setting, conjugation is better preserved than other aspects of bilirubin disposition, such as canalicular excretion. Various drugs, including pregnanediol, novobiocin, chloramphenicol, and gentamicin, may produce unconjugated hyperbilirubinemia by inhibiting UGT1A1 activity. Finally, certain fatty acids and the progestational steroid 3a,20b-pregnanediol, identified in the breast milk but not the serum of mothers whose infants have excessive neonatal hyperbilirubinemia (breast milk jaundice), inhibit bilirubin conjugation. The pathogenesis of breast milk jaundice appears to differ from that of transient familial neonatal hyperbilirubinemia (Lucey-Driscoll syndrome), in which a UGT1A1 inhibitor is found in maternal serum. HEREDITARY DEFECTS IN BILIRUBIN CONJUGATION Three familial disorders characterized by differing degrees of unconjugated hyperbilirubinemia have long been recognized. The defining clinical features of each are described below (Table 294-1). While these disorders have been recognized for decades to reflect differing degrees of deficiency in the ability to conjugate bilirubin, recent advances in the molecular biology of the UGT1 gene complex have elucidated their interrelationships and clarified previously puzzling features. Crigler-Najjar Syndrome, Type I (CN-I) This disorder is characterized by striking unconjugated hyperbilirubinemia of about 340 to 765 umol/L (20 to 45 mg/dL) that appears in the neonatal period and persists for life. Other conventional hepatic biochemical tests such as serum aminotransferases and alkaline phosphatase are normal, and there is no evidence of hemolysis. Hepatic histology is also essentially normal except for the occasional presence of bile plugs within canaliculi. Bilirubin glucuronides are markedly reduced or absent from the nearly colorless bile, and there is no detectable constitutive expression of UGT1A1 activity in hepatic tissue. Neither UGT1A1 activity nor the serum bilirubin concentration responds to administration of phenobarbital or other enzyme inducers. In the absence of conjugation, unconjugated bilirubin accumulates in plasma, from which it is eliminated very slowly by alternative pathways that include direct passage into the bile and small intestine. These account for the small amounts of urobilinogen found in feces. No bilirubin is found in the urine. First described in 1952, the disorder is rare (estimated prevalence of 0.6 to 1.0 per million). Many patients are from geographically or socially isolated communities in which consanguinity is common, and pedigree analyses suggest an autosomal recessive
pattern of inheritance. The majority of patients (type IA) exhibit defects in the glucuronide conjugation of a spectrum of substrates in addition to bilirubin, including various drugs and other xenobiotics. These individuals have mutations in one of the common exons (2 to 5) of the UGT1 gene (Fig. 294-2). In a smaller subset (type IB), the defect is limited largely to bilirubin conjugation, and the causative mutation is in the bilirubin-specific exon A1. More than 30 different UGT1A1 mutations responsible forCN-Ihave been identified, including deletions, frameshifts, alterations in intronic splice donor and acceptor sites, and point mutations that introduce premature stop codons or alter critical aminoacids. Their common feature is that they all encode proteins with absent or, at most, traces of bilirubin-UDP-glucuronosyltransferase enzymatic activity. Prior to the availability of phototherapy, most patients withCN-Idied of bilirubin encephalopathy (kernicterus) in infancy or early childhood. A few lived as long as early adult life without overt neurologic damage, although more subtle testing usually indicated mild but progressive brain damage. In all such cases, in the absence of liver transplantation, death eventually supervened from late-onset bilirubin encephalopathy, which often followed a nonspecific febrile illness. Recent data suggest that the best hope for survival of a neurologically intact patient involves the following regimen: (1) about 12 h/d of phototherapy from birth throughout childhood, perhaps supplemented by exchange transfusion in the immediate neonatal period; (2) use of tin-protoporphyrin to blunt transient episodes of increased hyperbilirubinemia; and (3) early liver transplantation, prior to the onset of brain damage. In a single patient, transplantation with isolated allogeneic hepatocytes produced a clinically significant reduction in serum bilirubin concentration. Crigler-Najjar Syndrome, Type II (CN-II) Characterized by marked unconjugated hyperbilirubinemia in the absence of abnormalities of other conventional hepatic biochemical tests, hepatic histology, or hemolysis, this condition was recognized as a distinct entity in 1962. It differs fromCN-I in several specific ways (Table 294-1). (1) Although there is considerable overlap, average bilirubin concentrations are lower in CN-II; (2) accordingly, CN-II is only infrequently associated with kernicterus; (3) bile is deeply colored and bilirubin glucuronides are present, with a striking, characteristic increase in monoglucuronides; (4) UGT1A1 in liver is usually present at reduced levels (typically £10% of normal) but may be undetectable by less sensitive older assays; (5) while typically detected in infancy, hyperbilirubinemia was not recognized in some cases until later in life, and in one instance, until age 34. As with CN-I, most CN-II cases exhibit abnormalities in the conjugation of other compounds, such as salicylamide and menthol, but in some instances the defect appears limited to bilirubin. Reduction of serum bilirubin concentrations by more than 25% in response to enzyme inducers such as phenobarbital distinguishesCN-IIfromCN-I, although this response may not be elicited in early infancy and often is not accompanied by measurable UGT1A1 induction. Bilirubin concentrations during phenobarbital administration do not return to normal but are typically in the range of 51 to 86 umol/L (3 to 5 mg/dL). Although the incidence of kernicterus in CN-II is low, instances have occurred, not only in infants but in adolescents and adults, often in the setting of an intercurrent illness, fasting, or any other factor that temporarily raises the serum bilirubin concentration above baseline. For this reason, phenobarbital therapy is widely recommended, a single bedtime dose often sufficing to maintain clinically safe plasma bilirubin concentrations.
At least 10 different mutations of UGT1 associated withCN-IIhave been identified. Their common feature is that they encode for a bilirubin-UDP-glucuronosyltransferase with markedly reduced but detectable enzymatic activity. The spectrum of residual enzyme activity explains the spectrum of phenotypic severity of the resulting hyperbilirubinemia. Molecular analysis has established that a large majority of CN-II patients are either homozygotes or compound heterozygotes for CN-II mutations and that individuals carrying one mutated and one entirely normal allele have normal bilirubin concentrations. Possible inheritance in one case as a dominant negative mutation remains to be confirmed. Gilbert's Syndrome This syndrome is characterized by mild unconjugated hyperbilirubinemia, normal values for standard hepatic biochemical tests, and normal hepatic histology other than a modest increase of lipofuscin pigment in some patients. Serum bilirubin concentrations are most often95% of cases. Nonspecificity can confound immunoassays for anti-HCV, especially in persons with a low prior probability of infection, such as volunteer blood donors, or in persons with circulating rheumatoid factor, which can bind nonspecifically to assay reagents. A supplementary recombinant immunoblot assay (RIBA), in which serum is incubated with a nitrocellulose strip containing viral protein bands, can be used to establish the specific viral proteins to which anti-HCV is directed (see "Virology and Etiology," above). Such RIBA determinations are used routinely to confirm anti-HCV reactivity in blood donors, but determinations of HCV RNA have supplanted RIBA in many clinical settings. Assays for HCV RNA are the most sensitive tests for HCV infection and represent the "gold standard" in establishing a diagnosis of hepatitis C. HCV RNA can be detected even before acute elevation of aminotransferase activity and
before the appearance of anti-HCV in patients with acute hepatitis C. In addition, HCV RNA remains detectable indefinitely, continuously in most but intermittently in some, in patients with chronic hepatitis C (even detectable in some persons with normal liver tests, i.e., asymptomatic carriers). In the small minority of patients with hepatitis C who lack anti-HCV, a diagnosis can be supported by detection of HCV RNA. If all these tests are negative and the patient has a well-characterized case of hepatitis after percutaneous exposure to blood or blood products, a diagnosis of hepatitis caused by another agent, as yet unidentified, can be entertained. Amplification techniques are required to detectHCV RNA, and two are available. One is a branched-chain complementary DNA (bDNA) assay, in which the detection signal (a colorimetrically detectable enzyme bound to a complementary DNA probe) is amplified. The other is aPCRassay, in which the viral RNA is reverse transcribed to complementary DNA and then amplified by repeated cycles of DNA synthesis and polymerization. Both can be used as quantitative assays and a measurement of relative "viral load"; PCR, with a sensitivity of 102 to 103virions per milliliter is more sensitive than bDNA, with a sensitivity of 2´105. Determination of viral load is not a reliable marker of disease severity or prognosis but is helpful in predicting relative responsiveness to antiviral therapy. The same is true for determinations of HCV genotype (Chap. 297). A proportion of patients with hepatitis C have isolated anti-HBc in their blood, a reflection of a common risk in certain populations to multiple bloodborne hepatitis agents. The anti-HBc in such cases is almost invariably of the IgG class and usually representsHBVinfection in the remote past, rarely current HBV infection with low-level virus carriage. The presence ofHDVinfection can be identified by demonstrating intrahepatic HDV antigen or, more practically, an anti-HDV seroconversion (a rise in titer of anti-HDV or de novo appearance of anti-HDV). Circulating HDV antigen, also diagnostic of acute infection, is detectable only briefly, if at all. Because anti-HDV is often undetectable onceHBsAgdisappears, retrospective serodiagnosis of acute self-limited, simultaneousHBV and HDV infection is difficult. Early diagnosis of acute infection may be hampered by a delay of up to 30 to 40 days in the appearance of anti-HDV. When a patient presents with acute hepatitis and hasHBsAgand anti-HDVin serum, determination of the class of anti-HBc is helpful in establishing the relationship between infection withHBV and HDV. Although IgM anti-HBc does not distinguish absolutely between acute and chronic HBV infection, its presence is a reliable indicator of recent infection and its absence a reliable indicator of infection in the remote past. In simultaneous acute HBV and HDV infections, IgM anti-HBc will be detectable, while in acute HDV infection superimposed on chronic HBV infection, anti-HBc will be of the IgG class. Tests for the presence ofHDV RNA are useful for determining the presence of ongoing HDV replication and relative infectivity. Currently, probes for this marker are restricted to a limited number of research laboratories. Similarly, diagnostic tests for hepatitis E are confined to a small number of research laboratories. Liver biopsy is rarely necessary or indicated in acute viral hepatitis, except when there is
a question about the diagnosis or when there is clinical evidence suggesting a diagnosis of chronic hepatitis. A diagnostic algorithm can be applied in the evaluation of cases of acute viral hepatitis. A patient with acute hepatitis should undergo four serologic tests,HBsAg, IgM anti-HAV, IgM anti-HBc, and anti-HCV(Table 295-4). The presence of HBsAg, with or without IgM anti-HBc, represents HBV infection. If IgM anti-HBc is present, the HBV infection is considered acute; if IgM anti-HBc is absent, theHBVinfection is considered chronic. A diagnosis of acute hepatitis B can be made in the absence of HBsAg when IgM anti-HBc is detectable. A diagnosis of acute hepatitis A is based on the presence of IgM anti-HAV. If IgM anti-HAV coexists with HBsAg, a diagnosis of simultaneous HAV and HBV infections can be made; if IgM anti-HBc (with or without HBsAg) is detectable, the patient has simultaneous acute hepatitis A and B, and if IgM anti-HBc is undetectable, the patient has acute hepatitis A superimposed on chronic HBV infection. The presence of anti-HCV, if confirmable, supports a diagnosis of acute hepatitis C. Occasionally, testing for HCV RNA or repeat anti-HCV testing later during the illness is necessary to establish the diagnosis. Absence of all serologic markers is consistent with a diagnosis of "non-A, non-B, non-C" hepatitis, if the epidemiologic setting is appropriate. In patients with chronic hepatitis, initial testing should consist ofHBsAgand anti-HCV. Anti-HCV supports and HCV RNA testing establishes the diagnosis of chronic hepatitis C. If a serologic diagnosis of chronic hepatitis B is made, testing forHBeAgand anti-HBe is indicated to evaluate relative infectivity. Testing forHBV DNA in such patients provides a more quantitative and sensitive measure of the level of virus replication and, therefore, is very helpful during antiviral therapy (Chap. 297). In patients with hepatitis B, testing for anti-HDVis useful under the following circumstances: patients with severe and fulminant diseases, patients with severe chronic disease, patients with chronic hepatitis B who have acute hepatitis-like exacerbations, persons with frequent percutaneous exposures, and persons from areas where HDV infection is endemic. PROGNOSIS Virtually all previously healthy patients with hepatitis A recover completely from their illness with no clinical sequelae. Similarly, in acute hepatitis B, 95 to 99% of previously healthy adults have a favorable course and recover completely. There are, however, certain clinical and laboratory features that suggest a more complicated and protracted course. Patients of advanced age and with serious underlying medical disorders may have a prolonged course and are more likely to experience severe hepatitis. Initial presenting features such as ascites, peripheral edema, and symptoms of hepatic encephalopathy suggest a poorer prognosis. In addition, a prolongedPT, low serum albumin level, hypoglycemia, and very high serum bilirubin values suggest severe hepatocellular disease. Patients with these clinical and laboratory features deserve prompt hospital admission. The case-fatality rate in hepatitis A and B is very low (approximately 0.1%) but is increased by advanced age and underlying debilitating disorders. Among patients ill enough to be hospitalized for acute hepatitis B, the fatality rate is 1%. Hepatitis C occurring after transfusion is less severe during the acute phase than hepatitis B and is more likely to be anicteric; fatalities are rare, but the precise case-fatality rate is not known. In outbreaks of waterborne hepatitis E in India and Asia, the case-fatality rate is 1 to 2% and up to 10 to 20% in pregnant women. Patients with
simultaneous acute hepatitis B and hepatitis D do not necessarily experience a higher mortality rate than do patients with acute hepatitis B alone; however, in several recent outbreaks of acute simultaneousHBV andHDVinfection among injection drug users, the case-fatality rate has been approximately 5%. In the case of HDV superinfection of a person with chronic hepatitis B, the likelihood of fulminant hepatitis and death is increased substantially. Although the case-fatality rate for hepatitis D has not been defined adequately, in outbreaks of severe HDV superinfection in isolated populations with a high hepatitis B carrier rate, the mortality rate has been recorded in excess of 20%. COMPLICATIONS AND SEQUELAE A small proportion of patients with hepatitis A experience relapsing hepatitis weeks to months after apparent recovery from acute hepatitis. Relapses are characterized by recurrence of symptoms, aminotransferase elevations, occasionally jaundice, and fecal excretion ofHAV. Another unusual variant of acute hepatitis A is cholestatic hepatitis, characterized by protracted cholestatic jaundice and pruritus. Rarely, liver test abnormalities persist for many months, even up to a year. Even when these complications occur, hepatitis A remains self-limited and does not progress to chronic liver disease. During the prodromal phase of acute hepatitis B, a serum sickness-like syndrome characterized by arthralgia or arthritis, rash, angioedema, and rarely hematuria and proteinuria may develop in 5 to 10% of patients. This syndrome occurs before the onset of clinical jaundice, and these patients are often erroneously diagnosed as having rheumatologic diseases. The diagnosis can be established by measuring serum aminotransferase levels, which are almost invariably elevated, and serumHBsAg. As noted above,EMC is an immune-complex disease that can complicate hepatitis C. Attention has been drawn as well to associations between hepatitis C and such cutaneous disorders as porphyria cutanea tarda and lichen planus. A mechanism for these associations is unknown. The most feared complication of viral hepatitis is fulminant hepatitis (massive hepatic necrosis); fortunately, this is a rare event. Fulminant hepatitis is primarily seen in hepatitis B and D, as well as hepatitis E, but rare fulminant cases of hepatitis A occur primarily in older adults and in persons with underlying chronic liver disease. Hepatitis B accounts for more than 50% of fulminant hepatitis cases, a sizable proportion of which are associated withHDVinfection. Participation of HDV can be documented in approximately one-third of patients with acute fulminant hepatitis B and two-thirds of patients with fulminant hepatitis superimposed on chronic hepatitis B. Fulminant hepatitis is seen rarely in hepatitis C, but hepatitis E, as noted above, can be complicated by fatal fulminant hepatitis in 1 to 2% of all cases and in up to 20% of cases occurring in pregnant women. Patients usually present with signs and symptoms of encephalopathy that may evolve to deep coma. The liver is usually small and thePTexcessively prolonged. The combination of rapidly shrinking liver size, rapidly rising bilirubin level, and marked prolongation of the PT, even as aminotransferase levels fall, together with clinical signs of confusion, disorientation, somnolence, ascites, and edema, indicates that the patient has hepatic failure with encephalopathy. Cerebral edema is common; brainstem compression, gastrointestinal bleeding, sepsis, respiratory failure, cardiovascular collapse, and renal failure are terminal events. The mortality rate is exceedingly high (greater than 80% in patients with deep coma), but
patients who survive may have a complete biochemical and histologic recovery. If a donor liver can be located in time, liver transplantation may be life-saving in patients with fulminant hepatitis. It is particularly important to document the disappearance ofHBsAgafter apparent clinical recovery from acute hepatitis B. Before laboratory methods were available to distinguish between acute hepatitis and acute hepatitis-like exacerbations (spontaneous reactivations) of chronic hepatitis B, observations suggested that approximately 10% of patients remained HBsAg-positive for longer than 6 months after the onset of clinically apparent acute hepatitis B. Half these persons cleared the antigen from their circulations during the next several years, but the other 5% remained chronically HBsAg-positive. More recent observations suggest that the true rate of chronic infection after clinically apparent acute hepatitis B is as low as 1% in normal, immunocompetent, young adults. Earlier, higher estimates may have been biased by inadvertent inclusion of acute exacerbations in chronically infected patients; these patients, chronically HBsAg-positive before exacerbation, were unlikely to seroconvert to HBsAg-negative thereafter. Whether the rate of chronicity is 10 or 1%, such patients have anti-HBc in serum; anti-HBs is either undetected or detected at low titer against the opposite subtype specificity of the antigen (see "Laboratory Features," above). These patients may (1) be asymptomatic carriers, (2) have low-grade, mild chronic hepatitis, or (3) have moderate to severe chronic hepatitis with or without cirrhosis. The likelihood of becoming an HBsAg carrier after acuteHBVinfection is especially high among neonates, persons with Down's syndrome, chronically hemodialyzed patients, and immunosuppressed patients, including persons with HIV infection. Chronic hepatitis is an important late complication of acute hepatitis B occurring in a small proportion of patients with acute disease but more common in those who present with chronic infection without having experienced an acute illness (Chap. 297). Certain clinical and laboratory features suggest progression of acute hepatitis to chronic hepatitis: (1) lack of complete resolution of clinical symptoms of anorexia, weight loss, and fatigue and the persistence of hepatomegaly; (2) the presence of bridging or multilobular hepatic necrosis on liver biopsy during protracted, severe acute viral hepatitis; (3) failure of the serum aminotransferase, bilirubin, and globulin levels to return to normal within 6 to 12 months after the acute illness; and (4) the persistence ofHBeAgbeyond 3 months orHBsAgbeyond 6 months after acute hepatitis. Although acute hepatitis D infection does not increase the likelihood of chronicity of simultaneous acute hepatitis B, hepatitis D has the potential for contributing to the severity of chronic hepatitis B. Hepatitis D superinfection can transform asymptomatic or mild chronic hepatitis B into severe, progressive chronic hepatitis and cirrhosis; it also can accelerate the course of chronic hepatitis B. SomeHDVsuperinfections in patients with chronic hepatitis B lead to fulminant hepatitis. Although HDV andHBVinfections are associated with severe liver disease, mild hepatitis and even asymptomatic carriage have been identified in some patients, and the disease may become indolent beyond the early years of infection. After transfusion-associated acute hepatitis C, at least 50% of patients have abnormal biochemical liver tests for more than a year. In some experiences, the frequency of progression to chronicity after acute hepatitis C is as high as 70%. In most of these patients, liver histology is consistent with moderate to severe chronic hepatitis. Even among those who recover biochemically, the likelihood of
retaining circulatingHCV RNA is high. Thus, after acute HCV infection, the likelihood of remaining chronically infected approaches 85 to 90%. Although many patients with chronic hepatitis C have no symptoms, cirrhosis may develop in as many as 20% within 10 to 20 years of acute illness; in some series of cases, cirrhosis has been reported in as many as 50% of patients with chronic hepatitis C. Although chronic hepatitis C accounts for at least a quarter of cases of chronic liver disease and a quarter of patients undergoing liver transplantation for end-stage liver disease in the United States and Europe, in the majority of patients with chronic hepatitis C, morbidity and mortality are limited during the initial 20 years after the onset of infection. Progression of chronic hepatitis C may be influenced by hepatitis C genotype, age of acquisition, duration of infection, and immunosuppression, as well as by coexisting excessive alcohol use or other hepatitis virus infection. In contrast, neitherHAV norHEVcauses chronic liver disease. Rare complications of viral hepatitis include pancreatitis, myocarditis, atypical pneumonia, aplastic anemia, transverse myelitis, and peripheral neuropathy. Carriers ofHBsAg, particularly those infected in infancy or early childhood, have an enhanced risk of hepatocellular carcinoma. The risk of hepatocellular carcinoma is increased as well in patients with chronic hepatitis C, almost exclusively in patients with cirrhosis, and almost always after at least several decades, usually after three decades of disease (see Chap. 91). In children, hepatitis B may present rarely with anicteric hepatitis, a nonpruritic papular rash of the face, buttocks, and limbs, and lymphadenopathy (papular acrodermatitis of childhood or Gianotti-Crosti syndrome). DIFFERENTIAL DIAGNOSIS Viral diseases such as infectious mononucleosis; those due to cytomegalovirus, herpes simplex, and coxsackieviruses; and toxoplasmosis may share certain clinical features with viral hepatitis and cause elevations in serum aminotransferase and less commonly in serum bilirubin levels. Tests such as the differential heterophile and serologic tests for these agents may be helpful in the differential diagnosis ifHBsAg, anti-HBc, IgM anti-HAV, and anti-HCVdeterminations are negative. Aminotransferase elevations can accompany almost any systemic viral infection; other rare causes of liver injury confused with viral hepatitis are infections with Leptospira, Candida, Brucella, Mycobacteria, and Pneumocystis. A complete drug history is particularly important, for many drugs and certain anesthetic agents can produce a picture of either acute hepatitis or cholestasis (Chap. 296). Equally important is a past history of unexplained "repeated episodes" of acute hepatitis. This history should alert the physician to the possibility that the underlying disorder is chronic hepatitis. Alcoholic hepatitis also must be considered, but usually the serum aminotransferase levels are not as markedly elevated and other stigmata of alcoholism may be present. The finding on liver biopsy of fatty infiltration, a neutrophilic inflammatory reaction, and "alcoholic hyaline" would be consistent with alcohol-induced rather than viral liver injury. Because acute hepatitis may present with right upper quadrant abdominal pain, nausea and vomiting, fever, and icterus, it is often confused with acute cholecystitis, common duct stone, or ascending cholangitis. Patients with acute viral hepatitis may tolerate surgery poorly; therefore, it is important to exclude this diagnosis, and in confusing cases, a percutaneous liver biopsy may be necessary before laparotomy. Viral hepatitis in the elderly is often misdiagnosed as obstructive jaundice resulting from a common duct stone or carcinoma of the
pancreas. Because acute hepatitis in the elderly may be quite severe and the operative mortality high, a thorough evaluation including biochemical tests, radiographic studies of the biliary tree, and even liver biopsy may be necessary to exclude primary parenchymal liver disease. Another clinical constellation that may mimic acute hepatitis is right ventricular failure with passive hepatic congestion or hypoperfusion syndromes, such as those associated with shock, severe hypotension, and severe left ventricular failure. Also included in this general category is any disorder that interferes with venous return to the heart, such as right atrial myxoma, constrictive pericarditis, hepatic vein occlusion (Budd-Chiari syndrome), or venoocclusive disease. Clinical features are usually sufficient to distinguish between these vascular disorders and viral hepatitis. Acute fatty liver of pregnancy, cholestasis of pregnancy, eclampsia, and the HELLP syndrome (hemolysis, elevated liver tests, and low platelets) can be confused with viral hepatitis during pregnancy. Very rarely, malignancies metastatic to the liver can mimic acute or even fulminant viral hepatitis. Occasionally, genetic or metabolic liver disorders (e.g., Wilson's disease,a1-antitrypsin deficiency) are confused with viral hepatitis. TREATMENT Treatment of Acute Attack Although therapy has been developed for chronic hepatitis B and C (Chap. 297), opportunities for treating acute hepatitis caused byHBV orHCV are limited. In hepatitis B, among previously healthy adults who present with clinically apparent acute hepatitis, recovery occurs in approximately 99%; therefore, antiviral therapy is not likely to improve the rate of recovery and is not required. In rare instances of severe acute hepatitis B, treatment with a nucleoside analogue, such as lamivudine, at the 100-mg/d oral dose used to treat chronic hepatitis B (Chap. 297), has been attempted successfully. However, clinical trials have not been done to establish the efficacy of this approach, severe acute hepatitis B is not an approved indication for therapy, and the duration of therapy has not been determined. In typical cases of acute hepatitis C, recovery is rare, progression to chronic hepatitis is the rule, occurring in 85 to 90% of patients, and meta-analyses of small clinical trials suggest that antiviral therapy with interferon alpha (3 million units subcutaneously three times a week) is beneficial, reducing the rate of chronicity considerably by inducing sustained responses in 40% of patients. The duration of therapy and whether to add the nucleoside analogue ribavirin remain to be determined, but the most reasonable approach is to follow recommendations for treatment of chronic hepatitis C (Chap 297). Because of the marked reduction over the last two decades in the frequency of acute hepatitis C, opportunities to identify and treat patients with acute hepatitis C are rare indeed. Hospital epidemiologists, however, will encounter health workers who sustain hepatitis C-contaminated needle sticks; when monitoring forALTelevations and HCV RNA after these accidents identifies acute hepatitis C, therapy should be initiated. Notwithstanding these specific therapeutic considerations, in most cases of typical acute viral hepatitis, specific treatment generally is not necessary. Although hospitalization may be required for clinically severe illness, most patients do not require hospital care. Forced and prolonged bed rest is not essential for full recovery, but many patients will feel better with restricted physical activity. A high-calorie diet is desirable, and because many patients may experience nausea late in the day, the major caloric intake is best tolerated in the morning. Intravenous feeding is necessary in the acute stage if the patient has persistent vomiting and cannot maintain oral intake. Drugs capable of
producing adverse reactions such as cholestasis and drugs metabolized by the liver should be avoided. If severe pruritus is present, the use of the bile salt-sequestering resin cholestyramine will usually alleviate this symptom. Glucocorticoid therapy has no value in acute viral hepatitis. Even in severe cases associated with bridging necrosis, controlled trials have failed to demonstrate the efficacy of steroids. In fact, such therapy may be hazardous. Physical isolation of patients with hepatitis to a single room and bathroom is rarely necessary except in the case of fecal incontinence for hepatitis A and E or uncontrolled, voluminous bleeding for hepatitis B (with or without concomitant hepatitis D) and hepatitis C. Because most patients hospitalized with hepatitis A excrete little if anyHAV, the likelihood of HAV transmission from these patients during their hospitalization is low. Therefore, burdensome enteric precautions are no longer recommended. Although gloves should be worn when the bedpans or fecal material of patients with hepatitis A are handled, these precautions do not represent a departure from sensible procedure for all hospitalized patients. For patients with hepatitis B and hepatitis C, emphasis should be placed on blood precautions, i.e., avoiding direct, ungloved hand contact with blood and other body fluids. Enteric precautions are unnecessary. The importance of simple hygienic precautions, such as hand washing, cannot be overemphasized. Universal precautions that have been adopted for all patients apply to patients with viral hepatitis. Hospitalized patients may be discharged when there is substantial symptomatic improvement, a significant downward trend in the serum aminotransferase and bilirubin values, and a return to normal of thePT. Mild aminotransferase elevations should not be considered contraindications to the gradual resumption of normal activity. In fulminant hepatitis, the goal of therapy is to support the patient by maintenance of fluid balance, support of circulation and respiration, control of bleeding, correction of hypoglycemia, and treatment of other complications of the comatose state in anticipation of liver regeneration and repair. Protein intake should be restricted, and oral lactulose or neomycin administered. Glucocorticoid therapy has been shown in controlled trials to be ineffective. Likewise, exchange transfusion, plasmapheresis, human cross-circulation, porcine liver cross-perfusion, and hemoperfusion have not been proven to enhance survival. Meticulous intensive care is the one factor that does appear to improve survival. Orthotopic liver transplantation is resorted to with increasing frequency, with excellent results, in patients with fulminant hepatitis (Chap. 301). PROPHYLAXIS Because application of therapy for acute viral hepatitis is limited, and because antiviral therapy for chronic viral hepatitis is effective in only a proportion of patients (Chap. 297), emphasis is placed on prevention through immunization. The prophylactic approach differs for each of the types of viral hepatitis. In the past, immunoprophylaxis relied exclusively on passive immunization with antibody-containing globulin preparations purified by cold ethanol fractionation from the plasma of hundreds of normal donors. Currently, for hepatitis A and B, active immunization with vaccines is available as well. Hepatitis A Both passive immunization with immune globulin (IG) and active
immunization with a killed vaccine are available. All preparations of IG contain anti-HAVconcentrations sufficient to be protective. When administered before exposure or during the early incubation period, IG is effective in preventing clinically apparent hepatitis A. In some cases, IG does not abort infection but, by attenuating it, renders it inapparent. As a result, long-lasting "passive-active" immunity occurs; however, this is now considered to be the exception rather than the rule. For postexposure prophylaxis of intimate contacts (household, institutional) of persons with hepatitis A, the administration of 0.02 mL/kg is recommended as early after exposure as possible; it may be effective even when administered as late as 2 weeks after exposure. Prophylaxis is not necessary for casual contacts (office, factory, school, or hospital), for most elderly persons, who are very likely to be immune, or for those known to have anti-HAV in their serum. In day-care centers, recognition of hepatitis A in children or staff should provide a stimulus for immunoprophylaxis in the center and in the children's family members. By the time most common-source outbreaks of hepatitis A are recognized, it is usually too late in the incubation period for IG to be effective; however, prophylaxis may limit the frequency of secondary cases. For travelers to tropical countries, developing countries, and other areas outside standard tourist routes, IG prophylaxis had been recommended, before a vaccine became available. When such travel lasted less than 3 months, 0.02 mL/kg was given; for longer travel or residence in these areas, a dose of 0.06 mL/kg every 4 to 6 months was recommended. Administration of plasma-derived globulin is safe; it has not been associated with transmission of AIDS to recipients, and the AIDS virus (HIV) is inactivated by 25% alcohol, to which plasma is subjected during the cold ethanol fractionation process. Formalin-inactivated vaccines made from strains ofHAVattenuated in tissue culture have been shown to be safe, immunogenic, and effective in preventing hepatitis A. Hepatitis A vaccines are approved for use in persons who are at least 2 years old and appear to provide adequate protection 4 weeks after a primary inoculation. If it can be given within 4 weeks of an expected exposure, such as by travel to an endemic area, hepatitis A vaccine is the preferred approach to preexposure immunoprophylaxis. If travel is more imminent,IG (0.02 mL/kg) should be administered at a different injection site, along with the first dose of vaccine. Because vaccination provides long-lasting protection (protective levels of anti-HAV should last 20 years after vaccination), persons whose risk will be sustained (e.g., frequent travelers or those remaining in endemic areas for prolonged periods) should be vaccinated, and vaccine should supplant the need for repeated IG injections. Other groups who are candidates for hepatitis A vaccination include military personnel, populations with cyclic outbreaks of hepatitis A (e.g., Alaskan natives), employees of day-care centers, primate handlers, laboratory workers exposed to hepatitis A or fecal specimens, children in communities with a high frequency of hepatitis A, and patients with chronic liver disease. Because of an increased risk of fulminant hepatitis A -- observed in some experiences but not confirmed in others -- among patients with chronic hepatitis C, patients with chronic hepatitis C have been singled out as candidates for hepatitis A vaccination. Other populations whose recognized risk of hepatitis A is increased should be vaccinated, including men who have sex with men, injection drug users, and persons with clotting disorders who require frequent administration of clotting-factor concentrates. Recommendations for dose and frequency differ for the two approved vaccine preparations; all injections are intramuscular. For the hepatitis A vaccine manufactured by SmithKline Beecham (Havrix), adults (older than 18 years) should receive two
1.0-mL injections containing 1440 enzyme-linked immunoassay units (ELU) 6 to 12 months apart. Children age 2 to 18 years should receive three 0.5-mL injections containing 360 ELU at time zero, 6, and 12 months or two 0.5-mL injections containing 720 ELU 6 to 12 months apart. For the hepatitis A vaccine manufactured by Merck (Vaqta), adults (older than 17 years) should receive two 1.0-mL injections containing 50 units 6 months apart; children age 2 to 17 years should receive two 0.5-mL doses containing 25 units 6 to 18 months apart. Hepatitis A vaccine has been reported to be effective in preventing secondary household cases of acute hepatitis A, but its role in other instances of postexposure prophylaxis remains to be demonstrated. Hepatitis B Until 1982, prevention of hepatitis B was based on passive immunoprophylaxis either with standardIG, containing modest levels of anti-HBs, or hepatitis B immune globulin (HBIG), containing high-titer anti-HBs. The efficacy of standard IG has never been established and remains questionable; even the efficacy of HBIG, demonstrated in several clinical trials, has been challenged, and its contribution appears to be in reducing the frequency of clinical illness, not in preventing infection. The first vaccine for active immunization, introduced in 1982, was prepared from purified, noninfectious 22-nm spherical forms ofHBsAgderived from the plasma of healthy HBsAg carriers. In 1987, the plasma-derived vaccine was supplanted by a genetically engineered vaccine derived from recombinant yeast. The latter vaccine consists of HBsAg particles that are nonglycosylated but are otherwise indistinguishable from natural HBsAg; two recombinant vaccines are licensed for use in the United States. Current recommendations can be divided into those for preexposure and postexposure prophylaxis. For preexposure prophylaxis against hepatitis B in settings of frequent exposure (health workers exposed to blood, hemodialysis patients and staff, residents and staff of custodial institutions for the developmentally handicapped, injection drug users, inmates of long-term correctional facilities, promiscuous homosexual men as well as promiscuous heterosexual individuals, persons such as hemophiliacs who require long-term, high-volume therapy with blood derivatives, household and sexual contacts ofHBsAgcarriers, persons living in or traveling extensively in endemic areas, unvaccinated children under the age of 18, and unvaccinated children who are Alaskan natives, Pacific Islanders, or residents in households of first-generation immigrants from endemic countries), three intramuscular (deltoid, not gluteal) injections of hepatitis B vaccine are recommended at 0, 1, and 6 months. Pregnancy is not a contraindication to vaccination. In areas of lowHBVendemicity such as the United States, despite the availability of safe and effective hepatitis B vaccines, a strategy of vaccinating persons in high-risk groups has not been effective. The incidence of new hepatitis B cases continued to increase in the United States after introduction of vaccines; fewer than 10% of all targeted persons in high-risk groups have actually been vaccinated, and approximately 30% of persons with sporadic acute hepatitis B do not fall into any high-risk-group category. Therefore, to have an impact on the frequency of HBV infection in an area of low endemicity such as the United States, universal hepatitis B vaccination in childhood has been recommended. For unvaccinated children born after the implementation of universal infant vaccination, vaccination during early adolescence, at age 11 to 12 years, was recommended, and this recommendation has been extended to include all unvaccinated children age 0 to 18 years.
The two available recombinant hepatitis B vaccines are comparable, one containing 10 ug ofHBsAg(Recombivax-HB) and the other containing 20 ug of HBsAg (Engerix-B), and recommended doses for each injection vary for the two preparations. For Recombivax-HB, 2.5 ug is recommended for children1.5 g daily, leads to microvesicular fat deposits in the liver. Liver injury, which is often only one facet of the toxicity produced by the direct hepatotoxins, may go unrecognized until jaundice appears. In idiosyncratic drug reactions the occurrence of hepatitis is usually infrequent and unpredictable, the response is not dose-dependent, and it may occur at any time during or shortly after exposure to the drug. Extrahepatic manifestations of hypersensitivity, such as rash, arthralgias, fever, leukocytosis, and eosinophilia, occur in about one-quarter of patients with idiosyncratic hepatotoxic drug reactions; this observation and the unpredictability of idiosyncratic drug hepatotoxicity contributed to the hypothesis
that this category of drug reactions is immunologically mediated. More recent evidence, however, suggests that, in most cases, even idiosyncratic reactions represent direct hepatotoxity but are caused by drug metabolites rather than by the intact compound. Even the prototype of idiosyncratic hepatoxicity reactions, halothane hepatitis, and isoniazid hepatotoxicity, associated frequently with hypersensitivity manifestations, are now recognized to be mediated by toxic metabolites that damage liver cells directly. Currently, most idiosyncratic reactions are thought to result from differences in metabolic reactivity to specific agents; host susceptibility is mediated by the kinetics of toxic metabolite generation, which differs among individuals. Occasionally, however, the clinical features of an allergic reaction (prominent tissue eosinophilia, autoantibodies, etc.) are difficult to ignore. In vitro models have been described in which lymphocyte cytotoxicity can be demonstrated against rabbit hepatocytes altered by incubation with the potential offending drug. Furthermore, several instances of drug hepatotoxicity are associated with the appearance of autoantibodies, including a class of antibodies to liver-kidney microsomes, anti-LKM2, directed against a cytochrome P450 enzyme. Similarly, in selected cases, a drug or its metabolite has been shown to bind to a host cellular component forming a hapten; the immune response to this "neoantigen" is postulated to play a role in the pathogenesis of liver injury. Therefore, some authorities subdivide idiosyncratic drug hepatotoxicity into hypersensitivity (allergic) and "metabolic" categories. Several unusual exceptions notwithstanding, true drug allergy is difficult to support in most cases of idiosyncratic drug-induced liver injury. Idiosyncratic reactions lead to a morphologic pattern that is more variable than those produced by direct toxins; a single agent is often capable of causing a variety of lesions, although certain patterns tend to predominate. Depending on the agent involved, idiosyncratic hepatitis may result in a clinical and morphologic picture indistinguishable from that of viral hepatitis (e.g., halothane) or may simulate extrahepatic bile duct obstruction clinically with morphologic evidence of cholestasis. Drug-induced cholestasis ranges from mild to increasingly severe: (1) bland cholestasis with limited hepatocellular injury (e.g., estrogens, 17,a-substituted androgens); (2) inflammatory cholestasis (e.g., phenothiazines, amoxicillin-clavulanic acid, oxacillin, erythromcyin estolate); (3) sclerosing cholangitis (e.g., after intrahepatic infusion of the chemotherapeutic agent floxuridine for hepatic metastases from a primary colonic carcinoma); (4) disappearance of bile ducts, "ductopenic" cholestasis, similar to that observed in chronic rejection following liver transplantation (e.g., carbamazine, chlorpromazine, tricyclic antidepressant agents). Morphologic alterations may also include bridging hepatic necrosis (e.g., methyldopa), or, infrequently, hepatic granulomas (e.g., sulfonamides). Some drugs result in macrovesicular or microvesicular steatosis or steatohepatitis, which in some cases has been linked to mitochondrial dysfunciton and lipid peroxidation. Severe hepatotoxicity associated with steatohepatitis, most likely a result of mitochondrial toxicity, is being recognized with increasing frequency among patients receiving antiretroviral therapy with reverse transcriptase inhibitors (e.g., zidovudine, didanosine) or protease inhibitors (e.g., indinavir, ritonavir) for HIV infection. Not all adverse hepatic drug reactions can be classified as either toxic or idiosyncratic in type. For example, oral contraceptives, which combine estrogenic and progestational compounds, may result in impairment of hepatic tests and occasionally in jaundice. However, they do not produce necrosis or fatty change, manifestations of hypersensitivity are generally absent, and susceptibility to the development of oral
contraceptive-induced cholestasis appears to be genetically determined. Other instances of genetically determined drug hepatotoxicity have been identified. For example, approximately 10% of the population have an autosomally recessive trait associated with the absence of cytochrome P450 enzyme 2D6 and have impaired debrisoquine-4-hydroxylase enzyme activity. As a result, they cannot metabolize, and are at increased risk of hepatotoxicity resulting from, certain compounds such as desipramine, propranolol, and quinidine. Because drug-induced hepatitis is often a presumptive diagnosis and many other disorders produce a similar clinicopathologic picture, evidence of a causal relationship between the use of a drug and subsequent liver injury may be difficult to establish. The relationship is most convincing for the direct hepatotoxins, which lead to a high frequency of hepatic impairment after a short latent period. Idiosyncratic reactions may be reproduced, in some instances, when rechallenge, after an asymptomatic period, results in a recurrence of signs, symptoms, and morphologic and biochemical abnormalities. Rechallenge, however, is often ethically unfeasible, because severe reactions may occur. TREATMENT Treatment of toxic and drug-induced hepatic disease is largely supportive, except in acetaminophen hepatotoxicity (see below). In patients with fulminant hepatitis resulting from drug hepatotoxicity, liver transplantation may be life-saving (Chap. 301). Withdrawal of the suspected agent is indicated at the first sign of an adverse reaction. In the case of the direct toxins, liver involvement should not divert attention from renal or other organ involvement, which may also threaten survival. InTable 296-2, several classes of chemical agents are listed, together with examples of the pattern of liver injury produced by them. Certain drugs appear to be responsible for the development of chronic as well as acute hepatic injury. For example, oxyphenisatin, methyldopa, and isoniazid have been associated with moderate to severe chronic hepatitis, and halothane and methotrexate have been implicated in the development of cirrhosis. A syndrome resembling primary biliary cirrhosis has been described following treatment with chlorpromazine, methyl testosterone, tolbutamide, and other drugs. Portal hypertension in the absence of cirrhosis may result from alterations in hepatic architecture produced by vitamin A or arsenic intoxication, industrial exposure to vinyl chloride, or administration of thorium dioxide. The latter three agents have also been associated with angiosarcoma of the liver. Oral contraceptives have been implicated in the development of hepatic adenoma and, rarely, hepatocellular carcinoma and occlusion of the hepatic vein (Budd-Chiari syndrome). Another unusual lesion, peliosis hepatis (blood cysts of the liver), has been observed in some patients treated with anabolic steroids. The existence of these hepatic disorders expands the spectrum of liver injury induced by chemical agents and emphasizes the need for a thorough drug history in all patients with liver dysfunction. The following are the patterns of adverse hepatic reactions for some prototypic agents. ACETAMINOPHEN HEPATOTOXICITY (DIRECT TOXIN)
Acetaminophen has caused severe centrilobular hepatic necrosis when ingested in large amounts in suicide attempts or accidentally by children. A single dose of 10 to 15 g, occasionally less, may produce clinical evidence of liver injury. Fatal fulminant disease is usually (although not invariably) associated with ingestion of 25 g or more. Blood levels of acetaminophen correlate with the severity of hepatic injury (levels >300 ug/mL 4 h after ingestion are predictive of the development of severe damage; levels200 ug/mL measured at 4 h or>100 ug/mL at 8 h after ingestion), the administration of sulfhydryl
compounds (e.g., cysteamine, cysteine, or N-acetylcysteine) appears to reduce the severity of hepatic necrosis. These agents appear to act by providing a reservoir of sulfhydryl groups to bind the toxic metabolites or by stimulating synthesis and repletion of hepatic glutathione. Therapy should be begun within 8 h of ingestion but may be effective even if given as late as 24 to 36 h after overdose. Later administration of sulfhydryl compounds is of uncertain value. Routine use of N-acetylcysteine has reduced substantially the occurrence of fatal acetaminophen hepatotoxicity. When given orally, N-acetylcysteine is diluted to yield a 5% solution. A loading dose of 140 mg/kg is given, followed by 70 mg/kg every 4 h for 15 to 20 doses. Whenever a patient with potential acetaminophen hepatotoxicity is encountered, a local poison control center should be contacted. Treatment can be stopped when plasma acetominophen levels indicate that the risk of liver damage is low. Survivors of acute acetaminophen overdose usually have no evidence of hepatic sequelae. In a few patients, prolonged or repeated administration of acetaminophen in therapeutic doses appears to have led to the development of chronic hepatitis and cirrhosis. HALOTHANE HEPATOTOXICITY (IDIOSYNCRATIC REACTION) Administration of halothane, a nonexplosive fluorinated hydrocarbon anesthetic agent that is structurally similar to chloroform, results in severe hepatic necrosis in a small number of individuals, many of whom have previously been exposed to this agent. The failure to produce similar hepatic lesions reliably in animals, the rarity of hepatic impairment in human beings, and the delayed appearance of hepatic injury suggest that halothane is not a direct hepatotoxin but rather a sensitizing agent. However, manifestations of hypersensitivity are seen in1 (Table 298-2). Hyperbilirubinemia is common and is accompanied by modest increases in the alkaline phosphatase. Derangement in hepatocyte synthetic function indicates more serious disease. Hypoalbuminemia and coagulopathy are common in advanced liver injury. The mean corpuscular volume (MCV) and uric acid level are commonly elevated in chronic alcohol abuse. Measurement of the carbohydrate-deficient transferrin (CDT) is superior to the measurement of the gamma-glutamyl transpeptidase (GGTP) or MCV in identifying excessive drinking. Ultrasonography is useful in detecting fatty infiltration of the liver and determining liver size. The demonstration by ultrasound of portal vein flow reversal, ascites, and intra-abdominal collaterals indicates serious liver injury with less potential for complete reversal of liver disease. PROGNOSIS Critically ill patients with alcoholic hepatitis have short-term mortality rates approaching 70%. Severe alcoholic hepatitis is heralded by coagulopathy (prothrombin time >5 s), anemia, serum albumin concentrations below 2.5 mg/dL, serum bilirubin levels>8 mg/dL, renal failure, and ascites. A discriminant function calculated as 4.6 ´[prothrombin time - control(seconds)] + serum bilirubin (mg/dL) can identify patients with a poor prognosis (discriminant function >32). The presence of ascites, variceal hemorrhage, deep encephalopathy, or hepatorenal syndrome predicts a dismal prognosis. The pathologic stage of the injury can be helpful in predicting prognosis. Liver biopsy should be performed whenever possible to confirm the diagnosis, to establish potential reversibility of the liver disease, and to guide the therapeutic decisions. TREATMENT Complete abstinence from alcohol is the cornerstone in the treatment of alcoholic liver disease. Improved survival rates and the potential for reversal of histologic injury regardless of the initial clinical presentation are associated with total avoidance of alcoholic ingestion. Referral of patients to experienced alcohol counselors and/or alcohol treatment programs should be routine in the management of patients with alcoholic liver disease. Attention should be directed to the nutritional and psychosocial states during the evaluation and treatment periods. Because of data suggesting that the pathogenic mechanisms in alcoholic hepatitis involve cytokine release and the perpetuation of injury by immunologic processes, glucocorticoids have been extensively evaluated in the treatment of alcoholic hepatitis. Patients with severe alcoholic hepatitis, defined as a discriminant function >32, were given prednisone, 40 mg/d, or prednisolone, 32 mg/d, for 4 weeks followed by a steroid taper (Fig. 298-2). Exclusion criteria included active gastrointestinal bleeding, sepsis, renal failure, or pancreatitis. Because of inordinate surgical mortality rates and the high rates of recidivism after transplantation, patients with alcoholic hepatitis are not candidates for immediate liver transplantation. The transplant candidacy of these patients should be reevaluated after a defined period of sobriety. (Bibliography omitted in Palm version)
Back to Table of Contents
299. CIRRHOSIS AND ITS COMPLICATIONS - Raymond T. Chung, Daniel K. Podolsky Cirrhosis is a pathologically defined entity that is associated with a spectrum of characteristic clinical manifestations. The cardinal pathologic features reflect irreversible chronic injury of the hepatic parenchyma and include extensive fibrosis in association with the formation of regenerative nodules. These features result from hepatocyte necrosis, collapse of the supporting reticulin network with subsequent connective tissue deposition, distortion of the vascular bed, and nodular regeneration of remaining liver parenchyma. The central event leading to hepatic fibrosis is activation of the hepatic stellate cell. Upon activation by factors released by hepatocytes and Kupffer cells, the stellate cell assumes a myofibroblast-like conformation and, under the influence of cytokines such as transforming growth factor b(TGF-b), produces fibril-forming type I collagen. The precise point at which fibrosis becomes irreversible is unclear. The pathologic process should be viewed as a final common pathway of many types of chronic liver injury. Clinical features of cirrhosis derive from the morphologic alterations and often reflect the severity of hepatic damage rather than the etiology of the underlying liver disease. Loss of functioning hepatocellular mass may lead to jaundice, edema, coagulopathy, and a variety of metabolic abnormalities; fibrosis and distorted vasculature lead to portal hypertension and its sequelae, including gastroesophageal varices and splenomegaly. Ascites and hepatic encephalopathy result from both hepatocellular insufficiency and portal hypertension. Classification of the various types of cirrhosis based on either etiology or morphology alone is unsatisfactory. A single pathologic pattern may result from a variety of insults, while the same insult may produce several morphologic patterns. Nevertheless, most types of cirrhosis may be usefully classified by a mixture of etiologically and morphologically defined entities as follows: (1) alcoholic; (2) cryptogenic and posthepatitic; (3) biliary; (4) cardiac; and (5) metabolic, inherited, and drug-related. This chapter considers the various types of cirrhosis and their complications. ALCOHOLIC CIRRHOSIS Definition Alcoholic cirrhosis is only one of many consequences resulting from chronic alcohol ingestion, and it often accompanies other forms of alcohol-induced liver injury, including alcoholic fatty liver and alcoholic hepatitis (Chap. 298). Alcoholic cirrhosis, historically referred to as Laennec's cirrhosis, is the most common type of cirrhosis encountered in North America and many parts of western Europe and South America. It is characterized by diffuse fine scarring, fairly uniform loss of liver cells, and small regenerative nodules, and therefore it is sometimes referred to as micronodular cirrhosis. However, micronodular cirrhosis may also result from other types of liver injury (e.g., following jejunoileal bypass), and thus alcoholic cirrhosis and micronodular cirrhosis are not necessarily synonymous. Conversely, alcoholic cirrhosis may progress to macronodular cirrhosis with time. Etiology SeeChap. 298, "Alcoholic Liver Disease." Pathology and Pathogenesis With continued alcohol intake and destruction of hepatocytes, fibroblasts (including activated hepatic stellate cells that have transformed
into myofibroblasts with contractile properties) appear at the site of injury and deposit collagen. Weblike septa of connective tissue appear in periportal and pericentral zones and eventually connect portal triads and central veins. This fine connective tissue network surrounds small masses of remaining liver cells, which regenerate and form nodules. Although regeneration occurs within the small remnants of parenchyma, cell loss generally exceeds replacement. With continuing hepatocyte destruction and collagen deposition, the liver shrinks in size, acquires a nodular appearance, and becomes hard as "end-stage" cirrhosis develops. Although alcoholic cirrhosis is usually a progressive disease, appropriate therapy and strict avoidance of alcohol may arrest the disease at most stages and permit functional improvement. In addition, there is strong evidence that concomitant chronic hepatitis C virus (HCV) infection significantly accelerates development of alcoholic cirrhosis. Clinical Features Signs and Symptoms Alcoholic cirrhosis may be clinically silent, and many cases (10 to 40%) are discovered incidentally at laparotomy or autopsy. In many cases symptoms are insidious in onset, occurring usually after 10 or more years of excessive alcohol use and progressing slowly over subsequent weeks and months. Anorexia and malnutrition lead to weight loss and a reduction in skeletal muscle mass. The patient may experience easy bruising, increasing weakness, and fatigue. Eventually the clinical manifestations of hepatocellular dysfunction and portal hypertension ensue, including progressive jaundice, bleeding from gastroesophageal varices, ascites, and encephalopathy. The abrupt onset of one of these complications may be the first event prompting the patient to seek medical attention. In other cases, cirrhosis first becomes evident when the patient requires treatment of symptoms related to alcoholic hepatitis. A firm, nodular liver may be an early sign of disease; the liver may be either enlarged, normal, or decreased in size. Other frequent findings include jaundice, palmar erythema, spider angiomas, parotid and lacrimal gland enlargement, clubbing of fingers, splenomegaly, muscle wasting, and ascites with or without peripheral edema. Men may have decreased body hair and/or gynecomastia and testicular atrophy, which, like the cutaneous findings, result from disturbances in hormonal metabolism, including increased peripheral formation of estrogen due to diminished hepatic clearance of the precursor androstenedione. Testicular atrophy may reflect hormonal abnormalities or the toxic effect of alcohol on the testes. In women, signs of virilization or menstrual irregularities may occasionally be encountered. Dupuytren's contractures resulting from fibrosis of the palmar fascia with resulting flexion contracture of the digits are associated with alcoholism but are not specifically related to cirrhosis. Although the cirrhotic patient may stabilize if drinking is discontinued, over a period of years, the patient may become emaciated, weak, and chronically jaundiced. Ascites and other signs of portal hypertension may become increasingly prominent. Ultimately, most patients with advanced cirrhosis die in hepatic coma, commonly precipitated by hemorrhage from esophageal varices or intercurrent infection. Progressive renal dysfunction often complicates the terminal phase of the illness. Laboratory Findings In advanced alcoholic liver disease, abnormalities of laboratory tests are more common. Anemia may result from acute and chronic gastrointestinal
blood loss, coexistent nutritional deficiency (notably of folic acid and vitamin B12), hypersplenism, and a direct suppressive effect of alcohol on the bone marrow. Hemolytic anemia, presumably due to effects of hypercholesterolemia or erythrocyte membranes resulting in unusual spurlike projections (acanthocytosis), has been described in some alcoholics with cirrhosis. Mild or pronounced hyperbilirubinemia may be found, usually in association with varying elevations of serum alkaline phosphatase levels. Levels of serum AST (asparate aminotransferase) are frequently elevated, but levels>5ukat (300 units) are unusual and should prompt one to look for other coincident or complicating factors. In contrast to viral hepatitis, the serum AST is usually disproportionately elevated relative to ALT (alanine aminotransferase), i.e., AST/ALT ratio >2. This discrepancy in alcoholic liver disease may result from the proportionally greater inhibition of ALT synthesis by ethanol, which may be partially reversed by pyridoxal phosphate. The serum prothrombin time is frequently prolonged, reflecting reduced synthesis of clotting proteins, most notably the vitamin K-dependent factors (see "Coagulopathy," below). The serum albumin level is usually depressed, while serum globulins are increased. Hypoalbuminemia reflects in part overall impairment in hepatic protein synthesis, while hyperglobulinemia is thought to result from nonspecific stimulation of the reticuloendothelial system. Elevated blood ammonia levels in patients with hepatic encephalopathy reflect diminished hepatic clearance because of impaired liver function and shunting of portal venous blood around the cirrhotic liver into the systemic circulation (see "Hepatic Encephalopathy," below). A variety of metabolic disturbances may be detected. Glucose intolerance due to endogenous insulin resistance may be present; however, clinical diabetes is uncommon. Central hyperventilation may lead to respiratory alkalosis in patients with cirrhosis. Dietary deficiency and increased urinary losses lead to hypomagnesemia and hypophosphatemia. In patients with ascites and dilutional hyponatremia, hypokalemia may occur from increased urinary potassium losses due in part to hyperaldosteronism. Prerenal azotemia is also observed in such patients. Diagnosis Alcoholic cirrhosis should be strongly suspected in patients with a history of prolonged or excessive alcohol intake and physical signs of chronic liver disease. However, since only 10 to 15% of individuals with excessive alcohol intake develop cirrhosis, other causes and types of liver disease may have to be excluded. The clinical features and laboratory findings are usually sufficient to provide reasonable indication of the presence and extent of hepatic injury. Although a percutaneous needle biopsy of the liver is not usually necessary to confirm the typical findings of alcoholic hepatitis or cirrhosis, it may be helpful in distinguishing patients with less advanced liver disease from those with cirrhosis and in excluding other forms of liver injury such as viral hepatitis. Biopsy may also be helpful as a diagnostic tool in evaluating patients with clinical findings suggestive of alcoholic liver disease who deny alcohol intake. In patients with features of cholestasis, ultrasonography may be appropriate to exclude the presence of extrahepatic biliary obstruction. When the clinical status of an otherwise stable cirrhotic patient deteriorates without an obvious explanation, complicating conditions, such as infection, portal vein thrombosis, and hepatocellular carcinoma, should be sought.
Prognosis Abstinence from alcohol as well as early and appropriate medical care can decrease long-term morbidity and mortality and delay or prevent the appearance of further complications. Patients who have had a major complication of cirrhosis and who continue to drink have a 5-year survival of less than 50%. However, those patients who remain abstinent have a substantially better prognosis. In general, the overall outlook in patients with advanced liver disease remains poor; most of these patients eventually die as a result of massive variceal hemorrhage and/or profound hepatic encephalopathy. TREATMENT Alcoholic cirrhosis is a serious illness that requires long-term medical supervision and careful management. Therapy of the underlying liver disease is largely supportive. Specific treatment is directed at particular complications such as variceal bleeding and ascites (see below). While some studies suggest that administration of glucocorticoids in moderately large doses for 4 weeks is helpful in patients with severe alcoholic hepatitis and encephalopathy, these drugs have no role in the treatment of established alcoholic cirrhosis. While one study suggested a mortality benefit for the antifibrotic agent colchicine in alcoholic cirrhosis, it has not yet been reproduced; thus colchicine cannot be routinely recommended. The patient should be made to realize that there is no medication that will protect the liver against the effects of further alcohol ingestion. Therefore, alcohol should be absolutely forbidden. An important component of the complete care of such patients is encouragement to become involved in an appropriate alcohol counseling program. All medicines must be administered with caution in the patient with cirrhosis, especially those eliminated or modified through hepatic metabolism or biliary pathways. In particular, care must be taken to avoid overzealous use of drugs that may directly or indirectly precipitate complications of cirrhosis. For example, vigorous treatment of ascites with diuretics may result in electrolyte abnormalities or hypovolemia, which can lead to coma. Similarly, even modest doses of sedatives can lead to deepening encephalopathy. Aspirin should be avoided in patients with cirrhosis because of its effects on coagulation and gastric mucosa. Acetaminophen should be used with caution and in doses of less than 2 g/day. Patients who drink alcohol are more sensitive to the hepatotoxic effects of acetaminophen, probably due to increased metabolism of the drug to toxic intermediates and decreased glutathione levels. POSTHEPATITIC AND CRYPTOGENIC CIRRHOSIS Definition Posthepatitic or postnecrotic cirrhosis represents the final common pathway of many types of chronic liver disease. Coarsely nodular and multilobular cirrhosis are terms synonymous with posthepatitic cirrhosis. The term cryptogenic cirrhosis has been used interchangeably with postpathepatitic cirrhosis, but this designation should be reserved for those cases in which the etiology of cirrhosis is unknown (approximately 10% of all patients with cirrhosis). Etiology Posthepatitic cirrhosis is a morphologic term referring to a defined stage of advanced chronic liver injury of both specific and unknown (cryptogenic) causes. Epidemiologic and serologic evidence suggest that viral hepatitis (hepatitis B or hepatitis
C) may be an antecedent factor in from one-fourth to three-fourths of cases of apparently cryptogenic posthepatitic cirrhosis. In areas where hepatitis B virus (HBV) infection is endemic (e.g., Southeast Asia, sub-Saharan Africa), up to 15% of the population may acquire the infection in early childhood, and cirrhosis may ultimately develop in one-fourth of these chronic carriers. Although HBV infection is much less prevalent in the United States, it is relatively common among certain high-risk groups (e.g., persons with multiple sexual partners, especially men who have sex with men, injection drug users) and contributes to an increased incidence of cirrhosis. In the United States,HCVinfection accounts for many cases of cirrhosis following blood transfusions. Before routine screening of blood donors was introduced, hepatitis C occurred in 5 to 10% of blood recipients. Following infection, cirrhosis may ultimately develop in more than 20% of individuals after 20 years. More than half of patients who would previously have been designated as having cryptogenic chronic liver disease have evidence of HCV infection. Increasing recognition of the progressive nature of nonalcoholic steatohepatitis has revealed that a large portion of cases previously designated cryptogenic cirrhosis may be attributable to this disorder (Chap. 300). Posthepatitic cirrhosis may also develop in patients with autoimmune hepatitis (Chap. 297). The most common causes of cirrhosis in the United States, which ultimately lead to liver transplantation, include chronicHCVinfection, alcohol, primary biliary cirrhosis, primary sclerosing cholangitis, and nonalcoholic steatohepatitis (NASH). Less common causes of posthepatitic cirrhosis, including drugs and toxins, are listed inTable 299-1. Pathology The posthepatitic liver is typically shrunken in size, distorted in shape, and composed of nodules of liver cells separated by dense and broad bands of fibrosis. The microscopic picture is consistent with the gross impression. Posthepatitic cirrhosis is characterized morphologically by (1) extensive confluent loss of liver cells, (2) stromal collapse and fibrosis resulting in broad bands of connective tissue containing the remains of many portal triads, and (3) irregular nodules of regenerating hepatocytes, varying in size from microscopic to several centimeters in diameter. Clinical Features In patients with cirrhosis of known etiology in whom there is progression to a posthepatitic stage, the clinical manifestations are an extension of those resulting from the initial disease process. Usually clinical symptoms are related to portal hypertension and its sequelae, such as ascites, splenomegaly, hypersplenism, encephalopathy, and bleeding gastroesophageal varices. The hematologic and liver function abnormalities resemble those seen with other types of cirrhosis. In a few patients with posthepatitic cirrhosis, the diagnosis may be made incidentally at operation, at postmortem, or by a needle biopsy of the liver performed to investigate abnormal liver function tests or hepatomegaly. Diagnosis and Prognosis Posthepatitic cirrhosis should be suspected in patients with signs and symptoms of cirrhosis or portal hypertension. Needle or operative liver biopsies confirm the diagnosis, although nonuniformity of the pathologic process may result in sampling errors. The diagnosis of cryptogenic cirrhosis is reserved for those patients in whom no known etiology can be demonstrated. About 75% of patients have progressive disease despite supportive therapy and die within 1 to 5 years from complications, including variceal hemorrhage, hepatic encephalopathy, or
superimposed hepatocellular carcinoma. TREATMENT Management is usually limited to treatment of the complications of portal hypertension, including control of ascites, avoidance of drugs or excessive protein intake that may induce hepatic coma, and prompt treatment of infections (see below). In patients with asymptomatic cirrhosis, expectant management alone is appropriate. In those patients in whom posthepatitic cirrhosis has developed as a result of a treatable condition, therapy directed at the primary disorder may limit further progression (e.g., Wilson's disease, hemochromatosis). BILIARY CIRRHOSIS Biliary cirrhosis results from injury to or prolonged obstruction of either the intrahepatic or extrahepatic biliary system. It is associated with impaired biliary excretion, destruction of hepatic parenchyma, and progressive fibrosis. Primary biliary cirrhosis (PBC) is characterized by chronic inflammation and fibrous obliteration of intrahepatic bile ductules. Secondary biliary cirrhosis (SBC) is the result of long-standing obstruction of the larger extrahepatic ducts. Although primary and secondary biliary cirrhosis are separate pathophysiologic entities with respect to the initial insult, many clinical features are similar. PRIMARY BILIARY CIRRHOSIS Etiology and Pathogenesis The cause ofPBCremains unknown. Several observations suggest that a disordered immune response may be involved. PBC is frequently associated with a variety of disorders presumed to be autoimmune in nature, such as the syndrome of calcinosis, Raynaud's phenomenon, esophageal dysmotility, sclerodactyly, telangiectasia (CREST); the sicca syndrome (dry eyes and dry mouth); autoimmune thyroiditis; type 1 diabetes mellitus; and IgA deficiency. Most important, a circulating IgG antimitochondrial antibody (AMA) is detected in more than 90% of patients withPBC and only rarely in other forms of liver disease. It has been demonstrated that these autoantibodies recognize three to five inner mitochondrial membrane proteins identified as enzymes of the pyruvate dehydrogenase complex (PDC), the branched chain-ketoacid dehydrogenase complex (BCKDC), and thea-ketoglutarate dehydrogenase complex (KGDC). The major autoantigen in PBC (found in 90% of patients) has been identified as the 74-kDa E2 component of the PDC, dihydrolipoamide acetyltransferase. The antibodies are directed to a region essential for binding of a lipoic acid cofactor and inhibit the overall enzymatic activity of the PDC. Other AMA autoantibodies in PNC patients are directed to similar constituents of BCKDC and KGDC and also inhibit their enzymatic function. It remains unclear whether these properties have a direct pathogenetic role in the development of PBC. In addition to AMA, elevated serum levels of IgM and cryoproteins consisting of immune complexes capable of activating the alternative complement pathway are found in 80 to 90% of patients. Aberrant expression of major histocompatibility complex class II molecules has been found on biliary epithelium in association with PBC, suggesting that these cells may serve as antigen-presenting cells in this setting. Lymphocytes are prominent in the
portal regions and surround damaged bile ducts. These histologic findings resemble those noted in graft-versus-host disease following bone marrow transplantation and suggest that damage to bile ducts may be immunologically mediated, perhaps reflecting a defect in a suppressor cell population. Pathology PBC is divided into four stages based on morphologic findings. The earliest recognizable lesion (stage I), termed chronic nonsuppurative destructive cholangitis, is a necrotizing inflammatory process of the portal triads. It is characterized by destruction of medium and small bile ducts, a dense infiltrate of acute and chronic inflammatory cells, mild fibrosis, and occasionally, bile stasis. At times, periductal granulomas and lymph follicles are found adjacent to affected bile ducts. Subsequently, the inflammatory infiltrate becomes less prominent, the number of bile ducts is reduced, and smaller bile ductules proliferate (stage II). Progression over a period of months to years leads to a decrease in interlobular ducts, loss of liver cells, and expansion of periportal fibrosis into a network of connective tissue scars (stage III). Ultimately, cirrhosis, which may be micronodular or macronodular, develops (stage IV). Clinical Features Signs and Symptoms Many patients withPBC are asymptomatic, and the disease is initially detected on the basis of elevated serum alkaline phosphatase levels during routine screening. The majority of such patients remain asymptomatic for prolonged periods, although most ultimately develop progressive liver injury. Among patients with symptomatic disease, 90% are women age 35 to 60. Often the earliest symptom is pruritus, which may be either generalized or limited initially to the palms and soles. In addition, fatigue is commonly a prominent early symptom. After several months or years, jaundice and gradual darkening of the exposed areas of the skin (melanosis) may ensue. Other early clinical manifestations ofPBCreflect impaired bile excretion. These include steatorrhea and the malabsorption of lipid-soluble vitamins. Protracted elevation of serum lipids, especially cholesterol, leads to subcutaneous lipid deposition around the eyes (xanthelasmas) and over joints and tendons (xanthomas). Over a period of months to years, the itching, jaundice, and hyperpigmentation slowly worsen. Eventually, signs of hepatocellular failure and portal hypertension develop and ascites appears. Progression may be quite variable. Whereas a proportion of asymptomatic patients may show no signs of progression for a decade or longer, in others, death due to hepatic insufficiency may occur within 5 to 10 years after the first signs of the illness. Such decompensation is often precipitated by uncontrolled variceal hemorrhage or infection. Physical examination may be entirely normal in the early phase of the disease, when patients are asymptomatic or pruritus is the sole complaint. Later, there may be jaundice of varying intensity, hyperpigmentation of the exposed skin areas, xanthelasmas and tendinous and planar xanthomas, moderate to striking hepatomegaly, splenomegaly, and clubbing of the fingers. Bone tenderness, signs of vertebral compression, ecchymoses, glossitis, and dermatitis may all be noted. Clinical evidence of the sicca syndrome can be found in as many as 75% of patients, and serologic evidence of autoimmune thyroid disease in 25%. Other conditions encountered with increased frequency include rheumatoid arthritis,CRESTsyndrome, keratoconjunctivitis sicca, IgA
deficiency, type 1 diabetes mellitus, scleroderma, pernicious anemia, and renal tubular acidosis. Bone disease is often a significant problem encountered over the course of the disease. While osteomalacia occurs due to diminished vitamin D absorption, accelerated osteoporosis in this patient population (the majority of whom are postmenopausal women) is even more common. Laboratory Findings PBC is increasingly diagnosed at a presymptomatic stage, prompted by the finding of a twofold or greater elevation of the serum alkaline phosphatase during routine screening. Serum 5¢-nucleotidase activity andg-glutamyl transpeptidase levels are also elevated. In this setting, serum bilirubin is usually normal and aminotransferase levels minimally increased. The diagnosis is supported by a positiveAMA test (titer > 1:40). The latter is both relatively specific and sensitive; a positive test is found in over 90% of symptomatic patients and is present in fewer than 5% of patients with other liver diseases. As the disease evolves, the serum bilirubin level rises progressively and may reach 510 umol/L (30 mg/dL) or more in the final stages. Serum aminotransferase values rarely exceed 2.5 to 3.3 ukat (150 to 200 units). Hyperlipidemia is common, and a striking increase of the serum unesterified cholesterol is often noted. An abnormal serum lipoprotein (lipoprotein X) may be present in PBC but is not specific and appears in other cholestatic conditions. A deficiency of bile salts in the intestine leads to moderate steatorrhea and impaired absorption of the fat-soluble vitamins and hypoprothrombinemia. Patients with PBC have elevated liver copper levels, but this finding is not specific and is found in all disorders in which there is prolonged cholestasis. Diagnosis PBCshould be considered in middle-aged women with unexplained pruritus or an elevated serum alkaline phosphatase and in whom there may be other clinical or laboratory features of protracted impairment of biliary excretion. Although a positive serumAMAdetermination provides important diagnostic evidence, false-positive results do occur; therefore, liver biopsy should be performed to confirm the diagnosis. Rarely, the AMA test may be negative in patients with histologic features of PBC. Frequently, patients have antibodies to the E2 protein in tests using these specific antigens. In some cases with histologic features of PBC and a negative AMA, antinuclear or smooth-muscle antibodies are present (as in autoimmune hepatitis), and the designation autoimmune cholangitis is applied. The natural history of this entity, however, appears to resemble that of PBC. If the AMA test is negative, the biliary tract should be evaluated to exclude primary sclerosing cholangitis and remediable extrahepatic biliary tract obstruction, especially in view of the frequent presence of coexisting cholelithiasis. TREATMENT While there is no specific therapy for PBC, ursodiol has been shown to improve biochemical and histologic features and might improve survival, particularly liver transplantation-free survival (although this remains unproven). Ursodiol should be given in doses of 10 to 15 mg/kg per day, but lower doses are sometimes just as effective in reducing serum alkaline phosphatase and aminotransferase levels. Ursodiol should be given with food and can be taken in a single dose daily. Side effects are rare: gastrointestinal intolerance (bloating, indigestion) and skin rashes occur but are uncommon. Isolated instances of severe exacerbation of pruritus have been reported in
patients with advanced disease. Ursodiol probably works by replacing the endogenously produced hydrophobic bile acids with urosdeoxycholate, a hydrophilic and relatively nontoxic bile acid. Unfortunately, ursodiol does not prevent ultimate progression ofPBC, and the only established "cure" is liver transplantation. Results of liver transplantation for PBC are excellent, survival exceeding that for patients receiving transplantation for most other forms of end-stage liver disease. Recurrence of PBC after liver transplantation has been reported but is uncommon, and the recurrent disease is only slowly progressive. Most patients remainAMApositive after transplantation, and as many as 25% will have histologic features of PBC on liver biopsy after 5 years. Other therapies such as glucocorticoids, colchicine, methotrexate, azathioprine, cyclosporine, and tacrolimus have been reported as effective in small cases series, but none have shown to be effective in adequately controlled trials. Relief of symptoms is also an important part of management ofPBC. As noted, ursodiol may be helpful in controlling symptoms and improving the patient's sense of well-being. Although the mechanism of the protracted pruritus is not entirely clear, cholestyramine, an oral bile salt-sequestering resin, may be helpful in doses of 8 to 12 g/d to decrease both pruritus and hypercholesterolemia. Rifampin, opiate antagonists, ondansetron, plasmapheresis, and ultraviolet light have all been tried for control of pruritus, with varying results. Steatorrhea can be reduced by a low-fat diet and substituting medium-chain triglycerides for dietary long-chain triglycerides. Fat-soluble vitamins A and K should be given by parenteral injection at regular intervals to prevent or correct night blindness and hypoprothrombinemia, respectively. Zinc supplementation may be necessary if night blindness is refractory to vitamin A therapy. An important part of management of PBC and any cholestatic liver disease is assessment and treatment of osteoporosis and osteomalacia. Patients should be screened periodically by bone densitometry and treated as needed with calcium supplements, estrogen, and/or the newer bisphosphonate agents (e.g., alendronate). Progression of PBC leads to the typical complications of advanced liver disease (see below). SECONDARY BILIARY CIRRHOSIS Etiology SBCresults from prolonged partial or total obstruction of the common bile duct or its major branches. In adults, obstruction is most frequently caused by postoperative strictures or gallstones, usually with superimposed infectious cholangitis. Chronic pancreatitis may lead to biliary stricture and secondary cirrhosis. SBC is also an important complication of primary sclerosing cholangitis, a progressive immunologic disorder of the intrahepatic and extrahepatic biliary tree (Chap. 302). Patients with malignant tumors of the common bile duct or pancreas rarely survive long enough to develop SBC. In children, congenital biliary atresia and cystic fibrosis are common causes of SBC. Choledochal cysts, if unrecognized, may also be a rare cause of SBC. Pathology and Pathogenesis Unrelieved obstruction of the extrahepatic bile ducts leads to (1) bile stasis and focal areas of centrilobular necrosis followed by periportal necrosis, (2) proliferation and dilatation of the portal bile ducts and ductules, (3) sterile or infected cholangitis with accumulation of polymorphonuclear infiltrates around bile ducts, and (4) progressive expansion of portal tracts by edema and fibrosis.
Extravasation of bile from ruptured interlobular bile ducts into areas of periportal necrosis leads to the formation of "bile lakes" surrounded by cholesterol-rich pseudoxanthomatous cells. As in other forms of cirrhosis, injury is accompanied by regeneration in residual parenchyma. These changes gradually lead to a finely nodular cirrhosis. In general, at least 3 to 12 months is required for biliary obstruction to result in cirrhosis. Relief of the obstruction is frequently accompanied by biochemical and morphologic improvement. Clinical Features The symptoms, signs, and biochemical findings ofSBC are similar to those ofPBC. Jaundice and pruritus are usually the most prominent features. In addition, fever and/or right upper quadrant pain, reflecting bouts of cholangitis or biliary colic, are typical. The manifestations of portal hypertension are found only in advanced cases. SBC should be considered in any patient with clinical and laboratory evidence of prolonged obstruction to bile flow, especially when there is a history of previous biliary tract surgery or gallstones, bouts of ascending cholangitis, or right upper quadrant pain. Cholangiography (either percutaneous or endoscopic) usually demonstrates the underlying pathologic process. Liver biopsy, although not always necessary from a clinical standpoint, can document the development of cirrhosis. TREATMENT Relief of obstruction to bile flow, by either endoscopic or surgical means, is the most important step in the prevention and therapy ofSBC. Effective decompression of the biliary tract results in a significant improvement in both symptoms and survival, even in patients with established cirrhosis. When obstruction cannot be relieved, as in sclerosing cholangitis, antibiotics may be helpful acutely in controlling superimposed infection or, when administered on a chronic basis, as prophylactic therapy in suppressing recurring episodes of ascending cholangitis. Without relief of obstruction, there is a steady progression to end-stage cirrhosis and its terminal manifestations. CARDIAC CIRRHOSIS Definition Prolonged, severe right-sided congestive heart failure may lead to chronic liver injury and cardiac cirrhosis. The characteristic pathologic features of fibrosis and regenerative nodules distinguish cardiac cirrhosis from both reversible passive congestion of the liver due to acute heart failure and acute hepatocellular necrosis ("ischemic hepatitis" or "shock liver") resulting from systemic hypotension and hypoperfusion of the liver. Etiology and Pathology In right-sided heart failure, retrograde transmission of elevated venous pressure via the inferior vena cava and hepatic veins leads to congestion of the liver. Hepatic sinusoids become dilated and engorged with blood, and the liver becomes tensely swollen. With prolonged passive congestion and ischemia from poor perfusion secondary to reduced cardiac output, necrosis of centrilobular hepatocytes ensues and leads to fibrosis in these central areas. Ultimately, centrilobular fibrosis develops, with collagen extending outward in a characteristic stellate pattern from the central vein. Gross examination of the liver shows alternating red (congested) and pale (fibrotic) areas, a pattern often referred to as "nutmeg liver." Improvement in management of cardiac disorders, particularly advances in surgical treatment, has reduced the
frequency of cardiac cirrhosis. Clinical Features A range of abnormalities of liver function tests may be found, though none is uniformly present. The serum bilirubin is usually only mildly increased and may be predominantly either conjugated or unconjugated. Mild to moderate elevation in alkaline phosphatase level and prothrombin time prolongation are sometimes present. TheASTlevel is typically mildly elevated but may be transiently very high following a period of marked systemic hypotension (shock liver), when the clinical picture can mimic acute viral or drug-induced hepatitis. In cases of tricuspid insufficiency the liver may be pulsatile, but this finding disappears as cirrhosis develops. With prolonged right-sided heart failure the liver becomes enlarged, firm, and usually nontender. The signs and symptoms of heart failure usually overshadow the liver disease. Bleeding from esophageal varices is rare, but chronic encephalopathy may be prominent, with a waxing and waning course reflecting variations in the severity of right-sided heart failure. Ascites and peripheral edema, often primarily related to the underlying cardiac dysfunction, may be worsened by the superimposed liver disease. Diagnosis The presence of a firm, enlarged liver with signs of chronic liver disease in a patient with valvular heart disease, constrictive pericarditis, or cor pulmonale of long duration (>10 years) should suggest cardiac cirrhosis. Liver biopsy can confirm the diagnosis but is usually contraindicated because of coagulopathy or ascites. Coexistent chronic heart and liver disease should also raise the possibility of hemochromatosis (Chap. 345), amyloidosis (Chap. 319), or other infiltrative diseases. Budd-Chiari syndrome resulting from the occlusion of the hepatic veins or inferior vena cava may be confused with acute congestive hepatomegaly. In this condition the liver is grossly enlarged and tender, and severe intractable ascites is present. However, signs and symptoms of heart failure are notably absent. The most common cause is thrombosis of the hepatic veins, often in the setting of polycythemia rubra vera, myeloproliferative syndromes, paroxysmal nocturnal hemoglobinuria, oral contraceptive use, or other hypercoagulable states; it may also result from invasion of the inferior vena cava by tumor, such as renal cell or hepatocellular carcinoma. Idiopathic membranous obstruction of the inferior vena cava is the most common cause of this syndrome in Japan. Hepatic venography or liver biopsy showing centrilobular congestion and sinusoidal dilatation in the absence of right-sided heart failure establishes the diagnosis of Budd-Chiari syndrome. Venocclusive disease affecting the sublobular branches of the hepatic veins and the hepatic venues may result from hepatic irradiation, treatment with certain antineoplastic agents, or ingestion of pyrrolidizine alkaloids present in some herbal teas ("bush tea disease") and can mimic congestive hepatomegaly. TREATMENT Prevention or treatment of cardiac cirrhosis depends on the diagnosis and therapy of the underlying cardiovascular disorder. Improvement in cardiac function frequently results in improvement of liver function and stabilization of the liver disease. METABOLIC, HEREDITARY, DRUG-RELATED, AND OTHER TYPES OF CIRRHOSIS (See Table 299-1)
Cirrhosis or hepatitis may result from a wide variety of other processes encompassing the spectrum of etiologic factors listed in Table 299-2. Although some of these disorders have distinctive clinical or morphologic features, the manifestations of cirrhosis are largely independent of the underlying pathogenic mechanism. NONCIRRHOTIC FIBROSIS OF THE LIVER Several diseases, either congenital or acquired, may be associated with localized or generalized hepatic fibrosis. They are distinguished from cirrhosis by the absence of hepatocellular damage and the lack of nodular regenerative activity. The clinical manifestations in such cases are largely secondary to portal hypertension. The different types of these disorders are indicated inTable 299-2; with the exception of schistosomiasis in some regions of the world, all these conditions are relatively rare. MAJOR COMPLICATIONS OF CIRRHOSIS The clinical course of patients with advanced cirrhosis is often complicated by a number of important sequelae that are independent of the etiology of the underlying liver disease. These include portal hypertension and its consequences (e.g., gastroesophageal varices and splenomegaly), ascites, hepatic encephalopathy, spontaneous bacterial peritonitis, hepatorenal syndrome, and hepatocellular carcinoma. PORTAL HYPERTENSION Definition and Pathogenesis Normal pressure in the portal vein is low (5 to 10 mmHg) because vascular resistance in the hepatic sinusoids is minimal. Portal hypertension (>10 mmHg) most commonly results from increased resistance to portal blood flow. Because the portal venous system lacks valves, resistance at any level between the right side of the heart and splanchnic vessels results in retrograde transmission of an elevated pressure. Increased resistance can occur at three levels relative to the hepatic sinusoids: (1) presinusoidal, (2) sinusoidal, and (3) postsinudoidal. Obstruction in the presinusoidal venous compartment may be anatomically outside the liver (e.g., portal vein thrombosis) or within the liver itself but at a functional level proximal to the hepatic sinusoids so that the liver parenchyma is not exposed to the elevated venous pressure (e.g., schistosomiasis). Postsinusoidal obstruction may also occur outside the liver at the level of the hepatic veins (e.g., Budd-Chiari syndrome), the inferior vena cava, or, less commonly, within the liver (e.g., venocclusive disease). When cirrhosis is complicated by portal hypertension, the increased resistance is usually sinusoidal. While distinctions between pre-, post-, and sinusoidal processes are conceptually appealing, functional resistance to portal flow in a given patient may occur at more than one level. Portal hypertension may also arise from increased blood flow (e.g., massive splenomegaly or arteriovenous fistulas), but the low outflow resistance of the normal liver makes this a rare clinical problem. Cirrhosis is the most common cause of portal hypertension in the United States. Clinically significant portal hypertension is present in >60% of patients with cirrhosis. Portal vein obstruction is the second most common cause; it may be idiopathic or occur in association with cirrhosis, infection, pancreatitis, or abdominal trauma. Portal vein
thrombosis may develop in a variety of hypercoagulable states including polycythemia vera; essential thrombocythemia; deficiencies of protein C, protein S, or antithrombin III; resistance to activated protein C (factor V Leiden); and a mutation of the prothrombin gene (G20210A). Portal vein thrombosis may be idiopathic, though some of these patients may have a subclinical myeloproliferative disorder. Hepatic vein thrombosis (Budd-Chiari syndrome) and hepatic venoocclusive disease are relatively infrequent causes of portal hypertension (see above). Portal vein occlusion may result in massive hematemesis from gastroesophageal varices, but ascites is usually found only when cirrhosis is present. Noncirrhotic portal fibrosis (Table 299-2) accounts for only a few cases of portal hypertension. Clinical Features The major clinical manifestations of portal hypertension include hemorrhage from gastroesophageal varices, splenomegaly with hypersplenism, ascites, and acute and chronic hepatic encephalopathy. These are related, at least in part, to the development of portal-systemic collateral channels. The absence of valves in the portal venous system facilitates retrograde (hepatofugal) blood flow from the high-pressure portal venous system to the lower-pressure systemic venous circulation. Major sites of collateral flow involve the veins around cardioesophageal junction (esophagogastric varices), the rectum (hemorrhoids), retroperitoneal space, and the falciform ligament of the liver (periumbilical or abdominal wall collaterals). Abdominal wall collaterals appear as tortuous epigastric vessels that radiate from the umbilicus toward the xiphoid and rib margins (caput medusae). A frequent marker of the presence of cirrhosis in a patient being followed for chronic liver disease is a progressive decrease in platelet count. A low-normal platelet count can be the first clue to progression to cirrhosis. Ultimately, a marked decrease in platelets (to 30,000 to 60,000/uL) and white blood cells can occur. Diagnosis In patients with known liver disease, the development of portal hypertension is usually revealed by the appearance of splenomegaly, ascites, encephalopathy, and/or esophageal varices. Conversely, the finding of any of these features should prompt evaluation of the patient for the presence of underlying portal hypertension and liver disease. Varices are most reliably documented by fiberoptic esophagoscopy; their presence lends indirect support to the diagnosis of portal hypertension. Although rarely necessary, portal venous pressure may be measured directly by percutaneous transhepatic "skinny needle" catheterization or indirectly through transjugular cannulation of the hepatic veins. Both free and wedged hepatic vein pressure should be measured. While the latter is elevated in sinusoidal and postsinusoidal portal hypertension, including cirrhosis, this measurement is usually normal in presinusoidal portal hypertension. In patients in whom additional information is necessary (e.g., preoperative evaluation before portal-systemic shunt surgery) or when percutaneous catheterization is not feasible, mesenteric and hepatic angiography may be helpful. Particular attention should be directed to the venous phase to assess the patency of the portal vein and the direction of portal blood flow. TREATMENT Although treatment is usually directed toward a specific complication of portal hypertension, attempts are sometimes made to reduce the pressure in the portal venous
system. Surgical decompression procedures have been used for many years to lower portal pressure in patients with bleeding esophageal varices (see below). However, portal-systemic shunt surgery does not result in improved survival rates in patients with cirrhosis. Decompression can now be accomplished without surgery through the percutaneous placement of a portal-systemic shunt, termed a transjugular intrahepatic portosystemic shunt (TIPS).b-Adrenergic blockade with propranolol or nadolol reduces portal pressure through vasodilatory effects on both the splanchnic arterial bed and the portal venous system in combination with reduced cardiac output. Such therapy has been shown to be effective in preventing both a first variceal bleed and subsequent episodes after an initial bleed. Treatment of patients with clinically significant sequelae of portal hypertension, especially variceal bleeding, with doses of propranolol titrated to reduce the resting pulse by 25% is reasonable if no contraindications exist. Vigorous treatment of patients with alcoholic hepatitis and cirrhosis, chronic active hepatitis, or other liver diseases may lead to a fall in portal pressure and to a reduction in variceal size. In general, however, portal hypertension due to cirrhosis is not reversible. In appropriately selected patients, hepatic transplantation will be beneficial. VARICEAL BLEEDING Pathogenesis While vigorous hemorrhage may arise from any portal-systemic venous collaterals, bleeding is most common from varices in the region of the gastroesophageal junction. The factors contributing to bleeding from gastroesophageal varices are not entirely understood but include the degree of portal hypertension (>12 mmHg) and the size of the varices. Clinical Features and Diagnosis Variceal bleeding often occurs without obvious precipitating factors and usually presents with painless but massive hematemesis with or without melena. Associated signs range from mild postural tachycardia to profound shock, depending on the extent of blood loss and degree of hypovolemia. Because patients with varices may bleed just as frequently from other gastrointestinal lesions (e.g., peptic ulcer, gastritis), exclusion of other bleeding sources is important even in patients with prior variceal hemorrhage. Endoscopy is the best approach to evaluate upper gastrointestinal hemorrhage in patients with known or suspected portal hypertension. TREATMENT (SeeFig. 299-1) Variceal bleeding is a life-threatening emergency. Prompt estimation and vigorous replacement of blood loss to maintain intravascular volume are essential and take precedence over diagnostic studies and more specific intervention to stop the bleeding. However, excessive fluid administration can increase portal pressure with resultant further bleeding and should therefore be avoided. Replacement of clotting factors with fresh-frozen plasma is important in patients with coagulopathy. Patients are best managed in an intensive care unit and require close monitoring of central venous or pulmonary capillary wedge pressures, urine output, and mental status. Only when the patient is hemodynamically stable should attention be directed toward specific diagnostic studies (especially endoscopy) and other therapeutic modalities to prevent further or recurrent bleeding.
About half of all episodes of variceal hemorrhage cease without intervention, although the risk of rebleeding is very high. The medical management of acute variceal hemorrhage includes the use of vasoconstrictors (somatostatin/octreotide or vasopressin), balloon tamponade, and endoscopic banding of varices or endoscopic sclerosis of varices (sclerotherapy). Intravenous infusion of vasopressin at a rate of 0.1 to 0.4 U/min results in generalized vasoconstriction leading to diminished blood flow in the portal venous system. Intravenous infusion of vasopressin is as effective as selective intraarterial administration. Control of bleeding can be achieved in up to 80% of cases, but bleeding recurs in more than half after the vasopressin is tapered and discontinued. Furthermore, a number of serious side effects, including cardiac and gastrointestinal tract ischemia, acute renal failure, and hyponatremia, may be associated with vasopressin therapy. Concurrent use of venodilators such as nitroglycerin as an intravenous infusion or isosorbide dinitrate sublingually may enhance the effectiveness of vasopressin and reduce complications. Somatostatin and its analogue, octreotide, are direct splanchnic vasoconstrictors. In some studies somatostatin, given as an initial 250-ug bolus followed by constant infusion (250 ug/h), has been found to be as effective as vasopressin. Octreotide at doses of 50 to 100 ug/h is also effective. These agents are preferable to vasopressin, offering equivalent efficacy with fewer complications. If bleeding is too vigorous or endoscopy is not available, balloon tamponade of the bleeding varices may be accomplished with a triple-lumen (Sengstaken-Blakemore) or four-lumen (Minnesota) tube with esophageal and gastric balloons. Because of the high risk of aspiration, endotracheal intubation should be performed prior to placing one of these tubes. After the tube is introduced into the stomach, the gastric balloon is inflated and pulled back into the cardia of the stomach. If bleeding does not stop, the esophageal balloon is inflated for additional tamponade. Complications occur in 15% or more of patients and include aspiration pneumonitis as well as esophageal rupture. Where available, endoscopic intervention should be employed as the first line of treatment to control bleeding acutely (Chaps. 44 and283). Over the past 18 years, endoscopic sclerosis of esophageal varices has been extensively employed. In this procedure, the varices are injected with one of several sclerosing agents via a needle-tipped catheter passed through the endoscope. After endoscopic identification of varices as the presumed source of bleeding, sclerotherapy controls acute bleeding in up to 90% of cases. In addition, repeated sclerotherapy can be performed until obliteration of all varices is accomplished in an effort to prevent recurrent bleeding. While available data support the efficacy of sclerotherapy in controlling bleeding acutely and in decreasing rebleeding rates, repeated sclerotherapy has not been documented to prolong survival. Mucosal ulceration resulting from injection of the caustic sclerosant may occur and result in further hemorrhage or stenosis. More recently, endoscopic band ligation, in which esophageal varices are ligated and strangulated with endoscopically placed small elastic O-rings, has gained favor. Band ligation has proven to be at least as effective as sclerotherapy in controlling acute variceal bleeding and preventing rebleeding. Because it has been associated with fewer treatment-related complications, band ligation is recommended for long-term obliteration of varices that have bled. Although prophylactic sclerosis or banding of esophageal varices in the absence of proven bleeding cannot yet be recommended, one report suggests that banding may be more effective than beta-blockade in primary prevention of variceal bleeding in high-risk
patients. The effectiveness of nonselective b-adrenergic blocking agents (e.g., propranolol) in the management of acute variceal bleeding is limited due to concomitant hypotension resulting from hypovolemia. However, a number of studies suggest they may be of value in secondary prevention of recurrent variceal hemorrhage. Moreover, prophylactic treatment with nonselective beta blockers (propranolol or nadolol) in patients with large ("high-risk") varices that have never bled appears to decrease the incidence of bleeding and prolong survival. Thus, endoscopic screening for varices in patients with cirrhosis is desirable; some have suggested this should be repeated every other year. Patients with portal hypertension without specific contraindications should be given propranolol in doses that produce a 25% reduction in the resting heart rate or the hepatic venous pressure gradient (HVPG), where available. Propranolol may also prevent recurrent bleeding from severe portal hypertensive gastropathy in patients with cirrhosis. The optimal combination of endoscopic and pharmacologic therapy for prevention of recurrent hemorrhage remains to be established and is the subject of ongoing trials. Surgical treatment of portal hypertension and variceal bleeding involves the creation of a portal-systemic shunt to permit decompression of the portal system. Two types of portal systemic shunts have been used: nonselective shunts, to decompress the entire portal system, and selective shunts, intended to decompress only the varices while maintaining blood flow to the liver itself. Nonselective shunts include end-to-side or side-to-side portacaval and proximal splenorenal anastomoses; selective shunts include the distal splenorenal shunt. Nonselective shunts are more likely to be complicated by encephalopathy than selective shunts. Emergency portal-systemic nonselective shunts may control acute hemorrhage, but such surgery is usually used only as a last resort because early operative mortality can be high. The role of portal-systemic shunt surgery after initial control of bleeding by nonoperative means is also uncertain. Surgically created shunts effectively reduce the risk of recurrent hemorrhage, but the overall mortality of patients undergoing such surgery is comparable to that of unoperated patients. Although patients who have undergone portal-systemic surgery succumb to recurrent bleeding less commonly than unoperated patients, this improvement is counterbalanced by increased morbidity from encephalopathy and death from progressive liver failure. Increasingly, therapeutic portal-systemic shunts have been reserved for patients who experience further bleeding despite serial endoscopic sclerotherapy or band ligation. InTIPS, a technique developed to create a portal-systemic shunt by a percutaneous approach, an expandable metal stent is advanced to the hepatic veins under angiographic guidance and then through the substance of the liver to create a direct portacaval channel. This technique offers an alternative to surgery for refractory bleeding due to portal hypertension. However, stents frequently undergo stenosis or occlude over a period of months, prompting the need for a second TIPS or an alternative approach. Encephalopathy may be encountered after TIPS just as in the surgical shunts and is especially problematic in the elderly and those patients with preexisting encephalopathy. TIPS should be reserved for those individuals who fail endoscopic or medical management and are poor surgical risks. TIPS may have a useful role as a "bridge" for those patients with end-stage cirrhosis awaiting liver transplantation. Procedures such as esophageal transection have also been advocated
for the management of acute variceal bleeding, but their efficacy remains unproven. Even though recent trials found that esophageal transection was as effective as endoscopic sclerotherapy, transection is usually considered a last resort. The management of bleeding gastric fundal varices, either alone or in conjunction with esophageal varices, is more problematic, since sclerotherapy and banding are generally not effective. Vasoactive pharmacologic therapy should be instituted, butTIPS or shunt surgery should be considered because of high failure and rebleeding rates. For isolated gastric varices, splenic vein thrombosis should be specifically sought, since splenectomy is curative. Portal Hypertensive Gastropathy Although variceal hemorrhage is the most commonly encountered bleeding complication of portal hypertension, many patients will develop a congestive gastropathy due to the venous hypertension. In this condition, identified by endoscopic examination, the mucosa appears engorged and friable. Indolent mucosal bleeding occurs rather than the brisk hemorrhage typical of a variceal source. b-Adrenergic blockade with propranolol (reducing splanchnic arterial pressure as well as portal pressure) is sometimes effective in ameliorating this condition. H2receptor antagonists or other agents useful in the treatment of peptic disease are usually not helpful. SPLENOMEGALY Definition and Pathogenesis Congestive splenomegaly is common in patients with severe portal hypertension. Rarely, massive splenomegaly from nonhepatic disease leads to portal hypertension due to increased blood flow in the splenic vein. Clinical Features Although usually asymptomatic, splenomegaly may be massive and contribute to the thrombocytopenia or pancytopenia of cirrhosis. In the absence of cirrhosis, splenomegaly in association with variceal hemorrhage should suggest the possibility of splenic vein thrombosis. TREATMENT Splenomegaly usually requires no specific treatment, although massive enlargement of the spleen may occasionally necessitate splenectomy at the time of shunt surgery. However, it should be noted that splenectomy without an accompanying shunt may actually increase portal pressure, and portal vein thrombosis may result from splenectomy. Splenectomy may also be indicated if splenomegaly is the cause rather than the result of portal hypertension (as in splenic vein thrombosis). Thrombcytopenia alone is rarely severe enough to necessitate removal of the spleen. Splenectomy should be avoided in a patient eligible for liver transplantation. ASCITES Definition Ascites is the accumulation of excess fluid within the peritoneal cavity. It is most frequently encountered in patients with cirrhosis and other forms of severe liver disease, but a number of other disorders may lead to either transudative or exudative ascites (Chap. 46).
Pathogenesis The accumulation of ascitic fluid represents a state of total-body sodium and water excess, but the event that initiates this imbalance is unclear. Three theories have been proposed (Fig. 299-2). The "underfilling" theory suggests that the primary abnormality is inappropriate sequestration of fluid within the splanchnic vascular bed due to portal hypertension and a consequent decrease in effective circulating blood volume. According to this theory, an apparent decrease in intravascular volume (underfilling) is sensed by the kidney, which responds by retaining salt and water. The "overflow" theory suggests that the primary abnormality is inappropriate renal retention of salt and water in the absence of volume depletion. A third and more recent theory, the peripheral arterial vasodilation hypothesis, may unify the earlier theories and accounts for the constellation of arterial hypotension and increased cardiac output in association with high levels of vasoconstrictor substances that are routinely found in patients with cirrhosis and ascites. Again, sodium retention is considered secondary to arterial vascular underfilling and the result of a disproportionate increase of the vascular compartment due to arteriolar vasodilation rather than from decreased intravascular volume. According to this theory, portal hypertension results in splanchnic arteriolar vasodilation, mediated by nitric oxide, and leading to underfilling of the arterial vascular space and baroreceptor-mediated stimulation of renin-angiotensin, sympathetic output, and antidiuretic hormone release. Regardless of the initiating event, a number of factors contribute to accumulation of fluid in the abdominal cavity (Fig. 299-2). Elevated levels of serum epinephrine and norepinephrine have been well documented. Increased central sympathetic outflow is found in patients with cirrhosis and ascites but not in those with cirrhosis alone. Increased sympathetic output results in diminished natriuresis by activation of the renin-angiotensin system and diminished sensitivity to atrial natriuretic peptide. Portal hypertension plays an important role in the formation of ascites by raising hydrostatic pressure within the splanchnic capillary bed. Hypoalbuminemia and reduced plasma oncotic pressure also favor the extravasation of fluid from plasma to the peritoneal cavity, and thus ascites is infrequent in patients with cirrhosis unless both portal hypertension and hypoalbuminemia are present. Hepatic lymph may weep freely from the surface of the cirrhotic liver due to distortion and obstruction of hepatic sinusoids and lymphatics and contributes to ascites formation. In contrast to the contribution of transudative fluid from the portal vascular bed, hepatic lymph may weep into the peritoneal cavity even in the absence of marked hypoproteinemia because the endothelial lining of the hepatic sinusoids is discontinuous. This mechanism may account for the high protein concentration present in the ascitic fluid of some patients with venoocclusive disease or the Budd-Chiari syndrome. Renal factors also play an important role in perpetuating ascites. Patients with ascites fail to excrete a water load in a normal fashion. They have increased renal sodium reabsorption by both proximal and distal tubules, the latter due largely to increased plasma renin activity and secondary hyperaldosteronism. Insensitivity to circulating atrial natriuretic peptide, often present in elevated concentrations in patients with cirrhosis and ascites, may be an important contributory factor in many patients. This insensitivity has been documented in those patients with the most severely impaired sodium excretion, who typically also exhibit low arterial pressure and marked overactivity of the renin-aldosterone axis. Renal vasoconstriction, perhaps resulting from increased serum
prostaglandin or catecholamine levels, may also contribute to sodium retention. Recently a role for endothelin, a potent vasoconstrictor peptide, has been proposed. While elevated levels have been reported by some, this has not been observed by others. As discussed inChap. 46, ascites may arise in a number of clinical settings in addition to cirrhosis and portal hypertension. Although historically ascites was classified as either transudative or exudative, similar to the characterization of pleural fluids, this schema has limitations. Instead, the serum-ascites albumin gradient (SAAG) provides a better classification than total protein content or other parameters. In cirrhosis, the serum albumin concentration is usually at least 10 g/L (1 g/dL) higher than that of the ascitic fluid, thus yielding a high SAAG [³11 g/L (³1.1 g/dL)], reflecting indirectly the abnormally high hydrostatic pressure gradient between the portal bed and the ascitic compartment. Conversely, the presence of a low SAAG [8.0) fluid per day containing about 20 enzymes and zymogens. The pancreatic secretions provide the enzymes needed to effect the major digestive activity of the gastrointestinal tract and provide an optimal pH for the function of these enzymes. REGULATION OF PANCREATIC SECRETION The exocrine pancreas is influenced by intimately interacting hormonal and neural systems. Gastric acid is the stimulus for the release of secretin, a peptide with 27 amino acids. Sensitive radioimmunoassay studies for secretin suggest that the pH threshold for its release from the duodenum and jejunum is 4.5. Secretin stimulates the secretion of pancreatic juice rich in water and electrolytes. Release of cholecystokinin (CCK) from the duodenum and jejunum is largely triggered by long-chain fatty acids, certain essential amino acids (tryptophan, phenylalanine, valine, methionine), and gastric acid itself. CCK evokes an enzyme-rich secretion from the pancreas. Gastrin, although it has the same terminal tetrapeptide as CCK, is a weak stimulus for pancreatic enzyme output. The parasympathetic nervous system (via the vagus nerve) exerts significant control over pancreatic secretion. Secretion evoked by secretin and CCK depends on permissive roles of vagal afferent and efferent pathways. This is particularly true for enzyme secretion, whereas water and bicarbonate secretion is heavily dependent on the hormonal effects of secretin and CCK. Also, vagal stimulation effects the release of vasoactive intestinal peptide (VIP), a secretin agonist. Bile salts also stimulate pancreatic secretion, thereby integrating the functions of the biliary tract, pancreas, and small intestine. Somatostatin acts on multiple sites to induce inhibition of pancreatic secretion. The appropriate roles of other peptides, such as peptide YY, pancreastatin, gastrin-releasing peptide, pituitary adenylate cyclase-activating polypeptide, calcitonin gene-related peptide, and galanin are still being defined. Nitric oxide is an important neurotransmitter in the regulation of pancreatic exocrine secretion, although its mechanism of action has not been fully elucidated. WATER AND ELECTROLYTE SECRETION Although sodium, potassium, chloride, calcium, zinc, phosphate, and sulfate are found in pancreatic secretions, bicarbonate is the ion of primary physiologic importance. In the acini and in the ducts, secretin causes the cells to add water and bicarbonate to the fluid. In the ducts, an exchange occurs between bicarbonate and chloride. There is a good correlation between the maximal bicarbonate output after stimulation with secretin and the pancreatic mass. The bicarbonate output of 120 to 300 mmol/d helps neutralize gastric acid and creates the appropriate pH for the activity of the pancreatic enzymes.
ENZYME SECRETION The pancreas secretes amylolytic, lipolytic, and proteolytic enzymes. Amylolytic enzymes, such as amylase, hydrolyze starch to oligosaccharides and to the disaccharide maltose. The lipolytic enzymes include lipase, phospholipase A, and cholesterol esterase. Bile salts inhibit lipase in isolation; but colipase, another constituent of pancreatic secretion, binds to lipase and prevents this inhibition. Bile salts activate phospholipase A and cholesterol esterase. Proteolytic enzymes include endopeptidases (trypsin, chymotrypsin), which act on internal peptide bonds of proteins and polypeptides; exopeptidases (carboxypeptidases, aminopeptidases), which act on the free carboxyl- and amino-terminal ends of peptides, respectively; and elastase. The proteolytic enzymes are secreted as inactive precursors (zymogens). Ribonucleases (deoxyribonucleases, ribonuclease) are also secreted. Although pancreatic enzymes usually are secreted in parallel, nonparallel secretion can occur as a result of exocytosis from heterogeneous sources in the pancreas. Enterokinase, an enzyme found in the duodenal mucosa, cleaves the lysine-isoleucine bond of trypsinogen to form trypsin. Trypsin then activates the other proteolytic zymogens in a cascade phenomenon. All pancreatic enzymes have pH optima in the alkaline range. AUTOPROTECTION OF THE PANCREAS Autodigestion of the pancreas is prevented by the packaging of proteases in precursor form and by the synthesis of protease inhibitors. These protease inhibitors are found in the acinar cell, the pancreatic secretions, and the alpha1- and alpha2-globulin fractions of plasma. EXOCRINE-ENDOCRINE RELATIONSHIPS Insulin appears to be needed locally for secretin andCCK to promote exocrine secretion; thus, it acts in a permissive role for these two hormones. ENTEROPANCREATIC AXIS AND FEEDBACK INHIBITION Pancreatic enzyme secretion is controlled, at least in part, by a negative feedback mechanism induced by the presence of active serine proteases in the duodenum. To illustrate, perfusion of the duodenal lumen with phenylalanine causes a prompt increase in plasmaCCKlevels as well as increased secretion of chymotrypsin. However, simultaneous perfusion with trypsin blunts both responses. Conversely, perfusion of the duodenal lumen with protease inhibitors actually leads to enzyme hypersecretion. The available evidence supports the concept that the duodenum contains a peptide called CCK releasing factor (CCK-RF) that is involved in stimulating CCK release. Two peptides, luminal CCK-RF and diazepam-binding inhibitor, have been found that may be the CCK-RF. Serine proteases inhibit pancreatic secretion by acting on CCK-RF. It appears that serine proteases inhibit pancreatic secretion by acting on a CCK-releasing peptide in the lumen of the small intestine. ACUTE PANCREATITIS GENERAL CONSIDERATIONS
Pancreatic inflammatory disease may be classified as (1) acute pancreatitis and (2) chronic pancreatitis. The pathologic spectrum of acute pancreatitis varies from edematous pancreatitis, which is usually a mild and self-limited disorder, to necrotizing pancreatitis, in which the degree of pancreatic necrosis correlates with the severity of the attack and its systemic manifestations. The term hemorrhagic pancreatitis is less meaningful in a clinical sense because variable amounts of interstitial hemorrhage can be found in pancreatitis as well as in other disorders such as pancreatic trauma, pancreatic carcinoma, and severe congestive heart failure. The incidence of pancreatitis varies in different countries and depends on cause, e.g., alcohol, gallstones, metabolic factors, and drugs (Table 304-1). In the United States, for example, acute pancreatitis is related to alcohol ingestion more commonly than to gallstones; in England, the opposite obtains. There are 185,000 new cases of acute pancreatitis per year in the United States. ETIOLOGY AND PATHOGENESIS There are many causes of acute pancreatitis (Table 304-1), but the mechanisms by which these conditions trigger pancreatic inflammation have not been identified. Alcoholic patients with pancreatitis may represent a special subset, since most alcoholics do not develop pancreatitis. The list of identifiable causes is growing, and it is likely that pancreatitis related to viral infections, drugs, and as yet undefined factors is more common than heretofore recognized. Approximately 2 to 5% of cases of acute pancreatitis are drug-related (Table 304-1). Drugs cause pancreatitis either by a hypersensitivity reaction or by the generation of a toxic metabolite, although in some cases it is not clear which of these mechanisms is operative. Autodigestion is one pathogenetic theory, according to which pancreatitis results when proteolytic enzymes (e.g., trypsinogen, chymotrypsinogen, proelastase, and phospholipase A) are activated in the pancreas rather than in the intestinal lumen. A number of factors (e.g., endotoxins, exotoxins, viral infections, ischemia, anoxia, and direct trauma) are believed to activate these proenzymes. Activated proteolytic enzymes, especially trypsin, not only digest pancreatic and peripancreatic tissues but also can activate other enzymes, such as elastase and phospholipase. The active enzymes then digest cellular membranes and cause proteolysis, edema, interstitial hemorrhage, vascular damage, coagulation necrosis, fat necrosis, and parenchymal cell necrosis. Cellular injury and death result in the liberation of activated enzymes. In addition, activation and release of bradykinin peptides and vasoactive substances (e.g., histamine) are believed to produce vasodilation, increased vascular permeability, and edema. Thus, a cascade of events culminates in the development of acute necrotizing pancreatitis. The autodigestion theory has largely eclipsed two older theories. First, according to the "common channel" theory, the existence of a common anatomic channel for pancreatic secretions and bile permits reflux of bile into the pancreatic duct, which results in activation of pancreatic enzymes. (Actually, a common channel with free communication
between the common bile duct and the main pancreatic duct is infrequently encountered.) The second theory is that obstruction and hypersecretion are pivotal in the development of pancreatitis. Obstruction of the main pancreatic duct, however, produces pancreatic edema but generally not pancreatitis. A recent hypothesis to explain the intrapancreatic activation of zymogens is that they become activated by lysosomal hydrolases in the pancreatic acinar cell itself. In two different types of experimental pancreatitis, it has been demonstrated that digestive enzymes and lysosomal hydrolases become admixed; as a result, the latter can activate the former in the acinar cell. In vitro, lysosomal enzymes such as cathepsin B can activate trypsinogen, and trypsin can activate the other protease precursors. It is still not clear, however, whether the human acinar cell can provide the pH (about 3.0) necessary for activation of trypsinogen by lysosomal hydrolases. It is now believed that ischemia/hypoperfusion can alone result in activation of trypsinogen and pancreatic injury. CLINICAL FEATURES Abdominal pain is the major symptom of acute pancreatitis. Pain may vary from a mild and tolerable discomfort to severe, constant, and incapacitating distress. Characteristically, the pain, which is steady and boring in character, is located in the epigastrium and periumbilical region and often radiates to the back as well as to the chest, flanks, and lower abdomen. The pain is frequently more intense when the patient is supine, and patients often obtain relief by sitting with the trunk flexed and knees drawn up. Nausea, vomiting, and abdominal distention due to gastric and intestinal hypomotility and chemical peritonitis are also frequent complaints. Physical examination frequently reveals a distressed and anxious patient. Low-grade fever, tachycardia, and hypotension are fairly common. Shock is not unusual and may result from (1) hypovolemia secondary to exudation of blood and plasma proteins into the retroperitoneal space (a "retroperitoneal burn"); (2) increased formation and release of kinin peptides, which cause vasodilation and increased vascular permeability; and (3) systemic effects of proteolytic and lipolytic enzymes released into the circulation. Jaundice occurs infrequently; when present, it usually is due to edema of the head of the pancreas with compression of the intrapancreatic portion of the common bile duct. Erythematous skin nodules due to subcutaneous fat necrosis may occur. In 10 to 20% of patients, there are pulmonary findings, including basilar rales, atelectasis, and pleural effusion, the latter most frequently left-sided. Abdominal tenderness and muscle rigidity are present to a variable degree, but, compared with the intense pain, these signs may be unimpressive. Bowel sounds are usually diminished or absent. A pancreatic pseudocyst may be palpable in the upper abdomen. A faint blue discoloration around the umbilicus (Cullen's sign) may occur as the result of hemoperitoneum, and a blue-red-purple or green-brown discoloration of the flanks (Turner's sign) reflects tissue catabolism of hemoglobin. The latter two findings, which are uncommon, indicate the presence of a severe necrotizing pancreatitis. LABORATORY DATA The diagnosis of acute pancreatitis is usually established by the detection of an
increased level of serum amylase. Values threefold or more above normal virtually clinch the diagnosis if overt salivary gland disease and gut perforation or infarction are excluded. However, there appears to be no definite correlation between the severity of pancreatitis and the degree of serum amylase elevation. After 48 to 72 h, even with continuing evidence of pancreatitis, total serum amylase values tend to return to normal. However, pancreatic isoamylase and lipase levels may remain elevated for 7 to 14 days. It will be recalled that amylase elevations in serum and urine occur in many conditions other than pancreatitis (seeTable 303-2). Importantly, patients with acidemia (arterial pH £7.32) may have spurious elevations in serum amylase. In one study, 12 of 33 patients with acidemia had elevated serum amylase, but only 1 had an elevated lipase value; in 9, salivary-type amylase was the predominant serum isoamylase. This finding explains why patients with diabetic ketoacidosis may have marked elevations in serum amylase without any other evidence of acute pancreatitis. Serum lipase activity increases in parallel with amylase activity, and measurement of both enzymes increases the diagnostic yield. An elevated serum lipase or trypsin value is usually diagnostic of acute pancreatitis; these tests are especially helpful in patients with nonpancreatic causes of hyperamylasemia (seeTable 303-2). Markedly increased levels of peritoneal or pleural fluid amylase [>1500 nmol/L (> 5000 U/dL)] are also helpful, if present, in establishing the diagnosis. Leukocytosis (15,000 to 20,000 leukocytes per microliter) occurs frequently. Patients with more severe disease may show hemoconcentration with hematocrit values exceeding 50% because of loss of plasma into the retroperitoneal space and peritoneal cavity. Hyperglycemia is common and is due to multiple factors, including decreased insulin release, increased glucagon release, and an increased output of adrenal glucocorticoids and catecholamines. Hypocalcemia occurs in approximately 25% of patients, and its pathogenesis is incompletely understood. Although earlier studies suggested that the response of the parathyroid gland to a decrease in serum calcium is impaired, subsequent observations have failed to confirm this idea. Intraperitoneal saponification of calcium by fatty acids in areas of fat necrosis occurs occasionally, with large amounts (up to 6.0 g) dissolved or suspended in ascitic fluid. Such "soap formation" also may be significant in patients with pancreatitis, mild hypocalcemia, and little or no obvious ascites. Hyperbilirubinemia [serum bilirubin>68 umol/L (> 4.0 mg/dL)] occurs in approximately 10% of patients. However, jaundice is transient, and serum bilirubin levels return to normal in 4 to 7 days. Serum alkaline phosphatase and aspartate aminotransferase (AST) levels are also transiently elevated and parallel serum bilirubin values. Markedly elevated serum lactic dehydrogenase (LDH) levels [>8.5 umol/L (> 500 U/dL)] suggest a poor prognosis. Serum albumin is decreased to £30 g/L (£3.0 g/dL) in about 10% of patients; this sign is associated with more severe pancreatitis and a higher mortality rate (Table 304-2). Hypertriglyceridemia occurs in 15 to 20% of patients, and serum amylase levels in these individuals are often spuriously normal (Chap. 303). Most patients with hypertriglyceridemia and pancreatitis, when subsequently examined, show evidence of an underlying derangement in lipid metabolism which probably antedated the pancreatitis (see below). Approximately 25% of patients have hypoxemia (arterial PO2£ 60 mmHg), which may herald the onset of adult respiratory distress syndrome. Finally, the electrocardiogram is occasionally abnormal in acute pancreatitis with ST-segment and T-wave abnormalities simulating myocardial ischemia.
Although one or more radiologic abnormalities are found in over 50% of patients, the findings are inconstant and nonspecific. The chief value of conventional x-rays [chest films; kidney, ureter, and bladder (KUB) studies] in acute pancreatitis is to help exclude other diagnoses, especially a perforated viscus. Upper gastrointestinal tract x-rays have been superseded by ultrasonography and computed tomography (CT). A CT scan may confirm the clinical impression of acute pancreatitis even in the face of normal serum amylase levels. Importantly, CT is quite helpful in indicating the severity of acute pancreatitis and the risk of morbidity and mortality (see below). Sonography and radionuclide scanning [N-p-isopropylacetanilide-iminodiacetic acid (PIPIDA) scan; hepatic 2,6-dimethyliminodiacetic acid (HIDA) scan] are useful in acute pancreatitis to evaluate the gallbladder and biliary tree.*Radiologic studies useful in the diagnosis of acute pancreatitis are discussed in Chap. 303 and listed in Table 303-1. DIAGNOSIS Any severe acute pain in the abdomen or back should suggest acute pancreatitis. The diagnosis is usually entertained when a patient with a possible predisposition to pancreatitis presents with severe and constant abdominal pain, nausea, emesis, fever, tachycardia, and abnormal findings on abdominal examination. Laboratory studies frequently reveal leukocytosis, an abnormal appearance on x-rays of the abdomen and chest, hypocalcemia, and hyperglycemia. The diagnosis is usually confirmed by the finding of an elevated level of serum amylase and/or lipase. Not all the above features have to be present for the diagnosis to be established. The differential diagnosis should include the following disorders: (1) perforated viscus, especially peptic ulcer; (2) acute cholecystitis and biliary colic; (3) acute intestinal obstruction; (4) mesenteric vascular occlusion; (5) renal colic; (6) myocardial infarction; (7) dissecting aortic aneurysm; (8) connective tissue disorders with vasculitis; (9) pneumonia; and (10) diabetic ketoacidosis. A penetrating duodenal ulcer usually can be identified by upper gastrointestinal x-rays and/or endoscopy. A perforated duodenal ulcer is readily diagnosed by the presence of free intraperitoneal air. It may be difficult to differentiate acute cholecystitis from acute pancreatitis, since an elevated serum amylase may be found in both disorders. Pain of biliary tract origin is more right-sided and gradual in onset, and ileus is usually absent; sonography and radionuclide scanning are helpful in establishing the diagnosis of cholelithiasis and cholecystitis. Intestinal obstruction due to mechanical factors can be differentiated from pancreatitis by the history of colicky pain, findings on abdominal examination, and x-rays of the abdomen showing changes characteristic of mechanical obstruction. Acute mesenteric vascular occlusion is usually evident in elderly debilitated patients with brisk leukocytosis, abdominal distention, and bloody diarrhea, in whom paracentesis shows sanguineous fluid and arteriography shows vascular occlusion. Serum as well as peritoneal fluid amylase levels are increased, however, in patients with intestinal infarction. Systemic lupus erythematosus and polyarteritis nodosa may be confused with pancreatitis, especially since pancreatitis may develop as a complication of these diseases. Diabetic ketoacidosis is often accompanied by abdominal pain and elevated total serum amylase levels, thus closely mimicking acute pancreatitis. However, the serum lipase and pancreatic isoamylase levels are not elevated in diabetic ketoacidosis. COURSE OF THE DISEASE AND COMPLICATIONS
It is important to identify patients with acute pancreatitis who have an increased risk of dying. Ranson and Imrie have used multiple prognostic criteria and have demonstrated that there is an increased mortality rate when three or more risk factors are identifiable either at the time of admission to the hospital or during the initial 48 h of hospitalization (Table 304-2). Recent studies indicate that obesity is a major risk factor for severe pancreatitis, presumably because the increased deposits of peripancreatic fat in such patients may predispose them to more extensive pancreatic and peripancreatic necrosis. The acute physiology and chronic health evaluation scoring system (APACHE II) uses the worst values of 12 physiologic measurements plus age and previous health status and provides a good description of illness severity for a wide range of common diseases; this score also correlates with outcome. Prospective studies have compared APACHE II with multiple prognostic criteria, i.e., Ranson and Imrie scores, in predicting the severity of acute pancreatitis. On admission, APACHE II identified approximately two-thirds of severe attacks, and after 48 h, the prognostic accuracy of APACHE II is comparable with that of Ranson and Imrie's scoring system. The drawbacks of APACHE II are (1) its complexity, (2) the requirement of a computer for scoring, and (3) standardization regarding peak values and cutoff scores. McMahon and colleagues have shown that the presence of a "toxic broth" or dark (hemorrhagic) fluid in abdominal pancreatitis is also an important prognostic indicator in acute pancreatitis. These multiple-factor scoring systems are difficult to use and have not been embraced consistently by clinicians. There is a great need for a reliable, simple biochemical test that consistently predicts outcome in patients with acute pancreatitis. Three candidate markers that show great promise are C-reactive protein, serum granulocyte elastase, and urinary trypsinogen activation peptide (TAP). The key indicators of a severe attack of acute pancreatitis are also listed inTable 304-2. Importantly, the presence of any one of these factors is associated with an increased risk of complications, and the presence of any two, with a 20 to 30% mortality rate. The high mortality rate of such severely ill patients is due in large part to infection and warrants intensive radiologic intervention and monitoring and/or a combination of radiologic and surgical means, as discussed in detail below. The local and systemic complications of acute pancreatitis are listed in Table 304-3. In the first 2 to 3 weeks after pancreatitis patients frequently develop an inflammatory mass, which may be due to pancreatic necrosis (with or without infection) or may represent an abscess or pseudocyst (see below). Systemic complications include pulmonary, cardiovascular, hematologic, renal, metabolic, and central nervous system abnormalities. Pancreatitis and hypertriglyceridemia constitute an association in which cause and effect remain incompletely understood. However, several reasonable conclusions can be drawn. First, hypertriglyceridemia can precede and apparently cause pancreatitis. Second, the vast majority (>80%) of patients with acute pancreatitis do not have hypertriglyceridemia. Third, almost all patients with pancreatitis and hypertriglyceridemia have preexisting abnormalities in lipoprotein metabolism. Fourth, many of the patients with this association have persistent hypertriglyceridemia after recovery from pancreatitis and are prone to recurrent episodes of pancreatitis. Fifth, any factor (e.g., drugs or alcohol) that causes an abrupt increase in serum triglycerides to levels greater than 11 mmol/L (1000 mg/dL) can precipitate a bout of pancreatitis that can be associated with significant complications and even become fulminant. To avert the risk of triggering pancreatitis, a fasting serum triglyceride measurement should be
obtained before estrogen replacement therapy is begun in postmenopausal women. Fasting levels less than 300 mg/dL pose no risk, whereas levels greater than 750 mg/dL are associated with a high probability of developing pancreatitis. Finally, patients with a deficiency of apolipoprotein CII have an increased incidence of pancreatitis; apolipoprotein CII activates lipoprotein lipase, which is important in clearing chylomicrons from the bloodstream. Purtscher's retinopathy, a relatively unusual complication, is manifested by a sudden and severe loss of vision in a patient with acute pancreatitis. It is characterized by a peculiar funduscopic appearance with cotton-wool spots and hemorrhages confined to an area limited by the optic disk and macula; it is believed to be due to occlusion of the posterior retinal artery with aggregated granulocytes. The two most common causes of acute pancreatitis are alcoholism and biliary tract disease; other causes are listed in Table 304-1. The risk of acute pancreatitis in patients with at least one gallstone smaller than 5 mm in diameter is fourfold greater than that in patients with larger stones. However, after a conventional workup, a specific cause is not identified in about 30% of patients. It is important to note that ultrasound examinations fail to detect gallstones, especially microlithiasis and/or sludge, in 4 to 7% of patients. In one series of 31 patients diagnosed initially as having idiopathic acute pancreatitis, 23 were found to have occult gallstone disease. Thus, approximately two-thirds of patients with recurrent acute pancreatitis without an obvious cause actually have occult gallstone disease due to microlithiasis. Examination of duodenal aspirates in such cases often reveals cholesterol crystals, which confirm the diagnosis. Other diseases of the biliary tree and pancreatic ducts that can cause acute pancreatitis include choledochocele, ampullary tumors, pancreas divisum, and pancreatic duct stones, stricture, and tumor. Approximately 2% of patients with pancreatic carcinoma present with acute pancreatitis. Pancreatitis in Patients with AIDS The incidence of acute pancreatitis is increased in patients with AIDS for two reasons: (1) the high incidence of infections involving the pancreas, such as infections with cytomegalovirus, Cryptosporidium, and the Mycobacterium avium complex; and (2) the frequent use by patients with AIDS of medications such as didanosine, pentamidine, and trimethoprim-sulfamethoxazole (Chap. 309). TREATMENT In most patients (approximately 85 to 90%) with acute pancreatitis, the disease is self-limited and subsides spontaneously, usually within 3 to 7 days after treatment is instituted. Conventional measures include (1) analgesics for pain, (2) intravenous fluids and colloids to maintain normal intravascular volume, (3) no oral alimentation, and (4) nasogastric suction to decrease gastrin release from the stomach and prevent gastric contents from entering the duodenum. Recent controlled trials, however, have shown that nasogastric suction offers no clear-cut advantages in the treatment of mild to moderately severe acute pancreatitis. Its use, therefore, must be considered elective rather than mandatory. It has been demonstrated thatCCK-stimulated pancreatic secretion is almost abolished
in four different experimental models of acute pancreatitis. This finding probably explains why drugs to block pancreatic secretion in acute pancreatitis have failed to have any therapeutic benefit. For this and other reasons, anticholinergic drugs are not indicated in acute pancreatitis. In addition to nasogastric suction and anticholinergic drugs, other therapies designed to "rest the pancreas" by inhibiting pancreatic secretion have not changed the course of the disease. Although antibiotics have been used in the treatment of acute pancreatitis, randomized, prospective trials have shown no benefit from their use in acute pancreatitis of mild to moderate severity. However, current experimental evidence favors the use of prophylactic antibiotics in severe acute pancreatitis. Results of four contemporary randomized clinical trials restricted to patients with prognostically severe acute pancreatitis have demonstrated an improved outcome, i.e., reduced rate of infection and/or mortality, associated with the antibiotic treatment. The carbapenem group of antibiotics, including imipenem, has a very broad spectrum including activity against Pseudomonas, Staphylococcus, and Enterococcus; and these agents penetrate well into pancreatic tissue. Furthermore, because secondary infection of necrotic pancreatic tissue (abscess, pseudocyst or obstructed biliary passages, ascending cholangitis complicating choledocholithiasis) contributes to many of the late deaths from pancreatitis, appropriate antibiotic therapy of established infections is quite important. Several other drugs have been evaluated by prospective controlled trials and found ineffective in the treatment of acute pancreatitis. The list, by no means complete, includes glucagon, H2blockers, protease inhibitors such as aprotinin, glucocorticoids, calcitonin, nonsteroidal anti-inflammataory drugs (NSAIDs) and lexiplafant, a platelet-activating factor inhibitor. A recent meta-analysis of somatostatin, ocreotide, and the antiprotease gabexate methylate in therapy of acute pancreatitis suggested (1) a reduced mortality rate but no change in complications with octreotide, and (2) no effect on the mortality rate but reduced pancreatic damage with gabexate. Intraabdominal Candida infection during acute necrotizing pancreatitis is increasing in frequency and is associated with an increased mortality rate. In one representatitve trial, intraabdominal Candida infection was found in 13 of 37 cases and was associated with a mortality rate fourfold greater than that associated with intraabdominal bacterial infection alone. Given the impact of Candida infection on the mortality rate in acute necrotizing pancreatitis and the apparent benefit of prophylactic chemotherapy, these data suggest earlier use of fungicides. ACT scan, especially a contrast-enhanced dynamic CT (CECT) scan, provides valuable information on the severity and prognosis of acute pancreatitis (Fig. 304-1 andTable 304-4). In particular, a CECT scan allows estimation of the presence and extent of pancreatic necrosis. Recent studies suggest that the likelihood of prolonged pancreatitis or a serious complication is negligible when the CT severity index is 1 or 2 and low with scores of 3 to 6. However, patients with scores of 7 to 10 had a 92% morbidity rate and a 17% mortality rate. Necrosis is present in 20 to 30% of patients. Those with necrosis have a morbidity rate>20%, whereas those without necrosis have a morbidity rate 30 g/L (>3.0 g/dL)] and a markedly elevated level of amylase. The fluid in true pancreatic ascites usually has an amylase concentration of >20,000 U/L as a result of the ruptured duct or leaking pseudocyst. Lower amylase elevations may be found in the peritoneal fluid of patients with acute pancreatitis. In addition,ERCPoften demonstrates passage of contrast material from a major pancreatic duct or a pseudocyst into the peritoneal cavity. As many as 15% of patients with pseudocysts have concurrent pancreatic ascites. The differential diagnosis should include intraperitoneal carcinomatosis, tuberculous peritonitis, constrictive pericarditis, and Budd-Chiari syndrome.
If the pancreatic duct disruption is posterior, an internal fistula may develop between the pancreatic duct and the pleural space, producing a pleural effusion, which is usually left-sided and often massive. This complication often requires thoracentesis or chest tube drainage. Treatment usually requires the use of nasogastric suction and parenteral alimentation to decrease pancreatic secretion. In addition, paracentesis is performed to keep the peritoneal cavity free of fluid and, it is hoped, to effect sealing of the leak. The long-acting somatostatin analogue octreotide, which inhibits pancreatic secretion, is useful in cases of pancreatic ascites and pleural effusion. If ascites continues to recur after 2 to 3 weeks of medical management, the patient should be operated on after pancreatography to define the anatomy of the abnormal duct. A disrupted main pancreatic duct can also be treated effectively by stenting. Patients in whomERCPidentifies two or more sites of extravasation are unlikely to respond to conservative management and/or stenting. CHRONIC PANCREATITIS AND PANCREATIC EXOCRINE INSUFFICIENCY GENERAL AND ETIOLOGIC CONSIDERATIONS Chronic inflammatory disease of the pancreas may present as episodes of acute inflammation in a previously injured pancreas or as chronic damage with persistent pain or malabsorption. The causes of relapsing chronic pancreatitis are similar to those of acute pancreatitis (Table 304-1), except that there is an appreciable incidence of cases of undetermined origin. In addition, the pancreatitis associated with gallstones is predominantly acute or relapsing-acute in nature. A cholecystectomy is almost always performed in patients after the first or second attack of gallstone-associated pancreatitis. Patients with chronic pancreatitis may present with persistent abdominal pain, with or without steatorrhea; some (~15%) present with steatorrhea and no pain. Patients with chronic pancreatitis in whom there is extensive destruction of the pancreas (less than 10% of exocrine function remaining) have steatorrhea and azotorrhea. Among American adults, alcoholism is the most common cause of clinically apparent pancreatic exocrine insufficiency, while cystic fibrosis is the most frequent cause in children. In up to 25% of American adults with chronic pancreatitis, the cause is not known; that is, they have idiopathic chronic pancreatitis. Mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene have been documented in patients with idiopathic chronic pancreatitis. It has been estimated that in patients with idiopathic pancreatitis the frequency of a single CFTR mutation is 11 times the expected frequency and the frequency of two mutant alleles is 80 times the expected frequency. The results of sweat chloride testing are not diagnostic of cystic fibrosis in these patients. However, these patients have functional evidence of a defect in CFTR-mediated ion transport in nasal epithelium. It is suggested that up to 25% of patients with idiopathic chronic pancreatitis may have abnormalities of the CFTR gene. The therapeutic and prognostic implication of these findings remain to be determined. In other parts of the world, severe protein-calorie malnutrition is a common cause.Table 304-5 lists other causes of pancreatic exocrine insufficiency, but they are relatively uncommon.
PATHOPHYSIOLOGY The events that initiate an inflammatory process in the pancreas are still not well understood, and the many hypotheses will not be reviewed here. In the case of alcohol-induced pancreatitis, it has been suggested that the primary defect may be the precipitation of protein (inspissated enzymes) in the ducts. The resulting ductal obstruction could lead to duct dilation, diffuse atrophy of the acinar cells, fibrosis, and eventual calcification of some of the protein plugs. However, the fact that some alcoholic patients with recurrent acute pancreatitis show no evidence of chronic pancreatitis does not support this hypothesis. In fact, experimental and clinical observations have shown that alcohol has direct toxic effects on the pancreas. While patients with alcohol-induced pancreatitis generally consume large amounts of alcohol, some consume very little (50 g/d or less). Thus prolonged consumption of "socially acceptable" amounts of alcohol is compatible with the development of pancreatitis. In addition, the finding of extensive pancreatic fibrosis in patients who died during their first attack of clinical acute alcohol-induced pancreatitis supports the concept that such patients already have chronic pancreatitis. CLINICAL FEATURES Patients with relapsing chronic pancreatitis may present with symptoms identical to those of acute pancreatitis, but pain may be continuous, intermittent, or absent. The pathogenesis of this pain is poorly understood. Although the classic description is of epigastric pain radiating through the back, the pain pattern is often atypical; the pain may be worst in the right or left upper quadrant of the back or may be diffuse throughout the upper abdomen; it may even be referred to the anterior chest or flank. Characteristically it is persistent, deep-seated, and unresponsive to antacids. It often is worsened by ingestion of alcohol or a heavy meal (especially one rich in fat). Often the pain is severe enough to necessitate the frequent use of narcotics. Weight loss, abnormal stools, and other signs or symptoms suggestive of malabsorption (seeTable 286-5) are common in chronic pancreatitis. However, clinically apparent deficiencies of fat-soluble vitamins are surprisingly rare. The physical findings in these patients are usually not impressive, so that there is a disparity between the severity of the abdominal pain and the physical signs (other than some abdominal tenderness and mild temperature elevation). DIAGNOSTIC EVALUATION (See also Chap. 303) In contrast to relapsing acute pancreatitis, the serum amylase and lipase levels are usually not elevated in chronic pancreatitis. Elevations of serum bilirubin and alkaline phosphatase levels may indicate cholestasis secondary to chronic inflammation around the common bile duct (Fig. 304-3). Many patients demonstrate impaired glucose tolerance, and some have an elevated fasting blood glucose level. The classic triad of pancreatic calcification, steatorrhea, and diabetes mellitus usually establishes the diagnosis of chronic pancreatitis and exocrine pancreatic insufficiency but is found in less than one-third of chronic pancreatitis patients. Accordingly, it is often necessary to perform an intubation test such as the secretin stimulation test, which
usually gives abnormal results when 60% or more of pancreatic exocrine function has been lost. Approximately 40% of patients with chronic pancreatitis have cobalamin (vitamin B12) malabsorption, which can be corrected by the administration of oral pancreatic enzymes. There is usually a marked excretion of fecal fat (Chap. 286), which can be reduced by the administration of oral pancreatic enzymes. The serum trypsinogen (Chap. 303) and the D-xylose urinary excretion test are useful in patients with "pancreatic steatorrhea," since the trypsinogen level will be abnormal, and D-xylose excretion usually is normal. A decreased serum trypsinogen level strongly suggests severe pancreatic exocrine insufficiency. The radiographic hallmark of chronic pancreatitis is the presence of scattered calcification throughout the pancreas (Fig. 304-3). Diffuse pancreatic calcification indicates that significant damage has occurred and obviates the need for a secretin test. While alcohol is by far the most common cause, pancreatic calcification also may be seen in cases of severe protein-calorie malnutrition, hereditary pancreatitis, posttraumatic pancreatitis, hyperparathyroidism, islet cell tumors, and idiopathic chronic pancreatitis. A large prospective study has shown convincingly that pancreatic calcification decreases or even disappears spontaneously in one-third of patients with severe chronic pancreatitis; this outcome may also follow ductal decompression. Pancreatic calcification is a dynamic process that is incompletely understood. Sonography,CT, andERCPgreatly aid the diagnosis of pancreatic disease. In addition to excluding pseudocysts and pancreatic cancer, sonography and CT may show calcification or dilated ducts associated with chronic pancreatitis (Fig. 304-4). ERCP is the only major technique that provides a direct view of the pancreatic duct. In patients with alcohol-induced pancreatitis, ERCP may reveal a pseudocyst missed by sonography or CT. COMPLICATIONS OF CHRONIC PANCREATITIS The complications of chronic pancreatitis are protean. Cobalamin (vitamin B12) malabsorption occurs in 40% of patients with alcohol-induced chronic pancreatitis and in virtually all with cystic fibrosis. It is consistently corrected by the administration of pancreatic enzymes (containing proteases). It may be due to excessive binding of cobalamin by cobalamin-binding proteins other than intrinsic factor, which ordinarily are destroyed by pancreatic proteases and therefore do not compete with intrinsic factor for cobalamin binding. Although most patients show impaired glucose tolerance, diabetic ketoacidosis and coma are uncommon. Similarly, end-organ damage (retinopathy, neuropathy, nephropathy) is also uncommon, and the appearance of these complications should raise the question of concomitant genetic diabetes mellitus. A nondiabetic retinopathy, peripheral in location and secondary to vitamin A and/or zinc deficiency, is common in these patients. Effusions containing high concentrations of amylase may occur into the pleural, pericardial, or peritoneal space. Gastrointestinal bleeding may occur from peptic ulceration, gastritis, a pseudocyst eroding into the duodenum, or ruptured varices secondary to splenic vein thrombosis due to inflammation of the tail of the pancreas. Icterus may occur, caused either by edema of the head of the pancreas, which compresses the common bile duct, or by chronic cholestasis secondary to a chronic inflammatory reaction around the intrapancreatic portion of the common bile duct (Fig. 304-3). The chronic obstruction may lead to
cholangitis and ultimately to biliary cirrhosis. Subcutaneous fat necrosis may appear as tender red nodules on the lower extremities. Bone pain may be secondary to intramedullary fat necrosis. Inflammation of the large and small joints of the upper and lower extremities may occur. The incidence of pancreatic carcinoma is increased in patients with chronic pancreatitis who have been followed for 2 or more years. Twenty years after the diagnosis of chronic pancreatitis, the cumulative risk of pancreatic carcinoma is 4%. Perhaps the most common and troublesome complication is addiction to narcotics. TREATMENT Therapy for patients with chronic pancreatitis is directed toward two major problems -pain and malabsorption. Patients with intermittent attacks of pain are treated essentially like those with acute pancreatitis (see above). Patients with severe and persistent pain should avoid alcohol completely and avoid large meals rich in fat. Since the pain is often severe enough to require frequent use of narcotics (and hence addiction), a number of surgical procedures have been developed for pain relief.ERCPallows the surgeon to plan the operative approach. If there is a stricture of the pancreatic duct, a local resection may ameliorate the pain. Unfortunately, isolated localized strictures are not common. In most patients with alcohol-induced disease, the pancreas is diffusely involved, and surgically correctible localized ductal disease is rare. When there is primary ductal obstruction and dilation, ductal decompression may provide effective pain palliation. Short-term pain relief may be achieved in up to 80% of patients, while long-term pain relief occurs in approximately 50%. In some of these patients, however, pain relief can be achieved only by resecting 50 to 95% of the gland. Although pain relief is achieved in three-quarters of these patients, they tend to develop pancreatic endocrine and exocrine insufficiency and must be treated with pancreatic enzyme replacement therapy. It is important to screen patients carefully, for such radical surgery is contraindicated in those who are severely depressed or suicidal or who continue to drink. Procedures such as splanchnicectomy, celiac ganglionectomy, and nerve blocks usually bring only temporary relief and are not recommended. Endoscopic treatment of chronic pancreatitis may involve sphincterotomy of the minor or major pancreatic sphincter, dilatation of strictures, removal of calculi, or stenting of the ventral or dorsal pancreatic duct. Although many of these techniques are technically impressive, none has been subjected to a randomized trial in patients with chronic pancreatitis. In addition, significant complications -- acute pancreatitis, pancreatic abscess, damage to the pancreatic duct, and death -- have occurred in up to 36% of patients after stent placement. Three double-blind trials have demonstrated that administration of pancreatic enzymes decreases abdominal pain in selected patients with chronic pancreatitis. In these trials, approximately 75% of the patients evaluated experienced pain relief. The patients most likely to respond are those with mild to moderate exocrine pancreatic dysfunction, as evidenced by an abnormal secretin test, normal fat absorption, and minimal abnormalities onERCPexamination. These clinical observations seem to fit with data from human beings and experimental animals demonstrating a negative feedback regulation for pancreatic exocrine secretion controlled by the amount of proteases within the lumen of the proximal small intestine. It seems reasonable to use the following approach for patients with severe, persistent, or continuous abdominal pain thought to
be caused by chronic pancreatitis. After other causes of abdominal pain (peptic ulcer, gallstones, etc.) have been excluded, a pancreatic sonogram should be done. If no mass is found, a secretin test may be performed, because its results usually are abnormal in cases of chronic pancreatitis with pain. If the results are abnormal (i.e., decreased bicarbonate concentration or volume output), a 3- to 4-week trial of pancreatic enzyme administration is appropriate. Eight conventional tablets or capsules are taken at meals and at bedtime. There are a number of studies suggesting that patients may have small-duct chronic pancreatitis and chronic abdominal pain with a normal appearance on radiographic evaluations (ultrasound,CT, ERCP) but abnormal results on hormone stimulation tests (secretin test) and/or abnormal pancreatic histology. Such minimal-change chronic pancreatitis may respond well to pancreatic enzyme therapy (non-enteric-coated) for relief of abdominal pain. If no relief is obtained, and especially if the volume secreted during the secretin test is very low, ERCP should be performed. If a pseudocyst or a localized ductal obstruction is found, surgery should be considered. A patient who has dilated ducts may be a candidate for a surgical ductal decompression procedure. This procedure provides short-term relief in up to 80% of patients, although long-term results are closer to 50%. Some studies have shown octreotide to be effective in decreasing abdominal pain in patients with severe large-duct disease. If no surgically remediable lesion is found and severe pain continues despite abstinence from alcohol, subtotal pancreatic resection may be necessary. The treatment of malabsorption rests on the use of pancreatic enzyme replacement therapy. Diarrhea and steatorrhea are usually improved by this treatment, although the steatorrhea may not be completely corrected. The major problem is delivering enough active enzyme into the duodenum. Steatorrhea could be abolished if 10% of the normal amount of lipase could be delivered to the duodenum at the proper time. This concentration of lipase cannot be achieved with the current preparations of pancreatic enzymes, even if the latter are given in large doses. The reason for these poor results may be that lipase is inactivated by gastric acid, that food empties from the stomach faster than do the pancreatic enzymes, and that batches of commercially available pancreatic extracts vary in enzyme activity. For the usual patient, two or three enteric-coated capsules or eight conventional (non-enteric-coated) tablets of a potent enzyme preparation should be administered with meals. Some patients using conventional tablets require adjuvant therapy to improve enzyme replacement treatment. H2receptor antagonists, sodium bicarbonate, and proton pump inhibitors are effective adjuvants. Antacids containing calcium carbonate or magnesium hydroxide are not effective and may actually result in increased steatorrhea. Several publications have reported colonic strictures in patients with cystic fibrosis receiving extraordinarily high doses of high-potency pancreatic enzyme preparations. Such lesions have not been reported in adults with chronic pancreatitis. Supportive measures include diet restriction and pain medications. The diet should be moderate in fat (30%), high in protein (24%), and low in carbohydrate (40%). Restriction of long-chain triglyceride intake can help patients who do not respond satisfactorily to pancreatic enzyme therapy. Use of foods containing mainly medium-chain fatty acids, which do not require lipase for digestion, may be beneficial. Nonnarcotic analgesics should be emphasized. Patients taking narcotic drugs for pain relief often become addicted and continue to have pain.
Patients with severe exocrine pancreatic insufficiency secondary to alcohol who continue to drink have a high mortality rate (in one series, 50% of patients who were followed for 5 to 12 years died during this period) and significant morbidity (weight loss, lassitude, vitamin deficiency, and narcotic addiction). Chronic pancreatitis carries significant medical and social costs. A recent study found that pancreatitis led to retirement in 11% of patients with the disease, accounting for 45% of all retirements. In 87% of patients with chronic pancreatitis unable to maintain gainful employment, alcoholism was a contributing factor. Patients with chronic pancreatitis also use substantial medical resources. In 1987 in the United States, this diagnosis accounted for 122,000 recorded outpatient visits and 56,000 hospital admissions. Pain may abate if progressive severe exocrine insufficiency continues. Patients who abstain from alcohol and use vigorous replacement therapy for maldigestion-malabsorption do reasonably well. HEREDITARY PANCREATITIS Hereditary pancreatitis is a rare disease that is similar to chronic pancreatitis except for an early age of onset and evidence of hereditary factors (involving an autosomal dominant gene with incomplete penetrance). A genome-wide search using genetic linkage analysis identified the hereditary pancreatitis gene on chromosome 7. An R117H mutation in the cationic trypsinogen gene occurs in most of the families with hereditary pancreatitis that have been studied. Molecular modeling predicts the formation of hydrolysis-resistant trypsin that could lead to pancreatic autodigestion. These patients have recurring attacks of severe abdominal pain which may last from a few days to a few weeks. The serum amylase and lipase levels may be elevated during acute attacks but usually are normal. Patients frequently develop pancreatic calcification, diabetes mellitus, and steatorrhea, and, in addition, they have an increased incidence of pancreatic carcinoma. Such patients often require ductal decompression for pain relief. Abdominal complaints in relatives of patients with hereditary pancreatitis should raise the question of pancreatic disease. PANCREATIC ENDOCRINE TUMORS *Pancreatic endocrine tumors are summarized in Table 304-6 and are discussed in Chap. 93. OTHER CONDITIONS ANNULAR PANCREAS When the ventral pancreatic anlage fails to migrate correctly to make contact with the dorsal anlage, the result may be a ring of pancreatic tissue encircling the duodenum. Such an annular pancreas may cause intestinal obstruction in the neonate or the adult. Symptoms of postprandial fullness, epigastric pain, nausea, and vomiting may be present for years before the diagnosis is entertained. The radiographic findings are symmetric dilation of the proximal duodenum with bulging of the recesses on either side of the annular band, effacement but not destruction of the duodenal mucosa, accentuation of the findings in the right anterior oblique position, and lack of change on
repeated examinations. The differential diagnosis should include duodenal webs, tumors of the pancreas or duodenum, postbulbar peptic ulcer, regional enteritis, and adhesions. Patients with annular pancreas have an increased incidence of pancreatitis and peptic ulcer. Because of these and other potential complications, the treatment is surgical even if the condition has been present for years. Retrocolic duodenojejunostomy is the procedure of choice, although some surgeons advocate Billroth II gastrectomy, gastroenterostomy, and vagotomy. PANCREAS DIVISUM Pancreas divisum occurs when the embryologic ventral and dorsal pancreatic anlagen fail to fuse, so that pancreatic drainage is accomplished mainly through the accessory papilla. Pancreas divisum is the most common congenital anatomic variant of the human pancreas. Current evidence indicates that this anomaly does not predispose to the development of pancreatitis in the great majority of patients who harbor it. However, the combination of pancreas divisum and a small accessory orifice could result in dorsal duct obstruction. The challenge is to identify this subset of patients with dorsal duct pathology. Cannulation of the dorsal duct byERCP is not as easily done as is cannulation of the ventral duct. Patients with pancreatitis and pancreas divisum demonstrated by ERCP should be treated with conservative measures. In many of these patients, pancreatitis is idiopathic and unrelated to the pancreas divisum. Endoscopic or surgical intervention is indicated only when the above methods fail. If marked dilation of the dorsal duct can be demonstrated, surgical ductal decompression should be performed. The appropriate therapy for patients without dilation of the dorsal duct is not yet defined. It should be stressed that the ERCP appearance of pancreas divisum -- i.e., a small-caliber ventral duct with an arborizing pattern -- may be mistaken as representing an obstructed main pancreatic duct secondary to a mass lesion. MACROAMYLASEMIA In macroamylasemia, amylase circulates in the blood in a polymer form too large to be easily excreted by the kidney. Patients with this condition demonstrate an elevated serum amylase value, a low urinary amylase value, and a Cam/Ccrratio of less than 1%. The presence of macroamylase can be documented by chromatography of the serum. The prevalence of macroamylasemia is 1.5% of the nonalcoholic general adult hospital population. Usually macroamylasemia is an incidental finding and is not related to disease of the pancreas or other organs. Macrolipasemia has now been documented in a few patients with cirrhosis or non-Hodgkin's lymphoma. In these patients, the pancreas appeared normal on ultrasound andCTexamination. Lipase was shown to be complexed with immunoglobulin A. Thus, the possibility of both macroamylasemia and macrolipasemia should be considered in patients with elevated blood levels of these enzymes. ACKNOWLEDGEMENT This chapter represents a revised version of a chapter by Dr. Norton J. Greenberger, Dr. Phillip P. Toskes, and Dr. Kurt J. Isselbacher that was in the previous editions of this textbook.
(Bibliography omitted in Palm version) Back to Table of Contents
PART TWELVE -DISORDERS OF THE IMMUNE SYSTEM, CONNECTIVE TISSUE, AND JOINTS SECTION 1 - DISORDERS OF THE IMMUNE SYSTEM 305. INTRODUCTION TO THE IMMUNE SYSTEM - Barton F. Haynes, Anthony S. Fauci DEFINITIONS ·Adaptive immune system -- recently evolved system of immune responses mediated by T and B lymphocytes. Immune responses by these cells are based on specific antigen recognition by clonotypic receptors that are products of genes that rearrange during development and throughout the life of the organism. Additional cells of the adaptive immune system include various types of antigen-presenting cells. · Antibody -- B cell-produced molecules encoded by genes that rearrange during B cell development consisting of immunoglobulin heavy and light chains that together form the central component of the B cell receptor for antigen. Antibody can exist as B cell surface antigen-recognition molecules or as secreted molecules in plasma and other body fluids. · Antigens -- foreign or self molecules that are recognized by the adaptive and innate immune systems resulting in innate immune cell triggering, T cell activation, and/or B cell antibody production. · Antimicrobial peptides -- small peptides25% of the adult population aged 15 to 49 is HIV-infected. In addition, among high-risk individuals (e.g., commercial sex workers, patients attendingSTDclinics) who live in urban areas of sub-Saharan Africa, seroprevalence is now>50% in many countries. The epidemic in Asian countries, particularly India and Thailand, has lagged temporally behind that in Africa; however, the number of new cases in this region is accelerating rapidly, and the magnitude of the epidemic is projected to exceed that of sub-Saharan Africa in the early part of the twenty-first century. The estimated number of cases in China is still relatively small; however, the potential exists for a major expansion of the epidemic in that nation of over 1 billion people. The major mode of transmission of HIV worldwide is unquestionably heterosexual sex; this is particularly true and has been so since the begining of the epidemic in developing countries, where the numbers of infected men and women are approximately equal. The epidemic in most developed countries was first introduced among homosexual men and, to a greater or lesser degree (depending on the individual country), amongIDUs. In this regard, the total numbers of AIDS cases in those countries still reflect a high proportion of cases among these high-risk groups. However, in most developed countries, including the United States (see below), there has been a gradual shift such that among new cases of AIDS, there is a greater prevalence among heterosexuals and IDUs than among homosexual men. AIDS IN THE UNITED STATES
AIDS has had and will continue to have an extraordinary public health impact in the United States. As of January 1, 2000, >724,600 cumulative cases of AIDS had been reported in adults and adolescents in the United States (Table 309-4) and approximately 425,000 AIDS-related deaths had been reported. It is the fifth leading cause of death among Americans aged 25 to 44 (Fig. 309-11), having dropped from first within the past few years. The death rate from AIDS declined 42% from 1996 to 1997 and 18% from 1997 to 1998. This trend is due to several factors including the improved prophylaxis and treatment of opportunistic infections, the growing experience among health professions in caring for HIV-infected individuals, improved access to health care, and the decrease in infections due to saturational effects and prevention efforts. However, the most influential factor clearly has been the increased use of potent antiretroviral drugs, generally administered in a combination of three or four agents, usually including a protease inhibitor (see below). When one looks at the totality of data collected from the beginning of the epidemic, approximately one-half of cases are among men who have had sex with men. However, over the past few years, the numbers of newly reported cases of AIDS among other groups, includingIDUs and heterosexuals, have surpassed the numbers of newly reported cases among men who have had sex with men. The proportion of new cases of AIDS per year attributed to heterosexual contact has increased dramatically over the past 15 years in the United States (Fig. 309-8). Women are increasingly affected; the proportion of AIDS cases in the United States reported among adult and adolescent females has increased from 99.5%. Most diagnostic laboratories use a commercial EIA kit that contains antigens from both HIV-1 and HIV-2 and thus are able to detect either. These kits use both natural and recombinant antigens and are continuously updated to increase their sensitivity to newly discovered species, such as group O viruses (Fig. 309-6). EIA tests are generally scored as positive (highly reactive), negative (nonreactive), or indeterminate (partially reactive). While the EIA is an extremely sensitive test, it is not optimal with regard to specificity. This is particularly true in studies of low-risk individuals, such as volunteer blood donors. In this latter population, only 10% of EIA-positive individuals are subsequently confirmed to have HIV infection. Among the factors associated with false-positive EIA tests are antibodies to class II antigens, autoantibodies, hepatic disease, recent influenza vaccination, and acute viral infections. For these reasons, anyone suspected of having HIV infection based upon a positive or inconclusive EIA result must have the result confirmed with a more specific assay. The most commonly used confirmatory test is the western blot (Fig. 309-23). This assay takes advantage of the fact that multiple HIV antigens of different, well-characterized molecular weights elicit the production of specific antibodies. These antigens can be separated on the basis of molecular weight, and antibodies to each component can be detected as discrete bands on the western blot. A negative western blot is one in which no bands are present at molecular weights corresponding to HIV gene products. In a patient with a positive or indeterminateEIA and a negative western blot, one can conclude with certainty that the EIA reactivity was a false positive. On the other hand, a western blot demonstrating antibodies to products of all three of the major genes of HIV (gag, pol, and env) is conclusive evidence of infection with HIV. Criteria established by the U.S. Food & Drug Administration (FDA) in 1993 for a positive western blot state that a result is considered positive if antibodies exist to two of the three HIV proteins: p24, gp41, and gp120/160. Using these criteria, approximately 10% of all blood donors deemed positive for HIV-1 infection lacked an antibody band to the pol gene product p31. Some 50% of these blood donors were subsequently found to be false positives. Thus, the absence of the p31 band should increase the suspicion that one may be dealing with a false-positive test result. In this setting it is prudent to obtain additional confirmation with an RNA-based test and/or a follow-up western blot. By definition, western blot patterns of reactivity that do not fall into the positive or negative categories are considered "indeterminate." There are two possible explanations for an indeterminate western blot result. The most likely explanation in a low-risk individual is that the patient being tested has antibodies that cross-react with one of the proteins of HIV. The most common patterns of cross-reactivity are antibodies that react with p24 and/or p55. The least likely explanation in this setting is that the individual is infected with HIV and is in the process of mounting a classic antibody response. In either instance, the western blot should be repeated in 1 month to determine whether or not the indeterminate pattern is a pattern in evolution. In addition, one may attempt to confirm a diagnosis of HIV infection with the p24 antigen capture assay or one of the tests for HIV RNA (discussed below). While the western blot is an excellent confirmatory test for HIV infection in patients with a positive or indeterminate EIA, it is a poor screening test. Among individuals with a negative EIA andPCR for HIV, 20 to 30% may
show one or more bands on western blot. While these bands are usually faint and represent cross-reactivity, their presence creates a situation in which other diagnostic modalities [such as DNA PCR, RNA PCR, the (b)DNA assay, or p24 antigen capture] must be employed to ensure that the bands do not indicate early HIV infection. A guideline for the use of these serologic tests in attempting to make a diagnosis of HIV infection is depicted in Fig. 309-24. In patients in whom HIV infection is suspected, the appropriate initial test is theEIA. If the result is negative, unless there is strong reason to suspect early HIV infection (as in a patient exposed within the previous 3 months), the diagnosis is ruled out and retesting should be performed only as clinically indicated. If the EIA is indeterminate or positive, the test should be repeated. If the repeat is negative on two occasions, one can assume that the initial positive reading was due to a technical error in the performance of the assay and that the patient is negative. If the repeat is indeterminate or positive, one should proceed to the HIV-1 western blot. If the western blot is positive, the diagnosis is HIV-1 infection. If the western blot is negative, the EIA can be assumed to have been a false positive for HIV-1 and the diagnosis of HIV-1 infection is ruled out. It would be prudent at this point to perform specific serologic testing for HIV-2 following the same type of algorithm. If the western blot for HIV-1 is indeterminate, it should be repeated in 4-6 weeks; in addition, one may proceed to a p24 antigen capture assay, HIV-1 RNA assay, or HIV-1 DNAPCR and specific serologic testing for HIV-2. If the p24 and HIV RNA assays are negative and there is no progression in the western blot, a diagnosis of HIV-1 is ruled out. If either the p24 or HIV-1 RNA assay is positive and/or the HIV-1 western blot shows progression, a tentative diagnosis of HIV-1 infection can be made and later confirmed with a follow-up western blot demonstrating a positive pattern. As mentioned above, a variety of laboratory tests are available for the direct detection of HIV or its components (Table 309-9;Fig. 309-25). These tests may be of considerable help in making a diagnosis of HIV infection when the western blot results are indeterminate. In addition, the tests detecting levels of HIV RNA can be used to determine prognosis and to assess the response to antiretroviral therapies. The simplest of the direct detection tests is the p24 antigen capture assay. This is anEIA-type assay in which the solid phase consists of antibodies to the p24 antigen of HIV. It detects the viral protein p24 in the blood of HIV-infected individuals where it exists either as free antigen or complexed to anti-p24 antibodies. Overall, approximately 30% of individuals with untreated HIV infection have detectable levels of free p24 antigen. This increases to about 50% when samples are treated with a weak acid to dissociate antigen-antibody complexes. Throughout the course of HIV infection, an equilibrium exists between p24 antigen and anti-p24 antibodies. During the first few weeks of infection, before an immune response develops, there is a brisk rise in p24 antigen levels (Fig. 309-22). After the development of anti-p24 antibodies, these levels decline. Late in the course of infection, when circulating levels of virus are high, p24 antigen levels also increase, particularly when detected by techniques involving dissociation of antigen-antibody complexes. This assay has its greatest use as a screening test for HIV infection in patients suspected of having the acute HIV syndrome, as high levels of p24 antigen are present prior to the development of antibodies. In addition, it is currently routinely used along with the HIV EIA assay to screen blood donors in the United States for evidence of HIV infection. Its utility as an assay is decreasing with the increased use of the reverse transcriptase PCR (RT-PCR) and
bDNA technique for direct detection of HIV RNA. The ability to measure and monitor levels of HIV RNA in the plasma of patients with HIV infection has been of extraordinary value in furthering our understanding of the pathogenesis of HIV infection and in providing a diagnostic tool in settings where measurements of anti-HIV antibodies may be misleading, such as in acute infection and neonatal infection. Two assays are predominantly used for this purpose. They are theRT-PCR(Amplicor) and the bDNA (Quantiplex). It should be pointed out that the only test approved by theFDA at this time for the measurement of HIV RNA levels is the RT-PCR test. While this approval is limited to the use of the test for determining prognosis, it is the general consensus that this test as well as the bDNA test are also of value for monitoring the effects of therapy and in making a diagnosis of HIV infection. In addition to these two commercially available tests, the DNA PCR is also employed by research laboratories for making a diagnosis of HIV infection by amplifying HIV proviral DNA from peripheral blood mononuclear cells. The commercially available RNA detection tests have a sensitivity of 40 to 50 copies of HIV RNA per milliliter of plasma, while the DNA PCR tests can detect proviral DNA at a frequency of one copy per 10,000 to 100,000 cells. Thus, these tests are extremely sensitive. One frequent consequence of a high degree of sensitivity is some loss of specificity, and false-positive results have been reported with each of these techniques. For this reason, a positiveEIA with a confirmatory western blot remains the "gold standard" for a diagnosis of HIV infection, and the interpretation of other test results must be done with this in mind. In theRT-PCRtechnique, following DNAase treatment, a cDNA copy is made of all RNA species present in plasma. Insofar as HIV is an RNA virus, this will result in the production of DNA copies of the HIV genome in amounts proportional to the amount of HIV RNA present in plasma. This proviral DNA is then amplified and characterized using standardPCRtechniques, employing primer pairs that can distinguish genomic cDNA from messenger cDNA. The bDNA assay involves the use of a solid-phase nucleic acid capture system and signal amplification through successive nucleic acid hybridizations to detect small quantities of HIV RNA. Both tests can achieve a tenfold increase in sensitivity to 40 to 50 copies of HIV RNA per milliliter with a preconcentration step in which plasma undergoes ultracentrifugation to pellet the viral particles. In addition to being a diagnostic and prognostic tool, RT-PCR is also useful for amplifying defined areas of the HIV genome for sequence analysis and has become an important technique for studies of sequence diversity and microbial resistance to antiretroviral agents. In patients with a positive or indeterminateEIA test and an indeterminate western blot, and in patients in whom serologic testing may be unreliable (such as patients with hypogammaglobulinemia or advanced HIV disease), these tests provide valuable tools for making a diagnosis of HIV infection. They should only be used for diagnosis when standard serologic testing has failed to provide a definitive result. LABORATORY MONITORING OF PATIENTS WITH HIV INFECTION The epidemic of HIV infection and AIDS has provided the clinician with new challenges for integrating clinical and laboratory data to effect optimal patient management. The close relationship between clinical manifestations of HIV infection and CD4+ T cell count has made measurement of the latter a routine part of the evaluation of HIV-infected individuals. Determinations of CD4+ T cell counts and measurements of the levels of
HIV RNA in serum or plasma provide a powerful set of tools for determining prognosis and monitoring response to therapy. While the CD4+ T cell count provides information on the current immunologic status of the patient, the HIV RNA level predicts what will happen to the CD4+ T cell count in the near future, and hence provides an important piece of prognostic information. CD4+ T Cell Counts The CD4+ T cell count is the laboratory test generally accepted as the best indicator of the immediate state of immunologic competence of the patient with HIV infection. This measurement, which is the product of the percent of CD4+ T cells (determined by flow cytometry) and the total lymphocyte count [determined by the white blood cell count (WBC) and the differential count] has been shown to correlate very well with the level of immunologic competence. Patients with CD4+ T cell counts 1.4 mg/dL)], hypertension, nephrotic syndrome (24-h urine protein excretion>2.6 g), anemia [hemoglobin< 124 g/L(2000/uL with more than 75% polymorphonuclear leukocytes is highly characteristic of inflammatory arthritis, although not diagnostic ofRA. Total hemolytic complement, C3, and C4 are markedly diminished in synovial fluid
relative to total protein concentration as a result of activation of the classic complement pathway by locally produced immune complexes. RADIOGRAPHIC EVALUATION Early in the disease, roentgenograms of the affected joints are usually not helpful in establishing a diagnosis. They reveal only that which is apparent from physical examination, namely, evidence of soft tissue swelling and joint effusion. As the disease progresses, abnormalities become more pronounced, but none of the radiographic findings is diagnostic ofRA. The diagnosis, however, is supported by a characteristic pattern of abnormalities, including the tendency toward symmetric involvement. Juxtaarticular osteopenia may become apparent within weeks of onset. Loss of articular cartilage and bone erosions develop after months of sustained activity. The primary value of radiography is to determine the extent of cartilage destruction and bone erosion produced by the disease, particularly when one is monitoring the impact of therapy with disease-modifying drugs or surgical intervention. Other means of imaging bones and joints, including99m Tc bisphosphonate bone scanning and magnetic resonance imaging, may be capable of detecting early inflammatory changes that are not apparent from standard radiography but are rarely necessary in the routine evaluation of patients with RA. CLINICAL COURSE AND PROGNOSIS The course ofRA is quite variable and difficult to predict in an individual patient. Most patients experience persistent but fluctuating disease activity, accompanied by a variable degree of joint abnormalities and functional impairment. After 10 to 12 years, fewer than 20% of patients will have no evidence of disability or joint abnormalities. Within 10 years, approximately 50% of patients will have work disability. A number of features are correlated with a greater likelihood of developing joint abnormalities or disabilities. These include the presence of more than 20 inflamed joints, a markedly elevated erythrocyte sedimentation rate, radiographic evidence of bone erosions, the presence of rheumatoid nodules, high titers of serum rheumatoid factor, the presence of functional disability, persistent inflammation, advanced age at onset, the presence of comorbid conditions, low socioeconomic status or educational level, or the presence of HLA-DRB1*0401 or -DRB*0404. The presence of one or more of these implies the presence of more aggressive disease with a greater likelihood of developing progressive joint abnormalities and disability. Persistent elevation of the erythrocyte sedimentation rate, disability, and pain on longitudinal follow-up are good predictors of work disability. Patients who lack these features have more indolent disease with a slower progression to joint abnormalities and disability. The pattern of disease onset does not appear to predict the development of disabilities. Approximately 15% of patients with RA will have a short-lived inflammatory process that remits without major disability. These individuals tend to lack the aforementioned features associated with more aggressive disease. Several features of patients withRAappear to have prognostic significance. Remissions of disease activity are most likely to occur during the first year. White females tend to have more persistent synovitis and more progressively erosive disease than males. Persons who present with high titers of rheumatoid factor, C-reactive protein, and
haptoglobin also have a worse prognosis, as do individuals with subcutaneous nodules or radiographic evidence of erosions at the time of initial evaluation. Sustained disease activity of more than 1 year's duration portends a poor outcome, and persistent elevation of acute-phase reactants appears to correlate strongly with radiographic progression. A large proportion of inflamed joints manifest erosions within 2 years, whereas the subsequent course of erosions is highly variable; however, in general, radiographic damage appears to progress at a constant rate in patients with RA. Foot joints are affected more frequently than hand joints. Despite the decrease in the rate of progressive joint damage with time, functional disability, which develops early in the course of the disease, continues to worsen at the same rate, although the most rapid rate of functional loss occurs within the first 2 years of disease. The median life expectancy of persons with RAis shortened by 3 to 7 years. Of the 2.5-fold increase in mortality rate, RA itself is a contributing feature in 15 to 30%. The increased mortality rate seems to be limited to patients with more severe articular disease and can be attributed largely to infection and gastrointestinal bleeding. Drug therapy may also play a role in the increased mortality rate seen in these individuals. Factors correlated with early death include disability, disease duration or severity, glucocorticoid use, age at onset, and low socioeconomic or educational status. DIAGNOSIS The mean delay from disease onset to diagnosis is 9 months. This is often related to the nonspecific nature of initial symptoms. The diagnosis ofRA is easily made in persons with typical established disease. In a majority of patients, the disease assumes its characteristic clinical features within 1 to 2 years of onset. The typical picture of bilateral symmetric inflammatory polyarthritis involving small and large joints in both the upper and lower extremities with sparing of the axial skeleton except the cervical spine suggests the diagnosis. Constitutional features indicative of the inflammatory nature of the disease, such as morning stiffness, support the diagnosis. Demonstration of subcutaneous nodules is a helpful diagnostic feature. Additionally, the presence of rheumatoid factor, inflammatory synovial fluid with increased numbers of polymorphonuclear leukocytes, and radiographic findings of juxtaarticular bone demineralization and erosions of the affected joints substantiate the diagnosis. The diagnosis is somewhat more difficult early in the course when only constitutional symptoms or intermittent arthralgias or arthritis in an asymmetric distribution may be present. A period of observation may be necessary before the diagnosis can be established. A definitive diagnosis ofRAdepends predominantly on characteristic clinical features and the exclusion of other inflammatory processes. The isolated finding of a positive test for rheumatoid factor or an elevated erythrocyte sedimentation rate, especially in an older person with joint pains, should not itself be used as evidence of RA. In 1987, the American College of Rheumatology developed revised criteria for the classification ofRA(Table 312-1). These criteria demonstrate a sensitivity of 91 to 94% and a specificity of 89% when used to classify patients with RA compared with control subjects with rheumatic diseases other than RA. Although these criteria were developed as a means of disease classification for epidemiologic purposes, they can be useful as
guidelines for establishing the diagnosis. Failure to meet these criteria, however, especially during the early stages of the disease, does not exclude the diagnosis. Moreover, in patients with early arthritis, the criteria do not discriminate effectively between patients who subsequently develop persistent, disabling, or erosive disease and those who do not. TREATMENT General Principles The goals of therapy ofRA are (1) relief of pain, (2) reduction of inflammation, (3) protection of articular structures, (4) maintenance of function, and (5) control of systemic involvement. Since the etiology of RA is unknown, the pathogenesis is not completely delineated, and the mechanisms of action of many of the therapeutic agents employed are uncertain, therapy remains largely empirical. None of the therapeutic interventions is curative, and therefore all must be viewed as palliative, aimed at relieving the signs and symptoms of the disease. The various therapies employed are directed at nonspecific suppression of the inflammatory or immunologic process in the hope of ameliorating symptoms and preventing progressive damage to articular structures. Management of patients withRAinvolves an interdisciplinary approach, which attempts to deal with the various problems that these individuals encounter with functional as well as psychosocial interactions. A variety of physical therapy modalities may be useful in decreasing the symptoms of RA. Rest ameliorates symptoms and can be an important component of the total therapeutic program. In addition, splinting to reduce unwanted motion of inflamed joints may be useful. Exercise directed at maintaining muscle strength and joint mobility without exacerbating joint inflammation is also an important aspect of the therapeutic regimen. A variety of orthotic and assistive devices can be helpful in supporting and aligning deformed joints to reduce pain and improve function. Education of the patient and family is an important component of the therapeutic plan to help those involved become aware of the potential impact of the disease and make appropriate accommodations in life-style to maximize satisfaction and minimize stress on joints. Medical management ofRAinvolves five general approaches. The first is the use of aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) and simple analgesics to control the symptoms and signs of the local inflammatory process. These agents are rapidly effective at mitigating signs and symptoms, but they appear to exert minimal effect on the progression of the disease. Recently, specific inhibitors of the isoform of cyclooxygenase (Cox) that is upregulated at inflammatory sites (Cox-2) have been developed. Cox-2-specific inhibitors (CSIs) have been shown to be as effective as classic NSAIDs, which inhibit both isoforms of Cox, but to cause significantly less gastroduodenal ulceration. The second line of therapy involves use of low-dose oral glucocorticoids. Although low-dose glucocorticoids have been widely used to suppress signs and symptoms of inflammation, recent evidence suggests that they may also retard the development and progression of bone erosions. Intraarticular glucocorticoids can often provide transient symptomatic relief when systemic medical therapy has failed to resolve inflammation. The third line of agents includes a variety of agents that have been classified as the disease-modifying or slow-acting antirheumatic drugs. These agents appear to have the capacity to decrease elevated levels of acute-phase
reactants in treated patients and, therefore, are thought to modify the inflammatory component of RA and thus its destructive capacity. Recently, combinations of disease-modifying antirheumatic drugs (DMARDs) have shown promise in controlling the signs and symptoms of RA. A fourth group of agents are theTNF-aneutralizing agents, which have been shown to have a major impact on the signs and symptoms of RA. A fifth group of agents are the immunosuppressive and cytotoxic drugs that have been shown to ameliorate the disease process in some patients. Additional approaches have been employed in an attempt to control the signs and symptoms of RA. Substituting omega-3 fatty acids such as eicosapentaenoic acid found in certain fish oils for dietary omega-6 essential fatty acids found in meat has also been shown to provide symptomatic improvement in patients with RA. A variety of nontraditional approaches have also been claimed to be effective in treating RA, including diets, plant and animal extracts, vaccines, hormones, and topical preparations of various sorts. Many of these are costly, and none has been shown to be effective. However, belief in their efficacy ensures their continued use by some patients. Drugs Nonsteroidal Anti-Inflammatory Drugs Besides aspirin, manyNSAIDsare available to treatRA. As a result of the capacity of these agents to block the activity of theCoxenzymes and therefore the production of prostaglandins, prostacyclin, and thromboxanes, they have analgesic, anti-inflammatory, and antipyretic properties. In addition, the agents may exert other anti-inflammatory effects. These agents are all associated with a wide spectrum of toxic side effects. Some, such as gastric irritation, azotemia, platelet dysfunction, and exacerbation of allergic rhinitis and asthma, are related to the inhibition of cyclooxygenase activity, whereas a variety of others, such as rash, liver function abnormalities, and bone marrow depression, may not be. None of the NSAIDs has been shown to be more effective than aspirin in the treatment of RA. However, these nonaspirin drugs are associated with a lower incidence of gastrointestinal intolerance. None of the newer NSAIDs appears to show significant therapeutic advantages over the other available agents. In addition, there is no consistent advantage of any of these newer agents over the others with respect to the incidence or severity of toxic manifestations. Recent evidence indicates that two separate enzymes, Cox-1 and -2, are responsible for the initial metabolism of arachidonic acid into various inflammatory mediators. The former is constitutively present in many cells and tissues, including the stomach and the platelet, whereas the latter is specifically induced in response to inflammatory stimuli. Inhibition of Cox-2 accounts for the anti-inflammatory effects of NSAIDs, whereas inhibition of Cox-1 induces much of the mechanism-based toxicity. As the currently available NSAIDs inhibit both enzymes, therapeutic benefit and toxicity are intertwined.CSIshave now been approved for the treatment of RA. Clinical trials have shown that CSIs suppress the signs and symptoms of RA as effectively as classic Cox-nonspecific NSAIDs but are associated with a significantly reduced incidence of gastroduodenal ulceration. This suggests that CSIs might be considered instead of classic Cox-nonspecific NSAIDs, especially in persons with increased risk of NSAID-induced major upper gastrointestinal side effects, including persons over 65, those with a history of peptic ulcer disease, persons receiving glucocorticoids or anticoagulants, or those requiring high doses of NSAIDs.
Disease-Modifying Antirheumatic Drugs Clinical experience has delineated a number of agents that appear to have the capacity to alter the course ofRA. This group of agents includes methotrexate, gold compounds, D-penicillamine, the antimalarials, and sulfasalazine. Despite having no chemical or pharmacologic similarities, in practice these agents share a number of characteristics. They exert minimal direct nonspecific anti-inflammatory or analgesic effects, and thereforeNSAIDsmust be continued during their administration, except in a few cases when true remissions are induced with them. The appearance of benefit fromDMARDtherapy is usually delayed for weeks or months. As many as two-thirds of patients develop some clinical improvement as a result of therapy with any of these agents, although the induction of true remissions is unusual. In addition to clinical improvement, there is frequently an improvement in serologic evidence of disease activity, and titers of rheumatoid factor and C-reactive protein and the erythrocyte sedimentation rate frequently decline. Moreover, emerging evidence suggests that DMARDs actually retard the development of bone erosions or facilitate their healing. Furthermore, developing evidence suggests that early aggressive treatment with DMARDs may be effective at slowing the appearance of bone erosions. WhichDMARDshould be the drug of first choice remains controversial, and trials have failed to demonstrate a consistent advantage of one over the other. Despite this, methotrexate has emerged as the DMARD of choice because of its relatively rapid onset of action, its capacity to effect sustained improvement with ongoing therapy, and the high level of patient retention on therapy. Each of the DMARDs is associated with considerable toxicity, and therefore careful patient monitoring is necessary. Toxicity of the various agents also becomes important in determining the drug of first choice. Of note, failure to respond or development of toxicity to one DMARD does not preclude responsiveness to another. Thus, a similar percentage ofRApatients who have failed to respond to one DMARD will respond to another when it is given as the second disease-modifying drug. No characteristic features of patients have emerged that predict responsiveness to aDMARD. Moreover, the indications for the initiation of therapy with one of these agents are not well defined, although recently the trend has been to begin DMARD therapy early in the course of the disease, and data have begun to emerge to support the conclusion that this approach may slow the development of bone erosions, although this remains controversial. The folic acid antagonist methotrexate, given in an intermittent low dose (7.5 to 30 mg once weekly), is currently a frequently utilizedDMARD. Most rheumatologists recommend use of methotrexate as the initial DMARD, especially in individuals with evidence of aggressiveRA. Recent trials have documented the efficacy of methotrexate and have indicated that its onset of action is more rapid than other DMARDs, and patients tend to remain on therapy with methotrexate longer than they remain on other DMARDs because of better clinical responses and less toxicity. Long-term trials have indicated that methotrexate does not induce remission but rather suppresses symptoms while it is being administered. Maximal improvement is observed after 6 months of therapy, with little additional improvement thereafter. Major toxicity includes gastrointestinal upset, oral ulceration, and liver function abnormalities that appear to be dose-related and reversible and hepatic fibrosis that can be quite insidious, requiring liver biopsy for detection in its early stages. Drug-induced pneumonitis has also been
reported. Liver biopsy is recommended for individuals with persistent or repetitive liver function abnormalities. Concurrent administration of folic acid or folinic acid may diminish the frequency of some side effects without diminishing effectiveness. Glucocorticoid Therapy Systemic glucocorticoid therapy can provide effective symptomatic therapy in patients withRA. Low-dose (50% of patients and include epigastric fullness, burning pain in the epigastric or retrosternal regions, and regurgitation of gastric contents. These symptoms, most noticeable when the patient is lying flat or bending over, are due to the reduced tone of the gastroesophageal sphincter and to dilatation of the distal esophagus. Peptic esophagitis frequently occurs and may lead to strictures and narrowing of the lower esophagus. However, it seldom results in bleeding. Barrett's metaplasia may develop, but transition to adenocarcinoma is uncommon. Dysphagia, particularly of solid foods, may occur independent of other esophageal symptoms and is caused by loss of esophageal motility due to neuromuscular dysfunction. Manometry or cineradiography reveals decreased amplitude or disappearance of peristaltic waves in the lower two-thirds of the esophagus. Raynaud's phenomenon in the absence of a connective tissue disease is also associated with esophageal dysmotility. Later in the course of the illness, dilatation and atony of the lower portion of the esophagus as well as reflux are seen. With gastric involvement, barium studies show dilatation, atony, and delayed gastric emptying. Patients may complain of early satiety. Gastric outlet obstruction can also occur. Hypomotility of the small intestine produces symptoms of bloating and abdominal pain and may suggest an intestinal obstruction or paralytic ileus (pseudoobstruction). Malabsorption syndrome with weight loss, diarrhea, and anemia is due to bacterial overgrowth in the atonic intestine or possibly to obliteration of lymphatics by fibrosis. Roentgenographic features of the second and third portions of the duodenum and of the jejunum include dilatation, loss of the usual feathery pattern, and delayed disappearance of barium. Pneumatosis intestinalis occasionally occurs and appears as radiolucent cysts or linear streaks within the wall of the small intestine. Benign pneumoperitoneum may result from the rupture of these cysts. Involvement of the large intestine may cause chronic constipation and fecal impaction with episodes of bowel obstruction. A segment of atonic bowel may act as a fulcrum for intussuception to occur. Barium studies of the large intestine may show dilatation, atony, and large-mouth diverticula. Laxity of the anal sphincter may cause incontinence or rarely anal prolapse. Some patients may have gastrointestinal features ofSSc with little or no cutaneous or other organ involvement, referred to as SSc sine scleroderma. Vascular ectasia may develop in the stomach and intestine and can be the source of gastrointestinal bleeding. These dilated submucosal capillaries in the stomach appear on endoscopy as broad stripes -- hence the term "watermelon stomach." Pulmonary Features Pulmonary involvement occurs in at least two-thirds ofSScpatients and is now the leading cause of death in SSc, replacing renal disease,
which can usually be treated effectively. The most common symptom is exertional dyspnea, often accompanied by a dry, nonproductive cough. Bilateral basilar rales may be present. In the majority of patients, symptoms usually correlate with radiologic evidence of pulmonary fibrosis and with restrictive lung disease on pulmonary function tests. Pulmonary function tests are frequently abnormal and show a reduction in vital capacity and decreased lung compliance. Impairment of gas exchange is reflected by a low diffusing capacity and low PO2 with exercise. These abnormalities may be present even when the chest radiograph is normal. Chest film may show a pattern of linear densities, mottling, and honeycombing involving most prominently the lower two-thirds of the lung. Early interstitial pulmonary disease can be detected by high-resolution computed tomography (HRCT) and bronchoalveolar lavage (BAL). Active inflammatory alveolitis gives a "ground glass" appearance on HRCT. The recovery by BAL of increased numbers of cells, mostly alveolar macrophages accompanied by neutrophils or eosinophils, is evidence for alveolitis. Both interstitial fibrosis and vascular lesions are found in the lungs of patients withSSc. Interstitial pulmonary fibrosis may be the predominant lesion in patients with diffuse or limited cutaneous SSc. Patients with diffuse cutaneous involvement who have antitopoisomerase 1 antibodies are particularly at risk of developing severe pulmonary fibrosis. In the absence of significant interstitial fibrosis, a severe form of pulmonary arterial hypertension may develop after many years of disease in a subset of patients with limited cutaneous SSc. Fewer than 10% of patients will develop this complication, which is caused by narrowing and obliteration of pulmonary arteries and arterioles by intimal fibrosis and medial hypertrophy. Pulmonary hypertension is manifested initially by exertional dyspnea and eventually by the appearance of right-sided heart failure. Pulmonary artery pressure can be measured noninvasively by two-dimensional echocardiography. The prognosis is extremely poor with the development of pulmonary hypertension; the mean duration of survival is approximately 2 years. A less common pulmonary problem is aspiration pneumonia resulting from gastric reflux due to lower esophageal atony. Restriction of chest movement caused by extensive fibrotic skin involvement of the thorax rarely occurs. Superimposed bacterial or viral infection can be a serious complication in patients with pulmonary fibrosis. An increased frequency of alveolar cell and bronchogenic carcinoma is seen in patients with pulmonary fibrosis. Cardiac Features Primary cardiac involvement inSScincludes pericarditis with or without effusions, heart failure, and varying degrees of heart block or arrhythmias. The majority of patients with diffuse cutaneous SSc have cardiac abnormalities. Cardiomyopathy attributable to myocardial fibrosis appears in 1000/uL in most patients. The histologic findings on biopsy of skin, fascia, and superficial muscle are similar to those found in eosinophilic fasciitis. The clinical features of EMS may persist after L-tryptophan has been discontinued. EMS may run a chronic course, and response to therapy has been variable. Treatment has included glucocorticoids, antimalarial drugs, immunosuppressive drugs, and plasmapheresis. Prednisone was beneficial during the acute inflammatory phase of the disease in the majority of patients and resulted in resolution of pulmonary infiltrates, peripheral edema, and eosinophilia. In the later phase of the illness, no treatment was found to be of particular value. The pathogenesis of this disease is not known. A follow-up of patients 2 years after their onset of illness showed that most symptoms and physical findings had resolved or improved except for cognitive dysfunction, which became worse in approximately one-third of the patients, and peripheral neuropathy, which remained unchanged (Chap. 382). (Bibliography omitted in Palm version) Back to Table of Contents
314. SJOGREN'S SYNDROME - Haralampos M. Moutsopoulos DEFINITION Sjogren's syndrome is a chronic, slowly progressive autoimmune disease characterized by lymphocytic infiltration of the exocrine glands resulting in xerostomia and dry eyes. Approximately one-third of patients present with systemic manifestations. A small but significant number of the patients may develop malignant lymphoma. The disease can be seen alone (primary Sjogren's syndrome) or in association with other autoimmune rheumatic diseases (secondary Sjogren's syndrome) (Table 314-1). INCIDENCE AND PREVALENCE The disease affects predominantly middle-aged women (female-to-male ratio 9:1), although it occurs in all ages, including childhood. The prevalence of primary Sjogren's syndrome is approximately 0.5 to 1.0%. In addition, 30% of patients with autoimmune rheumatic diseases suffer from secondary Sjogren's syndrome. PATHOGENESIS Sjogren's syndrome is characterized by lymphocytic infiltration of the exocrine glands and B lymphocyte hyperreactivity, as illustrated by circulating autoantibodies. The latter is accompanied by an oligomonoclonal B cell process, which is characterized by serum and urine monoclonal light chains and cryoprecipitable monoclonal immunoglobulins. Sera of patients with Sjogren's syndrome often contain a number of autoantibodies directed against non-organ-specific antigens such as immunoglobulins (rheumatoid factors) and extractable nuclear and cytoplasmic antigens (Ro/SS-A, La/SS-B). Ro/SS-A autoantigen consists of three polypeptide chains (52, 54, and 60 kDa) in conjunction with RNAs, whereas the 48-kDa La/SS-B protein is bound to RNA III polymerase transcripts. The presence of autoantibodies to Ro/SS-A and La/SS-B antigens in Sjogren's syndrome is associated with earlier disease onset, longer disease duration, salivary gland enlargement, severity of lymphocytic infiltration of minor salivary glands, and certain extraglandular manifestations such as lymphadenopathy, purpura, and vasculitis. Antibodies toa-fordin (120 kDa), a salivary gland-specific protein, have recently been found in sera of patients with Sjogren's syndrome but not in sera of patients with other connective tissue diseases. Phenotypic and functional studies have shown that the predominant cell infiltrating the affected exocrine glands is the helper/inducer T cell with characteristics of memory cells. Both B and T infiltrating lymphocytes are activated, as illustrated by production of immunoglobulins with autoantibody activity, spontaneous release of interleukin 2, and expression on the T cell surface of activation markers such as class II HLA as well as costimulatory molecules and lymphocyte function-associated antigen 1. Macrophages and natural killer cells are rarely detected in infiltrates, while epithelial cells of the affected glands inappropriately express class II molecules and possess messages for c-myc protooncogene and proinflammatory cytokines. All these phenomena suggest that the epithelial cell of the exocrine glands in Sjogren's syndrome may act as an antigen-presenting cell. In contrast to infiltrating lymphocytes, these cells undergo
apoptotic death, resulting in exocrine gland dysfunction. Immunogenetic studies have demonstrated that HLA-B8, -DR3, and -DRw52 are prevalent in patients with primary Sjogren's syndrome as compared with the normal control population. Molecular analysis of HLA class II genes has revealed that patients with Sjogren's syndrome, regardless of their ethnic origin, are highly associated with the HLA DQA1*0501 allele. CLINICAL MANIFESTATIONS The majority of the patients with Sjogren's syndrome have symptoms related to diminished lacrimal and salivary gland function. In most patients, the primary syndrome runs a slow and benign course. The initial manifestations can be mucosal dryness or nonspecific, and 8 to 10 years elapse from the initial symptoms to full-blown development of the disease. The principal oral symptom of Sjogren's syndrome is dryness (xerostomia). Patients complain of difficulty in swallowing dry food, inability to speak continuously, a burning sensation, increase in dental caries, and problems in wearing complete dentures. Physical examination shows a dry, erythematous, sticky oral mucosa. There is atrophy of the filiform papillae on the dorsum of the tongue, and saliva from the major glands is either not expressible or is cloudy. Enlargement of the parotid or other major salivary glands occurs in two-thirds of patients with primary Sjogren's syndrome but is uncommon in those with the secondary syndrome. Diagnostic tests include sialometry, sialography, and scintigraphy. The labial minor salivary gland biopsy permits histopathologic confirmation of the focal lymphocytic infiltrates. Ocular involvement is the other major manifestation of Sjogren's syndrome. Patients usually complain of dry eyes, with a sandy or gritty feeling under the eyelids. Other symptoms include burning, accumulation of thick strands at the inner canthi, decreased tearing, redness, itching, eye fatigue, and increased photosensitivity. These symptoms are attributed to the destruction of corneal and bulbar conjunctival epithelium, defined as keratoconjunctivitis sicca. Diagnostic evaluation of keratoconjunctivitis sicca includes measurement of tear flow by Schirmer's I test and tear composition as assessed by the tear breakup time or tear lysozyme content. Slit-lamp examination of the cornea and conjunctiva after rose Bengal staining reveals punctate corneal ulcerations and attached filaments of corneal epithelium. Involvement of other exocrine glands occurs less frequently and includes a decrease in mucous gland secretions of the upper and lower respiratory tree, resulting in dry nose, throat, and trachea (xerotrachea), and diminished secretion of the exocrine glands of the gastrointestinal tract, leading to esophageal mucosal atrophy, atrophic gastritis, and subclinical pancreatitis. Dyspareunia due to dryness of the external genitalia and dry skin also may occur. Extraglandular (systemic) manifestations are seen in one-third of patients with Sjogren's syndrome (Table 314-2), while they are very rare in patients with Sjogren's syndrome associated with rheumatoid arthritis. These patients complain more often of easy fatigability, low-grade fever, Raynaud's phenomenon, myalgias, and arthralgias. Most
patients with primary Sjogren's syndrome experience at least one episode of nonerosive arthritis during the course of their disease. Manifestations of pulmonary involvement are frequent but rarely important clinically. Dry cough is the major manifestation that is attributed to small airway disease. Renal involvement includes interstitial nephritis, clinically manifested by hypostenuria and renal tubular dysfunction with or without acidosis. Untreated acidosis may lead to nephrocalcinosis. Glomerulonephritis is a rare finding that occurs in patients with systemic vasculitis, cryoglobulinemia, or systemic lupus erythematosus overlapping with Sjogren's syndrome. Vasculitis affects small and medium-sized vessels. The most common clinical features are purpura, recurrent urticaria, skin ulcerations, glomerulonephritis, and mononeuritis multiplex. Sensorineural hearing loss was found in one-half of patients with Sjogren's syndrome and correlated with the presence of anticardiolipin antibodies. It has been suggested that primary Sjogren's syndrome with vasculitis also may present with multifocal, recurrent, and progressive nervous system disease, such as hemiparesis, transverse myelopathy, hemisensory deficits, seizures, and movement disorders. Aseptic meningitis and multiple sclerosis also have been reported in these patients. Lymphoma is a well-known manifestation of Sjogren's syndrome that usually presents later in the illness. Persistent parotid gland enlargement, lymphadenopathy, cutaneous vasculitis, peripheral neuropathy, lymphopenia, and cryoglobulinemia are manifestations suggesting the development of lymphoma. Most lymphomas are extranodal, marginal zone B cell, and low grade. Salivary glands are the most common site of involvement. Routine laboratory tests reveal mild normochromic, normocytic anemia. An elevated erythrocyte sedimentation rate is found in approximately 70% of patients. DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS A European multicenter study has developed diagnostic criteria of Sjogren's syndrome (Table 314-3), which have been validated and present high specificity and sensitivity. A diagnostic algorithm is depicted inFig. 314-1. The differential diagnosis of Sjogren's syndrome includes other conditions that may cause dry mouth or eyes or parotid salivary gland enlargement (Table 314-4). Infections with HIV and hepatitis C virus (Chap. 309) and sarcoidosis (Chap. 318) appear to produce a clinical picture indistinguishable from that of Sjogren's syndrome (Table 314-5). TREATMENT Sjogren's syndrome remains fundamentally an incurable disease. Hence treatment is aimed at symptomatic relief and limiting the damaging local effects of chronic xerostomia and keratoconjunctivitis sicca by substitution of the missing secretions. The sicca complex is treated with fluid replacement supplied as often as necessary. To replace deficient tears, there are several readily available ophthalmic preparations (Tearisol; Liquifilm; 0.5% methylcellulose; Hypo Tears). It may be necessary for
severely affected patients to use these preparations as often as every 30 min. If corneal ulceration is present, eye patching and boric acid ointments are recommended. Certain drugs that may increase lacrimal and salivary hypofunction such as diuretics, antihypertensive drugs, and antidepressants should be avoided. Propionic acid gels may be used to treat vaginal dryness. Pilocarpine (5 mg thrice daily) given orally appears to improve sicca manifestations. Hydroxychloroquine (200 mg/day) is helpful for arthralgias. Glucocorticoids (1 mg/kg per day) or other immunosuppressive agents (i.e., cyclophosphamide) are indicated for the treatment of extraglandular manifestations, particularly when renal or severe pulmonary involvement and systemic vasculitis have been documented. (Bibliography omitted in Palm version) Back to Table of Contents
315. ANKYLOSING SPONDYLITIS, REACTIVE ARTHRITIS, AND UNDIFFERENTIATED SPONDYLOARTHROPATHY - Joel D. Taurog, Peter E. Lipsky The spondyloarthropathies are a group of disorders that share certain clinical features and an association with the HLA-B27 allele. These disorders include ankylosing spondylitis, Reiter's syndrome, reactive arthritis, psoriatic arthritis and spondylitis, enteropathic arthritis and spondylitis, juvenile-onset spondyloarthropathy, and undifferentiated spondyloarthropathy. The similarities in clinical manifestations and genetic predisposition suggest that these disorders share pathogenic mechanisms. Specific definitions and diagnostic criteria for the individual conditions will be provided in subsequent sections of this chapter. ANKYLOSING SPONDYLITIS Ankylosing spondylitis (AS) is an inflammatory disorder of unknown cause that primarily affects the axial skeleton; peripheral joints and extraarticular structures may also be involved. The disease usually begins in the second or third decade; the prevalence in men is approximately three times that in women. It is considered the prototype of the spondyloarthropathies. Older names include Marie-Strumpell disease or Bechterew's disease. EPIDEMIOLOGY ASshows a striking correlation with the histocompatibility antigen HLA-B27 and occurs worldwide roughly in proportion to the prevalence of this antigen (Chap. 306). In North American Caucasians, the general prevalence of B27 is 7%, whereas over 90% of patients with AS have inherited this antigen. The association with B27 is independent of disease severity. In population surveys, 1 to 6% of adults inheriting B27 have been found to haveAS. In contrast, in families of patients with AS, the prevalence is 10 to 30% among adult first-degree relatives inheriting B27. The concordance rate in identical twins is estimated to exceed 65%. It is currently believed that susceptibility to AS is determined almost entirely by genetic factors, with as yet unidentified allelic genes in addition to B27 comprising about two-thirds of the genetic component and B27 itself comprising about one-third. AS is strongly associated with inflammatory bowel disease (IBD), including both ulcerative colitis and Crohn's disease. IBD is a risk factor for AS independent of HLA-B27, although 50 to 75% of patients with both AS and IBD are B27 positive.*See also Chap. 287. PATHOLOGY The enthesis, the site of ligamentous attachment to bone, is thought to be the primary site of pathology inAS, particularly in the lesions around the pelvis and spine. Enthesitis is associated with prominent edema of the adjacent bone marrow and is often characterized by erosive lesions that eventually undergo ossification. Sacroiliitis is usually one of the earliest manifestations ofAS, with features of both
enthesitis and synovitis. The early lesions consist of subchondral granulation tissue containing lymphocytes, plasma cells, mast cells, macrophages, and chondrocytes; infiltrates of lymphocytes and macrophages in ligamentous and periosteal zones; and subchondral bone marrow edema. Synovitis follows and may progress to pannus formation. Islands of new bone formation can be found within the inflammatory infiltrates. Usually, the thinner iliac cartilage is eroded before the thicker sacral cartilage. The irregularly eroded, sclerotic margins of the joint are gradually replaced by fibrocartilage regeneration and then by ossification. Ultimately, the joint may be totally obliterated. This progression is evident by imaging techniques (see below). In the spine, early in the process there is inflammatory granulation tissue at the junction of the annulus fibrosus of the disk cartilage and the margin of vertebral bone. The outer annular fibers are eroded and eventually replaced by bone, forming the beginning of a bony excrescence called a syndesmophyte, which then grows by continued enchondral ossification, ultimately bridging the adjacent vertebral bodies. Ascending progression of this process leads to the "bamboo spine" observed radiographically. Other lesions in the spine include diffuse osteoporosis, erosion of vertebral bodies at the disk margin, "squaring" of vertebrae, and inflammation and destruction of the disk-bone border. Inflammatory arthritis of the apophyseal joints is common, with erosion of cartilage by pannus, often followed by bony ankylosis. Bone mineral density is significantly diminished in the spine and proximal femur early in the course of the disease, before the advent of significant immobilization. The mechanism for this is not known. Peripheral arthritis inAS can show synovial hyperplasia, lymphoid infiltration, and pannus formation, but the process lacks the exuberant synovial villi, fibrin deposits, ulcers, and accumulations of plasma cells seen in rheumatoid arthritis (Chap. 312). Central cartilaginous erosions caused by proliferation of subchondral granulation tissue are common in AS but rare in rheumatoid arthritis. Acute anterior uveitis (iritis) occurs in at least 20% of patients withAS. Few cases have been studied histologically, none at an early stage. After recurrent attacks, the iris shows nonspecific inflammatory changes, scarring, increased vascularity, and pigment-laden macrophages. Pupillary synechiae and cataract formation are common sequelae. Aortic insufficiency develops in a small percentage of cases. There is thickening of the aortic valve cusps and the aorta near the sinuses of Valsalva, with dense adventitial scar tissue and intimal fibrous proliferation. The scar tissue can extend into the ventricular septum with resultant heart block. Microscopic inflammatory lesions of the colon and ileocecal valve have been found in 25 to 50% of patients withAS, even in those lacking any clinical evidence ofIBD. IgA nephropathy has been reported with increased frequency. PATHOGENESIS The pathogenesis ofAS is incompletely understood. A number of features of the disease
implicate immune-mediated mechanisms, including elevated serum levels of IgA and acute-phase reactants, inflammatory histology, and close association with HLA-B27. The inflamed sacroiliac joint is infiltrated with CD4+ and CD8+ T cells and macrophages and shows high levels of tumor necrosis factora. Transforming growth factorb is detectable near the sites of new bone formation. No specific event or exogenous agent that triggers the onset of disease has been identified, although overlapping features with reactive arthritis andIBDsuggest that enteric bacteria may play a role. Elevated serum titers of antibodies to certain enteric bacteria, particularly Klebsiella pneumoniae, are common in AS patients, but no role for these antibodies in the pathogenesis of AS has been identified. Evidence that B27 plays a direct role is provided by the finding that rats transgenic for B27 spontaneously develop spondylitis, along with colitis, peripheral arthritis, and other lesions characteristic of the spondyloarthropathies (see below). Some evidence has accumulated for autoimmunity to the cartilage proteoglycan aggrecan, and particularly its G1 globulin domain and link protein.ASpatients have been found to have cellular immunity to these molecules, and mice immunized with the G1 domain develop spondylitis and discitis. Sharing of proteoglycan antigenic epitopes among the pathologic sites in the skeleton, uveal tract, and aorta in AS suggests a possible explanation for the distribution of pathologic sites in AS. CLINICAL MANIFESTATIONS The symptoms of the disease are usually first noticed in late adolescence or early adulthood; the median age in western countries is 23 in both genders. In 5% of patients, symptoms begin after age 40. The initial symptom is usually dull pain, insidious in onset, felt deep in the lower lumbar or gluteal region, accompanied by low-back morning stiffness of up to a few hours' duration that improves with activity and returns following periods of inactivity. Within a few months of onset, the pain has usually become persistent and bilateral. Nocturnal exacerbation of pain that forces the patient to rise and move around may be frequent. In some patients bony tenderness may accompany back pain or stiffness, while in others it may be the predominant complaint. Common sites include the costosternal junctions, spinous processes, iliac crests, greater trochanters, ischial tuberosities, tibial tubercles, and heels. Occasionally, bony chest pain is the presenting complaint. Arthritis in the hips and shoulders ("root" joints) occurs in 25 to 35% of patients, in many cases early in the disease course. Arthritis of peripheral joints other than the hips and shoulders, usually asymmetric, occurs in up to 30% of patients and can occur at any stage of the disease. Neck pain and stiffness from involvement of the cervical spine are usually relatively late manifestations. Occasional patients, particularly in the older age group, present with predominantly constitutional symptoms such as fatigue, anorexia, fever, weight loss, or night sweats. ASoften has a juvenile onset in developing countries. In these individuals, peripheral arthritis and enthesitis usually predominate, with axial symptoms supervening in late adolescence. The most common extraarticular manifestation is acute anterior uveitis, which can antedate the spondylitis. Attacks are typically unilateral, causing pain, photophobia, and
increased lacrimation. These tend to recur, often in the opposite eye. Cataracts and secondary glaucoma are not uncommon sequelae. Aortic insufficiency, sometimes producing symptoms of congestive heart failure, occurs in a few percent of patients, occasionally early in the course of the spinal disease. Third-degree heart block may occur alone or together with aortic insufficiency. The block is in the atrioventricular node in 95% of cases. Up to half the patients have inflammation in the colon or ileum. This is usually asymptomatic, but in 5 to 10% of patients with AS, frankIBD will develop. Initially, physical findings mirror the inflammatory process. The most specific findings involve loss of spinal mobility, with limitation of anterior and lateral flexion and extension of the lumbar spine and of chest expansion. Limitation of motion is usually out of proportion to the degree of bony ankylosis, reflecting muscle spasm secondary to pain and inflammation. Pain in the sacroiliac joints may be elicited either with direct pressure or with maneuvers that stress the joints, but these techniques are unreliable in discriminating inflammatory sacroiliitis. In addition, there is commonly tenderness upon palpation at the sites of symptomatic bony tenderness and paraspinous muscle spasm. The Schober test is a useful measure of flexion of the lumbar spine. The patient stands erect, with heels together, and marks are made directly over the spine 5 cm below and 10 cm above the lumbosacral junction (identified by a horizontal line between the posterosuperior iliac spines.) The patient then bends forward maximally, and the distance between the two marks is measured. The distance between the two marks increases 5 cm or more in the case of normal mobility and less than 4 cm in the case of decreased mobility. Chest expansion is measured as the difference between maximal inspiration and maximal forced expiration in the fourth intercostal space in males or just below the breasts in females. Normal chest expansion is 5 cm or greater. Limitation or pain with motion of the hips or shoulders is usually present if either of these joints is involved. Careful examination is also necessary to detect inflammatory disease of peripheral joints. It should be emphasized that early in the course of mild cases, symptoms may be subtle and nonspecific, and the physical examination may be completely normal. The course of the disease is extremely variable, ranging from the individual with mild stiffness and radiographically equivocal sacroiliitis to the patient with a totally fused spine and severe bilateral hip arthritis, possibly accompanied by severe peripheral arthritis and extraarticular manifestations. Pain tends to be persistent early in the disease and then to become intermittent, with alternating exacerbations and quiescent periods. In a typical severe untreated case with progression of the spondylitis to syndesmophyte formation, the patient's posture undergoes characteristic changes. The lumbar lordosis is obliterated with accompanying atrophy of the buttocks. The thoracic kyphosis is accentuated. If the cervical spine is involved, there may be a forward stoop of the neck. Hip involvement with ankylosis may lead to flexion contractures, compensated by flexion at the knees. The progression of the disease may be followed by measuring the patient's height, chest expansion, Schober test, and occiput-to-wall distance when the patient stands erect with the heels and back flat against the wall. Occasional individuals are encountered with advanced physical findings suggestive of long-standingAS who report having never had significant symptoms.
In some but not all studies, onset of the disease in adolescence correlates with a worse prognosis, but there is general agreement that early severe hip involvement is an indication of progressive disease. The disease in women tends to progress less frequently to total spinal ankylosis, although there is some evidence for an increased prevalence of isolated cervical ankylosis and peripheral arthritis in women. In industrialized countries, peripheral arthritis (distal to hips and shoulders) occurs overall in about 25% of patients, usually as a late manifestation, whereas in developing countries, the prevalence is much higher, with onset typically early in the disease course. Pregnancy has no consistent effect onAS, with symptoms improving, remaining the same, or deteriorating in about one-third of pregnant patients, respectively. The most serious complication of the spinal disease is spinal fracture, which can occur with even minor trauma to the rigid, osteoporotic spine. The cervical spine is most commonly involved. These fractures are often displaced and cause spinal cord injury. Cauda equina syndrome and slowly progressive upper pulmonary lobe fibrosis are rare complications of long-standingAS. The prevalence of aortic insufficiency and of cardiac conduction disturbances, including third-degree heart block, increases with prolonged disease. Subclinical pulmonary lesions and cardiac dysfunction may be relatively common. Prostatitis has been reported to have an increased prevalence in men with AS. Amyloidosis is only rarely associated (Chap. 319). Several validated measures of disease activity and functional outcome have recently been developed forAS. Despite the persistence of the disease, most patients remain gainfully employed. The effect of AS on survival is controversial. Some, but not all, studies have suggested that AS shortens life span, compared with the general population. Mortality attributable to AS is largely the result of spinal trauma, aortic insufficiency, respiratory failure, amyloid nephropathy, or complications of therapy such as upper gastrointestinal hemorrhage. LABORATORY FINDINGS No laboratory test is diagnostic of AS. In most ethnic groups, the HLA-B27 gene is present in approximately 90% of patients with AS. Most, but not all, patients with active disease have an elevated erythrocyte sedimentation rate and an elevated level of C-reactive protein. A mild normochromic, normocytic anemia may be present. Patients with severe disease may show an elevated alkaline phosphatase level. Elevated serum IgA levels are common. Rheumatoid factor and antinuclear antibodies are largely absent unless caused by a coexistent disease. Synovial fluid from inflamed peripheral joints in AS is not distinctly different from that of other inflammatory joint diseases. In cases with restriction of chest wall motion, decreased vital capacity and increased functional residual capacity are common, but airflow measurements are normal and ventilatory function is usually well maintained. RADIOGRAPHIC FINDINGS Radiographically demonstrable sacroiliitis is usually present inAS. The earliest changes in the sacroiliac joints demonstrable by standard radiography are blurring of the cortical margins of the subchondral bone, followed by erosions and sclerosis. Progression of the erosions leads to "pseudowidening" of the joint space; as fibrous and then bony
ankylosis supervene, the joints may become obliterated radiographically. The changes and progression of the lesions are usually symmetric. Roentgenographic abnormalities generally appear in the sacroiliac joints before appearing elsewhere in the spine. In the lumbar spine, progression of the disease leads to straightening, caused by loss of lordosis, and reactive sclerosis, caused by osteitis of the anterior corners of the vertebral bodies with subsequent erosion, leading to "squaring" of the vertebral bodies. Progressive ossification of the superficial layers of the annulus fibrosus leads to eventual formation of marginal syndesmophytes, visible on plain films as bony bridges connecting successive vertebral bodies anteriorly and laterally. In mild cases, years may elapse before unequivocal sacroiliac abnormalities are evident on plain radiographs. Computed tomography (CT) and magnetic resonance imaging (MRI) can detect abnormalities reliably at an earlier stage than plain radiography. MRI has emerged as a highly sensitive and specific technique for identifying early intraarticular inflammation, cartilage changes, and underlying bone marrow edema in sacroiliitis (Fig. 315-1). In suspected cases in which conventional radiography does not reveal definite sacroiliac abnormalities or is undesirable (e.g., in young women or children), dynamic MRI is the procedure of choice for establishing a diagnosis of sacroiliitis. Reduced bone mineral density can be detected by dual-energy x-ray absorptiometry of the femoral neck and the lumbar spine. Falsely elevated readings related to spinal ossification can be avoided by using a lateral projection of the L3 vertebral body. DIAGNOSIS The diagnosis of earlyASbefore the development of irreversible deformity can be difficult to establish. Currently, modified New York criteria (1984) are widely used for diagnosis. These consist of the following: (1) a history of inflammatory back pain (see below); (2) limitation of motion of the lumbar spine in both the sagittal and frontal planes; (3) limited chest expansion, relative to standard values for age and sex; and (4) definite radiographic sacroiliitis. Using these criteria, the presence of radiographic sacroiliitis plus any one of the other three criteria is sufficient for a diagnosis of definite AS. These criteria may need to be further modified to include sacroiliitis demonstrated byMRI to increase their sensitivity. The presence of B27 is neither necessary nor sufficient for the diagnosis, but the B27 test can be helpful in patients with suggestive clinical findings who have not yet developed radiographic sacroiliitis. Moreover, the absence of B27 in a typical case ofASsignificantly increases the probability of coexistentIBD. ASmust be differentiated from numerous other causes of low-back pain, some of which are far more common than AS. The inflammatory back pain of AS is usually distinguished by the following five features: (1) age of onset below 40, (2) insidious onset, (3) duration greater than 3 months before medical attention is sought, (4) morning stiffness, and (5) improvement with exercise or activity. The most common causes of back pain other than AS are primarily mechanical or degenerative rather than
inflammatory and do not show these features. Less common metabolic, infectious, and malignant causes of back pain also must be differentiated from AS. Ochronosis can produce a phenotype that is clinically and radiographically similar to AS. Marked calcification and ossification of paraspinous ligaments occur in diffuse idiopathic skeletal hyperostosis (DISH). Although DISH is often categorized as a variant of osteoarthritis, diarthrodial joints are not involved. Ligamentous calcification and ossification are usually most prominent in the anterior spinal ligament and give the appearance of "flowing wax" on the anterior bodies of the vertebrae. However, a radiolucency may be seen between the newly deposited bone and the vertebral body, differentiating DISH from the marginal osteophytes in spondylosis. Intervertebral disk spaces are preserved, and sacroiliac and apophyseal joints appear normal, helping to differentiate DISH from spondylosis and fromAS, respectively. DISHoccurs in the middle-aged and the elderly and is more common in men than in women. Patients are frequently asymptomatic but may have stiffness. Radiographic changes are generally much more severe than might be predicted from the mild symptoms caused by DISH. TREATMENT There is no definitive treatment forAS. The principal goal of management is the conscientious participation by the patient in an exercise program designed to maintain functional posture and to preserve range of motion. There is evidence that exercise increases mobility and improves function. The proportion of patients wtih severe deformity has decreased markedly in recent decades, probably because of earlier diagnosis and widespread use of physical therapy. Smoking has been associated with a poor outcome and should be emphatically discouraged. Most patients require anti-inflammatory agents to achieve sufficient symptomatic relief to be able to remain functional and carry out the exercise program. It is not known whether drug treatment alone can alter the progression of the disease. Several nonsteroidal anti-inflammatory drugs (NSAIDs) have proved effective in reducing the pain and stiffness ofAS and are commonly used. Indomethacin is particularly effective as a 75-mg slow-release preparation taken once or twice daily. Although phenylbutazone, at doses of 200 to 400 mg/d, has been considered the most effective anti-inflammatory agent in AS, because of its greater potential for serious side effects such as aplastic anemia and agranulocytosis, its use in the United States is confined to patients with very severe disease whose symptoms do not respond at all to other agents. Recent controlled trials suggest that sulfasalazine, in doses of 2 to 3 g/d, is useful in reducing peripheral joint symptoms as well as reversing laboratory evidence of inflammation. Some studies have not shown it to benefit axial arthritis, and its effect on natural progression of the disease is unproven. The peripheral arthritis may also respond to the folic acid antagonist methotrexate.1No therapeutic role for gold, penicillamine, immunosuppressive drugs, or oral glucocorticoids has been documented in AS. Occasionally, intralesional or intraarticular glucocorticoid injections may be beneficial in patients with persistent enthesopathy or synovitis unresponsive to anti-inflammatory agents. Recent studies have suggested that symptomatic benefit can be achieved fromCT-guided glucocorticoid injections into the sacroiliac joints, but the
effects are not sustained. Anecdotal benefit has been reported for diverse agents such as pamidronate, thalidomide, pulse intravenous methylprednisolone, and tumor necrosis factor a antagonists. Controlled trials of these and other agents are needed, since for many patients current therapy is inadequate even for control of pain and stiffness. The most common indication for surgery in patients withAS is severe hip joint arthritis, the pain and stiffness of which are usually dramatically relieved by total hip arthroplasty. A small number of patients may benefit from surgical correction of extreme flexion deformities of the spine or of atlantoaxial subluxation. Attacks of iritis are usually effectively managed with local glucocorticoid administration in conjunction with mydriatic agents, although systemic glucocorticoids or even immunosuppressive drugs may be required in some cases. Coexistent cardiac disease may require pacemaker implantation and/or aortic valve replacement. 1Azathioprine,
methotrexate, and sulfasalazine have not been approved for this purpose by the U.S. Food and Drug Administration at the time of publication. REACTIVE ARTHRITIS AND UNDIFFERENTIATED SPONDYLOARTHROPATHY Reactive arthritis (ReA) refers to acute nonpurulent arthritis complicating an infection elsewhere in the body. In recent years, the term has been used primarily to refer to spondyloarthropathies following enteric or urogenital infections and occurring predominantly in individuals with the histocompatibility antigen HLA-B27. Included in this category is the constellation of clinical findings formerly commonly called Reiter's syndrome.*Other forms of reactive and infection-related arthritis not associated with B27 and showing a different spectrum of clinical features, such as rheumatic fever or Lyme disease, are discussed in Chaps. 235 and 176. HISTORIC BACKGROUND The association of acute arthritis with episodes of diarrhea or urethritis has been recognized for centuries. A large number of cases during World Wars I and II focused attention on the triad of arthritis, urethritis, and conjunctivitis, which became known as Reiter's syndrome, often occurring with additional mucocutaneous lesions. The identification of bacterial species capable of triggering the clinical syndrome and the finding that up to 85% of the patients possess the B27 antigen have led to the unifying concept ofReA as a clinical syndrome triggered by specific etiologic agents in a genetically susceptible host. A similar spectrum of clinical manifestations can be triggered by enteric infection with any of several Shigella, Salmonella, Yersinia, and Campylobacter species, by genital infection with Chlamydia trachomatis; and possibly by other agents as well. Although Reiter's syndrome can be said to represent one part of the spectrum of the clinical manifestations ofReA, particularly that induced by Shigella or Chlamydia, the term is now largely of historic interest only. Since most patients with spondyloarthropathy do not have the classic features of Reiter's syndrome, it has become customary to employ the term reactive arthritis, regardless of whether or not there is evidence for a triggering infection. For the purposes of this chapter, the use of ReA will be restricted to those cases of spondyloarthropathy in which there is at least
presumptive evidence for a related antecedent infection. Patients with clinical features of ReA who lack both evidence of an antecedent infection and the classic findings of Reiter's syndrome (urethritis, arthritis, conjunctivitis) will be considered to have undifferentiated spondyloarthropathy, which is discussed at the end of this chapter. EPIDEMIOLOGY LikeAS,ReAoccurs predominantly in individuals who have inherited the B27 gene; in most series, 60 to 85% of patients are B27 positive. In epidemics of arthritogenic bacterial infection, e.g., S. flexneri, it has been estimated that ReA develops in ~20% of exposed B27-positive individuals. In families with multiple cases of AS or ReA, the two conditions have been said to "breed true," i.e., to be uncommonly found together within an individual family. Whether this is caused by genetic or environmental factors is not known. The disease is most common in individuals 18 to 40 years of age, but it can occur both in children over 5 years of age and in older adults. The sex ratio inReAfollowing enteric infection is nearly 1:1, whereas venereally acquired ReA is predominantly confined to men. The overall prevalence and incidence of ReA are difficult to assess because of the variable prevalence of the triggering infections and genetic susceptibility factors in different populations. For example, in Olmsted County, MN, the incidence was estimated as 3.5 cases per 100,000 population per year. In contrast, in a population with a high rate of genitourinary and/or gastrointestinal infections such as urban homosexual and bisexual men, the prevalence may approach 1 per 1000. A particularly severe form of peripheral spondyloarthropathy has been described in patients with AIDS (Chap. 309). Most of these patients are HLA-B27 positive, but HIV infection per se is not an independent risk factor for spondyloarthropathy. PATHOLOGY Synovial histology is similar to that of other inflammatory arthropathies. Enthesitis is a common clinical finding inReA; the histology of this lesion resembles that ofAS. Microscopic histopathologic evidence of inflammation has occasionally been noted in the colon and ileum of patients with postvenereal ReA, but much less commonly than in postenteric ReA. The skin lesions of keratoderma blenorrhagica, which is associated mainly with venereally acquired ReA, are histologically indistinguishable from psoriatic lesions. ETIOLOGY AND PATHOGENESIS The first bacterial infection noted to be causally related toReA was S. flexneri. An outbreak of shigellosis among Finnish troops in 1944 resulted in numerous cases of ReA. Of the four species S. sonnei, S. boydii, S. flexneri, and S. dysenteriae, S. flexneri has most often been implicated in cases of ReA, both sporadic and epidemic. S. sonnei, although responsible for the majority of cases of shigellosis in the United States, has only rarely been implicated in cases of ReA. Other bacteria that have been definitively identified as triggers ofReAinclude several
Salmonella spp., Y. enterocolitica, C. jejuni, and C. trachomatis. There is suggestive evidence implicating several other microorganisms, including Y. pseudotuberculosis, Clostridium difficile, and Ureaplasma urealyticum. Chlamydia pneumoniae, a respiratory pathogen, has also recently been implicated in triggering ReA. There are also numerous isolated reports of acute arthritis preceded by other bacterial, viral, or parasitic infections, but whether the microorganisms involved are actual triggers of ReA remains to be determined. It has not been determined whetherReAoccurs by the same pathogenic mechanism following infection with each of these microorganisms, nor has the mechanism been fully elucidated in the case of any one of the known bacterial triggers. Most, if not all, of the triggering organisms produce lipopolysaccharide (LPS) and share a capacity to attack mucosal surfaces, to invade host cells, and survive intracellularly. Antigens from Chlamydia, Yersinia, Salmonella, and Shigella have been shown to be present in the synovium and/or synovial fluid leukocytes of patients with ReA for long periods following the acute attack. In ReA triggered by Y. enterocolitica, bacterial LPS and heat shock protein antigens have been found in peripheral blood cells years after the triggering infection. In the case of C. trachomatis, synovial persistence of microbial DNA and RNA suggests the presence of viable organisms, despite uniform failure to culture the organism from these specimens. There is thus evidence that ReA, at least in some cases, may be a form of chronic infection, rather than solely "reactive." T cells that specifically respond to antigens of the inciting organism are typically found in inflamed synovium but not in peripheral blood of patients with ReA. These T cells are predominantly CD4+, but CD8+ B27-restricted bacteria-specific cytolytic T cells have also been isolated in Yersinia- and C. trachomatis-induced ReA. Specific peptide antigens from these organisms have been identified as dominant T cells epitopes. Unlike the synovial CD4 T cells in rheumatoid arthritis, which are predominantly of the TH1 phenotype, those in ReA also show a TH2 phenotype. It is likely that antigen-specific T cells play an important role in the pathogenesis of ReA, but the precise mechanisms remain to be determined. The role of HLA-B27 inReA also remains to be determined. Transgenic rats with high expression of B27 spontaneously develop a multiple organ system inflammatory disease affecting the gut, peripheral and axial joints, male genital tract, and skin that resembles these human conditions clinically and histologically. When raised in a germ-free environment, the B27 rats do not develop gut or joint inflammation, but the skin and genital lesions are not prevented. These findings suggest that bacteria are necessary, and normal gut bacteria are sufficient, to induce B27-related joint inflammation. In both the rat and human diseases, it remains to be determined whether the primary process is an autoimmune response against host tissues or an immune response against antigens of the triggering organism that have disseminated to the target tissues, and the specific role of B27 itself remains to be determined. A potentially very informative converse observation, in which humans develop a disease process resembling one first described in rats, is the recent finding that 0.4 to 0.8% of individuals treated with intravesicular bacillus Calmette-Guerin for bladder cancer develop reactive arthritis, and 60% of these patients are B27 positive. The process closely mimics adjuvant-induced arthritis in rats given complete Freund's adjuvant, first described over 40 years ago, which is currently thought to be mediated by CD4+ T cells specific for mycobacterial heat shock protein.
An intriguing in vitro finding indicates that the presence of HLA-B27 significantly prolongs the intracellular survival of Y. enterocolitica, and S. enteritides in human and mouse cell lines. A unifying hypothesis suggests that prolonged intracellular bacterial survival, promoted by B27, other factors, or both, permits trafficking of infected leukocytes from the site of primary infection to joints, where a T cell response to persistent bacterial antigens promotes arthritis. Evidence exists supporting each step of this scheme. CLINICAL FEATURES The clinical manifestations ofReAconstitute a spectrum that ranges from an isolated, transient monarthritis to severe multisystem disease. In the majority of cases, a careful history will elicit some evidence of an antecedent infection 1 to 4 weeks before the onset of symptoms of the reactive disease. However, in a sizable minority, no clinical or laboratory evidence of an antecedent infection can be found. In many cases of presumed venereally acquired reactive disease, there is a history of a recent new sexual partner, even in the absence of laboratory evidence of infection. Constitutional symptoms are common, including fatigue, malaise, fever, and weight loss. The musculoskeletal symptoms are usually acute in onset. Arthritis is usually asymmetric and additive, with involvement of new joints occurring over a period of a few days to 1 to 2 weeks. The joints of the lower extremities, especially the knee, ankle, and subtalar, metatarsophalangeal, and toe interphalangeal joints, are the most common sites of involvement, but the wrist and fingers can be involved as well. The arthritis is usually quite painful, and tense joint effusions are not uncommon, especially in the knee. Dactylitis, or "sausage digit," a diffuse swelling of a solitary finger or toe, is a distinctive feature of bothReA and psoriatic arthritis (Chap. 324). It is not specific, however, in that it is also seen in polyarticular gout and sarcoidosis. Tendinitis and fasciitis are particularly characteristic lesions, producing pain at multiple insertion sites, especially the Achilles insertion, the plantar fascia, and sites along the axial skeleton. Spinal and low-back pain are quite common and may be caused by insertional inflammation, muscle spasm, acute sacroiliitis, or, presumably, arthritis in intervertebral articulations. Urogenital lesions may occur throughout the course of the disease. In males, urethritis may be marked or relatively asymptomatic and may be either an accompaniment of the triggering infection or a result of the reactive phase of the disease. Prostatitis is also common. Similarly, in females, cervicitis or salpingitis may be caused either by the infectious trigger or by the sterile reactive process. Ocular disease is common, ranging from transient, asymptomatic conjunctivitis to an aggressive anterior uveitis that occasionally proves refractory to treatment and may result in blindness. Mucocutaneous lesions are frequent. Oral ulcers tend to be superficial, transient, and often asymptomatic. The characteristic skin lesions, keratoderma blenorrhagica, consist of vesicles that become hyperkeratotic, ultimately forming a crust before disappearing. They are most common on the palms and soles but may occur elsewhere as well. In
patients with HIV infection, these lesions are often extremely severe and extensive, to the point of dominating the clinical picture (Chap. 309). Lesions on the glans penis, termed circinate balanitis, are common; these consist of vesicles that quickly rupture to form painless superficial erosions, which in circumcised individuals can form crusts similar to those of keratoderma blenorrhagica. Nail changes are common and consist of onycholysis, distal yellowish discoloration, and/or heaped-up hyperkeratosis. Less frequent or rare manifestations ofReAinclude cardiac conduction defects, aortic insufficiency, central or peripheral nervous system lesions, and pleuropulmonary infiltrates. Long-term follow-up studies suggest that some joint symptoms persist in 30 to 60% of patients withReA. Recurrences of the acute syndrome are common, and as many as 25% of patients either become unable to work or are forced to change occupations because of persistent joint symptoms. Chronic heel pain is often a particularly distressing symptom. Some aspects of ankylosing spondylitis are also common sequelae (see below). In some but not all studies, HLA-B27-positive patients have shown a worse outcome than B27-negative patients. The extent to which the long-term prognosis varies with different inciting agents is not known. However, patients with Yersinia-induced arthritis appear to have less chronic disease than those whose initial episode follows epidemic shigellosis. LABORATORY AND RADIOGRAPHIC FINDINGS The erythrocyte sedimentation rate is usually elevated during the acute phase of the disease. Mild anemia may be present, and acute-phase reactants tend to be increased. Synovial fluid is nonspecifically inflammatory, showing an elevated white cell count with a predominance of neutrophils. In most ethnic groups, 50 to 75% of the patients are B27 positive. It is unusual for the triggering infection to persist at the site of primary mucosal infection through the time of onset of the reactive disease, but it may occasionally be possible to culture the organism, e.g., in the case of Yersinia- or Chlamydia-induced disease. Serologic evidence of a recent infection may be present, such as a marked elevation of antibodies to Yersinia, Salmonella, or Chlamydia. In early or mild disease, radiographic changes may be absent or confined to juxtaarticular osteoporosis. With long-standing persistent disease, marginal erosions and loss of joint space can be seen in affected joints. Periostitis with reactive new bone formation is characteristic of the disease, as it is with all the spondyloarthropathies. Spurs at the insertion of the plantar fascia are common. Sacroiliitis and spondylitis may be seen as late sequelae. The sacroiliitis is more commonly asymmetric than inAS, and the spondylitis, rather than ascending symmetrically from the lower lumbar segments, can begin anywhere along the lumbar spine. The syndesmophytes may be coarse and nonmarginal, arising from the middle of a vertebral body, a pattern rarely seen in primary AS. Progression to spinal fusion as a sequela ofReA is uncommon. DIAGNOSIS
ReAis a clinical diagnosis, there being no definitively diagnostic laboratory test or radiographic finding. The diagnosis should be entertained in any patient with an acute inflammatory, asymmetric, additive arthritis or tendinitis. The evaluation of such a patient should include careful questioning regarding possible antecedent triggering events such as an episode of diarrhea or dysuria. On physical examination, careful attention must be paid to the distribution of the joint and tendon involvement and to possible sites of extraarticular involvement, such as the eyes, mucous membranes, skin, nails, and genitalia. Synovial fluid aspiration and analysis may be helpful in excluding septic or crystal-induced arthritis. Culture or serology may help to identify a triggering infection. The role of molecular methods of microbial detection has not been established (see below). Although typing for B27 is not needed to secure the diagnosis in clear-cut cases, it may have prognostic significance in terms of severity, chronicity, and the propensity for spondylitis and uveitis. Furthermore, it can be helpful diagnostically in atypical cases, a positive test increasing and a negative test decreasing the probability ofReA. It is particularly important to differentiateReA from disseminated gonococcal disease, both of which can be venereally acquired and associated with urethritis (Chap. 147). Gonococcal arthritis and tenosynovitis tend to involve both upper and lower extremities equally, whereas in ReA lower extremity symptoms usually predominate. Back pain is common in ReA but is not a feature of gonococcal disease, whereas the vesicular skin lesions characteristic of disseminated gonococcal disease are not found in ReA. A positive gonococcal culture from the urethra or cervix does not exclude a diagnosis of ReA; however, culturing gonococci from blood, skin lesion, or synovium establishes the diagnosis of disseminated gonococcal disease. Polymerase chain reaction (PCR) technology has recently been used in the diagnosis of infections with Neisseria gonorrheae and with C. trachomatis. Occasionally, the only definitive way to distinguish the two is through a therapeutic trial of antibiotics. ReAshares many features in common with psoriatic arthropathy, including the asymmetry of the arthritis, a propensity for "sausage digits" and nail involvement, an association with uveitis, and skin lesions of similar histology (Chap. 324). However, psoriatic arthritis is usually gradual in onset, the arthritis tends to affect primarily the upper extremities, and there is less associated periarthritis. Psoriatic arthritis is not associated with mouth ulcers or urethritis, or, usually, with bowel symptoms. Although psoriatic arthropathy shows some distinctive radiographic features that are not found in ReA, these occur only late in the disease and are of little help diagnostically. Only psoriatic spondylitis, not the peripheral arthritis, is associated with B27, about 50% of patients being positive. Occasional patients, usually B27 positive, following what appears to be a typical episode of ReA, will develop typical psoriasis and persistent arthritis such that the two entities become indistinguishable. Undifferentiated spondyloarthropathy, or simply "spondyloarthropathy," is diagnosed in patients who lack evidence of an antecedent infection that might triggerReA and who do not meet criteria forAS but who show clinical features of these disorders. TREATMENT
Most patients withReA are benefitted to some degree byNSAIDs, although rarely are symptoms of the acute arthritis completely ameliorated, and some patients fail to respond at all. Indomethacin, 75 to 150 mg/d in divided doses, is the initial treatment of choice. Other NSAIDs may be tried, with phenylbutazone, 100 mg tid or qid, being the NSAID of last resort, to be used only in severe, refractory cases because of its potentially serious side effects. It is unclear whether antibiotics have a role in the therapy ofReA. One controlled study suggested that prolonged administration of a long-acting tetracycline may accelerate recovery from Chlamydia-induced ReA, but subsequent results have been less encouraging, and therapy for other bacterial triggers of ReA has shown little or no benefit. However, there is evidence that prompt, appropriate antibiotic treatment of acute chlamydial urethritis may prevent subsequent ReA. Currently, expert opinion supports the use of antibiotic therapy in established urogenital ReA but not in gastrointestinal ReA. Two recent multicenter trials have suggested that sulfasalazine, up to 3 g/d in divided doses, may be beneficial to patients with persistentReA.1Patients with debilitating symptoms refractory toNSAIDand sulfasalazine therapy may respond to immunosuppressive agents such as azathioprine, 1 to 2 mg/kg per day, or to methotrexate, 7.5 to 15 mg per week. Systemic glucocorticoids are not generally recommended but in rare instances may be helpful in mobilizing a severely affected bedridden patient. Antimalarials, gold, and penicillamine are not useful in the treatment of ReA. Trials of new agents proven useful in rheumatoid arthritis, such as COX-2 inhibitors, leflunomide, and tumor necrosis factor a inhibitors, remain to be implemented. Tendinitis and other enthesitic lesions occasionally may benefit from intralesional glucocorticoids. Uveitis may require aggressive treatment with glucocorticoids to prevent serious sequelae. Skin lesions ordinarily require only symptomatic treatment. In patients with HIV infection andReA, many of whom have severe skin lesions, the skin lesions in particular appear to respond to systemic treatment with anti-retroviral agents (Chap. 309). Cardiac complications are managed conventionally; management of neurologic complications is symptomatic. Patients need to be educated about the nature of the disease and the factors that predispose to its recurrence. Comprehensive management includes counseling of patients in the avoidance of sexually transmitted disease and exposure to enteropathogens, as well as appropriate use of physical therapy, vocational counseling, and continued surveillance for long-term complications such as ankylosing spondylitis. UNDIFFERENTIATED AND JUVENILE-ONSET SPONDYLOARTHROPATHY It is not uncommon for clinicians to encounter patients, usually young adults, who do not haveIBD or psoriasis, lack evidence of an antecedent triggering infection, and do not have the classic triad of Reiter's syndrome or meet criteria for ankylosing spondylitis, who nonetheless present with some features of one or more of the spondyloarthropathies discussed above. For example, a patient may present with inflammatory synovitis of one knee, Achilles tendinitis, and dactylitis of one digit ("sausage digit"), or sacroiliitis in the absence of other criteria forAS. It is now common
to consider such patients as having undifferentiated spondyloarthropathy, or simply spondyloarthropathy. Other terms for this condition have included seronegative oligoarthritis, undifferentiated oligoarthritis, and the now-outmoded incomplete Reiter's syndrome. There is strong evidence that some, perhaps most, of these patients haveReA in which the triggering infection remains clinically silent. In some other cases, the patient subsequently develops IBD or psoriasis or the process eventually meets criteria for ankylosing spondylitis. Approximately half the patients with undifferentiated spondyloarthropathy are HLA-B27 positive, and thus the absence of B27 is not useful in establishing or excluding the diagnosis. In juvenile-onset spondyloarthropathy, which begins most commonly in boys (60 to 80%) between ages 7 and 16, an asymmetric, predominantly lower extremity oligoarthritis and enthesitis without extraarticular features is the typical mode of presentation. The prevalence of B27 in this condition, which has been termed the SEA syndrome (seronegative, enthesopathy, arthropathy), is approximately 80%. Many, but not all, of these patients go on to develop typical ankylosing spondylitis in late adolescence or adulthood. Management of undifferentiated spondyloarthropathy is similar to that of the other spondyloarthropathies, withNSAIDsand physical therapy forming the mainstays of treatment. Textbooks of pediatrics should be consulted for information on management of juvenile-onset spondyloarthropathy. An algorithm for the diagnosis of the spondyloarthropathies in adults is presented in Fig. 315-2. (Bibliography omitted in Palm version) Back to Table of Contents
316. BEHCET'S SYNDROME - Haralampos M. Moutsopoulos DEFINITION Behcet's syndrome is a multisystem disorder presenting with recurrent oral and genital ulcerations as well as ocular involvement. Internationally agreed diagnostic criteria have been proposed (Table 316-1). PREVALENCE, PATHOGENESIS, AND PATHOLOGY The disease has a worldwide distribution. The prevalence of Behcet's syndrome ranges from 1:10,000 in Japan to 1:500,000 in North America and Europe. It affects mainly young adults, with males having more severe disease than females. The etiology and pathogenesis of this syndrome remain obscure; vasculitis is the main pathologic lesion with a tendency to venous thrombus formation, and circulating autoantibodies to human oral mucous membrane are found in approximately 50% of the patients. Familial occurrence has been sporadically reported; and in patients from eastern Mediterranean countries and Japan, the disease appears to be linked to HLA-B5 (B51) alloantigens. CLINICAL FEATURES The recurrent aphthous ulcerations are a sine qua non for the diagnosis. The ulcers are usually painful, shallow or deep with a central yellowish necrotic base, appear singly or in crops, and are located anywhere in the oral cavity. The ulcers persist for 1 to 2 weeks and subside without leaving scars. The genital ulcers resemble the oral ones. Skin involvement includes folliculitis, erythema nodosum, an acne-like exanthem, and infrequently vasculitis. Nonspecific skin inflammatory reactivity to any scratches or intradermal saline injection (pathergy test) is a common and specific manifestation. Eye involvement is the most dreaded complication, since it occasionally progresses rapidly to blindness. The eye disease is usually present at the onset but also may develop within the first few years. In addition to iritis, posterior uveitis, retinal vessel occlusions, and optic neuritis can be seen in some patients with the syndrome. Hypopyon uveitis, which is considered the hallmark of Behcet's syndrome, is in fact a rare manifestation. The arthritis of Behcet's syndrome is not deforming and affects the knees and ankles. Superficial or deep peripheral vein thrombosis is seen in one-fourth of the patients. Pulmonary emboli are a rare complication. The superior vena cava is obstructed occasionally, producing a dramatic clinical picture. Arterial involvement occurs infrequently and presents with aortitis or peripheral arterial aneurysm and arterial thrombosis. Pulmonary artery vasculitis presenting with dyspnea, cough, chest pain, hemoptysis, and infiltrates on chest roentgenograms has been reported recently in 5% of patients.
Central nervous system involvement is found more frequently in patients from northern Europe and the United States. The most common lesions are benign intracranial hypertension, a multiple sclerosis-like picture, pyramidal involvement, and psychiatric disturbances. Gastrointestinal involvement is reported in patients from Japan and includes mucosal ulcerations of the gut. Laboratory findings are mainly nonspecific indices of inflammation such as leukocytosis and elevated erythrocyte sedimentation rate as well as C-reactive protein levels; antibodies to human oral mucosa are also found. TREATMENT The severity of the syndrome usually abates with time. Apart from the patients with neurologic complications, the life expectancy seems to be normal, and the only serious complication is blindness. Mucous membrane involvement may respond to topical glucocorticoids in the form of mouthwash or paste. In more serious cases thalidomide (100 mg/d) is effective. Thrombophlebitis is treated with aspirin, 325 mg/d. Colchicine or interferona can be beneficial for the arthritis of the syndrome. Uveitis and central nervous system involvement require systemic glucocorticoid therapy (prednisone, 1 mg/kg per day) and azathioprine, 2 to 3 mg/kg per day, or cyclosporine, 5 to 10 mg/kg per day. Early initiation of azathioprine tends to favorably affect the long-term prognosis of Behcet's syndrome. (Bibliography omitted in Palm version) Back to Table of Contents
317. THE VASCULITIS SYNDROMES - Anthony S. Fauci DEFINITION Vasculitis is a clinicopathologic process characterized by inflammation of and damage to blood vessels. The vessel lumen is usually compromised, and this is associated with ischemia of the tissues supplied by the involved vessel. A broad and heterogeneous group of syndromes may result from this process, since any type, size, and location of blood vessel may be involved. Vasculitis and its consequences may be the primary or sole manifestation of a disease; alternatively, vasculitis may be a secondary component of another primary disease. Vasculitis may be confined to a single organ such as the skin, or it may simultaneously involve several organ systems. CLASSIFICATION OF VASCULITIC SYNDROMES A major feature of the vasculitic syndromes as a group is the fact that there is a great deal of heterogeneity at the same time as there is considerable overlap among them. This has led to both difficulty and confusion with regard to the categorization of these diseases. The classification scheme listed inTable 317-1 takes into account this heterogeneity and overlap and will serve as a matrix to emphasize the fact that certain syndromes are predominantly systemic in nature and almost invariably lead to irreversible organ system dysfunction and even death if untreated, while others are usually localized to the skin and rarely result in irreversible dysfunction of vital organs. The distinguishing and overlapping features of the diseases listed inTable 317-1, which justify this classification scheme, will be discussed below. PATHOPHYSIOLOGY AND PATHOGENESIS Generally, most of the vasculitic syndromes are assumed to be mediated at least in part by immunopathogenic mechanisms (Table 317-2). However, evidence to this effect is for the most part indirect and may reflect epiphenomena as opposed to true causality. Pathogenic Immune-Complex Formation Vasculitis is generally considered within the broader category of immune-complex diseases that include serum sickness and certain of the connective tissue diseases of which systemic lupus erythematosus (Chap. 311) is the prototype. Although deposition of immune complexes in vessel walls is the most widely accepted pathogenic mechanism of vasculitis, the causal role of immune complexes has not been clearly established in most of the vasculitic syndromes. Circulating immune complexes need not result in deposition of the complexes in blood vessels with ensuing vasculitis, and many patients with active vasculitis do not have demonstrable circulating or deposited immune complexes. The actual antigen contained in the immune complex has only rarely been identified in vasculitic syndromes. In this regard, hepatitis B antigen has been identified in both the circulating and deposited immune complexes in a subset of patients with systemic vasculitis, most notably within the polyarteritis nodosa group (see below). Essential mixed cryoglobulinemia has been associated with hepatitis C virus infection; hepatitis C virions and hepatitis C virus antigen-antibody complexes have been identified in the cryoprecipitates of these patients. An association between persistent parvovirus B19 infection and certain vasculitides has been reported; however, the pathogenic mechanisms related to this
association are unclear. The mechanisms of tissue damage in immune complex-mediated vasculitis resemble those described for serum sickness. In this model, antigen-antibody complexes are formed in antigen excess and are deposited in vessel walls whose permeability has been increased by vasoactive amines such as histamine, bradykinin, and leukotrienes released from platelets or from mast cells as a result of IgE-triggered mechanisms. The deposition of complexes results in activation of complement components, particularly C5a, which is strongly chemotactic for neutrophils. These cells then infiltrate the vessel wall, phagocytose the immune complexes, and release their intracytoplasmic enzymes, which damage the vessel wall. As the process becomes subacute or chronic, mononuclear cells infiltrate the vessel wall. The common denominator of the resulting syndrome is compromise of the vessel lumen with ischemic changes in the tissues supplied by the involved vessel. Antineutrophil Cytoplasmic Antibodies (ANCA) ANCA are antibodies directed against certain proteins in the cytoplasm of neutrophils. They are present in a high percentage of patients with systemic vasculitis, particularly Wegener's granulomatosis, as well as in patients with microscopic polyangiitis and in patients with necrotizing and crescentic glomerulonephritis. There are two major categories of ANCA based on different targets for the antibodies. The terminology of cytoplasmic (c) ANCA refers to the diffuse, granular cytoplasmic staining pattern observed by immunofluorescence microscopy when serum antibodies bind to indicator neutrophils. Proteinase-3, the 29-kDa neutral serine proteinase present in neutrophil azurophilic granules is the major c-ANCA antigen. More than 90% of patients with typical Wegener's granulomatosis and active glomerulonephritis have a positive c-ANCA titer. The terminology of perinuclear (p) ANCA refers to the more localized perinuclear or nuclear staining pattern of the indicator neutrophils. The major target for p-ANCA is the enzyme myeloperoxidase; other targets of p-ANCA include elastase, cathepsin G, lactoferrin, lysozyme, and bactericidal/permeability-increasing protein. p-ANCA have been reported to occur in variable percentages of patients with microscopic polyangiitis, polyarteritis nodosa, Churg-Strauss syndrome, crescentic glomerulonephritis, and Goodpasture's syndrome as well as in association with nonvasculitic entities such as certain rheumatic and nonrheumatic autoimmune diseases, inflammatory bowel disease, certain drugs, and infections such as endocarditis and bacterial airway infections in patients with cystic fibrosis. It is unclear why patients with these vasculitis syndromes developANCA, whereas ANCA are rare in other inflammatory diseases. However, once ANCA are present, there are a number of in vitro observations that suggest feasible mechanisms whereby these antibodies can contribute to the pathogenesis of the vasculitis syndromes. When neutrophils are in the resting state, proteinase-3 exists in the azurophilic granules of the cytoplasm, apparently inaccessible to serum antibodies. However, when neutrophils are primed by tumor necrosis factor (TNF)a or interleukin (IL)1, proteinase-3 translocates to the cell membrane where it can interact with extracellular ANCA. The neutrophils then degranulate and produce reactive oxygen species that can cause tissue damage. Endothelial cells also translocate their cytoplasmic proteinase-3 to the cell membrane upon priming with TNF-a, IL-1, or interferon (IFN)g, thus rendering them susceptible to interaction with ANCA and leading possibly to tissue damage due to
complement-mediated cytotoxicity or antibody-dependent cellular cytotoxicity. Despite the attractiveness of these in vitro data, there is no conclusive evidence that ANCA are directly involved in the pathogenesis of the vasculitis syndromes, and they may represent merely an epiphenomenon; in fact, a number of clinical and laboratory observations argue against a primary pathogenic linkage. Patients may have vasculitis in the absence of ANCA; the absolute height of the antibody titers does not correlate well with disease activity; and patients with vasculitis, particularly Wegener's granulomatosis, in remission may continue to have high c-ANCA titers for years. Thus, their role in the pathogenesis of systemic vasculitis remains an open question. Pathogenic T Lymphocyte Responses and Granuloma Formation In addition to the classic immune complex-mediated mechanisms of vasculitis as well asANCA, other immunopathogenic mechanisms may be involved in damage to vessels. The most prominent of these are delayed hypersensitivity and cell-mediated immune injury as reflected in the histopathologic feature of granulomatous vasculitis. However, immune complexes themselves may induce granulomatous responses. Vascular endothelial cells can express HLA class II molecules following activation by cytokines such as IFN-g. This allows these cells to participate in immunologic reactions such as interaction with CD4+ T lymphocytes in a manner similar to antigen-presenting macrophages. Endothelial cells can secreteIL-1 which may activate T lymphocytes and initiate or propagate in situ immunologic processes within the blood vessel. In addition, IL-1 andTNF-aare potent inducers of endothelial-leukocyte adhesion molecule 1 (ELAM-1) and vascular cell adhesion molecule 1 (VCAM-1), which may enhance the adhesion of leukocytes to endothelial cells in the blood vessel wall. Other mechanisms such as direct cellular cytotoxicity or antibody directed against vessel components or antibody-dependent cellular cytotoxicity have been suggested in certain types of vessel damage. However, there is no convincing evidence to support their causal contribution to the pathogenesis of any of the recognized vasculitic syndromes. It is unknown why certain individuals develop vasculitis in response to certain antigenic stimuli, whereas others do not. However, it is likely that a number of factors are involved in the ultimate expression of a vasculitic syndrome. These include the genetic predisposition and the regulatory mechanisms associated with immune response to certain antigens. When immune complexes are involved in the pathogenic process, the ability of the reticuloendothelial system to clear circulating complexes from the blood, the size and physicochemical properties of immune complexes, the relative degree of turbulence of blood flow, the intravascular hydrostatic pressure in different vessels, and the preexisting integrity of the vessel endothelium likely explain why only certain types of immune complexes cause vasculitis and why the vasculitic process is selective for only certain vessels in individual patients. Approach to the Patient Given the heterogeneous nature of the vasculitis syndromes, workup of a patient with suspected vasculitis should follow a series of progressive steps that establish the diagnosis of vasculitis, determine where possible the category of the vasculitis syndrome (Table 317-1), and determine the pattern and extent of disease activity. This information should then be utilized to determine the choice of therapeutic options (Fig. 317-1). This approach is of considerable importance since several of the vasculitis
syndromes require aggressive therapy with glucocorticoids and immunosuppressive agents, while other syndromes usually resolve spontaneously and require symptomatic treatment only. Vasculitis is often suspected on clinical and laboratory grounds (see individual syndromes below). Depending on the individual category of vasculitis, measurement ofANCAtiters may be helpful in this regard. However, a diagnosis of a vasculitis syndrome should not be made nor should treatment be initiated on the basis of a positive ANCA titer alone. The definitive diagnosis of vasculitis is made upon biopsy of involved tissue. The yield of "blind" biopsies of organs with no subjective or objective evidence of involvement is very low and should be avoided. When syndromes such as classic polyarteritis nodosa, Takayasu's arteritis, or the polyangiitis overlap syndrome are suspected, angiogram of organs with suspected involvement should be performed. However, angiograms should not be performed routinely when patients present with localized cutaneous vasculitis with no clinical indication of visceral involvement. The constellation of clinical, laboratory, biopsy, and radiographic findings usually allows proper categorization to a specific syndrome, and therapy where appropriate should be initiated according to this information (see individual syndromes below). If an offending antigen that precipitates the vasculitis is recognized, the antigen should be removed where possible. If the syndrome resolves, no further action should be taken. If disease activity continues, treatment should be initiated. If the vasculitis is associated with an underlying disease such as an infection, neoplasm, or connective tissue disease, the underlying disease should be treated. If the syndrome resolves, no further action should be taken. If the syndrome does not resolve or if there is no recognizable underlying disease and the vasculitis persists, treatment should be initiated according to the category of the vasculitis syndrome. Treatment options will be considered under the individual syndromes, and general principles of therapy will be considered at the end of the chapter. SYSTEMIC NECROTIZING VASCULITIS POLYARTERITIS NODOSA AND MICROSCOPIC POLYANGIITIS Definition Classic polyarteritis nodosa (PAN) was described in 1866 by Kussmaul and Maier. It is a multisystem, necrotizing vasculitis of small and medium-sized muscular arteries in which involvement of the renal and visceral arteries is characteristic. Classic PAN does not involve pulmonary arteries, although bronchial vessels may be involved; granulomas, significant eosinophilia, and an allergic diathesis are not part of the classic syndrome. The term microscopic polyangiitis (microscopic polyarteritis) was introduced into the literature by Davson in 1948. The Chapel Hill Consensus Conference on the Nomenclature of Systemic Vasculitis held in 1992 officially adopted the term to connote a necrotizing vasculitis with few or no immune complexes (pauci-immune) affecting small vessels (capillaries, venules, or arterioles). Since necrotizing arteritis involving small and medium-sized arteries may also be present, it shares features with classic PAN except that glomerulonephritis is very common in microscopic polyangiitis, and pulmonary capillaritis often occurs. Incidence and Prevalence It is difficult to establish an accurate incidence of these diseases because of the fact that many reports ofPANactually have included both classic PAN and microscopic polyangiitis as well as other related vasculitides. Both
diseases are uncommon, but classic PAN is felt to be more uncommon than microscopic polyangiitis. The mean age of onset of both PAN and microscopic polyangiitis is approximately 50 years of age, and males are slightly more frequently affected than females in both diseases. Pathophysiology and Pathogenesis The vascular lesion in classicPAN is a necrotizing inflammation of small and medium-sized muscular arteries. The lesions are segmental and tend to involve bifurcations and branchings of arteries. They may spread circumferentially to involve adjacent veins. However, involvement of venules is not seen in classic PAN and, if present, suggests microscopic polyangiitis or the polyangiitis overlap syndrome (see below). In the acute stages of disease, polymorphonuclear neutrophils infiltrate all layers of the vessel wall and perivascular areas, which results in intimal proliferation and degeneration of the vessel wall. Mononuclear cells infiltrate the area as the lesions progress to the subacute and chronic stages. Fibrinoid necrosis of the vessels ensues with compromise of the lumen, thrombosis, infarction of the tissues supplied by the involved vessel, and, in some cases, hemorrhage. As the lesions heal, there is collagen deposition, which may lead to further occlusion of the vessel lumen. Aneurysmal dilatations up to 1 cm in size along the involved arteries are characteristic of classic PAN. Granulomas and substantial eosinophilia with eosinophilic tissue infiltrations are not characteristically found and suggest allergic angiitis and granulomatosis (see below). Multiple organ systems are involved, and the clinicopathologic findings reflect the degree and location of vessel involvement and the resulting ischemic changes. As mentioned above, pulmonary arteries are not involved in classicPAN, and bronchial artery involvement is uncommon, whereas pulmonary capillaritis occurs frequently in microscopic polyangiitis. The pathology in the kidney in classic PAN is predominantly that of arteritis without glomerulonephritis. In contrast, glomerulonephritis is very common in microscopic polyangiitis. In patients with significant hypertension, typical pathologic features of glomerulosclerosis may be seen alone or superimposed on lesions of glomerulonephritis. In addition, pathologic sequelae of hypertension may be found elsewhere in the body. The presence of hepatitis B antigenemia in approximately 20 to 30% of patients with systemic vasculitis, particularly of the classicPANtype, together with the isolation of circulating immune complexes composed of hepatitis B antigen and immunoglobulin, and the demonstration by immunofluorescence of hepatitis B antigen, IgM, and complement in the blood vessel walls, strongly suggest the role of immunologic phenomena in the pathogenesis of this disease. Hepatitis C infection has been reported in approximately 5% of patients with PAN; however, its pathogenic role in the vasculitis is unclear at present. Hairy cell leukemia can be associated with classic PAN; the pathogenic mechanisms of this association are unclear. Clinical and Laboratory Manifestations Nonspecific signs and symptoms are the hallmarks of classicPAN. Fever, weight loss, and malaise are present in over one-half of cases. Patients usually present with vague symptoms such as weakness, malaise, headache, abdominal pain, and myalgias. Specific complaints related to the vascular involvement within a particular organ system may also dominate the presenting clinical picture as well as the entire course of the illness (Table 317-3). In classic PAN, renal
involvement most commonly manifests as hypertension, renal insufficiency, or hemorrhage due to microaneurysms. In microscopic polyangiitis acute glomerulonephritis is the characteristic renal lesion. There are no diagnostic serologic tests for classicPAN. In over 75% of patients, the leukocyte count is elevated with a predominance of neutrophils. Eosinophilia is seen only rarely and, when present at high levels, suggests the diagnosis of allergic angiitis and granulomatosis. The anemia of chronic disease may be seen, and an elevated erythrocyte sedimentation rate (ESR) is almost always present. Other common laboratory findings reflect the particular organ involved. Hypergammaglobulinemia may be present, and up to 30% of patients have a positive test for hepatitis B surface antigen. PositiveANCAtiters (usually of the p-ANCA type) are found in a low percentage (90% of the glands before adrenal insufficiency appears. The adrenal is a frequent site for chronic granulomatous diseases, predominantly tuberculosis but also histoplasmosis, coccidioidomycosis, and cryptococcosis. In early series, tuberculosis was responsible for 70 to 90% of cases, but the most frequent cause now is idiopathic atrophy, and an autoimmune mechanism is probably responsible. Rarely, other lesions are encountered, such as adrenoleukodystrophy, bilateral hemorrhage, tumor metastases, HIV, cytomegalovirus (CMV), amyloidosis, adrenomyeloneuropathy, familial adrenal insufficiency, or sarcoidosis. Although half of patients with idiopathic atrophy have circulating adrenal antibodies, autoimmune destruction is probably secondary to cytotoxic T lymphocytes. Specific adrenal antigens to which autoantibodies may be directed include 21-hydroxylase (CYP21A2) and side chain cleavage enzyme but the significance of these antibodies in the pathogenesis of adrenal insufficiency is unknown. Some antibodies cause adrenal insufficiency by blocking the binding ofACTH to its receptors. Some patients also have antibodies to thyroid, parathyroid, and/or gonadal tissue (Chap. 339). There is also an increased incidence of chronic lymphocytic thyroiditis, premature ovarian failure, type 1 diabetes mellitus, and hypo- or hyperthyroidism. The presence of two or more of these autoimmune endocrine disorders in the same person defines the polyglandular autoimmune syndrome type II. Additional features include pernicious anemia, vitiligo, alopecia, nontropical sprue, and myasthenia gravis. Within families, multiple generations are affected by one or more of the above diseases. Type II polyglandular syndrome is the result of a mutant gene on chromosome 6 and is associated with the HLA alleles B8 and DR3. The combination of parathyroid and adrenal insufficiency and chronic mucocutaneous moniliasis constitutes type I polyglandular autoimmune syndrome. Other autoimmune diseases in this disorder include pernicious anemia, chronic active hepatitis, alopecia, primary hypothyroidism, and premature gonadal failure. There is no HLA association; this syndrome is inherited as an autosomal recessive trait. It is caused by mutations in autoimmune polyendocrinopathy candidiasis ectodermal dystrophy (APECED) located on chromosome 21q22.3. The gene encodes a transcription factor thought to be
involved in lymphocyte function. The type I syndrome usually presents during childhood, whereas the type II syndrome is usually manifested in adulthood. Clinical suspicion of adrenal insufficiency should be high in patients with AIDS (Chap. 309).CMVregularly involves the adrenal glands (so-called CMV necrotizing adrenalitis), and involvement with Mycobacterium avium-intracellulare, Cryptococcus, and Kaposi's sarcoma has been reported. Adrenal insufficiency in AIDS patients may not be manifest, but tests of adrenal reserve frequently give abnormal results. When interpreting tests of adrenocortical function, it is important to remember that medications such as rifampin, phenytoin, ketoconazole, megace, and opiates may cause or potentiate adrenal insufficiency. Adrenal hemorrhage and infarction occur in patients on anticoagulants and in those with circulating anticoagulants and hypercoagulable states, such as the antiphospholipid syndrome. There are several rare genetic causes of adrenal insufficiency that present primarily in infancy and childhood (see below). Clinical Signs and Symptoms Adrenocortical insufficiency caused by gradual adrenal destruction is characterized by an insidious onset of fatigability, weakness, anorexia, nausea and vomiting, weight loss, cutaneous and mucosal pigmentation, hypotension, and occasionally hypoglycemia (Table 331-7). Depending on the duration and degree of adrenal hypofunction, the manifestations vary from mild chronic fatigue to fulminating shock associated with acute destruction of the glands, as described by Waterhouse and Friderichsen. Asthenia is the cardinal symptom. Early it may be sporadic, usually most evident at times of stress; as adrenal function becomes more impaired, the patient is continuously fatigued, and bed rest is necessary. Hyperpigmentation may be striking or absent. It commonly appears as a diffuse brown, tan, or bronze darkening of parts such as the elbows or creases of the hand and of areas that normally are pigmented such as the areolae about the nipples. Bluish-black patches may appear on the mucous membranes. Some patients develop dark freckles, and irregular areas of vitiligo may paradoxically be present. As an early sign, tanning following sun exposure may be persistent. Arterial hypotension with postural accentuation is frequent, and blood pressure may be in the range of 80/50 or less. Abnormalities of gastrointestinal function are often the presenting complaint. Symptoms vary from mild anorexia with weight loss to fulminating nausea, vomiting, diarrhea, and ill-defined abdominal pain, which may be so severe as to be confused with an acute abdomen. Patients may have personality changes, usually consisting of excessive irritability and restlessness. Enhancement of the sensory modalities of taste, olfaction, and hearing is reversible with therapy. Axillary and pubic hair may be decreased in women due to loss of adrenal androgens. Laboratory Findings In the early phase of gradual adrenal destruction, there may be no demonstrable abnormalities in the routine laboratory parameters, but adrenal reserve
is decreased -- that is, while basal steroid output may be normal, a subnormal increase occurs after stress. Adrenal stimulation withACTHuncovers abnormalities in this stage of the disease, eliciting a subnormal increase of cortisol levels or no increase at all. In more advanced stages of adrenal destruction, serum sodium, chloride, and bicarbonate levels are reduced, and the serum potassium level is elevated. The hyponatremia is due both to loss of sodium into the urine (due to aldosterone deficiency) and to movement into the intracellular compartment. This extravascular sodium loss depletes extracellular fluid volume and accentuates hypotension. Elevated plasma vasopressin and angiotensin II levels may contribute to the hyponatremia by impairing free water clearance. Hyperkalemia is due to a combination of aldosterone deficiency, impaired glomerular filtration, and acidosis. Basal levels of cortisol and aldosterone are subnormal and fail to increase following ACTH administration. Mild to moderate hypercalcemia occurs in 10 to 20% of patients for unclear reasons. The electrocardiogram may show nonspecific changes, and the electroencephalogram exhibits a generalized reduction and slowing. There may be a normocytic anemia, a relative lymphocytosis, and a moderate eosinophilia. Diagnosis The diagnosis of adrenal insufficiency should be made only withACTHstimulation testing to assess adrenal reserve capacity for steroid production (see above for ACTH test protocols). In brief, the best screening test is the cortisol response 60 min after 250 ug of cosyntropin given intramuscularly or intravenously. Cortisol levels should exceed 495 nmol/L (18 ug/dL). If the response is abnormal, then primary and secondary adrenal insufficiency can be distinguished by measuring aldosterone levels from the same blood samples. In secondary, but not primary, adrenal insufficiency the aldosterone increment will be normal [³150 pmol/l (5 ng/dL)]. Furthermore, in primary adrenal insufficiency, plasma ACTH and associated peptides (b-LPT) are elevated because of loss of the usual cortisol-hypothalamic-pituitary feedback relationship, whereas in secondary adrenal insufficiency, plasma ACTH values are low or "inappropriately" normal (Fig. 331-11). Differential Diagnosis Since weakness and fatigue are common, diagnosis of early adrenocortical insufficiency may be difficult. However, the combination of mild gastrointestinal distress, weight loss, anorexia, and a suggestion of increased pigmentation makes it mandatory to performACTHstimulation testing to rule out adrenal insufficiency, particularly before steroid treatment is begun. Weight loss is useful in evaluating the significance of weakness and malaise. Racial pigmentation may be a problem, but a recent and progressive increase in pigmentation is usually reported by the patient with gradual adrenal destruction. Hyperpigmentation is usually absent when adrenal destruction is rapid, as in bilateral adrenal hemorrhage. The fact that hyperpigmentation occurs with other diseases may also present a problem, but the appearance and distribution of pigment in adrenal insufficiency are usually characteristic. When doubt exists, measurement of ACTH levels and testing of adrenal reserve with the infusion of ACTH provide clear-cut differentiation. TREATMENT All patients with adrenal insufficiency should receive specific hormone replacement. Like diabetics, these patients require careful education about the disease. Replacement therapy should correct both glucocorticoid and mineralocorticoid deficiencies.
Hydrocortisone (cortisol) is the mainstay of treatment. The dose for most adults (depending on size) is 20 to 30 mg/d. Patients are advised to take glucocorticoids with meals or, if that is impractical, with milk or an antacid, because the drugs may increase gastric acidity and exert direct toxic effects on the gastric mucosa. To simulate the normal diurnal adrenal rhythm, two-thirds of the dose is taken in the morning, and the remaining one-third is taken in the late afternoon. Some patients exhibit insomnia, irritability, and mental excitement after initiation of therapy; in these, the dosage should be reduced. Other situations that may necessitate smaller doses are hypertension and diabetes mellitus. Obese individuals and those on anticonvulsive medications may require increased dosages. Measurements of plasmaACTH or cortisol or of urine cortisol levels do not appear to be useful in determining optimal glucocorticoid dosages. Since the replacement dosage of hydrocortisone does not replace the mineralocorticoid component of the adrenal hormones, mineralocorticoid supplementation is usually needed. This is accomplished by the administration of 0.05 to 0.1 mg fludrocortisone per day by mouth. Patients should also be instructed to maintain an ample intake of sodium (3 to 4 g/d). The adequacy of mineralocorticoid therapy can be assessed by measurement of blood pressure and serum electrolytes. Blood pressure should be normal and without postural changes; serum sodium, potassium, creatinine, and urea nitrogen levels should also be normal. Measurement of plasma renin levels may also be useful in titrating the dose. In female patients with adrenal insufficiency, androgen levels are also low. Thus, some physicians believe that daily replacement with 25 to 50 mg ofDHEAorally may improve quality of life and skeletal density. Complications of glucocorticoid therapy, with the exception of gastritis, are rare at the dosages recommended for treatment of adrenal insufficiency. Complications of mineralocorticoid therapy include hypokalemia, hypertension, cardiac enlargement, and even congestive heart failure due to sodium retention. Periodic measurements of body weight, serum potassium level, and blood pressure are useful. All patients with adrenal insufficiency should carry medical identification, should be instructed in the parenteral self-administration of steroids, and should be registered with a medical alerting system. Special Therapeutic Problems During periods of intercurrent illness, especially in the setting of fever, the dose of hydrocortisone should be doubled. With severe illness it should be increased to 75 to 150 mg/d. When oral administration is not possible, parenteral routes should be employed. Likewise, before surgery or dental extractions, supplemental glucocorticoids should be administered. Patients should also be advised to increase the dose of fludrocortisone and to add salt to their otherwise normal diet during periods of strenuous exercise with sweating, during extremely hot weather, and with gastrointestinal upsets such as diarrhea. A simple strategy is to supplement the diet one to three times daily with salty broth (1 cup of beef or chicken bouillon contains 35 mmol of sodium). For a representative program of steroid therapy for the patient with adrenal insufficiency who is undergoing major surgery, seeTable 331-8. This schedule is designed so that on the day of surgery it will mimic the output of cortisol in normal individuals undergoing prolonged major stress (10 mg/h, 250 to 300 mg/d). Thereafter, if the patient is improving and is afebrile, the dose of hydrocortisone is tapered by 20 to
30% daily. Mineralocorticoid administration is unnecessary at hydrocortisone doses >100 mg/d because of the mineralocorticoid effects of hydrocortisone at such dosages. SECONDARY ADRENOCORTICAL INSUFFICIENCY ACTHdeficiency causes secondary adrenocortical insufficiency; it may be a selective deficiency, as is seen following prolonged administration of excess glucocorticoids, or it may occur in association with deficiencies of multiple pituitary hormones (panhypopituitarism) (Chap. 328). Patients with secondary adrenocortical hypofunction have many symptoms and signs in common with those having primary disease but are characteristically not hyperpigmented, since ACTH and related peptide levels are low. In fact, plasma ACTH levels distinguish between primary and secondary adrenal insufficiency, since they are elevated in the former and decreased to absent in the latter. Patients with total pituitary insufficiency have manifestations of multiple hormone deficiencies. An additional feature distinguishing primary adrenocortical insufficiency is the near-normal level of aldosterone secretion seen in pituitary and/or isolated ACTH deficiencies (Fig. 331-11). Patients with pituitary insufficiency may have hyponatremia, which can be dilutional or secondary to a subnormal increase in aldosterone secretion in response to severe sodium restriction. However, severe dehydration, hyponatremia, and hyperkalemia are characteristic of severe mineralocorticoid insufficiency and favor a diagnosis of primary adrenocortical insufficiency. Patients receiving long-term steroid therapy, despite physical findings of Cushing's syndrome, develop adrenal insufficiency because of prolonged pituitary-hypothalamic suppression and adrenal atrophy secondary to the loss of endogenousACTH. These patients have two deficits, a loss of adrenal responsiveness to ACTH and a failure of pituitary ACTH release. They are characterized by low blood cortisol and ACTH levels, a low baseline rate of steroid excretion, and abnormal ACTH and metyrapone responses. Most patients with steroid-induced adrenal insufficiency eventually recover normal hypothalamic-pituitary-adrenal responsiveness, but recovery time varies from days to months. The rapid ACTH test provides a convenient assessment of recovery of hypothalamic-pituitary-adrenal function. Because the plasma cortisol concentrations after injection of cosyntropin and during insulin-induced hypoglycemia are usually similar, the rapid ACTH test assesses the integrated hypothalamic-pituitary-adrenal function (see "Tests of Pituitary-Adrenal Responsiveness," above). Some investigators suggest using the low-dose (1 ug) ACTH test for suspected secondary ACTH deficiency. Additional tests to assess pituitary ACTH reserve include the standard metyrapone and insulin-induced hypoglycemia tests. Glucocorticoid therapy in patients with secondary adrenocortical insufficiency does not differ from that for the primary disorder. Mineralocorticoid therapy is usually not necessary, as aldosterone secretion is preserved. ACUTE ADRENOCORTICAL INSUFFICIENCY Acute adrenocortical insufficiency may result from several processes. On the one hand, adrenal crisis may be a rapid and overwhelming intensification of chronic adrenal insufficiency, usually precipitated by sepsis or surgical stress. Alternatively, acute hemorrhagic destruction of both adrenal glands can occur in previously well subjects. In
children, this event is usually associated with septicemia with Pseudomonas or meningococcemia (Waterhouse-Friderichsen syndrome). In adults, anticoagulant therapy or a coagulation disorder may result in bilateral adrenal hemorrhage. Occasionally, bilateral adrenal hemorrhage in the newborn results from birth trauma. Hemorrhage has been observed during pregnancy, following idiopathic adrenal vein thrombosis, and as a complication of venography (e.g., infarction of an adenoma). The third and most frequent cause of acute insufficiency is the rapid withdrawal of steroids from patients with adrenal atrophy owing to chronic steroid administration. Acute adrenocortical insufficiency may also occur in patients with congenital adrenal hyperplasia or those with decreased adrenocortical reserve when they are given drugs capable of inhibiting steroid synthesis (mitotane, ketoconazole) or of increasing steroid metabolism (phenytoin, rifampin). Adrenal Crisis The long-term survival of patients with adrenocortical insufficiency depends largely on the prevention and treatment of adrenal crisis. Consequently, the occurrence of infection, trauma (including surgery), gastrointestinal upsets, or other stresses necessitates an immediate increase in hormone. In untreated patients, preexisting symptoms are intensified. Nausea, vomiting, and abdominal pain may become intractable. Fever may be severe or absent. Lethargy deepens into somnolence, and hypovolemic vascular collapse ensues. In contrast, patients previously maintained on chronic glucocorticoid therapy may not exhibit dehydration or hypotension until they are in a preterminal state, since mineralocorticoid secretion is usually preserved. In all patients in crisis, a precipitating cause should be sought. TREATMENT Treatment is directed primarily toward repletion of circulating glucocorticoids and replacement of the sodium and water deficits. Hence an intravenous infusion of 5% glucose in normal saline solution should be started with a bolus intravenous infusion of 100 mg hydrocortisone followed by a continuous infusion of hydrocortisone at a rate of 10 mg/h. An alternative approach is to administer a 100-mg bolus of hydrocortisone intravenously every 6 h. However, only continuous infusion maintains the plasma cortisol constantly at stress levels [>830 nmol/L (30 ug/dL)]. Effective treatment of hypotension requires glucocorticoid replacement and repletion of sodium and water deficits. If the crisis was preceded by prolonged nausea, vomiting, and dehydration, several liters of saline solution may be required in the first few hours. Vasoconstrictive agents (such as dopamine) may be indicated in extreme conditions as adjuncts to volume replacement. With large doses of steroid, e.g., 100 to 200 mg hydrocortisone, the patient receives a maximal mineralocorticoid effect, and supplementary mineralocorticoid is superfluous. Following improvement, the steroid dosage is tapered over the next few days to maintenance levels, and mineralocorticoid therapy is reinstituted if needed (Table 331-8). HYPOALDOSTERONISM Isolated aldosterone deficiency accompanied by normal cortisol production occurs in association with hyporeninism, as an inherited biosynthetic defect, postoperatively following removal of aldosterone-secreting adenomas, during protracted heparin or heparinoid administration, in pretectal disease of the nervous system, and in severe
postural hypotension. The feature common to all forms hypoaldosteronism is the inability to increase aldosterone secretion appropriately in response to salt restriction. Most patients have unexplained hyperkalemia, which often is exacerbated by restriction of dietary sodium intake. In severe cases, urine sodium wastage occurs at a normal salt intake, whereas in milder forms, excessive loss of urine sodium occurs only with salt restriction. Most cases of isolated hypoaldosteronism occur in patients with a deficiency in renin production (so-called hyporeninemic hypoaldosteronism), most commonly in adults with diabetes mellitus and mild renal failure and in whom hyperkalemia and metabolic acidosis are out of proportion to the degree of renal impairment. Plasma renin levels fail to rise normally following sodium restriction and postural changes. The pathogenesis is uncertain. Possibilities include renal disease (the most likely), autonomic neuropathy, extracellular fluid volume expansion, and defective conversion of renin precursors to active renin. Aldosterone levels also fail to rise normally after salt restriction and volume contraction; this effect is probably related to the hyporeninism, since biosynthetic defects in aldosterone secretion usually cannot be demonstrated. In these patients, aldosterone secretion increases promptly afterACTHstimulation, but it is uncertain whether the magnitude of the response is normal. On the other hand, the level of aldosterone appears to be subnormal in relationship to the hyperkalemia. Hypoaldosteronism can also be associated with high renin levels and low or elevated levels of aldosterone (see below). Severely ill patients may also have hyperreninemic hypoaldosteronism; such patients have a high mortality rate (80%). Hyperkalemia is not present. Possible explanations for the hypoaldosteronism include adrenal necrosis (uncommon) or a shift in steroidogenesis from mineralocorticoids to glucocorticoids, possibly related to prolongedACTHstimulation. Before the diagnosis of isolated hypoaldosteronism is considered for a patient with hyperkalemia, "pseudohyperkalemia" (e.g., hemolysis, thrombocytosis) should be excluded by measuring the plasma potassium level. The next step is to demonstrate a normal cortisol response toACTHstimulation. Then, the response of renin and aldosterone levels to stimulation (upright posture, sodium restriction) should be measured. Low renin and aldosterone levels establish the diagnosis of hyporeninemic hypoaldosteronism. A combination of high renin levels and low aldosterone levels is consistent with an aldosterone biosynthetic defect or a selective unresponsiveness to angiotensin II. Finally, there is a condition that clinically and biochemically mimics hypoaldosteronism with elevated renin levels. However, the aldosterone levels are not low but high -- so-called pseudohypoaldosteronism. This inherited condition is caused by a mutation in the epithelial sodium channel (see below). TREATMENT The treatment is to replace the mineralocorticoid deficiency. For practical purposes, the oral administration of 0.05 to 0.15 mg fludrocortisone daily should restore electrolyte balance if salt intake is adequate (e.g., 150 to 200 mmol/d). However, patients with hyporeninemic hypoaldosteronism may require higher doses of mineralocorticoid to correct hyperkalemia. This need poses a potential risk in patients with hypertension,
mild renal insufficiency, or congestive heart failure. An alternative approach is to reduce salt intake and to administer furosemide, which can ameliorate acidosis and hyperkalemia. Occasionally, a combination of these two approaches is efficacious. GENETIC CONSIDERATIONS Glucocorticoid Diseases Congenital Adrenal Hyperplasia Congenital adrenal hyperplasia (CAH) is the consequence of recessive mutations that cause one of several distinct enzymatic defects (see below). Because cortisol is the principal adrenal steroid regulatingACTHelaboration and because ACTH stimulates adrenal growth and function, a block in cortisol synthesis may result in the enhanced secretion of adrenal androgens and/or mineralocorticids depending on the site of the enzyme block. In severe congenital virilizing hyperplasia, the adrenal output of cortisol may be so compromised as to cause adrenal deficiency despite adrenal hyperplasia. CAHis the most common adrenal disorder of infancy and childhood (Chap. 338). Partial enzyme deficiencies can be expressed after adolescence, predominantly in women with hirsutism and oligomenorrhea but minimal virilization. Late-onset adrenal hyperplasia may account for 5 to 25% of cases of hirsutism and oligomenorrhea in women, depending on the population. ETIOLOGY Enzymatic defects have been described in 21-hydroxylase (CYP21A2), 17a-hydroxylase/17,20-Lyase (CYP17), 11b-hydroxylase (CYP11B1), and in (3b-HSD2) (Fig. 331-2). Although the cDNAs for these enzymes have been cloned, the diagnosis of specific enzyme deficiencies with genetic techniques is not practical for routine use. CYP21A2 deficiency is closely linked to the HLA-B locus of chromosome 6 so that HLA typing and/or DNA polymorphism can be used to detect the heterozygous carriers and to diagnose affected individuals in some families (Chap. 306). The clinical expression in the different disorders is variable, ranging from virilization of the female (CYP2/A2) to feminization of the male (3b-HSD2) (Chap. 338). Adrenal virilization in the female at birth is associated with ambiguous external genitalia (female pseudohermaphroditism). Virilization probably begins after the fifth month of intrauterine development. At birth there may be enlarged genitalia in the male infant and enlargement of the clitoris, partial or complete fusion of the labia, and sometimes a urogenital sinus in the female. If the labial fusion is nearly complete, the female infant has external genitalia resembling a penis with hypospadias. In the postnatal period, CAH is associated with virilization in the female and isosexual precocity in the male. The excessive androgen levels result in accelerated growth, so that bone age exceeds chronologic age. Because epiphyseal closure is hastened by excessive androgens, growth stops, but truncal development continues, the characteristic appearance being a short child with a well-developed trunk. The most common form ofCAH (95% of cases) is a result of impairment of CYP21A2. In addition to cortisol deficiency, aldosterone secretion is decreased in approximately one-third of the patients. Thus, with CYP21A2 deficiency, adrenal virilization occurs with or without a salt-losing tendency due to aldosterone deficiency (Fig. 331-2).
CYP11B1 deficiency causes a "hypertensive" variant ofCAH. Hypertension and hypokalemia occur because of the impaired conversion of 11-deoxycorticosterone to corticosterone, resulting in the accumulation of 11-deoxycorticosterone, a potent mineralocorticoid. The degree of hypertension is variable. Increased shunting again occurs into the androgen pathway. CYP17 deficiency is characterized by hypogonadism, hypokalemia, and hypertension. This rare disorder causes decreased production of cortisol and shunting of precursors into the mineralocorticoid pathway with hypokalemic alkalosis, hypertension, and suppressed plasma renin activity. Usually, 11-deoxycorticosterone production is elevated. Because CYP17 hydroxylation is required for biosynthesis of both adrenal androgens and gonadal testosterone and estrogen, this defect is associated with sexual immaturity, high urinary gonadotropin levels, and low urinary 17-ketosteroid excretion. Female patients have primary amenorrhea and lack of development of secondary sexual characteristics. Because of deficient androgen production, male patients have either ambiguous external genitalia or a female phenotype (male pseudohermaphroditism). Exogenous glucocorticoids can correct the hypertensive syndrome, and treatment with appropriate gonadal steroids results in sexual maturation. With 3b-HSD2 deficiency, conversion of pregnenolone to progesterone is impaired, so that the synthesis of both cortisol and aldosterone is blocked, with shunting into the adrenal androgen pathway via 17a-hydroxypregnenolone andDHEA. Because DHEA is a weak androgen, and because this enzyme deficiency is also present in the gonad, the genitalia of the male fetus may be incompletely virilized or feminized. Conversely, in the female, overproduction ofDHEA may produce partial virilization. DIAGNOSIS The diagnosis ofCAHshould be considered in infants having episodes of acute adrenal insufficiency or salt-wasting or with hypertension. The diagnosis is further suggested by the finding of hypertrophy of the clitoris, fused labia, or a urogenital sinus in the female or of isosexual precocity in the male. In infants and children with a CYP21A2 defect, increased urine 17-ketosteroid excretion and increased plasmaDHEAsulfate levels are typically associated with an increase in the blood levels of 17-hydroxyprogesterone and the excretion of its urinary metabolite pregnanetriol. Demonstration of elevated levels of 17-hydroxyprogesterone in amniotic fluid at 14 to 16 weeks of gestation allows prenatal detection of affected female infants. The diagnosis of a salt-losing form ofCAH due to defects in CYP21A2 is suggested by episodes of acute adrenal insufficiency with hyponatremia, hyperkalemia, dehydration, and vomiting. These infants and children often crave salt and have laboratory findings indicating deficits in both cortisol and aldosterone secretion. With the hypertensive form ofCAH due to CYP11B1 deficiency, 11-deoxycorticosterone and 11-deoxycortisol accumulate. The diagnosis is confirmed by demonstrating increased levels of 11-deoxycortisol in the blood or increased amounts of tetrahydro-11-deoxycortisol in the urine. Elevation of 17-hydroxyprogesterone levels does not imply a coexisting CYP21A2 deficiency. Very high levels of urineDHEAwith low levels of pregnanetriol and of cortisol metabolites
in urine are characteristic of children with 3b-HSD2 deficiency. Marked salt-wasting may also occur. Adults with late-onset adrenal hyperplasia (partial deficiency of CYP21A2, CYP11B1, or 3b-HSD2) are characterized by normal or moderately elevated levels of urinary 17-ketosteroids and plasmaDHEAsulfate. A high basal level of a precursor of cortisol biosynthesis (such as 17-hydroxyprogesterone, 17-hydroxypregnenolone, or 11-deoxycortisol), or elevation of such a precursor afterACTHstimulation, confirms the diagnosis of a partial deficiency. Measurement of steroid precursors 60 min after bolus administration of ACTH is usually sufficient. Adrenal androgen output is easily suppressed by the standard low-dose (2 mg) dexamethasone test. TREATMENT Patients with CAHhave a fundamental defect of cortisol deficiency with resultant excessiveACTHsecretion, producing hyperplasia of the adrenal glands and causing additional shunting into the precursor steroid pathways. Therapy in these patients consists of daily administration of glucocorticoids to suppress pituitary ACTH secretion. Because of its cost and intermediate half-life, prednisone is the drug of choice except in infants, in whom hydrocortisone is usually used. In adults with late-onset adrenal hyperplasia, the smallest single bedtime dose of a long- or intermediate-acting glucocorticoid that suppresses pituitary ACTH secretion should be administered. The amount of steroid required by children with CAH is approximately 1 to 1.5 times the normal cortisol production rate of 27 to 35 umol (10 to 13 mg) of cortisol per square meter of body surface per day and is given in divided doses two or three times per day. The dosage schedule is governed by repetitive analysis of the urinary 17-ketosteroids, plasmaDHEAsulfate, and/or precursors of cortisol biosynthesis. Skeletal growth and maturation must also be monitored closely, as overtreatment with glucocorticoid replacement therapy retards linear growth. Receptor Mutations Much less common thanCAH are three syndromes secondary to mutation(s) in a key receptor involved in adrenal function. Isolated glucocorticoid deficiency is a rare autosomal recessive disease secondary to a mutation in theACTHreceptor. Usually mineralocorticoid function is normal. Adrenal insufficiency is manifest within the first 2 years of life usually as hyperpigmentation, convulsions, and/or frequent episodes of hypoglycemia. In some patients the adrenal insufficiency is associated with achalasia and alacrima -- Allgrove's, or triple A, syndrome. However, in some triple A syndrome patients, no mutation in the ACTH receptor has been identified, suggesting that a distinct genetic abnormality causes this syndrome. Adrenal hypoplasia congenita is a rare X-linked disorder caused by a mutation in the DAX1 gene located on the X chromosome. This gene encodes an orphan nuclear receptor that plays an important role in the development of the adrenal cortex and also the hypothalamic-pituitary-gonadal axis. Thus, patients present with signs and symptoms secondary to deficiencies of all three major adrenal steroids -- cortisol, aldosterone, and adrenal androgens -- as well as gonadotropin deficiency. Finally a rare cause of hypercortisolism without cushingoid stigmata is primary cortisol resistance due to mutations in the glucocorticoid receptor. The resistance is incomplete because patients do not exhibit signs of adrenal insufficiency. Thus, these three rare inherited disorders have in common an elevated ACTH. However, the clinical manifestations range from no
evidence of adrenal insufficiency to only cortisol deficiency (similar to secondary adrenal insufficiency) to a clinical picture indistinguishable from classic Addison's disease. Miscellaneous Conditions Adrenoleukodystrophy causes severe demyelination and early death in children, and adrenomyeloneuropathy is associated with a mixed motor and sensory neuropathy with spastic paraplegia in adults; both disorders are associated with elevated circulating levels of very long chain fatty acids and cause adrenal insufficiency. Autosomal recessive mutations in the steroidogenic acute regulatory (STAR) protein gene cause congenital lipoid adrenal hyperplasia (Chap. 338), which is characterized by adrenal insufficiency and defective gonadal steroidogenesis. Because STAR mediates cholesterol transport into the mitochondrion, mutations in the protein cause massive lipid accumulation in steroidogenic cells, ultimately leading to cell toxicity. Thus, these three rare inherited disorders have in common an elevated ACTH. However, the clinical manifestations range from no evidence of adrenal insufficiency to only cortisol deficiency (similar to secondary adrenal insufficiency) to a clinical picture indistinguishable from classic Addison's disease. MINERALOCORTICOID DISEASES Some forms ofCAH have a mineralocorticoid component (see above). Others are caused by a mutation in other enzymes or ion channels important in mediating or mimicking aldosterone's action. Hypermineralocorticoidism Low Plasma Renin Activity Rarely, hypermineralocorticoidism is due to a defect in cortisol biosynthesis, specifically 11- or 17-hydroxylation.ACTHlevels are increased, with a resultant increase in the production of the mineralocorticoid 11-deoxycorticosterone. Hypertension and hypokalemia can be corrected by glucocorticoid administration. The definitive diagnosis is made by demonstrating an elevation of precursors of cortisol biosynthesis in the blood or urine or by direct demonstration of the genetic defect. Glucocorticoid administration can also ameliorate hypertension or produce normotension even though a hydroxylase deficiency cannot be identified (Fig. 331-9). These patients have normal to slightly elevated aldosterone levels that do not suppress in response to saline but do suppress in response to 2 days of dexamethasone (2 mg/d). The condition is inherited as an autosomal dominant trait and is termed glucocorticoid-remediable aldosteronism (GRA). This entity is secondary to a chimeric gene duplication whereby the 11-bhydroxylase gene promoter (which is under the control ofACTH) is fused to the aldosterone synthase coding sequence. Thus, aldosterone synthase activity is ectopically expressed in the zona fasciculata and is regulated by ACTH, in a fashion similar to the regulation of cortisol secretion. Screening for this defect is best performed by assessing the presence or absence of the chimeric gene. Because the abnormal gene may be present in the absence of hypokalemia, its frequency as a cause of hypertension is unknown. Individuals with suppressed plasma renin levels and juvenile-onset hypertension or a history of early-onset hypertension in first-degree relatives should be screened for this disorder. Early hemorrhagic stroke also occurs in GRA-affected individuals.
Glucocorticoid-remediable hyperaldosteronism documented by genetic analysis may be treated with glucocorticoid administration or antimineralocorticoids, e.g., spironolactone, triamterene, or amiloride. Glucocorticoids should be used only in small doses to avoid inducing iatrogenic Cushing's syndrome. A combination approach is often necessary. High Plasma Renin Activity Bartter's syndrome is characterized by severe hyperaldosteronism (hypokalemic alkalosis) with moderate to marked increases in renin activity and hypercalciuria, but normal blood pressure and no edema; this disorder usually begins in childhood. Renal biopsy shows juxtaglomerular hyperplasia. The pathogenesis involves a defect in the renal conservation of sodium or chloride. The renal loss of sodium is thought to stimulate renin secretion and aldosterone production. Hyperaldosteronism produces potassium depletion, and hypokalemia further elevates prostaglandin production and plasma renin activity. In some cases, the hypokalemia may be potentiated by a defect in renal conservation of potassium. Increased production of prostaglandins is probably not a primary abnormality, since administration of inhibitors of prostaglandin synthesis reverses the features only temporarily (Chap. 276). Bartter's syndrome is caused by a mutation in the renal Na-K-2Cl co-transporter gene. Gitelman's syndrome is an autosomal recessive trait characterized by renal salt wasting and as a result, as in Bartter's syndrome, activation of the renin-angiotensin-aldosterone system. As a consequence affected individuals have low blood pressure, low serum potassium, low serum magnesium, and high serum bicarbonate. In contrast to Bartter's syndrome, urinary calcium excretion is reduced. Gitelman's syndrome results from loss-of-function mutations of the renal thiazide-sensitive Na-Cl co-transporter. Increased Mineralocorticoid Action Liddle's syndrome is a rare autosomal dominant disorder that mimicks hyperaldosteronism. The defect is in the genes encoding theb or g subunits of the epithelial sodium channel. Both renin and aldosterone levels are low, owing to the constitutively activated sodium channel and the resulting excess sodium reabsorption in the renal tubule. A rare autosomal recessive cause of hypokalemia and hypertension is 11b-HSDII deficiency, in which cortisol cannot be converted to cortisone and hence binds to theMR and acts as a mineralocorticoid. This condition, also termed apparent mineralocorticoid excess syndrome, is caused by a defect in the gene encoding the renal isoform of this enzyme, 11b-HSD II. Patients can be identified either by documenting an increased ratio of cortisol to cortisone in the urine or by genetic analysis. Patients with the 11b-HSD deficiency syndrome can be treated with small doses of dexamethasone. Although dexamethasone is a potent glucocorticoid that suppressesACTH and endogenous cortisol production, it binds less well to the mineralocorticoid receptor than does cortisol. The ingestion of candies or chewing tobacco containing certain forms of licorice produces a syndrome that mimics primary aldosteronism. The component of such agents that causes sodium retention is glycyrrhizinic acid, which inhibits the 11b-HSDII and hence allows cortisol to act as a mineralocorticoid and cause sodium retention, expansion of the extracellular fluid volume, hypertension, depressed plasma renin levels, and suppressed aldosterone levels. The diagnosis is established or excluded by
a careful history. Decreased Mineralocorticoid Production or Action In patients with these conditions, disorders of aldosterone biosynthesis or action are associated with high renin levels, salt wasting, and hyperkalemia. The aldosterone levels may be low or elevated. In patients with a deficiency in aldosterone biosynthesis, the transformation of corticosterone into aldosterone is impaired, owing to a mutation in the aldosterone synthase (CYP11B2) gene. These patients have low to absent aldosterone secretion, elevated plasma renin levels, and elevated levels of the intermediates of aldosterone biosynthesis (corticosterone and 18-hydroxycorticosterone). Pseudohypoaldosteronism type I (PHA-I) is an autosomal recessive disorder that is seen in the neonatal period and is characterized by salt wasting, hypotension, hyperkalemia, and high renin and aldosterone levels. In contrast to the gain-of-function mutations in the epithelial sodium channel (ENaC) in Liddle's syndrome, mutations in PHA-I result in loss of ENaC function. NONSPECIFIC CLINICAL USE OF ADRENAL STEROIDS The widespread use of glucocorticoids emphasizes the need for a thorough understanding of the metabolic effects of these agents. Before adrenal hormone therapy is instituted, the expected gains should be weighed against undesirable effects. HOW SERIOUS IS THE DISORDER? In a patient who has unexplained shock or in whom other measures have failed, the physician need not hesitate to employ high-dose steroid therapy. In contrast, one should exercise restraint in administering steroids to a patient with early rheumatoid arthritis for whom physiotherapy, anti-inflammatory agents, disease-modifying agents, and general medical care have not been tried (Chap. 312). HOW LONG WILL GLUCOCORTICOID THERAPY BE REQUIRED? The use of intravenous steroids for 24 to 48 h for a life-threatening situation such as status asthmaticus or pseudotumor cerebri has few or no contraindications, in contrast to the initiation of chronic steroid therapy for asthma, arthritis, or psoriasis. In the latter instances, the almost certain development of some degree of Cushing's syndrome must be weighed against the potential benefit. These side effects should be minimized by a careful choice of steroid preparations, alternate-day or interrupted therapy; the use of topical steroids, e.g., inhaled, intranasal, or dermal, whenever possible; and the judicious use of supplementary adjuvants. WHICH PREPARATION IS BEST? Several considerations should be taken into account in deciding which steroid preparation to use. 1. The biologic half-life. The rationale behind alternate-day therapy is to decrease the metabolic effects of the steroids for a significant part of each 48 h period while still
producing a pharmacologic effect durable enough to be effective. Too long a half-life would defeat the first purpose, and too short a half-life would defeat the second. In general, the more potent the steroid, the longer its biologic half-life. 2. The importance of the mineralocorticoid effects of the steroid. Most synthetic steroids have less mineralocorticoid effect than hydrocortisone (Table 331-9). 3. The biologically active form of the steroid. Cortisone and prednisone have to be converted to biologically active metabolites before anti-inflammatory effects can occur. Because of this, in a condition for which steroids are known to be effective and when an adequate dose has been given without response, one should consider substituting hydrocortisone or prednisolone for cortisone or prednisone. 4. The cost of the medication. This is a serious consideration if chronic administration is planned. Prednisone is the least expensive of available steroid preparations. 5. The type of formulation. Topical steroids have the distinct advantage over oral steroids in reducing the likelihood of systemic side effects. In addition, some inhaled steroids have been designed to minimize side effects by increasing their hepatic inactivation if they are swallowed (Chap. 252). However, all topical steroids can be absorbed into the systemic circulation. EVALUATION OF PATIENTS PRIOR TO INITIATING STEROID THERAPY (SeeTable 331-10) Chronic Infection Three issues demand attention: (1) Any active infection, particularly tuberculosis, should be identified. (2) If tuberculosis is present, steroid therapy should be employed only in conjunction with antituberculous chemotherapy. The chest film and tuberculin test provide baseline information for future comparison. Since high-dose steroids may impair the tuberculin reaction, serial chest roentgenograms may be necessary. (3) Infection due to opportunistic pathogens should be constantly considered in patients on steroid therapy, especially when combined with other immunosuppressive agents. Diabetes Mellitus Prolonged glucocorticoid therapy may unmask or aggravate diabetes mellitus. The presence of diabetes mellitus or the demonstration of impaired glucose tolerance may affect the decision to institute glucocorticoid therapy. Osteoporosis Patients receiving long-term steroid therapy are at risk for osteoporosis. Indeed, osteoporosis with vertebral fractures or compression is a dreaded complication for patients at high risk (postmenopausal women, elderly men, patients with restricted physical activity, and especially organ transplant patients who are maintained on high doses to prevent rejection). Alternate-day or interrupted steroid therapy does not prevent the risk of this complication (Table 331-11). Adjunctive therapies with antiresorptive agents, such as parenteral or oral bisphosphonates, have been shown to be effective in treating the osteoporosis (Chap. 342). Bone mineral density should be assessed periodically with dual-energy X-ray absorptiometry (DEXA) scans. Peptic Ulcer, Gastric Hypersecretion, or Esophagitis In conventional doses
(equivalent to£15 mg prednisone per day), glucocorticoids probably do not cause peptic ulceration; whether higher doses cause ulcer disease is not established and probably depends on duration, dose of treatment, and predisposing factors such as hypoalbuminemia or cirrhosis and also whether subjects are concomitantly ingesting nonsteroidal anti-inflammatory agents (NSAIDs). However, even at conventional doses, patients with a history of ulcer may experience aggravation of symptoms while receiving glucocorticoids. Consequently, all individuals with a positive history or with known risk factors should be managed with a vigorous antiulcer program (antacids, H2receptor antagonists, or ATPase inhibitors) along with glucocorticoids. The development of anemia in a patient receiving glucocorticoids should suggest gastrointestinal bleeding as a cause. Hypertension or Cardiovascular Disease The capacity of many adrenal steroid preparations to promote sodium retention makes it necessary to exercise caution when using them in patients with preexisting hypertension or cardiovascular or renal disease. The use of preparations with minimal sodium-retaining activity, restriction of dietary sodium intake, and the use of diuretic agents and supplementary potassium salts minimize the mineralocorticoid effects of steroids. However, hypertension may be exacerbated by steroid-induced increases in renin substrate and angiotensin II levels and by reduction in vasodilator prostaglandin production. Steroids also accelerate atherogenesis by induction of hypertension, glucose intolerance, and unfavorable lipid profiles. Glucocorticoid-associated lipid abnormalities include hypertriglyceridemia, hypercholesterolemia, and increasedLDLcholesterol levels. Psychological Difficulties Steroid therapy may cause psychological disturbances. In general, these disturbances correlate better with the patient's personality than with the dose of hormone, although larger doses cause more serious reactions. There is no reliable method of predicting the psychological reaction to steroid therapy; moreover, previous tolerance of steroids does not necessarily ensure the safety of subsequent courses. Likewise, untoward psychological reactions on one occasion do not invariably mean that the patient will respond unfavorably to a second course. However, patients with depressive symptoms during a first course of steroids may benefit from prophylactic treatment prior to a second course. Sleeplessness is common and can be minimized by using the shorter-acting steroids and by prescribing the total daily amount as a single early-morning dose. ALTERNATE-DAY STEROID THERAPY The most effective way to minimize the cushingoid effects of glucocorticoids is to administer the total 48-h dose as a single dose of intermediate-acting steroid in the morning, every other day. If symptoms of the underlying disorder can be controlled by this technique, it offers distinct advantages. Three considerations deserve mention: (1) The alternate-day schedule may be approached through transition schedules that allow the patient to adjust gradually; (2) supplementary nonsteroid medications may be needed on the "off" day to minimize symptoms of the underlying disorder; and (3) many symptoms that occur during the off day (e.g., fatigue, joint pain, muscle stiffness or tenderness, and fever) may represent relative adrenal insufficiency rather than exacerbation of the underlying disease.
The alternate-day approach capitalizes on the fact that cortisol secretion and plasma levels normally are highest in the early morning and lowest in the evening. The normal pattern is mimicked by administering an intermediate-acting steroid in the morning (7 to 8 A.M.) (Table 331-9). Initially, the steroid program often requires daily or more frequent doses of steroid to achieve the desired anti-inflammatory or immunity-suppressing action. Only after this desired effect is achieved is an attempt made to switch to an alternate-day program. A number of schedules can be used for transferring from a daily to an alternate-day program. The key points to be considered are flexibility in arranging a program and the use of supportive measures on the off day. One may attempt a gradual transition to the alternate-day schedule rather than an abrupt changeover. One approach is to keep the steroid dose constant on one day and gradually reduce it on the alternate day. Alternatively, the steroid dose can be increased on one day and reduced on the alternate day. In any case, it is important to anticipate that some increase in pain or discomfort may occur in the 36 to 48 h following the last dose. WITHDRAWAL OF GLUCOCORTICOIDS FOLLOWING LONG-TERM USE It is possible to reduce gradually and eventually to discontinue a daily steroid dose, but under most circumstances withdrawal of steroids should be initiated by first implementing an alternate-day schedule. Patients who have been on an alternate-day program for a month or more experience less difficulty during termination regimens. The dosage is gradually reduced and finally discontinued after a replacement dosage has been reached (e.g., 5 to 7.5 mg prednisone). Complications rarely ensue unless undue stress is experienced, and patients should understand that for 1 year or longer after withdrawal from long-term high-dose steroid therapy, supplementary hormone should be given in the event of a serious infection, operation, or injury. A useful strategy in patients with symptoms of adrenal insufficiency on a tapering regimen is to measure plasma cortisol levels prior to the steroid dose. A level150mmol/L (150 meq/L), 0.45% saline should be used. After hemodynamic stability is achieved, the intravenous fluid administration is directed at reversing the free water deficit using hypotonic fluids (0.45% saline initially then 5% dextrose in water, D5W). The calculated free water deficit (which averages 9 to 10 L) should be reversed over the next 1 to 2 days (infusion rates of 200 to 300 mL/h of hypotonic solution). Potassium repletion is usually necessary and should be dictated by repeated measurements of the serum potassium. In patients taking diuretics, the potassium deficit can be quite large and may be accompanied by magnesium deficiency. Hypophosphatemia may occur during therapy and can be improved by using KPO4 and beginning nutrition. As inDKA, rehydration and volume expansion lower the plasma glucose initially, but insulin is eventually required. InNKHS, patients tend to be more sensitive to insulin than in DKA and dose requirements are not usually as large. A reasonable regimen for NKHS begins with an intravenous insulin bolus of 5 to 10 units followed by intravenous insulin at a constant infusion rate (3 to 7 units/h). As in DKA, glucose should be added to intravenous fluid when the plasma glucose falls to 13.9 mmol/L (250 mg/dL), and the insulin infusion rate should be decreased to 1 to 2 units/h. The insulin infusion should be continued until the patient has resumed eating and can be transferred to a subcutaneous insulin regimen. The patient should be discharged from the hospital on insulin, though some patients can later undergo a trial of oral glucose-lowering agents. CHRONIC COMPLICATIONS The chronic complications ofDMaffect many organ systems and are responsible for the majority of morbidity and mortality associated with the disease. Chronic complications can be divided into vascular and nonvascular complications (Table 333-7). The vascular complications of DM are further subdivided into microvascular (retinopathy, neuropathy, nephropathy) and macrovascular complications (coronary artery disease, peripheral vascular disease, cerebrovascular disease). Nonvascular complications include problems such as gastroparesis, sexual dysfunction, and skin changes. This division is rather arbitrary since it is likely that multiple pathogenic processes are involved in all forms of complications. The risk of chronic complications increases as a function of the duration of hyperglycemia; they usually become apparent in the second decade of hyperglycemia. Since type 2DM may have a long asymptomatic period of hyperglycemia, many individuals with type 2 DM have complications at the time of diagnosis.
The microvascular complications of both type 1 and type 2DMresult from chronic hyperglycemia. Randomized, prospective clinical trials involving large numbers of individuals with type 1 or type 2 DM have conclusively demonstrated that a reduction in chronic hyperglycemia prevents or reduces retinopathy, neuropathy, and nephropathy. Other incompletely defined factors also modulate the development of complications. For example, despite longstanding DM, some individuals never develop nephropathy or retinopathy. Many of these patients have glycemic control that is indistinguishable from those who develop microvascular complications. Because of these observations, it is suspected that a genetic susceptibility for developing particular complications exists. However, the genetic loci responsible for these susceptibilities have not yet been identified. Evidence implicating a causative role for chronic hyperglycemia in the development of macrovascular complications is less conclusive, but some results suggest a role for chronic hyperglycemia in the development of macrovascular disease. For example, coronary heart disease events and mortality are two to four times greater in patients with type 2DM. These events correlate with fasting and postprandial plasma glucose levels as well as with the HbA1c. Other factors (dyslipidemia and hypertension) also play important roles in macrovascular complications. MECHANISMS OF COMPLICATIONS Although chronic hyperglycemia is an important etiologic factor leading to complications ofDM, the mechanism(s) by which it leads to such diverse cellular and organ dysfunction is unknown. Three major theories, which are not mutually exclusive, have been proposed to explain how hyperglycemia might lead to the chronic complications of DM (Fig. 333-7). One hypothesis is that increased intracellular glucose leads to the formation of advanced glycosylation end products (AGEs) via the nonenzymatic glycosylaton of cellular proteins. Nonenzymatic glycosylation results from the interaction of glucose with amino groups on proteins. AGEs have been shown to cross-link proteins (e.g., collagen, extracellular matrix proteins), accelerate atherosclerosis, promote glomerular dysfunction, reduce nitric oxide synthesis, induce endothelial dysfunction, and alter extracellular matrix composition and structure. The serum level of AGEs correlates with the level of glycemia, and these products accumulate as glomerular filtration rate declines. A second hypothesis proposed to explain how chronic hyperglycemia leads to complications ofDM is based on the observation that hyperglycemia increases glucose metabolism via the sorbitol pathway. Intracellular glucose is predominantly metabolized by phosphorylation and subsequent glycolysis, but when intracellular glucose is increased, some glucose is converted to sorbitol by the enzyme aldose reductase. Increased sorbitol concentrations affect several aspects of cellular physiology (decreased myoinositol, altered redox potential) and may lead to cellular dysfunction. However, testing of this theory in humans, using aldose reductase inhibitors, has not demonstrated beneficial effects on clinical endpoints of retinopathy, neuropathy, or nephropathy.
A third hypothesis proposes that hyperglycemia increases the formation of diacylglycerol leading to activation of certain isoforms of protein kinase C (PKC), which, in turn, affect a variety of cellular events that lead toDM-related complications. For example, PKC activation by glucose alters the transcription of genes for fibronectin, type IV collagen, contractile proteins, and extracellular matrix proteins in endothelial cells and neurons in vitro. Growth factors appear to play an important role in DM-related complications. Vascular endothelial growth factor (VEGF) is increased locally in diabetic proliferative retinopathy and decreases after laser photocoagulation. Transforming growth factor b (TGF-b) is increased in diabetic nephropathy and appears to stimulate basement membrane production of collagen and fibronectin by mesangial cells. Other growth factors, such as platelet-derived growth factor, epidermal growth factor, insulin-like growth factor I, growth hormone, basic fibroblast growth factor, and even insulin, have been suggested to play a role in DM-related complications. Although hyperglycemia serves as the initial trigger for complications of diabetes, it is still unknown whether the same pathophysiologic processes are operative in all complications or whether certain processes predominate in certain organs. Finally, oxidative stress and free radical generation, as a consequence of the hyperglycemia, may also promote the development of complications. GLYCEMIC CONTROL AND COMPLICATIONS The Diabetes Control and Complications Trial (DCCT) provided definitive proof that reduction in chronic hyperglycemia can prevent many of the early complications of type 1DM. This large multicenter clinical trial randomized over 1400 individuals with type 1 DM to either intensive or conventional diabetes management, and then evaluated the development of retinopathy, nephropathy, and neuropathy. Individuals in the intensive diabetes management group received multiple administrations of insulin each day along with intense educational, psychological, and medical support. Individuals in the conventional diabetes management group received twice daily insulin injections and quarterly nutritional, educational, and clinical evaluation. The goal in the former group was normoglycemia; the goal in the latter group was prevention of symptoms of diabetes. Individuals in the intensive diabetes management group achieved a substantially lower HbA1c (7.2%) than individuals in the conventional diabetes management group (HbA1c of 9.0%). Results from theDCCTdemonstrated that improvement of glycemic control reduced nonproliferative and proliferative retinopathy (47% reduction), microalbuminuria (39% reduction), clinical nephropathy (54% reduction), and neuropathy (60% reduction). Improved glycemic control also slowed the progression of early diabetic complications. There was a nonsignificant trend in reduction of macrovascular events. The results of the DCCT predicted that individuals in the intensive diabetes management group would gain 7.7 additional years of sight, 5.8 additional years free from end-stage renal disease (ESRD), and 5.6 years free from lower extremity amputations. If all complications ofDM were combined, individuals in the intensive diabetes management group would experience 15.3 more years of life without significant microvascular or neurologic complications of DM as compared to individuals who received standard therapy. This translates into an additional 5.1 years of life expectancy for individuals in the intensive diabetes management group. The benefit of the improved glycemic control during the
DCCT persisted even after the study concluded and glycemic control worsened. The benefits of an improvement in glycemic control occurred over the entire range of HbA1c values (Fig. 333-8), suggesting that at any HbA1c level, an improvement in glycemic control is beneficial. Therefore, there is no threshold beneath which the HbA1c can be reduced and the complications ofDMprevented. The clinical implication of this finding is that the goal of therapy is to achieve an HbA1c level as close to normal as possible, without subjecting the patient to excessive risk of hypoglycemia. Considerable debate has emerged as to whether theDCCTfindings are applicable to individuals with type 2DM, in whom insulin resistance, hyperinsulinemia, and obesity predominate. Concerns have been raised that therapies associated with weight gain and additional insulin therapy may worsen underlying insulin resistance and hyperinsulinemia. Despite these concerns, most available data support extrapolation of the results of the DCCT to individuals with type 2 DM. The United Kingdom Prospective Diabetes Study (UKPDS) studied the course of>5000 individuals with type 2DM for>10 years. This complex and important study utilized multiple treatment regimens and monitored the effect of intensive glycemic control and risk factor treatment on the development of diabetic complications. Newly diagnosed individuals with type 2 DM were randomized to (1) intensive management using various combinations of insulin, a sulfonylurea, or metformin; or (2) conventional therapy using dietary modification and pharmacotherapy with the goal of symptom prevention. In addition, individuals were randomly assigned to different antihypertensive regimens. Individuals in the intensive treatment arm achieved an HbA1c of 7.0%, compared to a 7.9% HbA1c in the standard treatment group. The UKPDS demonstrated that each percentage point reduction in HbA1c was associated with a 35% reduction in microvascular complications, a 25% reduction in DM-related deaths, and a 7% reduction in all-cause mortality. As in theDCCT, there was a continuous relationship between glycemic control and development of complications. Although there was no statistically significant effect of glycemic control on cardiovascular complications, there was a 16% reduction in fatal and nonfatal myocardial infarctions. One of the major findings of theUKPDSwas the observation that strict blood pressure control significantly reduced both macro- and microvascular complications. In fact, the beneficial effects of blood pressure control were greater than the beneficial effects of glycemic control. Lowering blood pressure to moderate goals (144/82 mmHg) reduced the risk ofDM-related death, stroke, microvascular end points, retinopathy, and heart failure (risk reductions between 32 and 56%). Improved glycemic control did not conclusively reduce (nor worsen) cardiovascular mortality but was associated with improvement with lipoprotein risk profiles, such as reduced triglycerides and increased high-density lipoprotein (HDL). Similar reductions in the risks of retinopathy and nephropathy were also seen in a small trial of lean Japanese individuals with type 2DMrandomized to either intensive glycemic control or standard therapy with insulin (Kumamoto study). These results demonstrate the effectiveness of improved glycemic control in individuals of different ethnicity with a presumably different etiology of DM (i.e., phenotypically different from those in theDCCT andUKPDS).
The findings of theDCCT,UKPDS, and Kumamoto study support the idea that chronic hyperglycemia plays a causative role in the pathogenesis of diabetic microvascular complications. These landmark studies prove the value of metabolic control and emphasize the importance of (1) intensive glycemic control in all forms ofDM, and (2) early diagnosis and strict blood pressure control in type 2 DM. OPHTHALMOLOGIC COMPLICATIONS OF DIABETES MELLITUS DMis the leading cause of blindness between the ages of 20 and 74 in the United States. The gravity of this problem is highlighted by the finding that individuals with DM are 25 times more likely to become legally blind than individuals without DM. Blindness is primarily the result of progressive diabetic retinopathy and clinically significant macular edema. Diabetic retinopathy is classified into two stages: nonproliferative and proliferative. Nonproliferative diabetic retinopathy usually appears late in the first decade or early in the second decade of the disease and is marked by retinal vascular microaneurysms, blot hemorrhages, and cotton wool spots (seePlate IV-15). Mild nonproliferative retinopathy progresses to more extensive disease, characterized by changes in venous vessel caliber, intraretinal microvascular abnormalities, and more numerous microaneurysms and hemorrhages. The pathophysiologic mechanisms invoked in nonproliferative retinopathy include loss of retinal pericytes, increased retinal vascular permeability, alterations in retinal blood flow, and abnormal retinal microvasculature, all of which lead to retinal ischemia. The appearance of neovascularization in response to retinal hypoxia is the hallmark of proliferative diabetic retinopathy. These newly formed vessels may appear at the optic nerve and/or macula and rupture easily, leading to vitreous hemorrhage, fibrosis, and ultimately retinal detachment. Not all individuals with nonproliferative retinopathy develop proliferative retinopathy, but the more severe the nonproliferative disease, the greater the chance of evolution to proliferative retinopathy within 5 years. This creates a clear opportunity for early detection and treatment of diabetic retinopathy (discussed below). In contrast, clinically significant macular edema may appear when only nonproliferative retinopathy is present. Fluorescein angiography is often useful to detect macular edema, which is associated with a 25% chance of moderate visual loss over the next 3 years. Duration ofDM and degree of glycemic control are the best predictors of the development of retinopathy. Nonproliferative retinopathy is found in almost all individuals who have had DM for >20 years (25% incidence with 5 years, and 80% incidence with 15 years of type 1 DM). Although there is genetic susceptibility for retinopathy, it confers less influence on the development of retinopathy than either the duration of DM or the degree of glycemic control. TREATMENT The most effective therapy for diabetic retinopathy is prevention. Intensive glycemic control will greatly delay the development or slow the progression of retinopathy in individuals with either type 1 or type 2DM. Paradoxically, during the first 6 to 12 months of improved glycemic control, established diabetic retinopathy may transiently worsen.
Fortunately, this progression is temporary, and in the long term, improved glycemic control is associated with less diabetic retinopathy. Individuals with known retinopathy should be considered candidates for prophylactic photocoagulation when initiating intensive therapy. Once advanced retinopathy is present, improved glycemic control imparts less benefit, though adequate ophthalmologic care can prevent most blindness. Equally as important as glycemic control are regular, comprehensive eye examinations for all individuals with DM. Most diabetic eye disease can be successfully treated if detected early. Routine, nondilated eye examinations by the primary care provider or diabetes specialist are inadequate to detect diabetic eye disease properly. The treatment of diabetic eye disease requires an ophthalmologist experienced in these disorders. Laser photocoagulation is very successful in preserving vision. Proliferative retinopathy is usually treated with panretinal laser photocoagulation, whereas macular edema is treated with focal laser photocoagulation. Although exercise has not been conclusively shown to worsen proliferative diabetic retinopathy, most ophthalmologists advise individuals with advanced diabetic eye disease to limit physical activities associated with repeated Valsalva maneuvers. Aspirin therapy (650 mg/d) does not appear to influence the natural history of diabetic retinopathy, but studies of other antiplatelet agents are under way. RENAL COMPLICATIONS OF DIABETES MELLITUS Diabetic nephropathy is the leading cause ofESRD in the United States and a leading cause ofDM-related morbidity and mortality. Proteinuria in individuals with DM is associated with markedly reduced survival and increased risk of cardiovascular disease. Individuals with diabetic nephropathy almost always have diabetic retinopathy also. Like other microvascular complications, the pathogenesis of diabetic nephropathy is related to chronic hyperglycemia (Fig. 334-7). The mechanisms by which chronic hyperglycemia leads toESRD, though incompletely defined, involve the following: interaction of soluble factors (growth factors, angiotensin II, endothelin,AGEs), hemodynamic alterations in the renal microcirculation (glomerular hyperfiltration, increased glomerular capillary pressure), and structural changes in the glomerulus (increased extracellular matrix, basement membrane thickening, mesangial expansion, fibrosis). Some of these effects may be mediated through angiotensin receptors. Smoking accelerates the decline in renal function. The natural history of diabetic nephropathy is shown schematically inFig. 333-9 and is characterized by a fairly predictable pattern of events. Although this sequence of events was defined for individuals with type 1DM, a similar pattern is also likely in type 2 DM. Glomerular hyperfusion and renal hypertrophy occur in the first years after the onset of DM and are reflected by an increased glomerular filtration rate (GFR). During the first 5 years of DM, thickening of the glomerular basement membrane, glomerular hypertrophy, and mesangial volume expansion occur as the GFR returns to normal. After 5 to 10 years of type 1 DM, ~40% of individuals begin to excrete small amounts of albumin in the urine (microalbuminuria). Microalbuminuria is defined as 30 to 300 mg/d in a 24-h collection or 30 to 300 ug/mg creatinine in a spot collection. The appearance of microalbuminuria (incipient nephropathy) in type 1 DM is a very important predictor of progression to overt proteinuria (>300 mg/d). Blood pressure may rise slightly at this
point but usually remains in the normal range. Once overt proteinuria is present, there is a steady decline in GFR, and ~50% of individuals reachESRD in 7 to 10 years. The early pathologic changes and albumin excretion abnormalities are reversible with normalization of plasma glucose. However, once nephropathy becomes overt, the pathologic changes are likely irreversible. The nephropathy that develops in type 2DMdiffers from that of type 1 DM in the following respects: (1) microalbuminuria or overt nephropathy may be present when type 2 DM is diagnosed, reflecting its long asymptomatic period; (2) hypertension more commonly accompanies microalbuminuria or overt nephropathy in type 2 DM; and (3) microalbuminuria may be less predictive of progression to overt nephropathy in type 2 DM. Finally, it should be noted that albuminuria in type 2 DM may be secondary to factors unrelated to DM, such as hypertension, congestive heart failure, prostate disease, or infection. Other renal problems may also occur in individuals withDM. Type IV renal tubular acidosis (hyporeninemic hypoaldosteronism) occurs in many individuals with DM. These individuals develop a propensity to hyperkalemia, which may be exacerbated by medications [especially angiotensin-converting enzyme (ACE) inhibitors]. Patients with DM are predisposed to radiocontrast-induced nephrotoxicity. Individuals with DM undergoing radiographic procedures with contrast dye should be well hydrated before and after dye exposure, and the serum creatinine should be monitored for several days following the procedure. TREATMENT The optimal therapy for diabetic nephropathy is prevention. As part of comprehensive diabetes care, microalbuminuria should be detected at an early stage when effective therapies can be instituted. The recommended strategy for detecting microalbuminuria is outlined in Fig. 333-10. Interventions effective in slowing progression from microalbuminuria to overt nephropathy include: (1) near normalization of glycemia, (2) strict blood pressure control, and (3) administration ofACEinhibitors. Improved glycemic control reduces the rate at which microalbuminuria appears and progresses in both type 1 and type 2DM. However, once overt nephropathy exists, it is unclear whether improved glycemic control will slow progression of renal disease. During the phase of declining renal function, insulin requirements may fall as the kidney is a site of insulin degradation. Furthermore, glucose-lowering medications (sulfonylureas and metformin) may accumulate and are contraindicated in renal insufficiency. Many individuals with type 1 or type 2DMdevelop hypertension. Numerous studies in both type 1 and type 2 DM demonstrate the effectiveness of strict blood pressure control in reducing albumin excretion and slowing the decline in renal function. Blood pressure should be maintained at 11.1 mmol/L (200 mg/dL), oral agents are usually not efficacious and insulin therapy is required. Short-acting insulin may be required to supplement long-acting insulin in order to control postprandial glucose excursions. REPRODUCTIVE ISSUES Reproductive capacity in either men or women withDMappears to be normal. Menstrual cycles may be associated with alterations in glycemic control in women with DM. Pregnancy is associated with marked insulin resistance; the increased insulin requirements often precipitate DM and lead to the diagnosis ofGDM. Glucose, which at high levels is a teratogen to the developing fetus, readily crosses the placenta, but insulin does not. Thus, hyperglycemia or hypoglycemia from the maternal circulation may stimulate insulin secretion in the fetus. The anabolic and growth effects of insulin may result in macrosomia. GDM complicates approximately 4% of pregnancies in the United States. The incidence of GDM is greatly increased in certain ethnic groups, including African Americans and Hispanic Americans, consistent with a similar increased risk of type 2 DM. Current recommendations advise screening for glucose intolerance between weeks 24 and 28 of pregnancy in women with high risk for GDM (³25 years; obesity; family history of DM; member of an ethnic group such as Hispanic American, Native American, Asian American, African American, or Pacific Islander). Therapy for GDM is similar to that for individuals with pregnancy-associated diabetes and involvesMNT and insulin, if hyperglycemia persists. Oral glucose-lowering agents have not been approved for use during pregnancy. With current practices, the morbidity and mortality of the mother with GDM and the fetus are no different from those in the nondiabetic population. Individuals who develop GDM are at marked increased risk for developing type 2 DM in the future and should be screened periodically for DM. After delivery, glucose homeostasis should be reassessed in the mother. Most individuals with GDM revert to normal glucose tolerance, but some will continue to have overt diabetes or impairment of glucose tolerance. In addition, children of women with GDM appear to be at risk for obesity and glucose intolerance and have an increased risk of diabetes beginning in the later stages of adolescence. Pregnancy in individuals with known DMrequires meticulous planning and adherence to strict treatment regimens. Intensive diabetes management and normalization of the HbA1c are the standard of care for individuals with existing DM who are planning
pregnancy. The crucial period of glycemic control is extremely early following fertilization. The risk of fetal malformations is increased 4 to 10 times in individuals with uncontrolled DM at the time of conception. The goals are normal plasma glucose during the preconception period and throughout the periods of organ development in the fetus. LIPODYSTROPHIC DM (See alsoChap. 354) Lipodystrophy, or the loss of subcutaneous fat tissue, may be generalized in certain genetic conditions such as leprechaunism. Generalized lipodystrophy is associated with severe insulin resistance and is often accompanied by acanthosis nigricans and dyslipidemia. Localized lipodystrophy associated with insulin injections has been reduced considerably by the use of human insulin. Protease Inhibitors and Lipodystrophy Protease inhibitors used in the treatment of HIV disease (Chap. 309) have been associated with a centripetal accumulation of fat (visceral and abdominal area), accumulation of fat in the dorsocervical region, loss of extremity fat, decreased insulin sensitivity (elevations of the fasting insulin level and reduced glucose tolerance on intravenous glucose tolerance testing), and dyslipidemia. Although many aspects of the physical appearance of these individuals resemble Cushing's syndrome, derangements in cortisol secretion have not been found consistently and do not appear to account for this appearance. Although some individuals haveIGT, diabetes is not a common feature. The possibility remains that this is related to HIV infection by some undefined mechanism, since some features of the syndrome were observed before the introduction of protease inhibitors. Therapy for HIV-related lipodystrophy is not well established. (Bibliography omitted in Palm version) Back to Table of Contents
334. HYPOGLYCEMIA - Philip E. Cryer Hypoglycemia occurs most commonly as a result of treating patients with diabetes mellitus. However, a number of other disorders, including insulinoma, large mesenchymal tumors, end-stage organ failure, alcoholism, endocrine deficiencies, postprandial reactive hypoglycemic conditions, and inherited metabolic disorders, are also associated with hypoglycemia (Table 334-1). Hypoglycemia is sometimes defined as a plasma glucose level 1 cm in size are frequently associated with local or distant metastases. Measurement of the serum calcitonin level after calcium or pentagastrin injection makes it possible to diagnose this disorder when the likelihood of metastasis is low (see below). Pheochromocytoma occurs in approximately 50% of patients withMEN 2A and causes palpitations, nervousness, headaches, and sometimes sweating (Chap. 332). About half the tumors are bilateral, and>50% of patients who have had unilateral adrenalectomy develop a pheochromocytoma in the contralateral gland within a decade. A second feature of these tumors is a disproportionate increase in the secretion of epinephrine relative to norepinephrine. Capsular invasion is common, but malignant behavior is uncommon. Hyperparathyroidism occurs in 15 to 20% of patients, with the peak incidence in the third or fourth decade. The manifestations of hyperparathyroidism do not differ from those in other forms of primary hyperparathyroidism (Chap. 341), with nephrolithiasis being common. Diagnosis is established by finding hypercalcemia, hypophosphatemia, hypercalciuria, and an inappropriately high serum level of intact parathyroid hormone. Multiglandular parathyroid hyperplasia is the most common histologic finding, although with long-standing disease adenomatous changes may be superimposed on hyperplasia. Multiple Endocrine Neoplasia Type 2B The association ofMTC, pheochromocytoma, mucosal neuromas, and a marfanoid habitus is designatedMEN 2B. MTC in MEN 2B develops earlier and is more aggressive than in MEN 2A. Metastatic disease has been described prior to 1 year of age, and death commonly occurs in the second or third decade of life. However, the prognosis is not invariably bad even in patients with metastatic disease, as evidenced by a number of multigenerational families with this disease. Pheochromocytoma occurs in more than half ofMEN 2B patients and does not differ from that in MEN 2A. Hypercalcemia is rare in MEN 2B, and there are no well-documented examples of hyperparathyroidism. The mucosal neuromas and marfanoid body habitus are the most distinctive features and are recognizable in childhood. Neuromas are present on the tip of the tongue, under the eyelids, and throughout the gastrointestinal tract and are true neuromas, distinct from neurofibromas. Children may present with gastrointestinal symptoms,
including increased gas, intermittent obstruction, and diarrhea caused by neuromas. GENETIC CONSIDERATIONS Mutations of the RET proto-oncogene have been identified in 93 to 95% of patients withMEN 2 (Fig. 339-3). RET encodes a tyrosine kinase receptor that is normally activated by glial cell line-derived neurotropic factor. RET mutations induce constitutive activity of the receptor, explaining the autosomal dominant transmission of the disorder. Naturally occurring mutations localize to two regions of the RET tyrosine kinase receptor. The first is a cysteine-rich extracellular domain; point mutations in the coding sequence for one of five cysteines (codons 609, 611, 618, 620, or 634) cause amino acid substitutions that induce receptor dimerization and activation in the absence of its ligand. Codon 634 mutations occur in 80% ofMEN 2A kindreds and are most commonly associated with classic MEN 2A features (Figs. 339-3 and339-2); an arginine substitution at this codon accounts for half of all MEN 2A mutations. All reported families with MEN 2A and cutaneous lichen amyloidosis are consistently associated with a codon 634 mutation. Mutations of codons 609, 611, 618, or 620 occur in 10 to 15% of MEN 2A kindreds and are more commonly associated withFMTC(Fig. 339-3). Mutations in codons 609, 618, and 620 have also been identified in MEN 2A and in the Hirschsprung variant (Fig. 339-3). The second region of the RET tyrosine kinase that is mutated inMEN 2 is in the substrate recognition pocket at codon 918 (Fig. 339-3). This activating mutation is present in approximately 95% of patients with MEN 2B and accounts for 10 to 15% of all RET proto-oncogene mutations in MEN 2. Mutations of codon 883 and 22 have also been identified in a few patients with MEN 2B. From 3 to 5% of kindreds withFMTChave no identifiable mutation of either of these regions. In a few such kindreds mutations of codons 768, 790, 791, 804, and 891 have been identified (Fig. 339-3). Somatic mutations (found only in the tumor and not transmitted in the germline) of the RET proto-oncogene have been identified in sporadicMTC; 25 to 35% of sporadic tumors have codon 918 mutations, and somatic mutations in codons 630, 768, and 804 have also been identified (Fig. 339-3). Germline mutations of the RET proto-oncogene are present in about 6% of patients with apparent sporadic MTC, indicating that other family members may be at risk for the disease. TREATMENT Screening for Multiple Endocrine Neoplasia Type 2 Death fromMTC can be prevented by early thyroidectomy. The identification of RET proto-oncogene mutations and the application of DNA-based molecular diagnostic techniques to identify these mutations has simplified the screening process. During the initial evaluation of a kindred, a RET proto-oncogene analysis should be performed on an individual with provenMEN 2A. Establishment of the specific mutation in a kindred facilitates the subsequent analysis of other family members. Each family member at risk should be tested twice for the presence of the specific mutation; the second analysis should be
performed on a new DNA sample and, ideally, in a second laboratory to exclude sample mix-up or technical error (see endrcr06.mda.uth.tmc.edu for a list of laboratory testing sites). Individuals in a kindred with a known mutation who have two normal analyses can be excluded from further screening. There is general consensus that children with codon 883, 918, and 922 mutations, or those associated withMEN 2B, should have a total thyroidectomy and central lymph node dissection (level VI) performed during the first months of life or soon after identification of the syndrome. If local metastasis is discovered, a more extensive lymph node dissection (levels II to V) is generally indicated. In children with codon 611, 618, 620, 630, 634, and 891 mutations, thyroidectomy should be performed before the age of 6 years because of reports of local metastatic disease in children this age. Finally, there are kindreds with codon 609, 768, 790, 791, and 804 mutations where the phenotype ofMTCappears to be less aggressive. In these kindreds, and in those with rare mutations, two management approaches have been suggested in association with genetic counseling: (1) perform a total thyroidectomy with or without central node dissection at some arbitrary age (perhaps 6 to 12 years of age), or (2) continue annual or biannual provocative testing for calcitonin release with performance of total thyroidectomy with or without central neck dissection when the test becomes abnormal. The pentagastrin test involves measurement of serum calcitonin basally and 2, 5, 10, and 15 min after a bolus injection of 5 ug pentagastrin per kilogram body weight. Patients should be warned before injection of epigastric tightness, nausea, warmth, and tingling of extremities and reassured that the symptoms will last approximately 2 min. The recent unavailability of pentagastrin in the United States has led to use of a short calcium infusion, performed by obtaining a baseline serum calcitonin and then infusing 150 mg calcium salt intravenously over 10 min with measurement of serum calcitonin at 5, 10, 15, 30 min after initiation of the infusion. The RET proto-oncogene analysis should be performed in patients with suspectedMEN 2B to detect codon 883, 918, and 922 mutations, especially in newborn children where the diagnosis is suspected but the clinical phenotype is not fully developed. Other family members at risk for MEN 2B should also be tested because the mucosal neuromas can be subtle and not always identified. In the rare families with proven germline transmission of MTC but no identifiable RET proto-oncogene mutation, annual pentagastrin or calcium-pentagastrin testing should be performed on members at risk. Annual screening for pheochromocytoma in subjects with germline RET mutations should be performed by measuring basal plasma or 24-h urine catecholamines and metanephrines. The goal is to identify a pheochromocytoma before it causes significant symptoms or is likely to cause sudden death, an event most commonly associated with large tumors. Although there are kindreds withFMTC and specific RET mutations in which no pheochromocytomas have been identified (Fig. 339-3), it is not clear that a large enough experience has been gained to exclude pheochromocytoma screening in these individuals. Radiographic studies, such asMRI orCTscans, are generally reserved for individuals with abnormal screening tests or with symptoms suggestive of pheochromocytoma (Chap. 332). Women should be tested during pregnancy because undetected pheochromocytoma can cause maternal death during childbirth. Measurement of serum calcium and parathyroid hormone levels every 2 to 3 years
provides an adequate screen for hyperparathyroidism, except in those families in which hyperparathyroidism is a prominent component, where measurements should be made annually. Treatment of Medullary Thyroid Carcinoma MTC is a multicentric disorder. Total thyroidectomy with a central lymph node dissection should be performed in children who carry the mutant genes. Incomplete thyroidectomy leaves the possibility of later transformation of residual long-term C cells. The goal of early therapy is cure, and a strategy that does not accomplish this goal is short-sighted. Long-term follow-up studies indicate an excellent outcome with approximately 90% of children free of disease 15 to 20 years after surgery. In contrast, 15 to 25% of patients in whom the diagnosis is made on the basis of a palpable thyroid nodule die from the disease within 15 to 20 years. In adults withMTC>1 cm in size, metastases to regional lymph nodes are common. Total thyroidectomy with central lymph node dissection and selective dissection of other regional chains provide the best chance for cure. In patients with extensive local metastatic disease in the neck, external radiation may prevent local recurrence or reduce tumor mass but is not curative. Chemotherapy with combinations of adriamycin, vincristine, cyclophosphamide, and dacarbazine may provide palliation. Treatment of Pheochromocytoma The long-term goal for management of pheochromocytoma is to prevent death and cardiovascular complications. Improvements in radiographic imaging of the adrenals make direct examination of the apparently normal contralateral gland during surgery less important, and the rapid evolution of laparoscopic surgery has simplified management of early pheochromocytoma. The major question is whether to remove both adrenal glands or to remove only the affected adrenal at the time of primary surgery. Issues to be considered in making this decision include the possibility of malignancy (98% of it in the skeleton. The calcium of the mineral phase at the surface of the crystals is in equilibrium with that in theECF, but only a minor fraction of the total pool (~0.5%) is exchangeable. In normal adults plasma levels range from 2.2 to 2.6 mmol/L (8.8 to 10.4 mg/dL). The calcium in plasma is present as three forms: free ions, ions bound to plasma proteins, and, to a small extent, diffusible complexes. The concentration of free calcium ions, averaging 1.2 mmol/L (4.8 mg/dL), influences many cellular functions and is subjected to tight hormonal control, especially throughPTH(Chap. 341). The concentration of serum proteins is an important determinant of calcium ion concentration; most calcium ion is bound to albumin. Ionized calcium can be measured directly with the use of calcium-specific electrodes. If ionized calcium cannot be measured, certain approximations can be used to estimate the protein-bound and ionized fractions. One formula that approximates the amount of calcium bound to protein is
A simplified correction is sometimes used to assess whether the total serum calcium concentration is abnormal when serum proteins are low. The correction is to add 1 mg/dL to the serum calcium level for every 1 g/dL by which the serum albumin level is below 4.0 g/dL. If the serum calcium level, for example, is 7.8 mg/dL (a subnormal value) and the serum albumin level is only 3.0 g/dL, then the stated serum calcium level is corrected by adding 1 mg/dL; the corrected value of 8.8 mg/dL is within the normal range. The concentration of calcium ions in theECF is kept constant by processes that constantly add and remove calcium. Calcium enters the plasma via absorption from the intestinal tract and resorption of ions from the bone mineral. Calcium leaves the ECF via secretion into the gastrointestinal tract (~100 to 200 mg/d), urinary excretion (~50 to 300 mg/d), deposition in bone mineral, and losses in sweat (up to 100 mg/d). Bone resorption and formation are tightly coupled, with approximately 12 mmol (500 mg) calcium entering and leaving the skeleton daily (Fig. 340-3). Calcium ions inside the cell mediate a variety of cellular functions. The level of free calcium in the cell is very low, approximately 0.1 umol/L; thus, the gradient between plasma and intracellular free calcium is about 10,000 to 1. This gradient is tightly regulated by various channels and ion pumps. The average dietary calcium intake for most adults in the United States is approximately 15 to 20 mmol/d (0.6 to 0.8 g/d). However, with heightened awareness of the role of adequate calcium intake for the prevention of osteoporosis, many adults on
supplements have an intake of 20 to 37 mmol/d (0.8 to 1.5 g/d). Less than half of dietary calcium is absorbed in adults. Calcium absorption increases during periods of rapid growth in children, in pregnancy, and in lactation and decreases with advancing age. Most of the calcium is absorbed in the proximal small intestine, and the efficiency of absorption decreases in the more distal intestinal segments. Both active transport and diffusion-limited absorption are involved; the former is more important in the upper intestine and the latter in the lower intestine. Both processes are influenced by vitamin D (see below). All forms of calcium in the diet are not equally absorbed; calcium as the chloride is probably absorbed more efficiently than that in other preparations. Secretion of calcium into the intestinal lumen is constant and independent of absorption. If calcium availability in the diet is low [30 to 40%. The urinary calcium excretion of normal adults having an average calcium intake ranges between 2.5 and 10 mmol/d (100 and 400 mg/d). When the dietary calcium level is5 g/d will induce diarrhea. For severe hypophosphatemia, with serum phosphorus levels 1.6 mmol/L (5 mg/dL). In children, this level is much higher. The most common causes of hyperphosphatemia are acute and chronic renal failure (Table 340-2). In renal failure, the loss of tubular function impairs phosphorus excretion. This results in a cascade of events that can also affect calcium and phosphorus metabolism. The increase in serum phosphorus levels reduces serum calcium levels and the production of 1,25(OH)2D, leading to decreased intestinal calcium absorption and secondary hyperparathyroidism. Patients with pseudohypoparathyroidism and tumoral calcinosis also have decreased renal phosphorus clearance that results in hyperphosphatemia. Hypothyroidism reduces renal phosphorus excretion and may increase circulating concentrations of phosphorus. Vitamin D intoxication, due to excessive ingestion of either vitamin D or one of its analogues, can cause hyperphosphatemia along with hypercalcemia. Severe hypothermia, crush injuries, nontrauma rhabdomyolysis, tumoral calcinosis, and cytotoxic therapy of hematologic malignancies such as acute lymphoblastic leukemia can be associated with hyperphosphatemia. The serum phosphorus level can be artifactually elevated due to hemolysis of the blood sample. Thrombocytosis and multiple myeloma can cause spuriously elevated serum phosphorus levels due to thrombocytolysis. Laboratory and Clinical Findings A rapid elevation of serum phosphorus can cause hypocalcemia and symptoms of neuromuscular irritability and tetany. Chronic hyperphosphatemia in association with normocalcemia can result in nephrocalcinosis and soft tissue calcification. TREATMENT In addition to treating the underlying disorder, dietary phosphorus intake should be limited by restricting carbonated beverages containing phosphoric acid and decreasing
milk and dairy product consumption. The dietary intake of phosphorus should be between 600 and 1000 mg a day with modest protein restriction. For control of chronic hyperphosphatemia, usually in patients with chronic renal failure, oral aluminum hydroxide or aluminum carbonate gels are indicated. Prolonged use of aluminum-containing compounds is not recommended because of aluminum toxicity causing adynamic bone disease, proximal myopathy, encephalopathy, and anemia. When hyperphosphatemia is due to vitamin D intoxication, calcium salts are contraindicated because the high efficiency of calcium absorption can lead to severe hypercalcemia, soft tissue calcification, and nephrocalcinosis. MAGNESIUM METABOLISM Magnesium is the most abundant intracellular divalent cation. It is an essential cofactor for a multitude of enzymatic reactions that are important for the generation of energy from ATP. Approximately 30% of magnesium in the serum is protein-bound, 55% is ionized, and the remaining 15% is complexed. Like calcium, magnesium is bound to albumin, and it is the ionized fraction that is important for physiologic processes including neuromuscular function and maintenance of cardiovascular tone. The serum concentration of magnesium is tightly regulated within a narrow range of approximately 0.7 to 1.1 mmol/L (1.4 to 2.2 meq/L)(1.7 to 2.6 mg/dl) as a result of the efficient absorption of dietary magnesium by the small intestine and conservation of magnesium in the kidney. About 30% of dietary magnesium is absorbed in the small intestine, but this fraction increases markedly when intake is substantially reduced. Approximately 96% of filtered magnesium is reabsorbed along the nephron, and only 4%is excreted into the urine. Because there is no regulation of magnesium absorption in the distal tubule and because magnesium reabsorption is very efficient, an increase in distal delivery increases magnesium loss in the urine. HYPOMAGNESEMIA Although magnesium deficiency is a common clinical problem, serum magnesium levels are often overlooked or not measured in patients at risk for the disorder. Approximately 10% of patients admitted to city hospitals are hypomagnesemic, and up to 65% of patients in intensive care units may be magnesium-deficient. Hypomagnesemia is caused primarily by renal or gastrointestinal losses or decreased efficiency of intestinal magnesium absorption (Table 340-3). Reduced renal reabsorption due to loop diuretics and alcohol use is a common cause of hypomagnesemia. Because magnesium excretion is tightly coupled to sodium and calcium excretion, intravenous fluid therapy and volume-expanded states, such as primary hyperaldosteronism, may result in hypomagnesemia. Hypercalcemia and hypercalciuria decrease tubular reabsorption of magnesium. Osmotic diuresis in diabetes mellitus is one of the more common causes of hypomagnesemia. Vomiting and nasogastric suctioning can cause severe magnesium depletion because intestinal tract fluids contain ~0.5 mmol/L (1.2 mg/dL)(1 meq/L). Fluid loss from diarrhea can contain as much as 7.4 mmol/L (18 mg/dL)(15 meq/L). Consequently, ulcerative colitis, Crohn's disease, and intestinal or biliary fistulas can result in magnesium depletion. Hypomagnesemia is prevalent in alcoholics. Ethanol causes a transient loss
of magnesium in the urine. In most alcoholics, however, the magnesium deficit is modest. A more profound fall in serum magnesium levels may occur during alcohol withdrawal, where the decrease is associated with falls in levels of serum phosphate and potassium, probably due to shifts of these ions into intracellular compartments. The use of loop diuretics, as well as aminoglycosides, cisplatin, cyclosporine, and amphotericin B can increase renal loss of magnesium. The clinical manifestations of hypomagnesemia are similar to those of severe hypocalcemia. The signs and symptoms of hypomagnesemia include muscle weakness, prolonged PR and QT intervals, and cardiac arrhythmias. Positive Chvostek's and Trousseau's signs indicative of hypocalcemia are often positive in hypomagnesemic patients as well; carpopedal spasm can also occur with hypomagnesemia. Magnesium is important for effectivePTHsecretion as well as the renal and skeletal responsiveness to PTH; thus, hypomagnesemia is often associated with hypocalcemia due to impaired PTH secretion and function (Chap. 341). Low serum magnesium levels 40% of the adult population over the age of 50 are vitamin D deficient. Although the standard lower limit of the normal range for 25(OH)D is 10 ng/mL, several studies suggest that the cutoff for vitamin D deficiency should be increased to 20 ng/mL. This suggestion is based onPTHresponses to various serum levels of 25(OH)D. For example, elevated levels of PTH are frequently seen in individuals with 25(OH)D levels between 10 and 20 ng/mL, and administration of vitamin D (50,000 units of vitamin D once a week for 8 weeks) lowers PTH levels. Defects in skeletal mineralization may accompany these disturbances in vitamin D and mineral metabolism. Pain in the hips may result in an antalgic gait. Muscle weakness is often associated with osteomalacia but is difficult to distinguish from hesitancy to move because of skeletal pain. Proximal weakness may mimic that of primary muscle disorders and contribute to the waddling gait. Many factors, including secondary hyperparathyroidism, hypophosphatemia, and vitamin D deficiency, contribute to the myopathy. Fractures of the involved bones may occur with minimal trauma. When the ribs are involved, severe deformities may develop in the thorax, and the collapse of the vertebral bodies may produce loss of height. RADIOLOGIC FEATURES In rickets, the most prominent radiologic alteration is evident at the growth plate (physis) which is increased in thickness, cupped, and hazy at the metaphyseal border owing to decreased calcification of the hypertrophic zone and inadequate mineralization of the primary spongiosa. The trabecular pattern of the metaphysis is abnormal, the cortices of the diaphysis may be thinned, and the shafts may be bowed. In osteomalacia, a decrease in bone density is usually associated with loss of trabeculae and thinning of the cortices. Radiologic and bone densitometric changes are indistinguishable from those in osteoporosis (Chap. 342). Trabecular patterns may be blurred, producing a homogeneous "ground glass" appearance. Radiolucent bands, ranging from a few millimeters to several centimeters in length and usually oriented perpendicular to the surface of the bones, suggest the presence of osteomalacia. They are particularly common at the inner aspects of the femur (especially near the femoral neck), in the pelvis, in the outer edge of the scapula, in the upper fibula, and in the metatarsals. These radiolucent bands, called pseudofractures or Looser's zones, occur most often where major arteries cross the bones and are thought to be due to the pulsation of these vessels in the undermineralized area (Fig. 340-9). Increased rather than decreased density of bones may be observed in patients who have renal tubular disorders rather than vitamin D deficiency and may produce a striking thickening of the cortices and a trabecular pattern of the spongy bone. Despite the increase in bone mass per unit volume, the trabeculae are covered with thickened osteoid seams typical of osteomalacia. Similar findings may occur in patients with chronic renal failure. The reason for the hyperostosis is unknown; the bone is architecturally abnormal and is subject to fracture with minimal trauma. LABORATORY FINDINGS Changes in serum concentration of calcium, inorganic phosphorus, 25(OH)D, and
1,25(OH)2D vary with the different disorders. In vitamin D deficiency, whether due to dietary lack, inadequate sunlight exposure, or intestinal malabsorption, serum calcium levels are normal or low, whereas phosphorus and 25(OH)D levels are consistently low, the latter usually 3.2 mmol/L (13 mg/dL), calcification in kidneys, skin, vessels, lungs, heart, and stomach occurs and renal insufficiency may develop, particularly if blood phosphate levels are normal or elevated due to impaired renal function. Severe hypercalcemia, usually defined as ³3.7 to 4.5 mmol/L (15 to 18 mg/dL) can be a medical emergency; coma and cardiac arrest can occur. Except in malignancy-associated hypercalcemia, acute management of the hypercalcemia is usually successful prior to definitive therapy. The type of treatment is based on the severity of the hypercalcemia and the nature of associated symptoms. PRIMARY HYPERPARATHYROIDISM Natural History and Incidence Primary hyperparathyroidism is a generalized disorder of calcium, phosphate, and bone metabolism due to an increased secretion ofPTH. The elevation of circulating hormone usually leads to hypercalcemia and hypophosphatemia. There is great variation in the manifestations. Patients may present with multiple signs and symptoms, including recurrent nephrolithiasis, peptic ulcers, mental changes, and, less frequently, extensive bone resorption. However, with greater awareness of the disease and wider use of multiphasic screening tests, including blood calcium assays, the diagnosis is frequently made in patients who have no symptoms and minimal, if any, signs of the disease other than hypercalcemia and elevated levels of PTH. The manifestations may be subtle, and the disease may have a benign course for many years or a lifetime. Rarely, hyperparathyroidism develops or worsens abruptly and causes severe complications, such as marked dehydration and coma, so-called hypercalcemic parathyroid crisis. The annual incidence of the disease is estimated to be as high as 0.2% in patients>60, with an estimated prevalence, including undiscovered asymptomatic patients, of ³1%. The disease has a peak incidence between the third and fifth decades but occurs in young children and in the elderly. Etiology Solitary Adenomas The cause of hyperparathyroidism is one or more hyperfunctioning glands. The traditional view has been that a single abnormal gland is the cause in approximately 80% of patients; the abnormality in the gland is usually a benign neoplasm or adenoma and rarely a parathyroid carcinoma. Some surgeons and pathologists report that the enlargement of multiple glands is common; double adenomas are reported. In approximately 15% of patients, all glands are hyperfunctioning; chief cell parathyroid hyperplasia is usually hereditary and frequently associated with other endocrine abnormalities. Multiple Endocrine Neoplasia Hereditary hyperparathyroidism can occur without other endocrine abnormalities but is usually part of a multiple endocrine neoplasia syndrome
(Chap. 339).MEN 1 (Wermer's syndrome) consists of hyperparathyroidism and tumors of the pituitary and pancreas, often associated with gastric hypersecretion and peptic ulcer disease (Zollinger-Ellison syndrome). MEN 2A is characterized by pheochromocytoma and medullary carcinoma of the thyroid, as well as hyperparathyroidism; MEN 2B has additional associated features such as multiple neuromas but usually lacks hyperparathyroidism. Each of these MEN syndromes is transmitted in an autosomal dominant manner. Pathology Adenomas are most often located in the inferior parathyroid glands, but in 6 to 10% of patients, parathyroid adenomas may be located in the thymus, the thyroid, the pericardium, or behind the esophagus. Adenomas are usually 0.5 to 5 g in size but may be as large as 10 to 20 g (normal glands weigh 25 mg on average). Chief cells are predominant in both hyperplasia and adenoma. The adenoma is sometimes encapsulated by a rim of normal tissue. With chief cell hyperplasia, the enlargement may be so asymmetric that some involved glands appear grossly normal. If generalized hyperplasia is present, however, histologic examination reveals a uniform pattern of chief cells and disappearance of fat even in the absence of an increase in gland weight. Thus, microscopic examination of biopsy specimens of several glands is essential to interpret findings at surgery. When an adenoma is present, the other glands are usually normal and contain a normal distribution of all cell types (rather than only chief cells) and normal amounts of fat. Parathyroid carcinoma is usually not aggressive in character. Long-term survival without recurrence is common if at initial surgery the entire gland is removed without rupture of the capsule. Recurrent parathyroid carcinoma is usually slow-growing with local spread in the neck, and surgical correction of recurrent disease may be feasible. Occasionally, however, parathyroid carcinoma is more aggressive, with distant metastases (lung, liver, and bone) found at the time of initial operation. It may be difficult to appreciate initially that a primary tumor is carcinoma; increased numbers of mitotic figures and increased fibrosis of the gland stroma may precede invasion. The diagnosis of carcinoma is often made in retrospect. Hyperparathyroidism from a parathyroid carcinoma may be indistinguishable from other forms of primary hyperparathyroidism; a potential clue to the diagnosis, however, is provided by the degree of calcium elevation. Calcium values of 3.5 to 3.7 mmol/L (14 to 15 mg/dL) are frequent with carcinoma and may alert the surgeon to remove the abnormal gland with care to avoid capsular rupture. GENETIC CONSIDERATIONS Defects Associated with Hyperparathyroidism As in many other types of neoplasia, two fundamental types of genetic defects have been identified in parathyroid gland tumors: (1) overactivity of protooncogenes, and (2) loss of function of tumor suppressor genes. The former, by definition, can lead to uncontrolled cellular growth and function by activation (gain-of-function mutation) of a single allele of the responsible gene, whereas the latter requires loss of function of both allelic copies. Mutations in the MENIN gene locus on chromosome 11q13 are responsible for causingMEN 1; the normal allele of this gene fits the definition of a tumor suppressor gene. A mutation of one allele is inherited; loss of the other allele via somatic cell mutation leads to monoclonal expansion and tumor development in tissues such as the
parathyroids. In approximately 20% of sporadic parathyroid adenomas, the MENIN locus on chromosome 11 appears to be deleted, implying that the same defect responsible for MEN 1 can also cause the sporadic disease (Fig. 341-4A). Consistent with the Knudson hypothesis for two-step neoplasia in certain inherited cancer syndromes (Chap. 81), the earlier onset of hyperparathyroidism in the hereditary syndromes reflects the statistical probability of only one mutational event triggering the monoclonal outgrowth. In sporadic adenomas, typically occurring later in life, two different somatic events must occur before the MENIN gene is silenced. The MENIN gene codes for a novel protein consisting of 610 amino acids. The protein has a nuclear localization signal and appears to interact with the transcription factor Jun D. Most of the mutations are clearly of the inactivating type (nonsense, deletions); there is not, however, a good correlation between clinical features in different kindreds and the specific mutation detected (e.g., penetrance or age of onset of pituitary or pancreatic tumors). This is in contrast to the correlation between genotype and phenotype in MEN 2 (see below). Other presumptive antioncogenes involved in hyperparathyroidism include a gene mapped to chromosome 1p seen in 40% of sporadic parathyroid adenomas and a gene mapped to chromosome Xp11 in patients with secondary hyperparathyroidism and renal failure, who progressed to "tertiary" hyperparathyroidism, now known to reflect monoclonal outgrowths within previously hyperplastic glands. The Rb gene, a tumor suppressor gene located on chromosome 13q14, was initially associated with retinoblastomas but has since been implicated in many other forms of neoplasia including parathyroid carcinoma. Allelic deletion (with a presumed point mutation in the second allele) has been identified in all parathyroid carcinomas examined; there is also an abnormal staining pattern of the protein product of the gene. Allelic deletion is also seen in 10% of parathyroid adenomas, although the abnormal staining pattern of the Rb protein is not seen. Other gene loci on chromosome 13 may be involved in addition to the Rb locus. There are two rare syndromes associated with hyperparathyroidism that involve one or more genes located on chromosome 1q. The hereditary hyperparathyroidism jaw tumor (HPT-JT) syndrome shows an autosomal dominant inheritance pattern; the jaw tumors are benign, but the parathyroid pathology may involve carcinoma as well as adenoma. Parathyroid carcinoma may also appear in the other syndrome, familial isolated primary hyperparathyroidism (FIPH). Both syndromes have been mapped through linkage studies to the chromosome 1q21-q31 region. Certain findings have led to speculation that this chromosome region might contain a protooncogene rather than an antioncogene. In some parathyroid adenomas, activation of a protooncogene has been identified (Fig. 341-4B). A reciprocal translocation involving chromosome 11 has been identified that juxtaposes thePTH gene promoter upstream of a gene product termed PRAD-1, a cyclin D protein that plays a key role in normal cell division. This translocation is found in as many as 15% of parathyroid adenomas, usually in larger tumors. Targeted overexpression of cyclin D1 in the parathyroid glands of transgenic mice causes the development of hyperparathyroidism, consistent with the role of this cell cycle control
protein in parathyroid neoplasia. A mutated protooncogene, RET, is involved in each of the clinical variants ofMEN2 (Chap. 339). RET encodes a tyrosine kinase-type receptor; specific mutations lead to constitutive activity of the receptor, thereby explaining the autosomal dominant mode of transmission and the relatively early onset of neoplasia. Signs and Symptoms Half or more of patients with hyperparathyroidism are asymptomatic. In series in which patients are followed without operation, as many as 80% are classified as without symptoms. Manifestations of hyperparathyroidism involve primarily the kidneys and the skeletal system. Kidney involvement, due either to deposition of calcium in the renal parenchyma or to recurrent nephrolithiasis, was present in 60 to 70% of patients prior to 1970. With earlier detection, renal complications occur in90% of patients with primary hyperparathyroidism have asymptomatic hypercalcemia; symptoms of malignancy are usually present in cancer-associated hypercalcemia. Disorders other than hyperparathyroidism and malignancy cause 1 year, malignancy can usually be excluded as the cause. A striking feature of malignancy-associated hypercalcemia is the rapidity of the course, whereby signs and symptoms of the underlying malignancy are evident within months of the detection of hypercalcemia. A careful history of dietary supplements and drug use may suggest intoxication with vitamins D or vitamin A or the use of thiazides.
Although clinical considerations are helpful in arriving at the correct diagnosis of the cause of hypercalcemia, appropriate laboratory testing is essential for definitive diagnosis. The immunoassay forPTHshould separate hyperparathyroidism from all other causes of hypercalcemia. Patients with hyperparathyroidism have elevated PTH levels despite hypercalcemia, whereas patients with malignancy and the other causes of hypercalcemia (except for disorders mediated by PTH such as lithium-induced hypercalcemia) have levels of hormone below normal or undetectable. Assays based on the double-antibody method for PTH exhibit very high sensitivity (especially if serum calcium is simultaneously evaluated) and specificity for the diagnosis of primary hyperparathyroidism (Fig. 341-9). In summary,PTHvalues are elevated in >90% of parathyroid-related causes of hypercalcemia, undetectable or low in malignancy-related hypercalcemia, and undetectable or normal in vitamin D-related and high-bone-turnover causes of hypercalcemia. In view of the specificity of the PTH immunoassay and the high frequency of hyperparathyroidism in hypercalcemic patients, it is cost-effective to measure the PTH level in all hypercalcemic patients unless malignancy or a specific nonparathyroid disease is obvious. False-positive PTH assay results are rare. There are very rare reports of ectopic production of excess PTH by nonparathyroid tumors. Immunoassays forPTHrPare helpful in diagnosing certain types of malignancy-associated hypercalcemia. AlthoughFHH is parathyroid-related, the disease should be managed distinctively from hyperparathyroidism. Clinical features and the low urinary calcium excretion can help make the distinction. Because the incidence of malignancy and hyperparathyroidism both increase with age, they can coexist as two independent causes of hypercalcemia. 1,25(OH)2D levels are elevated in many (but not all) patients with primary hyperparathyroidism. In other disorders associated with hypercalcemia, concentrations of 1,25(OH)2D are low or, at the most, normal. However, this test is of low specificity and is not cost-effective, as not all patients with hyperparathyroidism have elevated 1,25(OH)2D levels, and not all nonparathyroid hypercalcemic patients have suppressed 1,25(OH)2D. Measurement of 1,25(OH)2D is, however, critically valuable in establishing the cause of hypercalcemia in sarcoidosis and certain B cell lymphomas. A useful general approach is outlined inFig. 341-8. If the patient is asymptomatic and there is evidence of chronicity to the hypercalcemia, hyperparathyroidism is almost certainly the cause. IfPTHlevels (usually measured at least twice) are elevated, the clinical impression is confirmed and little additional evaluation is necessary. If there is only a short history or no data as to the duration of the hypercalcemia, occult malignancy must be considered; if the PTH levels are not elevated, then a thorough workup must be undertaken for malignancy, including chest x-ray,CT of chest and abdomen, and bone scan. Immunoassays forPTHrPmay be especially useful in such situations. Attention should also be paid to clues for underlying hematologic disorders such as anemia, increased plasma globulin, and abnormal serum immunoelectrophoresis; bone scans can be negative in some patients with metastases, such as in multiple myeloma. Finally, if a patient with chronic hypercalcemia is asymptomatic and malignancy therefore seems unlikely on clinical grounds, but PTH values are not elevated, it is useful to search for other chronic causes of hypercalcemia, such as occult sarcoidosis.
TREATMENT Hypercalcemic States The approach to medical treatment of hypercalcemia varies with its severity. Mild hypercalcemia, 10 million individuals in the United States, but only 10 to 20% are diagnosed and treated. DEFINITION Osteoporosis is defined as a reduction of bone mass (or density) or the presence of a fragility fracture. This reduction in bone tissue is accompanied by deterioration in the architecture of the skeleton, leading to a markedly increased risk of fracture. Osteoporosis is defined operationally as a bone density that falls 2.5 standard deviations (SD) below the mean -- also referred to as a T-score of-2.5. Those who fall at the lower end of the young normal range (a T-score of >1 SD below the mean) have low bone density and are considered to be at increased risk of osteoporosis. EPIDEMIOLOGY In the United States, as many as 8 million women and 2 million men have osteoporosis (T-score 1900 women with low bone mass treated with alendronate (10 mg/d) versus placebo, the incidence of all nonvertebral fractures was reduced by ~47% after just 1 year. Alendronate (5 to 10 mg/d) should be given with a full glass of water before breakfast, as bisphosphonates are poorly absorbed. Because of the potential for esophageal irritation, alendronate is contraindicated in patients who have stricture or inadequate emptying of the esophagus. It is recommended that patients remain upright for at least 30 min after taking the medication to avoid esophageal irritation. Cases of esophagitis, esophageal ulcer, and esophageal stricture have been described, but the incidence appears to be low. In clinical trials, overall gastrointestinal symptomatology was no different with alendronate compared to placebo. Risedronate produces a dramatic reduction in bone turnover and an increase in bone mass. Controlled clinical trials have demonstrated >40% reduction in vertebral fracture risk over 3 years, accompanied by a 33% reduction in clinical nonspine fractures. Reports from several studies show a 40% reduction in hip fracture in patients with osteoporosis, with a somewhat greater effect in patients with prevalent vertebral fractures. Patients should take risedronate (5.0 mg orally) with a full glass of plain water [0.18 to 0.25 L (6 to 8 oz)], to facilitate delivery to the stomach, and should not lie down
for 30 min after taking the drug. The incidence of gastrointestinal side effects in these trials with risedronate was similar to that of placebo. Etidronate was the first bisphosphonate to be approved, initially for use in Paget's disease and hypercalcemia. This agent has also been used in osteoporosis trials of smaller magnitude than those performed for alendronate and risedronate. Etidronate probably has some efficacy against vertebral fracture when given as an intermittent cyclical regimen (2 weeks on, 2 1/2 months off). MODE OF ACTION Bisphosphonates are structurally related to pyrophosphates, compounds that are incorporated into bone matrix. Through mechanisms that remain to be fully elucidated, bisphosphonates specifically impair osteoclast function and reduce osteoclast number, in part by the induction of apoptosis. Recent evidence suggests that the nitrogen-containing bisphosphonates also inhibit protein prenylation, one of the end products in the mevalonic acid pathway. This effect disrupts intracellular protein trafficking and may ultimately lead to apoptosis. Bisphosphonates have very long retention in the skeleton and may exert long-term effects. Calcitonin Calcitonin is a polypeptide hormone produced by the thyroid gland (Chap. 341). Its physiologic role is unclear as no skeletal disease has been described in association with calcitonin deficiency or calcitonin excess. Calcitonins are approved by theFDAapproved for Paget's disease, hypercalcemia, and osteoporosis in women>5 years past menopause. Injectable calcitonin produces small increments in bone mass of the lumbar spine. However, difficulty of administration and frequent reactions, including nausea and facial flushing, make general use limited. In 1995, a nasal spray containing calcitonin (200 IU/d) was approved for treatment of osteoporosis in postmenopausal women. Several studies indicate that nasal calcitonin produces small increments in bone mass and a small reduction in new vertebral fractures in calcitonin-treated patients versus those on calcium alone. Calcitonin is not indicated for prevention of osteoporosis and is not sufficiently potent to prevent bone loss in early postmenopausal women. As mentioned above, calcitonin might have an analgesic effect on bone pain, both in the subcutaneous and possibly the nasal form. MODE OF ACTION Calcitonin suppresses osteoclast activity by direct action on the osteoclast calcitonin receptor. Osteoclasts exposed to calcitonin cannot maintain their active ruffled border, which normally maintains close contact with underlying bone. Calcitonin also affects osteoclast mobility and the movement of enzyme-containing cytoplasmic granules. Experimental Agents PARATHYROID HORMONE EndogenousPTH is an 84-amino-acid peptide that is largely responsible for calcium homeostasis (Chap. 341). Although chronic elevation of PTH, as occurs in hyperparathyroidism, is associated with bone loss (particularly cortical bone), PTH can also exert anabolic effects on bone. Consistent with this, some
observational studies have indicated that mild elevations in PTH are associated with maintenance of trabecular bone mass. On the basis of these findings, preclinical and early clinical studies have been performed using an exogenous PTH analogue (1-34 PTH). The first randomized controlled trial in postmenopausal women showed that PTH, when superimposed on ongoing estrogen therapy, produced substantial increments in bone mass (13% over a 3-year period compared to estrogen alone). This increment in bone mass was also associated with a reduction in risk of vertebral compression deformity (Fig. 342-10). More recent studies have confirmed the ability of combined treatment with estrogen and PTH to induce striking increases in bone mass. PTHuse may be limited by its mode of administration, which currently requires subcutaneous injection. Alternative modes of delivery are being investigated, including transdermal and inhalation routes. The optimal frequency of administration also remains to be established, and it is possible that PTH might also be effective when used in high doses, 1 month out of every 3. MODE OF ACTION Exogenously administeredPTHappears to have direct actions on osteoblast activity, with biochemical and histomorphometric evidence of de novo bone formation early in response to PTH, prior to activation of bone resorption. Subsequently, PTH activates bone remodeling but still appears to favor bone formation over bone resorption. PTH stimulatesIGF-I and collagen production and appears to increase osteoblast number by inhibiting apoptosis and stimulating replication. FLUORIDE Fluoride has been available for many years and is a potent stimulator of osteoprogenitor cells when studied in vitro. It has been used in multiple osteoporosis studies with conflicting results, in part related to use of varying doses and preparations. Fluoride produces marked effects on bone mass, especially in the spine, where gains of around 10% per year have been observed. However, despite increments in bone mass, there is no consistent effect of fluoride on vertebral or nonvertebral fracture, which might actually increase when high doses of fluoride are used. Furthermore, animal data suggest that there is reduced biomechanical strength when fluoride is incorporated into bone as fluoroapatite, with excess osteoid accumulation and evidence of woven rather than lamellar bone formation, especially at high doses. For these reasons, fluoride remains an experimental agent, despite its long history and multiple studies. OTHER POTENTIAL ANABOLIC AGENTS Several small studies of growth hormone (GH), alone or in combination with other agents, have not shown consistent or substantial positive effects on skeletal mass. Many of these studies are relatively short-term, and the effects of GH and the IGFsare still under investigation. Anabolic steroids, mostly derivatives of testosterone, act primarily as antiresorptive agents to reduce bone turnover but may also stimulate osteoblastic activity. Effects on bone mass remain unclear but appear weak, in general, and use is limited by masculinizing side effects. Several recent observational studies suggest that the statin drugs, currently used to treat hypercholesterolemia, may be associated with increased bone mass and reduced fractures, but there are not clinical trial data. Nonpharmacologic Approaches Protective pads worn around the outer thigh, which cover the trochanteric region of the hip can prevent hip fractures in elderly residents in nursing homes. The use of hip protectors is limited largely by compliance and comfort,
but new devices are being developed that may circumvent these problems and provide adjunctive treatments. Treatment Monitoring There are currently no well-accepted guidelines for monitoring treatment of osteoporosis. Because most osteoporosis treatments produce small or moderate bone mass increments on average, it is reasonable to considerBMD as a monitoring tool. As with any biologic or assay determination, there is precision error with repeated measurements. Changes must exceed ~4% in the spine and 6% in the hip to be considered significant in any individual. The hip is the preferred site due to larger surface area and greater reproducibility. Medication-induced increments may require several years to produce changes of this magnitude (if they do at all). Consequently, it can be argued that BMD should not be repeated at intervals 50 mutations identified in the type III procollagen gene include partial gene deletions, RNA splicing mutations, and single-base mutations that cause substitution of glycine by amino acids with bulkier side chains (Fig. 351-4). In effect, most of the mutations lead to synthesis of abnormal but partially functional proa1(III) chains that produce procollagen suicide or alter fibril formation by the same mechanisms that amplify the effects of mutations in the genes for type I procollagen. Similar mutations in type III procollagen can cause aortic aneurysms in some individuals without other evidence of EDS type IV,MFS, or other inherited disorders of connective tissue. Type VIEDS is caused by mutations in the gene that encodes lysyl hydrolase. In one series, all the patients were homozygous or compound heterozygotes for the mutated genes, and all the mutations caused profound deficiency of lysyl hydroxylase, a decrease in the hydroxylysine content of collagen, and a decrease in the cross-links in collagen fibers. The decrease in cross-links is explained by the observation that some cross-links are less stable if formed from lysine instead of hydroxylysine. Type VIIEDS is due to a defect in the conversion of procollagen to collagen caused either by mutations that make type I procollagen resistant to cleavage by procollagen N-proteinase or by mutations that decrease the activity of the enzyme. The type VIIA mutations alter the cleavage site in the proa1(I) chain, and the type VIIB mutations alter the cleavage site in the proa2(I) chain. Both types are dominantly inherited. Type VIIC is caused by mutations that decrease the activity of procollagen N-proteinase and is inherited as an autosomal recessive trait. In all three forms of type VII EDS, the persistence of the N-propeptide causes the formation of collagen fibrils that are thin and irregular. Since most patients do not have clinical osteopenia, the thin and irregular fibrils apparently suffice for the mineralization of bone but do not provide the necessary tensile strength for ligaments and joint capsules. However, some patients fracture easily
and are difficult to distinguish from variants of OI. The cause of type VIIIEDS is unknown. Type IX is a disorder of copper transport. The syndrome, also referred to as Menkes' syndrome, is due to an X-linked defect and is associated with cutis laxa, hypopigmentation, unusual hair ("kinky"), vascular aneurysms, neurologic degeneration, and mental retardation. Mutations in a gene coding for a copper-transporting ATPase cause the disease (Chaps. 348 and353). Type X EDS may be caused by defects in fibronectin, but no specific mutations have been defined. Diagnosis The diagnosis is based on clinical criteria. Biochemical assays and gene analyses for known molecular defects inEDS are difficult and time-consuming, but specific diagnostic tests should be available in the future for families in which the genes at fault have been defined. TREATMENT There is no specific therapy. Surgical repair and tightening of joint ligaments require careful evaluation of individual patients, as the ligaments frequently do not hold sutures. Patients with easy bruisability should be evaluated for other bleeding disorders. Patients with type IVEDS and members of their families should probably be evaluated at regular intervals by sonography and related techniques for early detection of aneurysms. Surgical repair of aneurysms may be difficult because of increased friability of tissues, and there is limited experience with elective surgery in such patients. Also, women with type IV EDS should be counseled about the increased risk of uterine rupture, bleeding, and other complications of pregnancy. CHONDRODYSPLASIAS (See alsoChap. 343) TheCDs are inherited skeletal disorders that cause dwarfism and abnormal body proportions. The category also includes some individuals with normal stature and body proportions who have features such as ocular changes or cleft palate that are common in more severe CDs. Many patients develop degenerative joint changes, and mild CD in adults may be difficult to differentiate from primary generalized osteoarthritis. Some authors refer to the disorders as "skeletal dysplasias," but CD is a more widely used term. Classification Over 150 distinct types and subtypes have been defined (Table 351-5) based on criteria such as "bringing death" (thanatophoric), causing "twisted" bones (diastrophic), affecting metaphyses (metaphyseal), affecting epiphyses (epiphyseal), and producing histologic changes such as an apparent increase in the fibrous material in the epiphyses (fibrochondrogenesis). Also, a number of eponyms are based on the first or most comprehensive case reports. Severe forms of the diseases produce gross distortions of most cartilaginous structures and of the eye. Mild forms are more difficult to classify. Among the features are cataracts, degeneration of the vitreous and retinal detachment, high forehead, hypoplastic facies, cleft palate, short extremities, and gross distortions of the epiphyses, metaphyses, and joint surfaces. Incidence Data on the frequency of mostCDs are not available, but the incidence of the
Stickler syndrome may be as high as 1 in 10,000. Therefore, the diseases are probably among the more common heritable disorders of connective tissue. Molecular Defects The first mutations shown to causeCDs were in the COL2A1 gene for type II collagen, the most abundant protein in cartilage. A number of mutations in this gene have now been reported in variants of CD ranging from mild to lethal (Fig. 351-5). A large fraction of patients with lethal CDs, a smaller number of patients with moderately severe CDs, and about 2% of families with early-onset generalized osteoarthritis have mutations in the same gene. However, similar phenotypes can also be caused by mutations in other genes, including genes for three other collagens, additional components of the cartilage matrix, growth factors, growth factor receptors, and transcription factors (see Table 351-6 for selected examples). The number of mutated genes reported does not necessarily reflect the incidence of such mutations in the diseases themselves but rather the complexity of the genes and the technical difficulties in searching the complete gene for mutations. Also, it reflects the availability of large families for DNA linkage analysis and the vigor with which investigators have pursued their interest in a given gene. It is likely that mutations in additional genes will be found. Mutations in the COL2A1 gene were first found in patients with severeCDscharacterized by gross deformities of bones and cartilage such as spondyloepiphyseal dysplasia congenita, hypochondrogenesis/achondrogenesis II, and the Kniest syndrome. However, mutations in the COL2A1 gene have been found in a few families in which few if any symptoms are present in childhood but in which joint stiffness, joint pain, and degenerative changes of osteoarthritis develop in midlife. The mutations in the COL2A1 gene are similar to the mutations in the genes for types I and III procollagens (Fig. 351-5), and the correlations between genotype and the severity of the phenotype are equally difficult. In addition, mutations that change a codon for a Y-position amino acid in the -Gly-X-Y- repeat sequence from an arginine to cystine were found in families with early-onset osteoarthritis and minimal evidence of CDs. Stickler syndrome and related syndromes are caused by mutations in three different genes: the COL2AI gene for type II collagen and the COL11A1 and COL11A2 genes for type XI collagen. A series of mutations that introduce premature terminal signals in the COL2A1 gene lead to classic Stickler syndrome. However, some patients with classic Stickler syndrome have glycine substitutions in COL11A1. RNA splicing mutations in the COL11A1 gene are found in patients with the Marshall syndrome, which is similar to classic Stickler syndrome but with milder eye changes and more severe hearing loss. Patients classified as having nonocular Stickler syndrome have RNA splicing mutations in the COL11A2 gene. Many individuals with the Schmid metaphysealCD, characterized by short stature, coxa vara, flaring metaphyses, and waddling gait, have mutations in the gene for the type X collagen, a short, network-forming collagen found primarily in the hypertrophic zone of endochondral cartilage. Mutations in the receptor for fibroblast growth factor 3 (FGFR-3) are present in most patients with achondroplasia, the most common cause of short-limbed dwarfism accompanied by macrocephaly and dysplasias of the metaphyses of long bones (Table 351-6). The same single-base mutation in the gene that converts glycine to arginine at position 380 is present in>90% of patients. Most patients represent sporadic new mutations, and this nucleotide change must be one of the most common recurring
mutations in the human genome. The mutation causes unregulated signal transduction through the receptor and inappropriate development of cartilage. Mutations that alter other domains of FGFR-3 have been found in patients with the more severe disorders hypochondroplasia and thanatophoric dysplasia and in a few families with a variant of craniosynostosis. However, most patients with craniosynostosis appear to have mutations in the related gene FGFR-2 gene. Mutations in the gene for the cartilage oligomeric matrix protein (COMP) have been found in patients with multiple epiphyseal dysplasia or pseudoachondroplasia, and in related syndromes characterized by short limbs and degenerative arthritis. However, some families with multiple epiphyseal dysplasia had a mutation in the gene for thea2(IX) or a3(IX) chain of type IX collagen (COL9A2 and COL9A3). All the known mutations in these two type IX collagen genes in patients with multiple epiphyseal dysplasia cause splicing out of the codons of exon 3. A mutation in the COL9A2 gene was also found in patients with the common condition of sciatica and herniations of vertebral discs. About 4% of Finnish probands with the phenotype had a single base substitution that converted a codon for glutamate to tryptophan in thea2(IX) chain of type IX collagen. Diagnosis The diagnosis of severe forms ofCD is made on the basis of the physical appearance, x-ray findings, histologic changes, and clinical course (Table 351-5). TREATMENT No definitive therapy is available. Symptomatic treatment is directed to secondary features such as degenerative arthritis. Many patients require joint replacement surgery and corrective surgery for cleft palate. The eyes should be monitored carefully for the development of cataracts and for the need for laser therapy to prevent retinal detachment. Patients should probably be advised to avoid obesity and contact sports. Counseling for the psychological problems of short stature is critical, and support groups have formed in many countries. Ultrasonography is sometimes successful for prenatal diagnosis but less frequently than withOI. Specific DNA tests are available for theCDscaused by mutations in the genes for types II, X, and XI collagens. MARFAN SYNDROME SevereMFS is characterized by a triad of features: (1) long, thin extremities frequently associated with other skeletal changes, such as loose joints and arachnodactyly; (2) reduced vision as the result of dislocations of the lenses (ectopia lentis); and (3) aortic aneurysms that typically begin at the base of the aorta. Classification The clinical diagnosis is frequently problematic because some affected members of families with MFSpresent with only one or two features of the typical clinical trial. Also, many patients present with one or two of the features of MFS without a family history, apparently because they represent sporadic mutations. Therefore, it is frequently difficult to determine on the basis of clinical data alone whether a patient with ectopia lentis or the characteristic body habitus of MFS is at risk for developing a life-threatening aortic aneurysm. The new DNA diagnostic tests for mutations in the fibrillin-1 and fibrillin-2 genes can probably resolve most, but not all, of these problems.
Most patients who are prone to develop an aortic aneurysm as a component of MFS can be identified by detection of mutations in the fibrillin-1 gene. Some of these patients develop aortic aneurysms because of a mutation in the fibrillin-1 gene without the skeletal or ocular changes characteristic of MFS. Patients with the rarer form of MFS that is characterized by contractural arachnodactyly instead of loose joints can usually be identified by detection of a mutation in the fibrillin-2 gene that is similar in structure to the gene for fibrillin-1. Preliminary data suggest that patients with mutations in the fibrillin-2 gene are not prone to develop aneurysms. However, affected members of some rare families with a mutation in the fibrillin-1 gene also do not develop aortic aneurysms, even though they may show the skeletal or ocular changes. Therefore, the DNA tests are most helpful if: (1) a mutation is detected in either of the two genes, and (2) informative data are available on the clinical symptoms that the same mutation produces in the patient's family or in other families with similar clinical features. Incidence and Inheritance MFS has an incidence of about 1 in 10,000 in most racial and ethnic groups. The disorder is inherited as an autosomal dominant trait; at least one-fourth of patients do not have an affected parent, and their cases are probably due to new mutations. Skeletal Changes Patients are usually tall compared with other members of the same family and have long limbs. The ratio of the upper segment (top of the head to top of the pubic ramus) to the lower segment (top of the pubic ramus to the floor) is usually 2 SDs below mean for age, race, and sex. The fingers and hands are long and slender and have a spider-like appearance (arachnodactyly). Many patients have severe chest deformities, including depression (pectus excavatum), protrusion (pectus carinatum), or asymmetry. Scoliosis is frequent and usually accompanied by kyphosis. High-arched palate and high pedal arches or pes planus are common. A few patients have severe joint hypermobility similar toEDS. Cardiovascular Changes Cardiovascular abnormalities are the major source of morbidity and mortality (Chap. 247). Mitral valve prolapse develops early in life and in about one-quarter progresses to mitral valve regurgitation of increasing severity because of redundancy of the leaflets, stretching of the chordae tendineae, and dilatation of the valvulae annulus. Dilatation of the root of the aorta and the sinuses of Valsalva are characteristic and ominous features of the disease that can develop at any age and in rare instances may be detected by echocardiography in utero. The rate of dilatation is unpredictable, but it can lead to aortic regurgitation, dissection of the aorta, and rupture. Dilatation is probably accelerated by physical and emotional stress, as well as by pregnancy. Ocular Changes Dislocations of the lens may be readily apparent, but diagnosis usually requires pupillary dilatation and slit-lamp examination. The displacement is usually not progressive but may contribute to the formation of cataracts. The ocular globe is frequently elongated, most patients are myopic, and some develop retinal detachment. A few patients have lattice degeneration and retinal tears; most have adequate vision. Associated Changes Striae may occur over the shoulders and buttocks. Otherwise the skin is normal. A number of patients develop spontaneous pneumothorax. Inguinal and
incisional hernias are common. Marked dilatation of the dural sac is seen frequently in computed tomography scans, but the condition is usually asymptomatic. Patients are typically thin with little subcutaneous fat, but adults may develop centripetal obesity. Molecular Defects Most patients with the classic features ofMFS are heterozygotes for mutations in a gene on chromosome 15 that encodes fibrillin-1, a glycoprotein of 350 kDa that is a major component of elastin-associated microfibrils. These microfibrils are abundant in large blood vessels and the suspensory ligaments of the lens. Mutations in the fibrillin-1 gene include missense, nonsense, in-frame deletions, and RNA splicing mutations. Many of the mutations are single amino acid substitutions in the epidemal growth factor-like domains of the molecule that may be involved with calcium binding. Mutations in the fibrillin-2 gene that cause the MFS variant characterized by contractures appear to follow a similar pattern. As with most genetic diseases, the nature and location of mutations in the genes are only an approximate guide to the severity of the phenotype unless the same mutation has been seen in other members of the same family or in similar unrelated patients. However, there is a clustering of mutations in the middle portion of the molecule of fibrillin-1 encoded by exons 23 to 32 that causes the most severe phenotype, referred to as neonatal lethal MFS. The function of fibrillin has not been defined, but the data suggest that fibrillin self-assembles into a fibrillar structure and that the conformation and surface properties of the entire molecule are critical for normal assembly. Therefore, the functional consequences of mutations that change the amino acid sequence of fibrillin may be similar to the effects of mutations that change the conformation of a fibrillar collagen. Diagnosis The diagnosis is easily established if the patient and other members of the family have dislocated lenses, aortic dilatation, and long and thin extremities together with kyphoscoliosis or other chest deformities. The diagnosis is frequently made if ectopia lentis and an aneurysm of the ascending aorta occur in the absence of a Marfan habitus or a positive family history. All patients in whom the diagnosis is suspected should have a slit-lamp examination and an echocardiogram. Also, homocystinuria (Table 351-3) should be ruled out by a negative cyanide-nitroprusside test for disulfides in the urine (Chap. 352). A few patients with types I, II, and IIIEDS have ectopia lentis but lack the Marfan habitus and instead have characteristic skin changes not present inMFS. Patients with familial aortic aneurysms tend to develop aneurysms at the base of the abdominal aorta. The location of the aneurysms, however, is somewhat variable, and the high incidence of aortic aneurysms in the general population (1 in 100) makes the differential diagnosis difficult unless other features of MFS are clearly present. A few families with familial aortic aneurysms have mutations in the gene for type III procollagen (Fig. 351-4). TREATMENT There is no established treatment, but several investigators have recommended use of propranolol or other b-adrenergic blocking agents to lower blood pressure and thereby delay or prevent aortic dilatation. Surgical replacement of the aorta, aortic valve, and mitral valve has been successful in some patients, and all patients should be followed carefully with echocardiography and other techniques for evaluation of cardiovascular changes (Chap. 247). Patients should probably be advised of the risks of severe physical and emotional stress and of pregnancy.
The scoliosis tends to be progressive and should be treated by mechanical bracing and physical therapy if>20° or by surgery if it progresses to>45°. Estrogen has been tried in girls with scoliosis, but the results are inconclusive. Dislocated lenses rarely require surgical removal, but patients should be followed closely for retinal detachment. Diagnostic tests based on detection of fibrillin defects in cultured skin fibroblasts or DNA analysis of the gene are now available from several laboratories. DISEASES RELATED TO ELASTIN As may be expected from the role of elastin in maintaining the elasticity of skin, mutations in the elastin gene cause cutis laxa, a rare and heterogeneous group of disorders characterized by skin that is both lax and inelastic. Three different frame-shift mutations were found in three unrelated families with dominant forms of the disease. Surprisingly, other mutations in the elastin gene produce phenotypes that primarily involve the aorta, whose elasticity also depends on the presence of elastin. A large deletion that includes the elastin gene and probably several adjacent genes causes the Williams syndrome, characterized by supravalvular aortic stenosis, growth retardation, characteristic facies, and an unusual mental phenotype of low intelligence quotient together with a high degree of sociability. EPIDERMOLYSIS BULLOSA EBconsists of a group of similar disorders in which the skin and related epithelial tissues break and blister as the result of minor trauma. As with most heritable disorders of connective tissues, the clinical manifestations range from lethal to mild. Classification Four types of EB are defined on the basis of the level at which blistering occurs:EBsimplex for blistering in the epidermis, EB hemidesmosomal for fissures between kerotinocytes and between kerotinocytes and the basal lamina, EB junctional for blistering in the dermal-epidermal junction, and EB dystrophica for blistering in the dermis (Table 351-7). Incidence The incidence ofEB in the United States is estimated to be 1 in 50,000. Molecular Defects The molecular basis of several specific variants ofEB has been defined. A series of patients with EB simplex were found to have mutations in either keratin 14 or keratin 5, two of the major keratins in basal epithelial cells. Patients with the related syndrome, epidermolytic icthyosis, have mutations in keratin 1 and keratin 10. The new disease phenotype of hemidesmosomal EB has three clinical variants that are caused by mutations in one of four genes (Table 351-7): (1) A generalized atrophic and benign form of EB is caused by mutations in the COL17A1 gene for type XVII collagen; (2) a variant with EB associated with pyloric atresia and other intestinal abnormalities is caused by mutations in either the gene for thea6 integrin (ITG A6) or the gene for theb4 integrin (ITG B4); and (3) another variant characterized by relatively mild blistering at birth but associated with late-onset muscular dystrophy is caused by mutations in the gene for plectin (PLEC-1). Junctional EB is caused by mutations in any one of three genes for laminin (LAMA-3, LAMB-3, LAMC-2). The most severe dystrophic
form of EB is caused by mutations in the gene for type VII collagen (COL7A1). Diagnosis The diagnosis is based on skin that readily breaks and forms blisters.EBsimplex and EB hemidesmosomal are generally milder than EB junctional or EB dystrophica. EB dystrophica variants usually cause large and prominent scars. Precise classification within subtypes usually requires electron microscopy. The treatment is symptomatic. ALPORT SYNDROME (See also Chap. 275) ASis an inherited disorder characterized by hematuria. Four forms of the disease are now recognized: (1) classic AS, which is inherited as an X-linked disorder with hematuria, sensorineural deafness, and conical deformation of the anterior surface of the lens (lenticonus); (2) a subtype of the X-linked form associated with diffuse leiomyomatosis; (3) an autosomal recessive form; and (4) an autosomal dominant form. The two autosomal forms can cause renal disease without deafness or lenticonus. Incidence The incidence ofAS is about 1 in 10,000 in the general population and as high as 1 in 5000 in some ethnic groups. About 80% of AS patients have the X-linked variant. Molecular Defects Electron microscopy of kidneys from patients with classicASdemonstrates that the glomerular basement membrane is up to five times thicker than normal and that the lamina densa is distorted and split. The X-linked and autosomal recessive forms are caused primarily by mutations in genes for thea3(IV), a4(IV),a5(IV), or a6(IV) chains of type IV collagen, a major component of basement membranes. The type IV collagen in most membranes consists primarily of a1(IV) anda2(IV) chains folded into a large, rodlike molecule with globular ends and a long triple-helical domain that is interrupted by short sequences that do not form triple helices. The molecules self-assemble through both the globular ends and the long triple-helical domain to form a complex three-dimensional network. The four additional achains of type IV collagen are minor components of basement membranes, similar in structure, and are probably incorporated into the same or similar molecules. The six genes for the proteins are arranged in tandem pairs on different chromosomes in a head-to-head orientation and with overlapping promoters, i.e., the a1(IV) and a2(IV) genes are head-to-head on chromosome 13q34, the a3(IV) anda4(IV) genes are on chromosome 2q35-37, and thea5(IV) and a6(IV) genes are on chromosome Xq22. An X-linked variant is caused by mutations in the COL4A5 gene, and the X-linked variant associated with leiomyomatosis is caused by deletions that involve both the COL4A5 gene and the nearby COL4A6 gene. The autosomal recessive variants are caused by mutations in either the COL4A3 or COL4A4 genes. The mutations responsible for the autosomal dominant variants are still unknown, but they have been mapped to the same locus as the COL4A3 and COL4A4 genes. Diagnosis The diagnosis of classicAS is based on X-linked inheritance of hematuria, sensorineural deafness, and lenticonus. Because of the X-linked transmission, women are usually less severely affected than men and are generally underdiagnosed. The hematuria progresses to nephritis and may cause renal failure in late adolescence in affected males and at older ages in some women. The sensorineural deafness is
primarily in the high-tone range. It can frequently be detected only by an audiogram and is usually not progressive. The lenticonus can occur without nephritis but is generally considered to be pathognomonic of classic AS. TREATMENT There is no known treatment, but renal transplantation is usually successful. (Bibliography omitted in Palm version) Back to Table of Contents
352. INHERITED DISORDERS OF AMINO ACID METABOLISM AND STORAGE Nicola Longo All polypeptides and proteins are polymers of amino acids. Eight amino acids, referred to as essential, cannot be synthesized by humans and must be obtained from dietary sources. The others are formed endogenously. Although most of the body's amino acids are "tied up" in proteins, small intracellular pools of free amino acids are in equilibrium with extracellular reservoirs in plasma, cerebrospinal fluid, and the lumina of the gut and kidney. Physiologically, amino acids are more than mere "building blocks" of proteins. Some (glycine, glutamate,g-aminobutyric acid) are neurotransmitters. Others (phenylalanine, tyrosine, tryptophan, glycine) are precursors of hormones, coenzymes, pigments, purines, or pyrimidines. Each has a unique degradative pathway by which its nitrogen and carbon components are used for the synthesis of other amino acids, carbohydrates, and lipids. More than 70 disorders of amino acid metabolism are now known. The catabolic and storage defects (approximately 60) discussed in this chapter far outnumber the transport abnormalities (approximately 10) considered in Chap. 353. Each of these disorders is rare -- the incidences range from 1 in 10,000 for cystinuria or phenylketonuria to 1 in 200,000 for homocystinuria or alkaptonuria. Collectively, however, they occur in perhaps 1 in 500 to 1 in 1000 live births. Almost all are transmitted as autosomal recessive traits. The features of inherited disorders of amino acid catabolism are summarized inTable 352-1. In general, these disorders are named for the compound that accumulates to highest concentration in blood (-emias) or urine (-urias). For many conditions (often called aminoacidopathies), the parent amino acid is found in excess; for others, generally referred to as organic acidemias, products in the catabolic pathway accumulate. Which compound(s) accumulates depends, of course, on the site of the enzymatic block, the reversibility of the reactions proximal to the lesion, and the availability of alternative pathways of metabolic "runoff." For some amino acids, such as the sulfur-containing or branched-chain molecules, defects have been described at nearly each step in the catabolic pathway. For others, only small numbers of defective reactions have been described. Biochemical and genetic heterogeneity is common. Four distinct forms of hyperphenylalaninemia, seven forms of homocystinuria, and seven types of methylmalonic acidemia are recognized. Such heterogeneity reflects the presence of an even larger array of molecular defects. The manifestations of these conditions differ widely (Table 352-1). Some, such as sarcosinemia or hyperprolinemia, produce no clinical consequences. At the other extreme, complete deficiency of ornithine transcarbamylase or of branched-chain keto acid dehydrogenase is lethal in the untreated neonate. Central nervous system (CNS) dysfunction, in the form of developmental retardation, seizures, alterations in sensorium, or behavioral disturbances, is present in more than half the disorders. Protein-induced vomiting, neurologic dysfunction, and hyperammonemia occur in many disorders of urea cycle intermediates. Metabolic ketoacidosis, often accompanied by hyperammonemia, is a frequent presenting finding in the disorders of branched-chain amino acid metabolism. Occasional disorders produce focal tissue or organ involvement such as liver disease, renal failure, cutaneous abnormalities, or ocular lesions.
The clinical manifestations in many of these conditions can be prevented or mitigated if diagnosis is achieved early and appropriate treatment (i.e., dietary protein or amino acid restriction or vitamin supplementation) is instituted promptly. For this reason, several aminoacidopathies and organic acidemias are routinely screened in newborns using an array of chemical and microbiologic techniques. Once a presumptive diagnosis is made, confirmation can be provided by direct enzyme assay on extracts of leukocytes, erythrocytes, or cultured fibroblasts. DNA-based testing is possible for several disorders including phenylketonuria, ornithine transcarbamylase deficiency, citrullinemia, gyrate atrophy of the retina, propionic acidemia, and methylmalonic acidemia. As additional mutations are defined, DNA-based analysis may allow better predictions of outcome and improved therapeutic plans. Several of these disorders (including branched-chain ketoaciduria, isovaleric acidemia, propionic acidemia, methylmalonic acidemia, homocystinuria, cystinosis, phenylketonuria, ornithine transcarbamylase deficiency, citrullinemia, argininosuccinic aciduria) can be diagnosed prenatally by chemical analysis of amniotic fluid or by chemical, enzymatic, or DNA-based studies of fresh or cultured amniotic fluid cells. In addition to predicting genotype and alleviating parental anxiety, prenatal diagnosis has led to improved treatment of affected newborns. The focus of this chapter is on selected disorders that illustrate the principles, properties, and problems presented by the disorders of amino acid metabolism. THE HYPERPHENYLALANINEMIAS DEFINITION The hyperphenylalaninemias (Table 352-1) result from impaired conversion of phenylalanine to tyrosine. The most common and clinically important is phenylketonuria, which is characterized by an increased concentration of phenylalanine in blood, increased concentrations of phenylalanine and its by-products (notably phenylpyruvate, phenylacetate, phenyllactate, and phenylacetylglutamine) in urine, and severe mental retardation if untreated in infancy. GENETIC CONSIDERATIONS Each of the hyperphenylalaninemias results from reduced activity of phenylalanine hydroxylase. In humans, this complete enzyme system is expressed only in liver. Phenylalanine and molecular oxygen are substrates, and a reduced pteridine, tetrahydrobiopterin, is a cofactor (Fig. 352-1). Tyrosine and dihydrobiopterin are the products of this catalytic system, the latter being reconverted to tetrahydrobiopterin by two enzymes, pterin-4a-carbinolamine dehydratase and dihydropteridine reductase. Abnormalities in phenylalanine metabolism are autosomal recessive traits that occur in about 1 in 10,000 births. Phenylketonuria type I is widely distributed among whites and Orientals. It is rare in blacks. Phenylalanine hydroxylase activity in obligate heterozygotes is low but higher than in affected homozygotes. Adult heterozygous carriers are clinically well but can be identified by an increased ratio of phenylalanine/tyrosine in plasma in the semifasting state. They may have transient
cognitive impairment after phenylalanine loads. Hyperphenylalaninemias are caused by mutations in the gene encoding phenylalanine hydroxylase (PAH) or in genes encoding enzymes involved in tetrahydrobiopterin synthesis or recycling. In the vast majority of patients, mutations occur in the PAH gene (causing phenylketonuria type I), and>300 mutations have been identified. Mutations causing a complete impairment of enzyme activity, such as the R408W, are associated with a more severe outcome requiring stringent dietary restriction of phenylalanine. Mutations causing a less complete deficiency of the enzyme, such as the I65T, are associated with milder forms of phenylketonuria. In fewer than 2% of patients with phenylketonuria, mutations occur in other genes, including dihydrobiopterin reductase (DHPR) (30%), 6-pyruvoyl-tetrahydropterin synthase (6-PTS) (60%), GTP cyclohydrolase I (GTP-CH) (5%), and pterin-4a-carbinolamine dehydratase (PCD) (5%). In these cases, the impairment in phenylalanine hydroxylation results from tetrahydrobiopterin deficiency due to blocks in the pathway by which tetrahydrobiopterin is synthesized from GTP (phenylketonuria type III and malignant hyperphenylalaninemia) or deficiency of dihydropterine reductase (phenylketonuria type II), the enzyme that regenerates tetrahydrobiopterin from dihydrobiopterin (Fig. 352-1). Tyrosine hydroxylase and tryptophan hydroxylase also require tetrahydrobiopterin. Their products (L-dopa and 5-hydroxytryptophan) are essential for the synthesis of neurotransmitters. Heterozygotes for these conditions do not have hyperphenylalaninemia, but carriers of mutations in GTP cyclohydrolase have a peculiar form of dystonia, transmitted as a dominant trait with variable expressivity and higher penetrance in females; it is exquisitely responsive to levodopa. Neurotransmitter levels are not altered in transient hyperphenylalaninemia (sometimes called transient phenylketonuria), which has been described in some patients with pterin-4a-carbinolamine dehydratase deficiency. Etiology and Pathogenesis Phenylalanine accumulation in blood and urine and reduced tyrosine formation are direct consequences of the impaired hydroxylation. In untreated phenylketonuria and in its tetrahydrobiopterin-deficient variants, plasma concentrations of phenylalanine become sufficiently high [1 mmol/L (16 mg/dL)] to activate alternative pathways of metabolism and lead to formation of phenylpyruvate, phenylacetate, phenyllactate, and other derivatives that are rapidly cleared by the kidney and excreted in urine. The severe brain damage is due to several consequences of phenylalanine accumulation: competitive inhibition of transport of other amino acids required for protein synthesis, impaired polyribosome formation or stabilization, reduced synthesis and increased degradation of myelin, and inadequate formation of norepinephrine and serotonin. Phenylalanine is a competitive inhibitor of tyrosinase, a key enzyme in the pathway of melanin synthesis. This block, plus reduced availability of the melanin precursor tyrosine, accounts for the hypopigmentation of hair and skin. Clinical Manifestations No abnormalities are apparent at birth, but untreated children with classic phenylketonuria fail to attain early developmental milestones, develop microcephaly, and demonstrate progressive impairment of cerebral function. Hyperactivity, seizures, and severe mental retardation are major clinical problems later in life. Electroencephalographic abnormalities; "mousy" odor of skin, hair, and urine (due to phenylacetate accumulation); and a tendency to hypopigmentation and eczema complete the devastating clinical picture. In contrast, affected children who are detected
at birth and treated promptly show none of these abnormalities. Children with tetrahydrobiopterin deficiency, however, suffer a worse clinical course. Seizures appear early, followed by progressive cerebral and basal ganglia dysfunction (rigidity, chorea, spasms, hypotonia). Most succumb to secondary infection within a few years despite early diagnosis and vigorous treatment. A number of women with phenylketonuria who have been treated since infancy have reached adulthood and become pregnant. If maternal phenylalanine levels are not strictly controlled before and during pregnancy, their offspring are at risk, even if they are heterozygous, for maternal phenylketonuria. Affected children have microcephaly and an increased risk of congenital defects. After birth, these children have severe neurodevelopmental delay and growth retardation. Diagnosis Plasma phenylalanine concentrations are usually normal at birth in the hyperphenylalaninemias but rise rapidly after institution of protein feedings. To prevent mental retardation, diagnosis and initiation of dietary treatment of classic phenylketonuria must occur before the child is 3 weeks of age. For this reason, most newborns in North America and Europe are screened by determinations of blood phenylalanine concentration using the Guthrie bacterial inhibition assay. Abnormal values are confirmed using quantitative analysis of plasma amino acids. Prenatal diagnosis of type I phenylketonuria is now feasible using DNA-based tests that detect specific mutations or polymorphic markers that are linked to the PAH gene. In newborns with type I phenylketonuria, plasma levels depend on the amount of phenylalanine in the diet and the degree of impairment of phenylalanine hydroxylase. Dietary phenylalanine restriction is usually instituted if blood phenylalanine levels are >250 umol/L (4 mg/dL). Careful monitoring of these infants reveals the degree of phenylalanine hydroxylase impairment and dictates the degree of dietary phenylalanine restriction. Deficiency of tetrahydrobiopterin, which occurs in 1 to 2% of newborns with increased blood phenylalanine, is excluded by screening the urinary pteridine profile and by assay of dihydropteridine reductase activity on dried blood specimens. Dihydropteridine reductase deficiency and the blocks in tetrahydrobiopterin synthesis can be detected in utero using assays on cultured amniocytes. Tetrahydrobiopterin deficiency manifests with hyperphenylalaninemia and progressive neurologic impairment despite prompt dietary restriction of phenylalanine. TREATMENT Phenylketonuria is the first inherited metabolic disease in which a strategy of reducing the accumulation of the offending metabolite prevented the dire clinical consequences. This was accomplished by a special diet low in phenylalanine and supplemented with tyrosine. Tyrosine becomes an essential amino acid in phenylalanine hydroxylase deficiency. Sufficient phenylalanine is provided for new protein synthesis and normal growth. This amount varies with age and requires frequent adjustments, especially early in life. Ordinarily, plasma phenylalanine concentrations are maintained between 120 and 360 umol/L (2 and 6 mg/dL). Such diet therapy must be instituted during the first 3 weeks of life. Even then, modestCNSdysfunction may occur with more deleterious
mutations or after excess protein intake. Because uncontrolled hyperphenylalaninemia results in brain damage, dietary restriction should be continued and monitored indefinitely, recognizing that transient phenylketonuria may not require lifelong therapy. An enteric-coated formulation of phenylalanine ammonia lyase, which degrades phenylalanine in the gut, is under investigation for its potential to reduce the strain imposed by the special diet. Children with tetrahydrobiopterin deficiency deteriorate despite dietary phenylalanine restriction. Such patients may be helped, however, by a regimen in which dietary phenylalanine restriction is combined with tetrahydrobiopterin supplements, levodopa, 5-hydroxytryptophan, and carbidopa. Finally, the deleterious consequences of maternal phenylketonuria can be minimized by continuing lifelong phenylalanine-restricted diets in females with phenylketonuria and assuring strict phenylalanine restriction prior to conception and throughout gestation. THE HOMOCYSTINURIAS (HYPERHOMOCYSTEINEMIAS) The homocystinurias are seven biochemically and clinically distinct disorders (Table 352-1), each characterized by increased concentration of the sulfur-containing amino acid homocystine in blood and urine. The most common form results from reduced activity of cystathionineb-synthase (CBS), an enzyme in the transsulfuration pathway that converts methionine to cysteine (Fig. 352-2). The other forms are the result of impaired conversion of homocysteine to methionine, a reaction catalyzed by homocysteine:methyltetrahydrofolate methyltransferase (also known as methionine synthase) and two essential cofactors, methyltetrahydrofolate and methylcobalamin (methyl-vitamin B12). Depending on the underlying disorder, some patients show chemical and, in some instances, clinical improvement following administration of specific vitamin supplements (pyridoxine, folate, or cobalamin) (Chap. 75). In classic homocystinuria, the levels of free homocystine in plasma increase and result in homocystinuria. Hyperhomocysteinemia refers to increased total plasma concentration of homocysteine in the sulfhydryl and disulfide form, free and protein-bound. Hyperhomocysteinemia, in the absence of significant homocystinuria, is found in individuals who are heterozygous or homozygous for certain genetic defects that impair folate or vitamin B12metabolism or cause cystathionine synthase deficiency. Changes of homocysteine levels are also observed with increasing age; in postmenopausal women; in patients with renal failure, hypothyroidism, leukemias, or psoriasis; and during therapy with drugs such as methotrexate, nitrous oxide, isoniazid, and some antiepileptic agents. Homocysteine acts as an atherogenic and thrombophilic agent. An increase in total plasma homocysteine represents an independent risk factor for coronary, cerebrovascular, and peripheral arterial disease as well as for deep-vein thrombosis (Chap. 241). Homocysteine is synergistic with hypertension and smoking, and it is additive with other risk factors that predispose to peripheral arterial disease. In addition, hyperhomocysteinemia and folate and vitamin B12deficiency have been associated with an increased risk of neural tube defects in pregnant women. CYSTATHIONINEb-SYNTHASE DEFICIENCY
Definition Deficiency of this enzyme leads to increased concentrations of methionine and homocystine in body fluids and to decreased concentrations of cysteine and cystine. Clinical hallmarks include dislocation of optic lenses (usually downward and medially), mental retardation, marfanoid habitus, osteoporosis, and thrombotic vascular disease. GENETIC CONSIDERATIONS The sulfur atom of the essential amino acid methionine is transferred to cysteine by the transsulfuration pathway (Fig. 352-2). In one of these steps, homocysteine condenses with serine to form cystathionine. This reaction is catalyzed by the pyridoxal phosphate-dependent enzymeCBS. Heterogenous mutations in the CBS gene are present in different families. The G307S mutation is associated with lack of response to pyridoxine, whereas the I278T mutation correlates with pyridoxine-responsiveness and a milder clinical phenotype. Homocystinuria is common in Ireland (1 in 60,000 births) but rare elsewhere (50. These disorders present as progressive and often symmetric muscle weakness. Patients usually report increasing difficulty with everyday tasks requiring the use of proximal muscles, such as getting up from a chair, climbing steps, stepping onto a curb, lifting objects, or combing hair. Fine-motor movements that depend on the strength of distal muscles, such as buttoning a shirt, sewing, knitting, or writing, are affected only late in the course ofPM andDM, but fairly early inIBM. Falling is common in IBM because of early involvement of the quadriceps muscle with buckling of the knees. Ocular muscles are spared, even in advanced, untreated cases; if these muscles are affected, the diagnosis of inflammatory myopathy should be in doubt. Facial muscles are unaffected in PM and DM, but mild facial muscle weakness occurs in up to 60% of patients with IBM. In all forms of inflammatory myopathy, pharyngeal and neck-flexor muscles are often involved, causing dysphagia or difficulty in holding up the head (neck drop). In advanced and rarely in acute cases, respiratory muscles may also be affected. Severe weakness, if untreated, is almost always associated with muscle wasting. Sensation remains normal. The tendon reflexes are preserved but may be absent in severely weakened or atrophied muscles, especially in IBM where atrophy of the quadriceps and the distal muscles is common. Myalgia and muscle tenderness may occur in a small number of patients, usually early in the diseaese and more often in DM than in PM. Weakness in PM and DM progresses subacutely over a period of weeks or months and rarely acutely; by contrast, IBM progresses very slowly, over years, and its course may simulate late-life muscular dystrophies (Chap. 383) or slowly progressive motor neuron disorders (Chap. 365). SPECIFIC FEATURES (Table 382-1) Polymyositis In most patients, the actual onset ofPM is not easily determined, and patients typically delay seeking medical advice for several months. This is in contrast toDM, in which the rash facilitates early recognition (see below). PM is a subacute inflammatory myopathy affecting adults, and rarely children, who do not have any of the following: rash, involvement of the extraocular and facial muscles, family history of a neuromuscular disease, history of exposure to myotoxic drugs or toxins, endocrinopathy, neurogenic disease, muscular dystrophy, biochemical muscle disorder (deficiency of a muscle enzyme), orIBM as excluded by muscle biopsy analysis (see
below). PM may occur either in isolation, in association with a systemic autoimmune or connective tissue disease, or with known viral or bacterial infection. D-Penicillamine and, on occasion, zidovudine (AZT) may also produce an inflammatory myopathy similar to PM. Dermatomyositis (Figs. 382-CD1,382-CD2and382-CD3)DM is a distinctive entity identified by a characteristic rash accompanying, or more often preceding, muscle weakness. The rash may consist of a heliotrope rash (blue-purple discoloration) on the upper eyelids with edema, a flat red rash on the face and upper trunk, and erythema of the knuckles with a raised violaceous scaly eruption (Gottron rash) that later results in scaling of the skin (see Plates IIE-63 andIIE-65). The erythematous rash can also occur on other body surfaces, including the knees, elbows, malleoli, neck and anterior chest (often in a V sign), or back and shoulders (shawl sign), and may worsen after sun exposure. In some patients the rash is pruritic, especially on the scalp, chest, and back. Dilated capillary loops at the base of the fingernails are also characteristic. The cuticles may be irregular, thickened, and distorted, and the lateral and palmar areas of the fingers may become rough and cracked, with irregular, "dirty" horizontal lines, resembling mechanic's hands. The weakness can be mild, moderate, or severe enough to lead to quadraparesis. At times, the muscle strength appears normal, hence the term dermatomyositis sine myositis. When muscle biopsy is performed in such cases, however, significant perivascular and perimysial inflammation is seen. In children, DM resembles the adult disease, except for more frequent extramuscular manifestations, as discussed later. A common early abnormality in children is "misery," defined as an irritable child who appears uncomfortable, has a red flush on the face, is fatigued, does not wish to socialize, and has a varying degree of proximal muscle weakness. A tiptoe gait due to flexion contracture of the ankles is also common. DMusually occurs alone but may overlap with scleroderma and mixed connective tissue disease. Fasciitis and thickening of the skin similar to that seen in chronic cases of DM have occurred in patients with the eosinophilia-myalgia syndrome associated with the ingestion of contaminated L-tryptophan. Inclusion Body Myositis In patients ³50,IBM is the most common of the inflammatory myopathies. It is often misdiagnosed asPM and suspected only retrospectively when a patient with presumed PM does not respond to therapy. Weakness and atrophy of distal muscles, especially foot extensors and deep finger flexors, occur in almost all cases of IBM and may be a clue to early diagnosis. Some patients present with falls because their knees collapse due to early quadriceps weakness. Others present with weakness in the small muscles of the hands, especially finger flexors, and complain of inability to hold certain objects, such as golf clubs, or perform certain tasks, such as turning keys or tying knots. On occasion, the weakness and accompanying atrophy can be asymmetric and selectively involve the quadriceps, iliopsoas, triceps, biceps, and finger flexors, resembling a lower motor neuron disease. Dysphagia is common, occurring in up to 60% of IBM patients, and may lead to episodes of choking. Sensory examination is generally normal; some patients have mildly diminished vibratory sensation at the ankles that presumably is age-related. The distal weakness does not represent motor neuron or peripheral nerve involvement but results from the myopathic process affecting distal muscles. The diagnosis is always made by the characteristic findings on the muscle biopsy, as discussed below. Disease progression is slow but steady, and most
patients require an assistive device such as cane, walker, or wheelchair within several years of onset. In at least 20% of cases,IBM is associated with systemic autoimmune or connective tissue diseases. Familial aggregation has also been noted in coaffected siblings with typical IBM; such cases have been designated as familial inflammatory IBM. This disorder is distinct from hereditary inclusion body myopathy (h-IBM), which describes a heterogeneous group of recessive and less frequently dominant, inherited syndromes. The h-IBMs are noninflammatory myopathies with clinical profiles distinct from sporadic IBM. A subset of h-IBM that spares the quadriceps muscle has emerged as a distinct entity. This disorder, originally described in Iranian Jews and now seen in many ethnic groups, is linked to chromosome 9p1. ASSOCIATED CLINICAL FINDINGS Extramuscular Manifestations In addition to the primary myopathy, a number of extramuscular manifestations may be present to a varying degree in patients withPM orDM: 1. Systemic symptoms, such as fever, malaise, weight loss, arthralgia, and Raynaud's phenomenon especially when inflammatory myopathy is associated with a connective tissue disorder. 2. Joint contractures, mostly in DM and especially in children. 3. Dysphagia and gastrointestinal symptoms due to involvement of the oropharyngeal striated muscles and upper esophagus. Dysphagia may be prominent in the active stages of DM and is frequent in IBM. Gastrointestinal ulcerations due to vasculitis and infection were common in children with DM before the use of immunosuppressive drugs. 4. Cardiac disturbances, including atrioventricular conduction defects, tachyarrythmias, dilated cardiomyopathy, and low ejection fraction. Congestive heart failure and myocarditis may also occur, either from the disease itself or from hypertension associated with long-term use of glucocorticoids. 5. Pulmonary dysfunction, due to primary weakness of the thoracic muscles, drug-induced pneumonitis (e.g., from methotrexate), or interstitial lung disease may cause dyspnea, nonproductive cough, and aspiration pneumonia. Interstitial lung disease may precede myopathy or occur early in the disease, and develops in up to 10% of patients with PM or DM. 6. Subcutaneous calcifications, sometimes extruding on the skin and causing ulcerations and infections, are seen in DM, primarily in children. Malignancies Although all the inflammatory myopathies can have a chance association with malignant lesions, especially in older age groups, the incidence of malignant conditions appears to be specifically increased only in patients withDM but not PMor IBM. The most common tumors associated with DM are ovarian cancer, breast cancer, melanoma, and colon cancer. The extent of the search that should be conducted for an
occult malignant neoplasm in adults with DM depends on the clinical circumstances. Tumors in these patients are usually uncovered by abnormal findings in the medical history and physical examination and not through an extensive radiologic blind search. Thus the weight of evidence argues against performing expensive, invasive, and nondirected tumor searches. When a suspected malignancy is not apparent, a complete annual physical examination with pelvic, breast, and rectal examinations; urinalysis; complete blood count; blood chemistry tests; and a chest film should suffice. Overlap The term overlap syndrome has been used loosely to describe the frequent association of inflammatory myopathies with connective tissue diseases. A well-characterized overlap syndrome occurs in patients withDM who also have manifestations of systemic sclerosis or mixed connective tissue disease, such as sclerotic thickening of the dermis, contractures, esophageal hypomotility, microangiopathy, and calcium deposits (Table 382-1). By contrast, signs of rheumatoid arthritis, systemic lupus erythematosus, or Sjogren's syndrome are very rare in patients with DM. Patients with the overlap syndrome of DM and systemic sclerosis may have a specific antinuclear autoantibody, the anti-PM/Scl, directed against a nucleolar-protein complex. PATHOGENESIS An autoimmune origin of these disorders is supported by their association with other systemic autoimmune, viral, or connective tissue diseases; the presence of various autoantibodies; their association with histocompatibility genes; the evidence of T cell-mediated myocytotoxicity or complement-mediated microangiopathy; and their response to immunotherapies. However, the specific muscle or capillary target antigens have not been identified, and the agents initiating self-sensitization are still unknown. Autoantibodies and Immunogenetics Various autoantibodies against nuclear antigens (antinuclear antibodies) and cytoplasmic antigens are found in up to 20% of patients with inflammatory myopathies. The antibodies to cytoplasmic antigens, present in4 to 6 weeks because of the development of tolerance and the risk of abuse and dependence. It is important to warn patients that concomitant usage of alcohol or other sedating drugs may result in neurotoxicity and impair their ability to function. An optimistic
approach that encourages the patient to clarify environmental precipitants, anticipate his or her reactions, and plan effective response strategies are essential elements of therapy. Adverse effects of benzodiazepines generally parallel their relative half-lives. Longer-acting agents, such as diazepam, chlordiazepoxide, flurazepam, and clonazepam, tend to accumulate active metabolites, with resultant sedation, impairment of cognition, and poor psychomotor performance. Shorter-acting compounds, such as alprazolam and oxazepam, can result in daytime anxiety, early morning insomnia, and with discontinuation, rebound anxiety and insomnia. Although patients develop tolerance to the sedative effects of benzodiazepines, they are less likely to habituate to the adverse psychomotor effects. Withdrawal from the longer half-life benzodiazepines can be accomplished through gradual, stepwise dose reduction (by ~10% every 1 to 2 weeks) over 6 to 12 weeks. It is usually more difficult to taper patients off shorter-acting benzodiazepines. Physicians may need to switch the patient to a benzodiazepine with a longer half-life or use an adjunctive medication, such as a beta blocker or carbamazepine, before attempting to discontinue the benzodiazepine. Withdrawal reactions vary in severity and duration; they can include depression, anxiety, delirium, lethargy, diaphoresis, tinnitus, autonomic arousal, unusual neuromuscular movements, and, rarely, seizures. Buspirone, an azaspirone, is a nonbenzodiazepine anxiolytic agent. It is nonsedating, does not lead to tolerance or dependence, does not interact with benzodiazepine receptors or alcohol, and has no abuse or disinhibition potential. However, it requires several weeks to take effect and requires thrice-daily dosing. Patients who were previously responsive to a benzodiazepine are unlikely to rate buspirone as equally effective, but patients with head injury or dementia who have symptoms of anxiety and/or agitation may do well with this agent. Administration of benzodiazepines to geriatric patients requires special care. Such patients have increased drug absorption; decreased hepatic metabolism, protein binding, and renal excretion; and an increased volume of distribution. These factors, together with the likely presence of comorbid medical illnesses and medication, dramatically increase the likelihood of toxicity. Iatrogenic psychomotor impairment can result in falls and fractures, confusional states, or motor vehicle accidents. If used, agents in this class should be started at the lowest possible dose, and results should be monitored closely. Benzodiazepines are contraindicated during pregnancy and breast-feeding. PHOBIC DISORDERS Clinical Manifestations The cardinal feature of phobic disorders is a marked and persistent fear of objects or situations, exposure to which results in an immediate anxiety reaction. The patient avoids the phobic stimulus, and this avoidance usually impairs occupational or social functioning. Panic attacks may be triggered by the phobic stimulus or may emerge spontaneously during the course of the illness. Unlike patients with other anxiety disorders, individuals with phobias experience anxiety only in specific situations. Common phobias include fear of closed spaces (claustrophobia), fear of blood, and fear of flying. Social phobia is distinguished by a specific fear of social or
performance situations in which the individual is exposed to unfamiliar individuals or to possible examination and evaluation by others. Examples include having to converse at a party, use public restrooms, and meet strangers. In each case, the affected individual is aware that the experienced fear is excessive and unreasonable given the circumstance. The specific content of a phobia may vary across gender, ethnic, and cultural boundaries. Phobic disorders are common, with a 1-year prevalence rate of 9% and a lifetime rate of 10 to 11%. Onset is typically in childhood to early adulthood. Familial aggregation may occur. In one study of female twins, concordance rates for agoraphobia, social phobia, and animal phobia was found to be 23% for monozygotic twins and 15% for dizygotic twins. Full criteria for diagnosis are usually satisfied first in adulthood, but behavioral avoidance of unfamiliar people, situations, or objects dating from early childhood is common. TREATMENT Recent controlled trials have documented the efficacy of several pharmacologic agents in the treatment of phobic disorders. Beta blockers (e.g., propranolol, 20 to 40 mg orally 2 h before the event) are particularly effective in the treatment of "performance anxiety" (but not general social phobia) and appear to achieve their benefit by preventing the occurrence of peripheral manifestations of anxiety, such as perspiration, tachycardia, palpitations, and tremor.MAOIsalleviate social phobia independently of their antidepressant activity, andSSRIsappear to be effective also. Benzodiazepines can be helpful in reducing fearful avoidance, but the chronic nature of phobic disorders limits their usefulness. Behaviorally focused psychotherapy is an important component of treatment, as relapse rates are high when medication is used as the sole treatment. Cognitive-behavioral strategies are the cornerstone of treatment; these are based upon the finding that distorted perceptions and interpretations of fear-producing stimuli play a major role in perpetuation of phobias. Individual and group therapy sessions teach the patient to identify specific negative thoughts associated with the anxiety-producing situation and help to reduce the patient's fear of loss of control. In desensitization therapy, hierarchies of feared situations are constructed and the patient is encouraged to pursue and master gradual exposure to the anxiety-producing stimuli. Patients with social phobia, in particular, have a high rate of comorbid alcohol abuse, as well as of other psychiatric conditions (e.g., eating disorders), necessitating the need for parallel management of each disorder if anxiety reduction is to be achieved. STRESS DISORDERS Clinical Manifestations Patients may develop anxiety after exposure to extreme traumatic events such as the threat of personal death or injury or the death of a loved one. The reaction may occur shortly after the trauma (acute stress disorder) or be delayed and subject to recurrence (PTSD) (Table 385-8). In both syndromes, individuals experience associated symptoms of detachment and loss of emotional responsivity. The patient may feel depersonalized and unable to recall specific aspects of the trauma,
though typically it is reexperienced through intrusions in thought, dreams, or flashbacks, particularly when cues of the original event are present. Patients often actively avoid stimuli that precipitate recollections of the trauma and demonstrate a resulting increase in vigilance, arousal, and startle response. Patients with stress disorders are at risk for the development of other anxiety, mood, and substance-related disorders. Between 5 and 10% of Americans will at some time in their life satisfy criteria for PTSD, with women more likely to be affected than men. Risk factors for the development ofPTSDinclude a past psychiatric history and personality characteristics of high neuroticism and extroversion. Studies of monozygotic and dizygotic twins showed a substantial influence of genetics on all symptoms associated with PTSD, with no evidence for an environment effect. Etiology and Pathophysiology It is hypothesized that inPTSDthere is excessive release of norepinephrine from the locus coeruleus in response to stress. Increased noradrenergic activity at locus coeruleus projection sites in hippocampus and amygdala theoretically facilitates encoding of fear-based memories. Greater sympathetic responses to cues associated with the traumatic event occurs in PTSD. TREATMENT Acute stress reactions are usually self-limited, and treatment typically involves the short-term use of benzodiazepines and supportive/expressive psychotherapy. The chronic and recurrent nature ofPTSD, however, requires a more complex approach employing drug and behavioral treatments.TCAssuch as imipramine and amitriptyline, theMAOIphenelzine, and theSSRIs(fluoxetine, sertraline, citalopram, paroxetine) can all reduce anxiety, symptoms of intrusion, and avoidance behaviors. Trazodone, a sedating antidepressant, is frequently used at night to help with insomnia (50 to 150 mg qhs). Carbamazepine, valproic acid, or alprazolam have also independently produced improvement in uncontrolled trials. There is frequent comorbidity with substance abuse, especially alcohol. Psychotherapeutic strategies are used in treatment ofPTSD to help the patient overcome avoidance behaviors and demoralization and master fear of recurrence of the trauma; therapies that encourage the patient to dismantle avoidance behaviors through stepwise focusing on the experience of the traumatic event are the most effective. OBSESSIVE-COMPULSIVE DISORDER Clinical Manifestations Obsessive-compulsive disorder (OCD) was previously considered a relatively rare condition, but recent epidemiologic data indicate a lifetime prevalence of 2 to 3% worldwide. OCD is characterized by obsessive thoughts and compulsive behaviors that impair everyday functioning. Fears of contamination and germs are common, as are handwashing, counting behaviors, and having to check and recheck such actions as whether a door is locked. The degree to which the disorder is disruptive for the individual varies, but in all cases obsessive-compulsive activities take up >1 h per day and are undertaken to relieve the anxiety triggered by the core fear. Patients often conceal their symptoms, usually because they are embarrassed by the content of their thoughts or the nature of their actions. Physicians must ask specific
questions regarding recurrent thoughts and behaviors, particularly if physical clues such as chafed and reddened hands or patchy hair loss (from repetitive hair pulling, or trichotillomania) are present. Tics are sometimes associated with OCD. OCD usually has a gradual onset, beginning in early adulthood, but childhood onset is not rare. The disorder usually has a waxing and waning course, but some cases may show a steady deterioration in psychosocial functioning. Etiology and Pathophysiology A genetic contribution toOCD is suggested by a higher monozygotic than dizygotic concordance rate and the fact that familial studies show an aggregation with Tourette's disorder. OCD is more common in males and in first-born children. The anatomy of obsessive-compulsive behavior is thought to involve a frontal-subcortical neural circuit involving the orbital frontal cortex, caudate nucleus, and globus pallidus. Neuroimaging studies have demonstrated a decrease in caudate nucleus volume, abnormalities in frontal lobe white matter, and increases in glucose metabolism in the orbital cortex of the frontal lobes and the head of the caudate nucleus. The caudate nucleus seems particularly involved in the acquisition and maintenance of habit and skill learning, and interventions that are successful in reducing obsessive-compulsive behaviors are paralleled by a comparable decrease in caudate glucose metabolic rate. TREATMENT Clomipramine, fluoxetine, and fluvoxamine are approved for the treatment ofOCD. Clomipramine is aTCA that is often tolerated poorly owing to significant anticholinergic and sedative side effects at the doses required to treat the illness (150 to 250 mg/d). Its efficacy in OCD is unrelated to its antidepressant activity. Fluoxetine (40 to 60 mg/d) and fluvoxamine (100 to 300 mg/d) are as effective as clomipramine and show a more benign side-effect profile. Fluvoxamine, a structurally uniqueSSRI, is metabolized through the hepatic P450 microsomal system (as is fluoxetine); it appears to inhibit the III A4 isoenzyme specifically and should not be given with other drugs that act on III A4, such as terfenadine and astemizole, because life-threatening cardiac arrhythmias may result. Only 50 to 60% of patients with OCD show an acceptable degree of improvement with pharmacotherapy alone. In treatment-resistant cases, augmentation with other serotonergic agents, such as buspirone, or with a neuroleptic or benzodiazepine may be beneficial. When a therapeutic response is achieved, long-duration maintenance therapy is usually indicated. For many individuals, particularly those with time-consuming compulsions, behavior therapy will result in as much improvement as that afforded by medication. Effective techniques include the gradual increase in exposure to stressful situations, maintenance of a diary to clarify stressors, and homework assignments that substitute new activities for their compulsive behavior. MOOD DISORDERS Mood disorders are characterized by a disturbance in the regulation of mood, behavior, and affect. Mood disorders are subdivided into (1) depressive disorders, (2) bipolar
disorders, and (3) depression in association with medical illness or alcohol and substance abuse (Chaps. 387 through 389). Depressive disorders are differentiated from bipolar disorders by the absence of a manic or hypomanic episode. The relationship between pure depressive syndromes and bipolar disorders is not well understood; depression occurs at increased frequency in families of bipolar individuals, but the reverse is not true. Depression in general is associated with high disability and societal cost; in the Global Burden of Disease Study conducted by the World Health Organization, unipolar major depression ranked fourth in percentage of disability-adjusted life years and was projected to rank second in the year 2020. DEPRESSION IN ASSOCIATION WITH MEDICAL ILLNESS Depression occurring in the context of medical illness is difficult to evaluate. Depressive symptomatology may reflect the psychological stress of coping with the disease, may be caused by the disease process itself or by the medications used to treat it, or may simply coexist in time with the medical diagnosis. Virtually every class of medication includes some agent that can induce depression. Antihypertensive drugs, anticholesterolemic agents, and antiarrhythmic agents are commonly used classes of medications that can trigger depressive symptoms. Among the antihypertensive agents, b-adrenergic blockers and, to a lesser extent, calcium channel blockers are the most likely to cause depressed mood. Iatrogenic depression should also be considered in patients receiving glucocorticoids, antimicrobials, systemic analgesics, antiparkinsonian medications, and anticonvulsants. To decide whether a causal relationship exists between pharmacologic therapy and a patient's change in mood, it is necessary to chart the chronology of symptoms and sometimes to undertake an empirical trial of an alternative medication. Between 20 and 30% of cardiac patients manifest a depressive disorder; an even higher percentage experience depressive symptomatology when self-reporting scales are used. Depressive symptoms following myocardial infarction impair rehabilitation and are associated with higher rates of mortality and medical morbidity. Depressed patients often show decreased variability in heart rate (an index of reduced parasympathetic nervous system activity), and this has been proposed as one mechanism by which depression may predispose individuals to ventricular arrhythmia and increased morbidity. AlthoughTCAshave been used to treat depression in individuals with cardiac disease for a number of years, and although the quinidine-like effect of tricyclics may be useful in patients with preexisting arrhythmias, TCAs are contraindicated in patients with preexisting bundle branch block. They may also paradoxically precipitate arrhythmias. Tricyclic-induced tachycardia is an additional concern in patients with congestive heart failure. Experience with theSSRIs is more limited, but thus far they appear not to induceECGchanges or adverse cardiac events. SSRIs may interfere with hepatic metabolism of anticoagulants, however, causing increased anticoagulation. Epidemiologic surveys of depression in patients with cancer show a wide variability in prevalence, as might be predicted by differences in tumor site, severity of illness, and type of medical or surgical intervention. There is an overall mean prevalence of 25%, but depression occurs in 40 to 50% of patients with cancers of the pancreas or oropharynx. Assessment of the validity of prevalence rates is complicated by the fact
that extreme cachexia may be misinterpreted as part of the symptom complex of depression. The higher prevalence of depression in patients with pancreatic cancer nevertheless persists when patients are compared to those with advanced gastric cancer. Initiation of antidepressant medication in cancer patients has been shown to improve quality of life as well as mood. Psychotherapeutic approaches, particularly group therapy, may have some effect on short-term depression, anxiety, and pain symptoms and on recurrence rates and long-term survival. In a study of female patients with metastatic breast cancer, patients in group therapy had longer survival than control patients. Depression occurs frequently in patients with neurologic disorders, particularly cerebrovascular disorders, Parkinson's disease, multiple sclerosis, and traumatic brain injury. Left-hemisphere strokes, particularly those involving the dorsal lateral frontal cortex, are most likely to cause depression. Both tricyclic andSSRIantidepressants are effective in the treatment of depression secondary to stroke, as are stimulant compounds and, in some patients,MAOIs. The reported prevalence of depression in patients with diabetes mellitus varies from 8 to 27%, with the severity of the mood state correlating with the physical symptoms of illness and the degree of hyperglycemia. Pharmacologic treatment of depression is complicated by antidepressant effects on the blood glucose level.MAOIscan induce hypoglycemia and weight gain.TCAs can lead to hyperglycemia and carbohydrate craving.SSRIs, like MAOIs, may cause a reduction in fasting plasma glucose, but they are easier to use and may also improve dietary and medication compliance. Hypothyroidism is frequently associated with features of depression, most commonly depressed mood and memory impairment. Hyperthyroid states may also present in a similar fashion, usually in geriatric populations. Improvement in mood usually follows normalization of thyroid function, but adjunctive antidepressant medication is sometimes required. Patients with subclinical hypothyroidism can also experience symptoms of depression and cognitive difficulty that respond to thyroid replacement. DEPRESSIVE DISORDERS Clinical Manifestations Major depression is defined as depressed mood on a daily basis for a minimum duration of 2 weeks (Table 385-9). An episode may be characterized by sadness, indifference or apathy, or irritability and is usually associated with change in neurovegetative functions, including sleep patterns, appetite and weight, motor agitation or retardation, fatigue, impairment in concentration and decision making, feelings of shame or guilt, and thoughts of death or dying. Patients with depression have a profound loss of pleasure in all enjoyable activities, exhibit early morning awakening, feel that the dysphoric mood state is qualitatively different from sadness, and often notice a diurnal variation in mood (worse in morning hours). Paradoxically, these more severe features predict a good response to antidepressant treatment. Approximately 15% of the population experiences a major depressive episode at some point in life, and 6 to 8% of all outpatients in primary care settings satisfy diagnostic criteria for the disorder. Depression is often undiagnosed, and, even more frequently, it is treated inadequately. If a physician suspects the presence of a major depressive
episode, the initial task is to determine whether it represents unipolar or bipolar depression or is one of the 10 to 15% of cases that are secondary to general medical illness or substance abuse. Physicians should also assess the risk of suicide by direct questioning, as patients are often reluctant to verbalize such thoughts without prompting. If specific plans are uncovered or if significant risk factors exist (e.g., a past history of suicide attempts, profound hopelessness, concurrent medical illness, substance abuse, or social isolation), the patient must be referred to a mental health specialist for immediate care. In evaluating suicidal risk the physician should specifically probe each of these areas in an empathic and hopeful manner, being sensitive to denial and possible minimization of distress. The presence of anxiety, panic, or agitation significantly increases near-term suicidal risk. Nearly 15% of patients whose depressive illness goes untreated will commit suicide; most will have sought help from a physician within 1 month of their death. In some depressed patients, the mood disorder does not appear to be episodic and is not clearly associated with either psychosocial dysfunction or change from the individual's usual experience in life. Dysthymic disorder consists of a pattern of chronic (at least 2 years), ongoing, mild depressive symptoms that are less severe and less disabling than those found in major depression; the two conditions are sometimes difficult to separate, however, and can occur together ("double depression"). Many patients who exhibit a profile of pessimism, disinterest, and low self-esteem respond to antidepressant treatment. Dysthymic disorder exists in ~5% of primary care patients. Studies of various cultures have shown that external manifestations of depression differ but the core symptoms remain the same. The incidence of depression increases with age; the disorder is approximately twice as prevalent in women as in men, regardless of age. These gender differences were previously believed to reflect sociocultural factors, but recent longitudinal twin studies indicate that the liability to major depression in adult women is largely genetic in origin, and that the effect of environmental factors is transitory and does not affect lifetime prevalence. The relationship between psychological stress, negative life events, and the onset of depressive episodes is complex. Negative life events can precipitate and contribute to depression, but recent data indicate that genetic factors influence the sensitivity of individuals to these stressful events. In most cases, both biologic and psychosocial factors are involved in the precipitation and unfolding of depressive episodes. The most potent stressors appear to involve death of a relative, assault, or severe marital or relationship problems. Unipolar depressive disorders usually have their onset in early adulthood, and recurrences over the course of a lifetime are likely. The best predictor of future risk is the number of past episodes; 50 to 60% of patients who have a first episode have at least one or two more episodes. Some patients experience multiple episodes that become more severe and frequent over time. The duration of an untreated episode varies greatly, ranging from a few months to ³1 year. The pattern of recurrence and clinical progression in a developing episode is also variable. Within an individual, there is often long-term stability in phenotype (presenting symptoms, frequency and duration of episodes). In a minority of patients, the severity of the depressive episode may progress to psychotic symptomatology; in elderly patients, depressive symptoms may be associated with confusion and mistaken for dementia (i.e., "pseudodementia"). A seasonal pattern of depression, called seasonal affective disorder, may manifest with
onset and remission of episodes at predictable times of the year. This disorder is more common in women, whose symptoms are anergy, fatigue, weight gain, hypersomnia, and episodic carbohydrate craving. The prevalence increases with distance from the equator, and mood improvement may occur by altering light exposure. Etiology and Pathophysiology The neurobiology of unipolar depression is poorly understood. Although evidence for genetic transmission is not as strong as in bipolar disorder, monozygotic twins have a higher concordance rate (46%) than dizygotic siblings (20%), with little evidence for any effect of a shared family environment. Parallels between the affective, motor, and cognitive dysfunctions seen in unipolar depression and those observed in diseases of the basal ganglia have suggested that neural networks involving prefrontal cortex and the basal ganglia may be involved. This hypothesis is supported by positron emission tomography (PET) studies of brain glucose metabolism that show a decrease in metabolic rate in the caudate nuclei and frontal lobes in depressed patients that returns to normal with recovery. Single-photon emission computed tomography (SPECT) studies show comparable changes in blood flow. Magnetic resonance imaging (MRI) findings in some patients include an increased frequency of subcortical white matter lesions. However, because these findings are more prevalent in patients with late onset of depressive illness, their significance remains unproven. A number of studies document increased ventricle-to-brain ratios in some patients with recurrent depression, but whether this finding is state-dependent or represents true cerebral atrophy is controversial. Postmortem examination of brains of suicide victims suggest altered noradrenergic activity, including increased binding toa 1-,a2-, andb-adrenergic receptors in the cerebral cortex and a decreased total number and density of noradrenergic neurons in the locus coeruleus. Involvement of the serotonin system is suggested by findings of reduced plasma tryptophan levels, a decreased cerebrospinal fluid level of 5-hydroxyindolacetic acid (the principal metabolite of serotonin in brain), and decreased platelet serotonergic transporter binding. An increase in brain serotonin receptors in suicide victims is also reported. Depletion of blood tryptophan, the amino acid precursor of serotonin, rapidly reverses the antidepressant benefit in depressed patients who have been successfully treated. However, a decrement in mood after tryptophan reduction is considerably less robust in untreated patients, indicating that, if presynaptic serotonergic dysfunction occurs in depression, it likely plays a contributing rather than a causal role. Neuroendocrine abnormalities that reflect the neurovegetative signs and symptoms of depression include (1) increased cortisol and corticotropin-releasing hormone (CRH) secretion, (2) an increase in adrenal size, (3) a decreased inhibitory response of glucocorticoids to dexamethasone, and (4) a blunted response of thyroid-stimulating hormone (TSH) level to infusion of thyroid-releasing hormone (TRH). Antidepressant treatment leads to normalization of these pituitary-adrenal abnormalities. Diurnal variations in symptom severity and alterations in circadian rhythmicity of a number of neurochemical and neurohumoral factors suggest that biologic differences may be secondary to a primary defect in regulation of biologic rhythms. Patients with major depression show consistent findings of a decrease in rapid eye movement (REM) sleep onset (REM latency), an increase in REM density, and, in some subjects, a decrease in stage IV delta slow-wave sleep.
Although antidepressant drugs result in a blockade of neurotransmitter uptake within hours, their therapeutic effects typically emerge over several weeks, implicating neuroadaptive changes in second messenger systems and transcription factors as possible mechanisms of action. TREATMENT Treatment planning requires coordination of short-term symptom remission with longer term maintenance strategies designed to prevent recurrence. The most effective intervention for achieving remission and preventing relapse is medication, but combined treatment, incorporating psychotherapy to help the patient cope with decreased self-esteem and demoralization, improves outcome (Fig. 385-1). About 40% of primary care patients with depression drop out of treatment and discontinue medication if symptomatic improvement is not noted within a month, unless additional support is provided. Outcome improves with (1) increased intensity and frequency of visits during the first 4 to 6 weeks of treatment, (2) supplemental educational materials, and (3) psychiatric consultation as indicated. Despite the widespread use ofSSRIs, there is no convincing evidence that this class of antidepressant is more efficacious thanTCAs. Between 60 and 70% of all depressed patients respond to any drug chosen, if it is given in a sufficient dose for 6 to 8 weeks. There is no ideal antidepressant; no current compound combines rapid onset of action, moderate half-life, a meaningful relationship between dose and blood level, a low side effect profile, minimal interaction with other drugs, and safety in overdose. A rational approach to selecting which antidepressant to use involves matching the patient's preference and medical history with the metabolic and side effect profile of the drug (Tables 385-4 and385-5). A previous response, or a family history of a positive response, to a specific antidepressant would suggest that that drug be tried first. Before initiating antidepressant therapy, the physician should evaluate the possible contribution of comorbid illnesses and consider their specific treatment. In individuals with suicidal ideation, particular attention should be paid to choosing a drug with a low toxicity if taken in overdose. The SSRIs and other newer antidepressant drugs are distinctly safer in this regard; nevertheless, the advantages of TCAs have not been completely superseded. The existence of generic equivalents make TCAs relatively cheap, and for several tricyclics, particularly nortriptyline, imipramine, and desipramine, well-defined relationships between dose, plasma level, and therapeutic response exist. The steady-state plasma level achieved for a given drug dose can vary more than tenfold between individuals. Plasma levels may help in understanding resistance to treatment and/or unexpected drug toxicity. The principal disadvantages of TCAs are antihistamine side effects (sedation) and anticholinergic side effects (constipation, dry mouth, urinary hesitancy, and blurred vision). Severe cardiac toxicity due to conduction block or arrhythmias can also occur but is uncommon at therapeutic levels. TCAs are probably contraindicated in patients with cardiovascular risk factors. Tricyclic agents are lethal in overdose, with desipramine carrying the greatest risk. Prescribing only a 10-day supply may be judicious. Most patients require a daily dose of 150 to 200 mg of imipramine or amitriptyline or its equivalent to achieve a therapeutic blood level of 150 to 300 ng/mL and a satisfactory remission; some patients show a partial effect at lower doses. Geriatric patients in particular may require a low starting dose and slow escalation. Ethnic differences in drug metabolism are significant; Hispanic, Asian, and African American patients generally require lower doses than
Caucasians to achieve a comparable blood level. Second-generation antidepressants include amoxapine, maprotiline, trazodone, and bupropion. Amoxapine is a dibenzoxazepine derivative that blocks norepinephrine and serotonin reuptake and has a metabolite that shows a degree of dopamine blockade. Long-term use of this drug carries a risk of tardive dyskinesia. Maprotiline is a potent noradrenergic reuptake blocker that has little anticholinergic effect but may produce seizures. Bupropion is a novel antidepressant whose mechanism of action is thought to involve enhancement of noradrenergic function. It has no anticholinergic, sedating, or orthostatic side effects and has a low incidence of sexual side effects. It may, however, be associated with aversive stimulant-like side effects, may lower seizure threshold, and has an exceptionally short half-life, requiring multiple dosing. An extended-release preparation is available. SSRIssuch as fluoxetine, sertraline, paroxetine, and citalopram cause a lower frequency of anticholinergic, sedating, and cardiovascular side effects but a possibly greater incidence of gastrointestinal complaints, sleep impairment, and sexual dysfunction than doTCAs. Akathisia, involving an inner sense of restlessness and anxiety, may also be more common, particularly during the first week of treatment. A serious concern, aside from drug interaction, is the risk of "serotonin syndrome," thought to result from hyperstimulation of brainstem 5HT1Areceptors and characterized by myoclonus, agitation, abdominal cramping, hyperpyrexia, hypertension, and potentially death. Combinations of serotonergic agonists should be monitored closely for this reason. Considerations such as half-life, compliance, toxicity, and drug-drug interactions may guide the choice of a particular SSRI. Fluoxetine and its principal active metabolite, norfluoxetine, for example, have a combined half-life of almost 7 days, resulting in a delay of 5 weeks before steady-state levels are achieved and a similar delay for complete drug excretion once its use is discontinued. All the SSRIs may impair sexual function, resulting in diminished libido, impotence, or difficulty in achieving orgasm. Sexual dysfunction frequently results in noncompliance and should be asked about specifically in patients using SSRIs. Sexual dysfunction can sometimes be ameliorated by lowering the dose, by instituting drug holidays over the weekend (two or three times a month), or by treatment with amantadine (100 mg tid), bethanechol (25 mg tid), or buspirone (10 mg tid). Paroxetine appears to be more anticholinergic than either fluoxetine or sertraline, and sertraline carries a lower risk of producing an adverse drug interaction than the other two. Rare side effects of SSRIs include vasospastic angina and alterations of prothrombin time. Citalopram is the most specific of currently available SSRIs and appears to have no specific inhibitory effects on the P450 system. Venlafaxine, like imipramine, blocks the reuptake of both norepinephrine and serotonin, but it produces relatively little in the way of traditional tricyclic side effects. Unlike theSSRIs, it has a relatively linear dose-response curve. Patients should be monitored for a possible increase in diastolic blood pressure, and multiple daily dosing is required because of the drug's short half-life. An extended-release form is available and has a somewhat lower incidence of gastrointestinal side effects. Nefazadone is a selective 5HT2receptor antagonist that also inhibits the presynaptic reuptake of serotonin and norepinephrine. Its side effects are similar to those of the SSRIs, and twice-daily dosing produces a steady state within 4 to 5 days. The drug is related structurally to trazodone, which is currently used more for its sedative than its antidepressant properties.
Nefazadone appears to produce a lower incidence of sexual side effects than do the SSRIs. Mirtazapine is a tetracyclic antidepressant that has a comparatively unique spectrum of activity. It increases noradrenergic and serotonergic neurotransmission through a blockade of centrala 2-adrenergic auto- and heteroreceptors and postsynaptic 5HT2 and 5HT3receptors. It is also strongly antihistaminic and, as such, may produce sedation at lower doses. With the exception of citalopram, each of theSSRIs, as well as nefazadone, may inhibit one or more cytochrome P450 enzymes (Table 385-5). Depending on the specific isoenzyme involved, the metabolism of a number of concomitantly administered medications can be dramatically affected. Fluoxetine and paroxetine, for example, by inhibiting 2D6, can cause dramatic increases in the blood level of type 1C antiarrhythmics, while sertraline and nefazadone, by acting on 3A4, may alter blood levels of terfenadine, carbamazepine, and astemizole. Because many of these compounds have a narrow therapeutic window and can cause iatrogenic ventricular arrhythmias at toxic levels, the possibility of an adverse drug interaction should be considered. Other treatment options include theMAOIsand electroconvulsive therapy. The MAOIs are highly effective, particularly in atypical depression, but the risk of hypertensive crisis following intake of tyramine-containing food or sympathomimetic drugs makes them inappropriate as first-line agents. Common side effects include orthostatic hypotension, weight gain, insomnia, and sexual dysfunction. MAOIs should not be used concomitantly withSSRIs, because of the risk of serotonin syndrome, or withTCAs, because of possible hyperadrenergic effects. Electroconvulsive therapy is at least as effective as medication, but its use is reserved for treatment-resistant cases and delusional depressions. Regardless of the medication chosen, the treatment response should be evaluated after approximately 2 months of therapy. Three-quarters of patients show an adequate response by this time, but if remission is inadequate, the patient should be questioned about medication compliance, and an increase in dose should be considered if side effects are not troublesome. If there is no improvement, consultation with or referral to a mental health specialist is advised. Strategies for treatment then include selection of an alternative drug, combinations of antidepressants, and/or adjunctive treatment with other classes of drugs, including lithium, thyroid hormone, and dopamine agonists. Patients whose response to anSSRIdisappears over time may benefit from the addition of buspirone (10 mg tid) or pindolol (2.5 mg tid) or small amounts of a tricyclic antidepressant such as desipramine (25 mg bid or tid). Once significant remission is achieved, drug treatment should be continued for at least 6 to 9 months to prevent relapse. In patients who have had two or more episodes of depression, indefinite maintenance treatment should be considered. It is essential to counsel patients about depression and the medications they are receiving. An educational approach is best, describing what is known about the depressive syndrome and how the medications may help. Advice about stress reduction, side effects, and expected length of treatment and cautions that alcohol may exacerbate depressive symptoms and impede drug response are helpful. Patients should be given time to describe their experience and the impact it has had on them,
their family, and their outlook. Occasional empathic silence may be as helpful for the treatment alliance as verbal reassurance. BIPOLAR DISORDER Clinical Manifestations Bipolar disorder is common, affecting approximately 3 million persons in the United States, but often difficult to diagnose. It is characterized by unpredictable swings in mood from mania (or hypomania) to depression. Some patients suffer only from recurrent attacks of mania, which in its pure form is associated with increased psychomotor activity; excessive social extroversion; decreased need for sleep; impulsivity and impairment in judgment; and expansive, grandiose, and sometimes irritable mood (Table 385-10). In severe mania, patients may experience delusions and paranoid thinking indistinguishable from schizophrenia. Half of patients with bipolar disorder present with a mixture of psychomotor agitation and activation with dysphoria, anxiety, and irritability. It may be difficult to distinguish mixed mania from agitated depression. In some bipolar patients (bipolar II disorder), the full criteria for mania are lacking, and the requisite recurrent depressions are separated by periods of mild activation and increased energy (hypomania). In cyclothymic disorder, there are numerous hypomanic periods, usually of relatively short duration, alternating with clusters of depressive symptoms that fail, either in severity or duration, to meet the criteria of major depression. The mood fluctuations are chronic and should be present for at least 2 years before the diagnosis is made. Manic episodes typically emerge over a period of days to weeks, but onset within hours is possible, usually in the early morning hours. An untreated episode of either depression or mania can be as short as several weeks or last as long as 8 to 12 months, and rare patients have an unremitting chronic course. The term rapid cycling is used for patients who have four or more episodes of either depression or mania in a given year. This pattern occurs in 15% of all patients, almost all of whom are women. In some cases, rapid cycling is linked to an underlying thyroid dysfunction and, in others, is iatrogenically triggered by prolonged antidepressant treatment. Although bipolar illness is associated with frequent episodic recurrence, it was once thought to have a favorable prognosis and outcome. More recent data, however, show that approximately half of patients with the disorder have sustained difficulties in work performance and psychosocial functioning. The most frequent age of onset for bipolar disorder is between 20 and 30 years of age, but many individuals report premorbid symptoms in late childhood or early adolescence. The prevalence is similar for men and women; women are likely to have more depressive and men more manic episodes over a lifetime. Differential Diagnosis The differential diagnosis of mania includes toxic effects of stimulant or sympathomimetic drugs as well as secondary mania induced by hyperthyroidism, AIDS, or neurologic disorders, such as Huntington's or Wilson's disease, or cerebrovascular accidents. Comorbidity with alcohol and substance abuse is common, either because of poor judgment and increased impulsivity or because of an attempt at self-medication. Etiology and Pathophysiology Evidence for a genetic predisposition to bipolar
disorder is significant. The concordance rate for monozygotic twin pairs approaches 80%, and segregation analyses are consistent with autosomal dominant transmission. Several chromosomal locations for the gene have been proposed in the past decade on the basis of linkage analysis in affected families. None, however, has yet received convincing confirmation. The pathophysiologic mechanisms underlying the profound and recurrent mood swings of bipolar disorder remain unknown. Cellular models of changes in membrane Na+ - and K+ -activated ATPase and proposals of disordered signal transduction mechanisms involving the phosphoinositol system and GTP-binding proteins have received the most attention. Alterations in glutamate regulation and in neuroprotective transcription factors are also being investigated as possible explanations for the therapeutic effects of lithium. Neurophysiologic studies suggest that patients with bipolar disorder have altered circadian rhythmicity. Lithium may exert its therapeutic benefit through a resynchronization of intrinsic rhythms keyed to the light/dark cycle (Chap. 27). Neuroimaging techniques have also identified a higher rate of subcortical white matter abnormalities in patients than in age-matched controls. TREATMENT (Table 385-11) Lithium carbonate is the mainstay of treatment in bipolar disorder, although sodium valproate is equally effective in acute mania. Carbamazepine is also efficacious. The response rate to lithium carbonate is 70 to 80% in acute mania, with beneficial effects appearing in 1 to 2 weeks. Lithium also has a prophylactic effect in prevention of recurrent mania, and, to a lesser extent, in the prevention of recurrent depression. A simple cation, lithium is rapidly absorbed from the gastrointestinal tract and remains unbound to plasma or tissue proteins. Some 95% of a given dose is excreted unchanged through the kidneys within 24 h. Serious side effects from lithium administration are rare, but minor complaints such as gastrointestinal discomfort, nausea, diarrhea, polyuria, weight gain, skin eruptions, alopecia, and edema are common. Over time, urine-concentrating ability may be decreased, but significant nephrotoxicity does not occur. In a small subset of patients in whom excessive polyuria occurs (>3000 mL/24 h), dose or schedule adjustments or the adjunctive use of diuretics should be considered. Lithium exerts an antithyroid effect by interfering with the synthesis and release of thyroid hormones. Approximately 5% of patients taking lithium for ³18 months develop hypothyroidism, with women more likely to be affected than men. Iatrogenic hypothyroidism should be ruled out in any patient who experiences a recurrence of depressive symptomatology during lithium treatment. More serious side effects include tremor, interference with concentration and memory, ataxia, dysarthria, and incoordination.ECGchanges of T wave flattening and conduction delays may occur. There is suggestive, but not conclusive, evidence that lithium is teratogenic, inducing cardiac malformations in the first trimester. In the treatment of acute mania, lithium is initiated at 300 mg bid or tid, and the dose is then increased by 300 mg every 2 to 3 days to achieve blood levels of 0.8 to 1.2 meq/L. Before initiating treatment the physician should obtain baseline measures of
electrolytes, creatinine, thyroid function, and a complete blood count (CBC). Because the therapeutic effect of lithium may not appear until 7 to 10 days of treatment, adjunctive usage of lorazepam (1 to 2 mg every 4 h) or clonazepam (0.5 to 1 mg every 4 h) may be beneficial to control agitation. Antipsychotics are indicated in patients with severe agitation who respond only partially to benzodiazepines. These agents should be discontinued in the transition to maintenance lithium therapy. Patients using lithium should be monitored closely, since the blood levels required to achieve a therapeutic benefit are close to those associated with neurotoxicity. Risk factors for neurotoxicity include concomitant medical illness, decrease in salt intake, or concurrent use of medications that may increase the serum level of lithium (neuroleptics, diuretics, and calcium channel blockers). Once stabilization is achieved, the lithium level can be monitored on a bimonthly basis, and thyroid and renal functions on a biannual basis, or more frequently if clinical change occurs. Valproic acid is an alternative in patients who cannot tolerate lithium or respond poorly to it. Valproic acid may be better than lithium for patients who have a rapid-cycling course (i.e., more than four episodes a year) or who present with a mixed or dysphoric mania. Valproic acid is usually started at 500 to 750 mg/d in divided doses. The dose is increased every several days to achieve blood levels in the range of 50 to 100 ug/mL, which typically are achieved at a dose of 1000 to 2500 mg/d. The most serious adverse effects of valproic acid are hepatoxicity, which may be fatal, and hyponatremia. Such cases are fortunately rare, but periodic monitoring of liver enzymes, particularly during the first 90 days of treatment, is indicated. Carbamazepine, although not formally approved by the U.S. Food and Drug Administration (FDA) for bipolar disorder, has clinical efficacy in the treatment of acute mania. Carbamazepine is initiated at 400 to 600 mg/d in divided doses, and the dose is increased to achieve a blood level of 4 to 12 mg/L. Carbamazepine may induce a benign leukopenia, but the risk of aplastic anemia is minimal. Nevertheless, it is wise to obtain aCBCperiodically. Preliminary evidence also suggests that other anticonvulsant agents such as gabapentin, lamotrigine, and topiramate may possess some therapeutic benefit. The recurrent nature of bipolar mood disorder necessitates maintenance treatment. Maintenance of blood lithium levels of at least 0.8 mg/L is important to achieve optimal prophylaxis. Compliance is frequently an issue and often requires enlistment and education of concerned family members to avoid relapse. Efforts to identify and limit psychosocial factors that may trigger episodes are important, as is an emphasis on life-style regularity. Antidepressant medications are sometimes required for the treatment of severe breakthrough depressions, but their use should generally be avoided during maintenance treatment because of the risk of precipitating mania or accelerating the cycle frequency. Loss of efficacy over time may be observed with any of the mood-stabilizing agents. In such situations, an alternative agent or combination therapy is usually helpful. Consensus guidelines for the treatment of acute mania and bipolar depression are described inTable 385-12.
SOMATOFORM DISORDERS CLINICAL MANIFESTATIONS Patients with multiple somatic complaints that cannot be explained by a known medical condition or by the effects of alcohol or of recreational or prescription drugs are seen commonly in primary care practice; one survey indicates a prevalence of 5%. The somatoform disorders include a variety of conditions that differ in terms of the specific symptoms that are present and in whether or not the symptoms are intentionally produced. In somatization disorder, the patient presents with multiple physical complaints referable to different organ systems (Table 385-13). Onset is usually before age 30, and the disorder is persistent. Formal diagnostic criteria require the recording of at least four pain, two gastrointestinal, one sexual, and one pseudoneurologic symptom. Patients with somatization disorder often present with dramatic complaints, but the complaints are inconsistent. Symptoms of comorbid anxiety and mood disorder are common and may be the result of drug interactions due to regimens initiated independently by different physicians. Patients with somatization disorder may be impulsive and demanding and frequently qualify for a formal comorbid psychiatric diagnosis. In conversion disorder, the symptoms focus on deficits that involve voluntary motor or sensory function and on psychological factors that initiate or exacerbate the medical presentation. Like somatization disorder, the deficit is not intentionally produced or simulated, as is the case in factitious disorder (malingering). In hypochondriasis, the essential feature is a belief of serious medical illness that persists despite reassurance and appropriate medical evaluation. As with somatization disorder, patients with hypochondriasis have a history of poor relationships with physicians stemming from their sense that they have been evaluated and treated inappropriately or inadequately. Hypochondriasis can be disabling in intensity and is persistent, with waxing and waning symptomatology. In factitious illnesses, the patient consciously and voluntarily produces physical symptoms of illness. The term Munchausen's syndrome is reserved for individuals with particularly dramatic, chronic, or severe factitious illness. In true factitious illness, the sick role itself is gratifying. A variety of signs, symptoms, and diseases have been either simulated or caused by factitious behavior, the most common including chronic diarrhea, fever of unknown origin, intestinal bleeding or hematuria, seizures, and hypoglycemia. Factitious disorder is usually not diagnosed until 5 to 10 years after its onset, and it can produce significant social and medical costs. In malingering, the fabrication derives from a desire for some external reward, such as a narcotic medication or disability reimbursement. TREATMENT Patients with somatization disorders are frequently subjected to multiple diagnostic testing and exploratory surgeries in an attempt to find their "real" illness. Such an approach is doomed to failure and does not address the core issue. Successful treatment is best achieved through behavior modification, in which access to the physician is tightly regulated and adjusted to provide a sustained and predictable level of support that is less clearly contingent on the patient's level of presenting distress. Visits can be brief and should not be associated with a need for a diagnostic or
treatment action. Although the literature is limited, some patients with somatization disorder may benefit from antidepressant treatment. Fluoxetine andMAOIshave both been found to be useful in reducing obsessive ruminations, dysphoria, and anxious preoccupation in patients with multiple somatic complaints. The treatment of factitious disorder is complicated in that any attempt to confront the patient usually only creates a sense of humiliation and causes the patient to abandon treatment from that caregiver. A better strategy is to introduce psychological causation as one of a number of possible explanations and to include factitious illness as an option in the differential diagnoses that are discussed. Without directly linking psychotherapeutic intervention to the diagnosis, the patient can be offered a face-saving means by which the pathologic relationship with the health care system can be examined and alternative approaches to life stressors developed. PERSONALITY DISORDERS CLINICAL MANIFESTATIONS Personality disorders are characteristic patterns of thinking, feeling, and interpersonal behavior that are relatively inflexible and cause significant functional impairment or subjective distress for the individual. The observed behaviors are not secondary to another mental disorder, nor are they precipitated by substance abuse or a general medical condition. This distinction is often difficult to make in clinical practice, as personality change may be the first sign of serious neurologic, endocrine, or other medical illness. Patients with frontal lobe tumors, for example, can present with changes in motivation and personality while the results of the neurologic examination remain within normal limits. Personality traits are stable over time and environmental situation and are recognizable in adolescence or early adult life. AlthoughDSM-IVportrays personality disorders as qualitatively distinct categories, there is an alternative perspective that personality characteristics vary as a continuum between normal functioning and formal mental disorder. Personality disorders have been grouped into three clusters that share similar attributes. Cluster A includes paranoid, schizoid, and schizotypal personality disorders. It includes individuals who are odd and eccentric and who maintain an emotional distance from others. Individuals have a restricted emotional range and remain socially isolated. Patients with schizotypal personality disorder frequently have unusual perceptual experiences and express magical beliefs about the external world. The essential feature of paranoid personality disorder is a pervasive mistrust and suspiciousness of others to an extent that is unjustified by available evidence. Cluster B disorders include antisocial, borderline, histrionic, and narcissistic types and describe individuals whose behavior is impulsive, excessively emotional, and erratic. Cluster C incorporates avoidant, dependent, and obsessive-compulsive personality types; enduring traits are anxiety and fear. The boundaries between cluster types are to some extent artificial, and many patients who meet criteria for one personality disorder also meet criteria for aspects of another. The risk of a comorbid major mental disorder is increased in patients who qualify for a diagnosis of personality disorder. TREATMENT
Historically, recommended treatment for personality disorders was long-term psychotherapy, in which the pathologic patterns of interaction with the world at large could be relived and examined through the corrective emotional experience of the controlled therapeutic relationship. More recently, the recognition that personality derives in part from biologically determined components of temperament has given rise to the empirical use of drugs to treat specific symptom clusters as well as any coexisting major mental disorder. Antidepressant medications and low-dose antipsychotic drugs have some efficacy in cluster A personality disorders, while anticonvulsant mood-stabilizing agents andMAOIsmay be considered for patients with cluster B diagnoses who show marked mood reactivity, behavioral dyscontrol, and/or rejection hypersensitivity. Anxious or fearful cluster C patients often have a response to medication that parallels that for patients with axis I anxiety disorders. In all cases, it is important for both the physician and the patient to have reasonable expectations as to the possible effect of the medication and any associated side effects. Beneficial responses may be subtle and observable only over time. SCHIZOPHRENIA CLINICAL MANIFESTATIONS Schizophrenia is a heterogeneous syndrome characterized by perturbations of language, perception, thinking, social activity, affect, and volition. There are no pathognomonic features. The syndrome commonly begins in late adolescence, has an insidious onset, and, classically, a poor outcome, progressing from social withdrawal and perceptual distortions to a state of chronic delusions and hallucinations. Patients may present with positive symptoms (such as conceptual disorganization, delusions, or hallucinations) or negative symptoms (loss of function, anhedonia, decreased emotional expression, impaired concentration, and diminished social engagement) and must have at least two of these for a 1-month period and continuous signs for at least 6 months to meet formal diagnostic criteria. "Negative" symptoms predominate in one-third of the schizophrenic population and are associated with a poor long-term outcome and a poor response to drug treatment. However, marked variability in the course and individual character of symptoms is typical. Schizophrenia can be classified according to the specific symptomatology present, although such distinctions do not correlate well with either course of illness or response to treatment, and many individuals have symptoms of more than one type. The four main symptom subtypes are catatonic, paranoid, disorganized, and residual. Catatonic-type describes patients whose clinical presentation is dominated by profound changes in motor activity, negativism, and echolalia or echopraxia. Paranoid-type describes patients who have a prominent preoccupation with a specific delusional system and who otherwise do not qualify as having disorganized-type disease, in which disorganized speech and behavior are accompanied by a superficial or silly affect. In residual-type disease, negative symptomatology exists in the absence of delusions, hallucinations, or motor disturbance. The diagnosis of schizophreniform disorder is reserved for patients who meet the symptom requirements but not the duration requirements for schizophrenia, and that of schizoaffective disorder is used for those whose symptoms of schizophrenia are independent of associated periods of mood
disturbance. Prognosis depends not on symptom severity but on the response to antipsychotic medication. Patients may present with acute rather than insidious onset of symptoms, and remission without recurrence does occur. About 10% of schizophrenic patients commit suicide. As currently defined, schizophrenia is present in 0.85% of individuals worldwide. Overall, lifetime prevalence is approximately 1 to 1.5%. The societal costs of schizophrenia are substantial. An estimated 300,000 episodes of acute schizophrenia occur annually, resulting in direct and indirect costs that have been estimated at >$33 billion. DIFFERENTIAL DIAGNOSIS For a diagnosis of schizophrenia to be made, the symptom complex must cause significant dysfunction in social or occupational domains and last for at least 6 months. The diagnosis is principally one of exclusion, requiring the absence of significant associated mood symptoms, any relevant medical condition, and substance abuse. Drug reactions that cause hallucinations, paranoia, confusion, or bizarre behavior may be dose-related or idiosyncratic; b-adrenergic blockers, clonidine, cycloserine, quinacrine, and procaine derivatives are most commonly associated with these symptoms. Drug causes should be ruled out in any case of newly emergent psychosis. The general neurologic examination in patients with schizophrenia is usually normal, but motor rigidity, tremor, and dyskinesias are noted in one-quarter of untreated patients. EPIDEMIOLOGY AND PATHOPHYSIOLOGY Epidemiologic surveys identify three principal risk factors for schizophrenia: (1) genetic susceptibility, (2) early developmental insults, and (3) winter birth. Family, twin, and adoption studies show that genetic factors are involved in at least a subset of individuals who develop schizophrenia. Using conservative diagnostic definitions, schizophrenia is observed in approximately 6.6% of all first-degree relatives of an affected proband. If both parents are affected, the risk for offspring is 40%. The concordance rate for monozygotic twins is 50%, compared to 10% for dizygotic twins. Examination of families in which aggregation of schizophrenia occurs has revealed an increased incidence of other psychotic and nonpsychotic psychiatric disorders as well, including schizoaffective disorder and schizotypal and schizoid personality disorders, the latter terms designating individuals who show a lifetime pattern of social and interpersonal deficits characterized by an inability to form close interpersonal relationships, eccentric behavior, and mild perceptual distortions. Some relatives and individuals with schizophrenia have been found to have distinctive patterns in expressing emotion, most often involving increased criticism, hostility, and emotional overinvolvement. There is evidence that environmental influences modulate genetic factors in the expression of schizophrenia, and, in sporadic cases, may serve as a sufficient cause. Gestational and birth complications, including Rh factor incompatibility, prenatal exposure to influenza during the second trimester, and prenatal nutritional deficiency have been implicated. Studies of monozygotic twins discordant for schizophrenia have reported neuroanatomic differences between affected and unaffected siblings, supporting a "two-strike" etiology involving both genetic susceptibility and an environmental insult. The latter might involve localized hypoxia during critical stages of
brain development. Neuroimaging and postmortem studies have identified a number of structural and functional abnormalities, including (1) enlargement of the lateral and third ventricles with associated cortical atrophy and sulcal enlargement; (2) volumetric reductions in the amygdala, hippocampus, right prefrontal cortex, and thalamus; (3) altered asymmetry of the planum temporale; and (4) decreases in neuronal metabolism in the thalamus and prefrontal cortex. Some, but not all, prospective studies record progressive reduction in hemispheric volume over years. Neuropathologic studies have reported changes in the size, orientation, and density of cells in the hippocampus and, in the prefrontal cerebral cortex, decreases in neuronal number and the density of interneurons in layer II as well as an increased density of pyramidal cells in layer V. These observations suggest that schizophrenia results from a disturbance in a cortical striatal-thalamic circuit resulting in deficits in sensory filtering and attentional behavior. Although the formal diagnostic requirements for schizophrenia are not usually met until early adult life, children who eventually develop the disorder may exhibit subtle deficits in motor function, cognition, and emotional expression from an early age. The hypothesized alterations in cortical neuronal circuitry are paralleled clinically by impairments in attention and cortical information processing, autonomic nervous system activation, and habituation. Schizophrenic individuals are highly distractible and demonstrate deficits in perceptual-motor speed, ability to shift attention, and filtering out of background stimuli. Event-related evoked potential studies of schizophrenia have defined a specific reduction in P300 amplitude to a novel stimulus, which implicates an impairment in cognitive processing. Impaired information processing is found in unaffected family members. Despite evidence for a genetic causation, the results of molecular genetic linkage studies in schizophrenia are inconclusive. Reports of linkage of schizophrenia to loci on chromosomes 1, 5, 6, 8, 11, and 22 and other regions have not been formally replicated and have led to larger scale association studies currently underway. The dopamine hypothesis of schizophrenia is based on the serendipitous discovery that agents that diminish dopaminergic activity have beneficial effects in reducing the acute symptoms and signs of psychosis, specifically agitation, anxiety, and hallucinations. Amelioration of delusions and social withdrawal is less dramatic. Thus far, however, evidence for increased dopaminergic activity is indirect. An increase in the activity of nigrostriatal and mesolimbic systems and a decrease in mesocortical tracts innervating the prefrontal cortex is hypothesized, although it is likely that other neurotransmitters, including serotonin, acetylcholine, glutamate, andGABAalso contribute to the pathophysiology of the illness. Involvement of excitatory amino acids is postulated, based on the finding that NMDA receptor antagonists and channel blockers, such as phencyclidine (PCP) and ketamine, produce characteristic signs of schizophrenia in normal individuals. TREATMENT Antipsychotic agents (Table 385-14) remain the cornerstone of acute and maintenance treatment of schizophrenia and are effective in the treatment of hallucinations,
delusions, and thought disorders, regardless of etiology. The exact mechanism of action remains incompletely understood, but dopaminergic receptor blockade in the limbic system and basal ganglia appears to be an essential element, since the clinical potencies of traditional antipsychotic drugs parallel their affinities for the D2receptor, and even the newer "atypical" agents exert some degree of D2receptor blockade. All neuroleptics induce expression of the immediate-early gene c-fos in the nucleus accumbens, a dopaminergic site connecting prefrontal and limbic cortices. The clinical efficacy of newer atypical neuroleptics, however, may involve D1, D3, and D4receptor blockade,a1- anda2-noradrenergic activity, and/or altering the relationship between 5HT2 and D2receptor activity. Conventional neuroleptics differ in their potency and side-effect profile. Older agents, such as chlorpromazine and thioridazine, are more sedating and anticholinergic and more likely to cause orthostatic hypotension, while higher potency antipsychotics, such as haloperidol, perphenazine, and thiothixene, carry a higher risk of inducing extrapyramidal side effects. The model atypical antipsychotic agent is clozapine, a dibenzodiazepine that has a greater potency in blocking the 5HT2than the D2receptor and a much higher affinity for the D4 than the D2receptor. Its principal disadvantage is risk of blood dyscrasia, requiring regular monitoring of theCBC. Unlike other antipsychotics, clozapine does not cause a rise in prolactin level. Approximately 30% of patients have a better antipsychotic response to these agents than to traditional neuroleptics, suggesting that they will increasingly displace the older-generation drugs. Clozapine appears to be the most effective member of this class; however, its side-effect profile makes it most appropriate for treatment-resistant cases. Clozapine increases the activity of the immediate-early gene c-fos in the prefrontal cortex, the neuroanatomic region having the highest concentration of D4receptors and an area thought to mediate the specific executive functions that are prominently impaired in schizophrenia. Risperidone, a benzisoxazole derivative, is more potent at 5HT2than D2receptor sites, like clozapine, but it also exerts significanta 2antagonism, a property that may contribute to its perceived ability to improve mood and increase motor activity. Risperidone is not as effective as clozapine in treatment-resistant cases but does not carry a risk of blood dyscrasia. Olanzapine is more similar neurochemically to clozapine but has a significant risk of inducing weight gain. Quetiapine is distinct in having a weak D2effect but potenta1 and histamine blockade. Conventional antipsychotic agents are effective in ~70% of patients presenting with a first episode. Improvement may be observed within hours or days, but full remission usually requires 6 to 8 weeks. The choice of agent depends principally on the side-effect profile and cost of treatment or on a past personal or family history of a favorable response to the drug in question. Atypical agents appear to be more effective in treating negative symptoms and improving cognitive function. Equivalent treatment response can usually be achieved with relatively low doses of any drug selected, i.e., 4 to 6 mg/d of haloperidol, 10 to 15 mg of olanzapine, or 4 to 6 mg/d of risperidone. Doses in this range result in>80% D2receptor blockade, and there is little evidence that higher doses increase either the rapidity or degree of response. Maintenance treatment requires careful attention to the possibility of relapse and monitoring for the development of a movement disorder. Intermittent drug treatment is less effective than regular dosing, but gradual dose reduction is likely to improve social functioning in many schizophrenic patients who have been maintained at high doses. If medications are completely
discontinued, however, the relapse rate is ~60% within 6 months. Long-acting injectable preparations (haloperidol decanoate and fluphenazine decanoate) are considered when noncompliance with oral therapy leads to relapses. In treatment-resistant patients, a transition to clozapine usually results in rapid improvement, but a prolonged delay in response in some cases necessitates a 6- to 9-month trial for maximal benefit to occur. Antipsychotic medications can cause a broad range of side effects, including lethargy, weight gain, postural hypotension, constipation, and dry mouth. Extrapyramidal symptoms such as dystonia, akathisia, and akinesia are also frequent with traditional agents and may contribute to poor compliance if not specifically addressed. Anticholinergic and parkinsonian symptoms respond well to trihexyphenidyl, 2 mg bid, or benztropine mesylate, 1 to 2 mg bid. Akathisia may respond to beta blockers. In rare cases, more serious and occasionally life-threatening side effects may emerge, including ventricular arrhythmias, gastrointestinal obstruction, retinal pigmentation, obstructive jaundice, and neuroleptic malignant syndrome (characterized by hyperthermia, autonomic dysfunction, muscular rigidity, and elevated creatine phosphokinase levels). The most serious adverse effects of clozapine are agranulocytosis, which has an incidence of 1%, and induction of seizures, which has an incidence of 10%. Weekly white blood cell counts are required, particularly during the first 3 months of treatment. A serious side effect of long-term use of the classic antipsychotic agents is tardive dyskinesia, characterized by repetitive, involuntary, and potentially irreversible movements of the tongue and lips (bucco-linguo-masticatory triad), and, in approximately half of cases, choreoathetoid movements of the limbs (Chap. 22). Tardive dyskinesia has an incidence of ~4% per year of exposure, and a maximal prevalence of ~20% in chronic patients treated with high-dose neuroleptics. The risk associated with the newer atypical agents is unknown but expected to be much less. The prevalence increases with age and with total dose and duration of drug administration, but unknown individual factors play the greatest part in determining risk. The cause of tardive dyskinesia is unknown, but evidence suggests that chronic neuroleptic treatment increases the formation of free radicals and perhaps damages mitochondrial energy metabolism. Vitamin E may reduce abnormal involuntary movements if given early in the syndrome. Drug treatment of schizophrenia is by itself insufficient. Psychoeducational efforts directed towards families and relevant community resources have proven to be necessary to maintain stability and optimize prognosis. A treatment model involving a multidisciplinary case-management team that seeks out and closely follows the patient in the community has proven particularly effective, not only in maintaining pharmacologic adherence but also in facilitating occupational achievement and interactions with welfare, legal, and primary medical care systems. ASSESSMENT AND EVALUATION OF VIOLENCE Primary care physicians may encounter situations in which familial, domestic, or societal violence is discovered or suspected. Such an awareness can carry legal and moral obligations; many state laws mandate reporting of child, spousal, and elder abuse. Physicians are frequently the first point of contact for both victim and abuser. Between 1
and 2 million older Americans and 1.5 million U.S. children are thought to experience some form of physical maltreatment each year. Spousal abuse is thought to be even more prevalent. A recent survey of internal medicine practices found that 5.5% of all female patients had experienced domestic violence in the previous year, and that these individuals were more likely to suffer from depression, anxiety, somatization disorder, and substance abuse and to have attempted suicide. When domestic violence is suspected, direct but nonjudgmental questioning should be pursued with each party separately -- "Do you feel safe at home?" and "If there's a disagreement or a conflict between the two of you, how is it worked out?" In addition to obvious and suggestive physical injury, individuals who are abused frequently express low self-esteem, vague somatic symptomatology, social isolation, and a passive feeling of loss of control. Although it is essential to treat these elements in the victim, the first obligation is to ensure that the perpetrator has taken responsibility for preventing any further violence. Substance abuse and/or dependence and serious mental illness in the abuser may contribute to the risk of harm and require direct intervention. Depending on the situation, law enforcement agencies, community resources such as support groups and shelters, and individual and family counseling can be appropriate components of a treatment plan. A safety plan should be formulated with the victim, in addition to the provision of information about abuse, its likelihood of recurrence, and its tendency to increase in severity and frequency. Antianxiety and antidepressant medications may sometimes be useful in treating the acute symptoms, but only if independent evidence for an appropriate psychiatric diagnosis exists. Antidepressants are generally not indicated when the diagnosis is linked to the social situation, such as an adjustment disorder with depressed mood. The most important element in treatment is the development of a supportive doctor-patient relationship that avoids further blame of the victim. In certain circumstances, a significant potential for societal violence may be discovered. Sympathetic, but direct, questioning about potential violent impulses, access to weapons, recreational drug use, and specific homicidal ideation is necessary and is sometimes therapeutic in its own right. The existence and possible contribution of such medical conditions as delirium and/or intoxication should be evaluated. Available disposition options for potentially violent patients include police custody, psychiatric hospitalization, and referral to home care, with involvement of family, friends, and caregivers. In deciding which treatment option is most appropriate, clinicians should endeavor to establish an empathic interaction with the patient, while avoiding interventions or stimuli that might precipitate or increase the risk of violent behavior. Formal verbal limit setting may be necessary if the patient reveals the existence of a weapon or becomes increasingly agitated or verbally abusive. Use of the least restrictive intervention is generally the best approach during the initial evaluation. MENTAL HEALTH PROBLEMS IN THE HOMELESS There is a high prevalence of mental disorders and substance abuse among homeless and impoverished people. The total number of homeless individuals in the United States is estimated at 2 to 3 million, one-third of whom qualify as having a serious mental disorder. Poor hygiene and nutrition, substance abuse, psychiatric illness, physical trauma, and exposure to the elements combine to make the provision of medical care a challenging enterprise. Only a minority of individuals receive formal mental health care; the main points of contact are outpatient medical clinics and emergency departments.
Primary care settings represent a critical site in which housing needs, treatment of substance dependence, and evaluation and treatment of psychiatric illness can most efficiently take place. Successful intervention is dependent on breaking down traditional administrative barriers to health care and recognizing the physical constraints and emotional costs imposed by homelessness. Simplifying health care instructions and follow-up, allowing frequent visits, and dispensing medications in limited amounts that require ongoing contact are possible techniques for establishing a successful therapeutic relationship. Child neglect, resulting in developmental delay and emotional difficulty in addition to other health problems, is unfortunately common and necessitates an effort to evaluate the well being of any offspring independently. (Bibliography omitted in Palm version) Back to Table of Contents
SECTION 6 -ALCOHOLISM AND DRUG DEPENDENCY 386. BIOLOGY OF ADDICTION - Robert O. Messing Drug addiction is a chronic, relapsing disorder characterized by compulsion to take a drug and loss of self-control in limiting drug intake. The American Psychiatric Association (DSM-IV) uses the term substance dependence instead of drug addiction and requires at least three of the following symptoms to be present for diagnosis: (1) tolerance; (2) withdrawal; (3) persistent desire or unsuccessful attempts to reduce use; (4) use in larger amounts than intended; (5) reduction in important social, occupational, or recreational activities because of drug use; (6) considerable time spent obtaining the substance; and (7) continued use despite health, social, or economic problems resulting from substance use. Substance abuse is a milder disorder characterized by repetitive drug use that results in social or economic distress. Experimental studies in humans and in animal models (rodents and also simpler organisms such as flies and worms) have begun to elucidate the cellular and molecular mechanisms that mediate the loss of control in drug taking that is the hallmark of addiction. This chapter will review current understanding of the neurobiology of drug abuse relevant to the specific substances discussed in subsequent chapters, namely alcohol (Chap. 387), opioids (Chap. 388), cocaine and marijuana (Chap. 389), and nicotine (Chap. 390). BEHAVIORAL RESPONSES TO DRUGS OF ABUSE Drugs of abuse produce euphoria, which is an emotional state characterized by intensely pleasant feelings. A major reason why users are motivated to seek and take more of a drug is because they perceive the experience as rewarding. Reinforcement refers to the ability of a drug to produce a pleasurable response that motivates the user to take the drug repeatedly. The powerful reinforcing and rewarding properties of abusable drugs can be measured by the tremendous effort experimental animals will expend, e.g., by pressing a lever multiple times, to obtain an oral or intravenous dose of a drug. Tolerance is a reduction in response to a drug after repeated use and is a normal, adaptive, physiologic response. Pharmacokinetic tolerance may arise through an increase in the rate of metabolism. For example, barbiturates induce hepatic microsomal enzymes resulting in more rapid metabolism. Pharmacodynamic tolerance results from drug-induced changes in cell signaling and gene expression. Behavioral sensitization is a process whereby repeated administration of a drug leads to a progressively stronger behavioral response. Sometimes called "reverse tolerance," it is often measured by examining drug-induced locomotor activation. It generally requires longer intervals between doses to develop than does tolerance. Both tolerance and sensitization can promote repeated drug use. Tolerance develops to the rewarding properties of most abusable drugs, requiring the user to employ higher doses to achieve a euphoric effect. Sensitization also promotes drug self-administration, since rodents will expend greater effort in lever-pressing for drugs to which they are sensitized. Physical dependence is an adaptive state that develops through resetting of homeostatic mechanisms to permit normal function despite the continued presence of a drug. When drug intake is abruptly terminated in a physically dependent individual, a
withdrawal syndrome emerges. The symptoms of withdrawal tend to be opposite to those seen during acute drug exposure. Thus, abstinence from alcohol and other sedative-hypnotics causes nervous system hyperactivity, whereas withdrawal from cocaine and other stimulants is characterized by fatigue, sedation, and depression. Withdrawal symptoms are the principal evidence for physical dependence. Like tolerance, physical dependence is a normal physiologic response to repeated drug exposure and does not necessarily indicate drug abuse or addiction. However, withdrawal can cause intensely negative, unpleasant emotions such as dysphoria, anxiety, and irritability. In animal studies employing intracranial self-stimulation, withdrawal is also associated with reduced brain reward function. Thus, it appears that drug withdrawal can act as a negative reinforcer that contributes to repeated drug use. Human patients prescribed opioids for treatment of pain may develop tolerance and physical dependence but rarely become addicted. Likewise, in experimental animals, establishing physical dependence is not sufficient to induce voluntary drug self-administration. Instead, it appears that animals and humans seek abusable drugs mainly for their positive reinforcing properties. In susceptible persons, repeated use of abusable drugs induces drug craving, a powerful motivational state in which the addict seeks the drug to the exclusion of other activities. Craving is a manifestation of psychological dependence on a drug and is most severe during acute abstinence. It is a long-lasting, conditioned response that may be evoked by environmental cues such as sights, smells, or situations associated with previous drug use, even after long periods of drug abstinence. Understanding the mechanisms that underlie susceptibility to drug craving is a critical task in addiction research. GENETIC FACTORS IN ADDICTION Genetic factors have been studied most extensively in alcoholism (Chap. 387). Patterns of inheritance in humans are most consistent with alcoholism being a polygenic disorder. Some genes confer a reduced risk for alcoholism. Approximately half of all individuals in Asian populations carry an allele of aldehyde dehydrogenase that encodes an isozyme with reduced enzymatic activity. After ingesting alcohol, they have increased blood levels of acetaldehyde and experience vasodilatation, tachycardia, hot sensations, and hypotension. They also report feeling intoxicated at very low doses of alcohol. Individuals expressing this isoenzyme rarely abuse alcohol. Other genetic factors predispose individuals to increased risk for alcoholism. A recent multicenter study of 105 families of alcoholics revealed evidence for susceptibility loci for alcohol dependence on chromosomes 1, 7, and possibly 2. An additional study of a Southwestern Native American population found evidence for genetic linkage to alcohol dependence on chromosomes 4 and 11. The identity of the genes associated with increased risk is not yet known. NEUROANATOMY OF DRUG REWARD Early studies of intracranial electrical self-stimulation in rodents identified key structures involved in drug reward and motivational aspects of drug dependence (Fig. 386-1). These include the midbrain ventral tegmental area (VTA), median forebrain bundle (MFB), nucleus accumbens, medial frontal cortex, amygdala, and lateral hypothalamus. Many of these structures are key components of a mesocorticolimbic dopamine system
that is now a principal focus of addiction research. Dopaminergic neurons in the VTA project axons via the MFB to the nucleus accumbens, amygdala, and frontal cortex. Reciprocal projections from g-aminobutyric acid (GABA)-containing neurons in the nucleus accumbens project back through the MFB onto VTA neurons. Other brain regions modulate this system through opioid peptide, GABA, serotonin, and glutamate inputs that interact with the VTA, nucleus accumbens, and other structures within the system. Certain nuclei within the mesocorticolimbic dopamine system share similarities in architecture, receptor expression, and connectivity with other brain regions. These related structures reside in the basal forebrain and include the central medial amygdala, bed nucleus of the stria terminalis, and shell of the nucleus accumbens. They appear to constitute a functional entity that has been called the "extended amygdala." The extended amygdala receives inputs from the hippocampus, basolateral amygdala, midbrain, and lateral hypothalamus and sends efferents to the ventral pallidum,VTA, and lateral hypothalamus. MOLECULAR TARGETS OF ABUSED DRUGS Dopamine released from presynaptic terminals ofVTAneurons in the nucleus accumbens is a major mediator of drug reward and reinforcement (Fig. 386-2A). Acute administration of all abusable drugs increases extracellular levels of dopamine in the shell of the nucleus accumbens. In addition, dopamine receptor antagonists injected into this region reduce drug self-administration in animals. Dopamine binds to a family of five G protein-coupled, seven-transmembrane receptors that can be grouped into two major classes, D1-like (D1 and D5receptors) and D2-like (D2, D3, and D4receptors). D1-like receptors activate adenylyl cyclase by coupling to the stimulatory G protein G s, whereas D2-like receptors inhibit adenylyl cyclase by coupling to inhibitory Giproteins. Despite opposing actions on adenylyl cyclase, both classes of dopamine receptors appear to mediate drug reinforcement. Experimental animals will self-administer D1-like and D2-like receptor agonists, and antagonists of D1, D2, and D3receptors decrease the reinforcing properties of cocaine. These receptor-specific responses most likely result from dopamine actions on different subpopulations of cells in the nucleus accumbens. Opioids, nicotine, psychostimulants, barbiturates, benzodiazepines, and cannabinoids elicit their acute behavioral effects by binding to specific seven-transmembrane neurotransmitter receptors (opioids and cannabinoids), neurotransmitter receptor-gated ion channels (nicotine, barbiturates, and benzodiazepines), or transporters (cocaine and amphetamines) on the plasma membrane of neuronal cells (Fig. 386-2). Ethanol interacts with several signaling proteins including serotonin 5HT-3 receptors, nicotinic receptors, voltage-gated calcium channels, and sodium-independent purine transporters, butGABAAreceptors and the N-methyl-D-aspartate (NMDA) subtype of glutamate receptors appear to be most sensitive to intoxicating concentrations of ethanol. As discussed below, drug actions at these targets lead to elevation of extracellular dopamine levels in the nucleus accumbens. This appears to be extremely important for the reinforcing properties of psychostimulants. For several other drugs of abuse, dopamine-independent pathways also contribute. Dopamine Transporter Much evidence indicates that the rewarding properties of
cocaine and amphetamine are due primarily to their ability to elevate extracellular dopamine levels in the nucleus accumbens. Specific populations of neurons within the nucleus accumbens are activated during cocaine self-administration in rodents. In addition, selective destruction of dopaminergic terminals within the nucleus accumbens or administration of dopamine receptor antagonists into that region eliminates cocaine self-administration. Cocaine and amphetamines act by altering transport of dopamine through plasma membrane dopamine transporters (DATs) in presynaptic nerve terminals (Fig. 386-2A). Reuptake of dopamine through DATs is the major mechanism for termination of dopaminergic neurotransmission. DAT is a 12-transmembrane glycoprotein and a member of the large sodium- and chloride-dependent transporter family, which also includes carriers forGABA, glycine, serotonin, norepinephrine, and other organic molecules. Cocaine binds to DATs and inhibits dopamine reuptake. Amphetamine causes intracellular release of dopamine from vesicles and reverse transport of dopamine through DATs. These actions serve to elevate levels of extracellular dopamine at dopaminergic synapses. Recent studies of mice lacking theDAT gene have revealed redundancy in systems mediating cocaine reinforcement. Psychostimulants fail to alter extracellular dopamine levels or induce locomotor activity in these mice. However, DAT-null mice can still be trained to press a lever to receive intravenous cocaine, suggesting that other genes can also mediate the reinforcing properties of cocaine. In addition to inhibiting DAT function, cocaine blocks reuptake of serotonin and norepinephrine. In DAT-null mice, residual binding of a cocaine analogue can be displaced by serotonin reuptake inhibitors, and cocaine stimulates neurons in brain regions with a high density of serotonergic fibers. Therefore, inhibition of serotonin uptake through plasma membrane serotonin transporters may also contribute to psychostimulant reward. GABAAReceptors The major inhibitory neurotransmitter in the nervous system isGABA. Binding of GABA to GABAAreceptors activates a Cl- current that is enhanced by benzodiazepines, barbiturates, and ethanol (Fig. 386-2B). Activation of this current maintains the neuronal plasma membrane close to its resting potential and thereby inhibits the generation of action potentials. GABAAreceptors appear to be pentameric membrane glycoproteins composed ofa, b,g, and possibly d peptide subunits. Fifteen subunits are known to be expressed in the mammalian central nervous system (six a, threeb, three g, oned, one e, oneq) and RNA splice variants have been identified. In rodents, the GABAAreceptor agonist muscimol substitutes for ethanol in tests of drug discrimination; when injected into the nucleus accumbens, muscimol terminates ethanol self-administration. These results suggest that the reinforcing properties of ethanol are mediated in part by ethanol's actions at GABAAreceptors in the nucleus accumbens. NMDA Receptors In the nervous system, excitatory synaptic activity evoked by glutamate and aspartate is mediated by neurotransmitter receptor-gated ion channels (ionotropic receptors) that regulate cation conductances and by G protein-coupled receptors (metabotropic receptors) that stimulate phosphoinositide hydrolysis. Ionotropic glutamate receptors have been subclassified based on activation by selective agonists into three groups:NMDAreceptors, high-affinity kainate receptors, anda-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors. There is a rich glutamatergic projection from neurons in the frontal cortex to the nucleus accumbens where fibers terminate mainly on medium spiny GABA-ergic neurons.
NMDA receptors in the nucleus accumbens appear to play a role in the rewarding properties of several drugs. Thus, phencyclidine and other NMDA antagonists have rewarding actions when administered in the nucleus accumbens. The rewarding properties of ethanol may also be partly due to inhibition of NMDA-activated calcium currents. In addition, NMDA receptors modulate responses to psychostimulants since administration of the NMDA receptor antagonist MK-801 prior to cocaine or amphetamine prevents sensitization to these drugs. Nicotinic Receptors Like other drugs of abuse, nicotine elicits dopamine release in the nucleus accumbens, and intravenous self-administration of nicotine is blocked by dopamine antagonists and by neurochemical lesions that destroy dopaminergic fibers in the nucleus accumbens. Nicotine increases dopamine release by activating nicotinic acetylcholine receptors (nAChRs) on cell bodies and nerve terminals of dopaminergicVTAneurons. Neuronal nAChRs are receptor-gated ion channels that allow entry of sodium into cells when acetylcholine is present. They appear to be pentameric complexes composed of different combinations of at least 10 different subunits. Targeted disruption of theb2subunit gene eliminates most high-affinity nicotine binding in the brain and prevents nicotine-induced dopamine release in the nucleus accumbens. In addition, mice that lack theb2subunit show attenuated nicotine self-administration, indicating thatb2-containing nAChRs mediate the reinforcing properties of nicotine. Opioid Receptors Morphine and other opioids activate opioid receptors, which are a family of seven-transmembrane, G protein-coupled receptors (Figs. 386-2C and 386-3). Three classes, u,d, and k, have been identified. The rewarding action of opioids appears to be mediated by activation of u receptors since opioid reinforcement is blocked by selective u receptor antagonists and by targeted disruption of the u receptor gene. Binding of opioid agonists to u receptors activates the G proteins G i and Go. This results in inhibition of adenylyl cyclase, thereby decreasing levels of the intracellular second messenger cyclic AMP and reducing activity of cyclic AMP-dependent protein kinase A (PKA). In addition, these G proteins activate voltage-gated potassium channels and inhibit voltage-gated calcium channels. The net result is suppression of electrical excitability in neurons expressing u receptors. GABA-containing interneurons in theVTAsuppress firing of dopaminergic VTA neurons that project to the nucleus accumbens. Opioids disinhibit these dopaminergic neurons by binding to u receptors expressed by the GABA-containing interneurons. This increases the firing rate of dopaminergic VTA neurons and promotes dopamine release in the nucleus accumbens. Rodents will self-administer opioids into both the VTA and the nucleus accumbens. Opioid self-administration into the nucleus accumbens occurs even after dopaminergic projections to that region are destroyed. Thus, dopamine-dependent mechanisms involving the VTA and dopamine-independent mechanisms in the nucleus accumbens both contribute to opioid reward. Opioid receptors also regulate ethanol consumption. Ethanol acutely inhibits opioid binding tod-opioid receptors, and chronic ethanol exposure increases the density of uand d-opioid receptors. Nonselective opioid antagonists reduce ethanol self-administration in animals. Several regions of the extended amygdala appear to mediate this response, although the central nucleus of the amygdala appears most important. In two independent clinical trials, the opioid receptor antagonist naltrexone, in
combination with counseling, reduced craving and relapse in abstinent alcoholics. Thus, opioid systems appear to modulate ethanol craving in addicted individuals. Cannabinoid Receptors The active ingredient of cannabis,D9-tetrahydrocannabinol (D-9-THC), and the endogenous cannabinoid anandamide bind two subtypes of G protein-coupled cannabinoid receptors. Studies with mutant mice lacking the CB1receptor gene have revealed that the CB1receptor is responsible for the reinforcing properties of cannabinoids. When administered intravenously, D-9-THC increases dopamine levels in the nucleus accumbens. This is blocked by cannabinoid receptor antagonists and by u-opioid receptor antagonists administered into theVTA. Conversely, morphine reinforcement and the severity of opioid withdrawal are reduced in mice that lack CB1receptors. These results suggest that opioids and cannabinoids share common signaling pathways in the brain and can interact to promote each other's reinforcing properties. ADAPTATION TO CHRONIC DRUG USE Repeated drug exposure elicits changes in neural function that lead to drug dependence and craving. Several mechanisms are being elucidated, including drug-induced alterations in receptor and ion channel function, intracellular signal transduction, gene expression, and synaptic connectivity. Chronic exposure to drugs of abuse can change the function of receptors and ion channels by altering their density, subunit composition, or coupling to signal transduction cascades. For example, chronic exposure to ethanol increases the density ofNMDAreceptors and L-type voltage-gated calcium channels in the brain. These changes contribute to neuronal hyperactivity observed during alcohol withdrawal since NMDA and L-type channel antagonists reduce signs of withdrawal in alcohol-dependent rodents deprived of ethanol. In addition, chronic exposure to ethanol decreasesGABAAreceptor function and abolishes potentiation by ethanol. Downregulation of GABAAreceptor function contributes to manifestations of alcohol withdrawal since benzodiazepines and barbiturates, which activate GABAAreceptors, are very helpful in reducing alcohol withdrawal symptoms. Chronic exposure to many drugs of abuse induces adaptive changes in neuronal signal transduction pathways. This has been most clearly demonstrated for opioids, which increase cyclic AMP signaling in the locus coeruleus after chronic administration (Fig. 386-3). The locus coeruleus is the principal adrenergic nucleus in the brain and regulates attention states and the autonomic nervous system. Hyperactivity of this nucleus has been implicated in opioid withdrawal. Chronic opioid exposure increases expression of adenylyl cyclase,PKA, and the cyclic AMP response element binding protein, CREB, which mediates cyclic AMP-dependent gene expression. These changes increase the intrinsic firing rate of neurons in the locus coeruleus, in part through activation of an inward sodium current. Thus, upregulation of cyclic AMP signaling opposes the acute inhibitory action of opioids on this pathway. Chronic exposure to cocaine, opioids, or ethanol upregulates cyclic AMP-mediated signaling in other brain regions, including theVTA and the nucleus accumbens. Upregulation of cyclic AMP signaling in the nucleus accumbens contributes to
psychostimulant tolerance since pharmacologic inhibition ofPKA or overexpression ofCREB in the nucleus accumbens decreases the rewarding properties of cocaine. Upregulation of cyclic AMP signaling may also account for supersensitivity of neurons in the nucleus accumbens to D1receptor agonists following chronic cocaine administration. Additional neuroadaptive changes in dopamine signaling can modify drug-seeking behavior in rodents. Rats can be readily trained to voluntarily self-administer intravenous cocaine for several hours. If during a course of self-administration, saline is substituted for cocaine, the rate of self-administration declines dramatically. However, exposure to a single intraperitoneal priming injection of cocaine or a D2-like receptor agonist causes the animal to resume lever pressing. In contrast, treatment with a D1-like receptor agonist blocks the ability of cocaine to reinstate drug-seeking behavior. It appears that D1-like agonists inhibit relapse in drug seeking. Therefore, they may prove to be useful in treatment of cocaine addiction. A current hypothesis views addiction as a form of learning mediated by maladaptive recruitment of memory systems involving limbic structures. Mechanisms that could contribute to such learned, long-term adaptation include drug-induced gene expression and synaptic reorganization. For example, the transcription factorCREB, which is activated by chronic drug use, has been implicated in models of learning. Repeated exposure to many drugs of abuse also causes prolonged activation of another transcription factor, Fos-related antigen, in the nucleus accumbens. Ethanol increases the number of dendritic spines on hippocampal pyramidal cells and somatosensory cortical neurons, whereas amphetamine increases dendritic length and the density of dendritic spines on neurons in the nucleus accumbens and prefrontal cortex. Such drug-induced changes in gene expression and neuronal connectivity could lead to long-term alterations in brain reward pathways that may underlie drug addiction in humans. (Bibliography omitted in Palm version) Back to Table of Contents
387. ALCOHOL AND ALCOHOLISM - Marc A. Schuckit The yearly cost of alcohol-related problems in the United States is as much as $300 billion, including accidents, health problems, lost productivity, crime, and treatment. There are more than 22,000 deaths from alcohol-related auto accidents per year, as well as almost 2 million nonfatal injuries and damage to almost 5 million vehicles. In addition, alcohol is responsible for almost 5% of missed work time, with a 25% decrease in work performance among heavy drinkers. Men and women who fulfill criteria for alcohol use disorders decrease their life span by approximately 15 years, with abuse and dependence responsible for almost 25% of premature deaths in men and 15% in women, figures that represent a three- to sixfold odds ratio of early death even among people with higher levels of education and socioeconomic functioning. PHARMACOLOGY AND NUTRITIONAL IMPACT OF ETHANOL Ethanol is a weakly charged molecule that moves easily through cell membranes, rapidly equilibrating between blood and tissues. The effects of drinking depend in part on the amount of ethanol consumed per unit of body weight; the level of alcohol in the blood is expressed as milligrams or grams of ethanol per deciliter (e.g., 100 mg/dL or 0.10 g/dL). A level of 0.02 to 0.03 results from the ingestion of one to two typical drinks. In round figures, 340 mL (12 oz) of beer, 115 mL (4 oz) of nonfortified wine, and 43 mL (1.5 oz) (a shot) of 80-proof beverage each contain approximately 10 g of ethanol; 0.5 L (1 pint) of 86-proof beverage contains approximately 160 g, and 1 L of wine contains approximately 80 g of ethanol. Congeners found in alcoholic beverages may contribute to body damage with heavy drinking; these include low-molecular-weight alcohols (e.g., methanol and butanol), aldehydes, esters, histamine, phenols, tannins, iron, lead, and cobalt. Ethanol is a central nervous system (CNS) depressant that decreases activity of neurons, although some behavioral stimulation is observed at low blood levels. This drug has cross-tolerance and shares a similar pattern of behavioral problems with other brain depressants, including the benzodiazepines and barbiturates. Alcohol is absorbed from mucous membranes of the mouth and esophagus (in small amounts), from the stomach and large bowel (in modest amounts), and from the proximal portion of the small intestine (the major site). The rate of absorption is increased by rapid gastric emptying; by the absence of proteins, fats, or carbohydrates (which interfere with absorption); by the absence of congeners; by dilution to a modest percentage of ethanol (maximum at about 20% by volume); and by carbonation (e.g., champagne). Between 2% (at low blood alcohol concentrations) and about 10% (at high blood alcohol concentrations) of ethanol is excreted directly through the lungs, urine, or sweat, but the greater part is metabolized to acetaldehyde, primarily in the liver. At least two metabolic routes, each with different optimal concentrations of ethanol (Km), result in the metabolism of approximately one drink per hour. The most important pathway occurs in the cell cytosol where alcohol dehydrogenase (ADH) produces acetaldehyde, which is then rapidly destroyed by aldehyde dehydrogenase (ALDH) in the cytosol and mitochondria. Each of these steps requires nicotinamide adenine dinucleotide (NAD) as a cofactor, and it is the increased ratio of the reduced cofactor (NADH) to NAD (NADH:NAD) that is responsible for many of the metabolic derangements observed
after drinking. A second pathway occurs in the microsomes of the smooth endoplasmic reticulum (the microsomal ethanol-oxidizing system, or MEOS), which is responsible for 10% or more of ethanol oxidation at high blood alcohol concentrations. One gram of ethanol has approximately 29.7 kJ (7.1 kcal) of energy, and a drink contains between 293.0 and 418.6 kJ (70 and 100 kcal) from ethanol and other carbohydrates. However, these are "empty" of nutrients such as minerals, proteins, and vitamins. In addition, alcohol interferes with absorption of vitamins in the small intestine and decreases their storage in the liver. These actions affect folate (folacin or folic acid), pyridoxine (B6), thiamine (B1), nicotinic acid (niacin, B3), and vitamin A. Heavy drinking can also produce low blood levels of potassium, magnesium, calcium, zinc, and phosphorus as a consequence of dietary deficiency and acid-base imbalances during excess alcohol ingestion or withdrawal. An ethanol load in a fasting, healthy individual is likely to produce transient hypoglycemia within 6 to 36 h, secondary to the acute actions of ethanol on gluconeogenesis. This can result in glucose intolerance until the alcoholic has abstained for 2 to 4 weeks. Alcohol ketoacidosis, probably reflecting a decrease in fatty acid oxidation coupled with poor diet or recurrent vomiting, should not be misdiagnosed as diabetic ketosis. With the former, patients show an increase in serum ketones along with a mild increase in glucose but a large anion gap, a mild to moderate increase in serum lactate, and a b-hydroxybutyrate/lactate ratio of between 2:1 and 9:1 (with normal being 1:1). BEHAVIORAL EFFECTS, TOLERANCE, AND DEPENDENCE The effects of any drug depend on the dose, the rate of increase in plasma, the concomitant presence of other drugs, and the past experience with the agent. With alcohol, an additional factor is whether blood alcohol levels are rising or falling; the effects are more intense during the former period. Even though "legal intoxication" requires a blood alcohol concentration of at least 80 to 100 mg/dL, behavioral, psychomotor, and cognitive changes are seen at levels as low as 20 to 30 mg/dL (i.e., after one to two drinks). Deep but disturbed sleep can be seen at twice the legal intoxication level, and death can occur with levels between 300 and 400 mg/dL. Beverage alcohol is probably responsible for more overdose deaths than any other drug. The intoxicating effects of alcohol appear to be due to actions at specific neurotransmitter receptors and transporters. Alcohol enhances g-aminobutyric acid A (GABAA) receptors, and inhibits N-methyl-D-asparate (NMDA) receptors (Chap. 386). In vitro studies suggest that additional effects involve inhibition of adenosine uptake and a translocation of the cyclic AMP-dependent protein kinase catalytic subunit from the cytoplasm to the nucleus. Neurons adapt quickly to these actions, and thus different effects may be present during chronic administration and withdrawal. At least three types of compensation develop after repeated exposure to the drug, producing tolerance of higher ethanol levels. First, after 1 to 2 weeks of daily drinking, metabolic or pharmacokinetic tolerance develops, with a 30% increase in the rate of
hepatic ethanol metabolism. This alteration disappears almost as rapidly as it develops. Second, cellular or pharmacodynamic tolerance develops through neurochemical changes that may also contribute to physical dependence. Third, individuals can learn to adapt their behavior so that they can function better than expected under drug influence (behavioral tolerance). The cellular changes caused by chronic ethanol exposure may not resolve for several weeks or longer following cessation of drinking. In the interim, the neurons require ethanol to function optimally, and the individual can be said to be physically dependent. This physical condition is distinct from psychological dependence, a concept indicating that the person is psychologically uncomfortable without the drug. THE EFFECTS OF ETHANOL ON BODY SYSTEMS While one to two drinks per day in an otherwise healthy and nonpregnant individual can have some beneficial effects, at higher doses alcohol is toxic to most body systems. Knowledge about the deleterious effects of alcohol helps the practicing physician to identify alcoholic patients. Signs and symptoms of ethanol abuse can be used to help motivate the patient to abstain. It is important to remember that the typical white- or blue-collar alcoholic functions at a fairly high level for years, and that not everyone develops each problem. CENTRAL NERVOUS SYSTEM Approximately 35% of drinkers may experience a blackout, an episode of temporary anterograde amnesia, in which the person forgets all or part of what occurred during a drinking evening. Another common problem, one seen after as few as one or two drinks, is that while alcohol can help someone to fall asleep, it also "fragments" the sleep pattern causing alterations between sleep stages and a deficiency in deep sleep. At the same time, alcohol diminishes rapid eye movement (REM) or dream sleep early in the evening, with resulting prominent and sometimes disturbing dreams later in the night. Finally, alcohol relaxes muscles in the pharynx, which can cause snoring and exacerbate sleep apnea, with symptoms of the latter in 75% of alcoholic men over age 60. An additional problem related to the acute effects of alcohol on most drinkers is the impairment in judgment, balance, and motor coordination that contributes to the high incidence and severity of accidents. At least half of individuals who experience severe physical trauma in an accident have evidence of substance-related impairment, a finding that is consistent with the fact that 40% of drinkers in the United States have at some time driven while intoxicated with alcohol and that 15% of flight crews have evidence of repeated heavy drinking. Regarding the latter, at least one study noted that pilot performance is still impaired 14 h after a blood alcohol concentration of 100 mg/dL, despite subsequent abstinence. The effect of alcohol on the nervous system is even more pronounced among alcohol-dependent individuals. Chronic intake of high doses of ethanol causes peripheral neuropathy in 5 to 15% of alcoholics, which is possibly related to thiamine deficiency. Patients complain of bilateral limb numbness, tingling, and paresthesias;
symptoms are more pronounced distally than proximally. The treatment is abstinence and thiamine supplementation. Wernicke's syndrome (ophthalmoparesis, ataxia, and encephalopathy) and Korsakoff's syndrome (alcohol-induced persisting amnestic disorder), are seen in the United States at a rate of approximately 50 per million people per year. These disorders are the result of thiamine deficiency in vulnerable individuals, possibly owing to interaction with a genetic transketolase deficiency. Korsakoff's syndrome presents as profound and persistent anterograde amnesia (inability to learn new material) and a milder retrograde amnesia. Additional symptoms can include impairment in visuospatial, abstract, and conceptual reasoning but with a normal intelligence quotient (IQ). Some patients demonstrate an acute onset of Korsakoff's syndrome in association with the neurologic stigmata seen with Wernicke's syndrome (e.g., sixth nerve palsy and ataxia), whereas others have a more gradual onset. With oral thiamine replacement (50 to 100 mg/d), only one-quarter of Korsakoff's patients achieve full recovery, one-half experience partial improvement, and one-quarter show no improvement, even after many months of supplementation.*Wernicke's syndrome is discussed in detail in Chap. 376. About 1% of alcoholics develop cerebellar degeneration, a syndrome of progressive unsteady stance and gait often accompanied by mild nystagmus. Atrophy of the cerebellar vermis is seen on brain computed tomography and magnetic resonance imaging scans, but the cerebrospinal fluid is usually normal. Treatment consists of abstinence and multiple vitamin supplementation, although improvement is often minimal. Alcoholics can show severe cognitive problems and impairment in recent and remote memory for weeks to months after an alcoholic binge. Increased size of the brain ventricles and cerebral sulci are seen in 50% or more of chronic alcoholics, but these changes are often reversible, returning toward normal after a year or more of abstinence. PermanentCNSimpairment (alcohol-induced persisting dementia) can develop and accounts for up to 20% of chronically demented patients. There is no single alcoholic dementia syndrome; rather, this label is used to describe patients who have apparently irreversible cognitive changes (possibly from diverse causes) in the midst of chronic alcoholism. Finally, almost every psychiatric syndrome can be seen temporarily during heavy drinking or subsequent withdrawal. These include intense sadness lasting for days to weeks in the midst of heavy drinking in 40% of alcoholics, which is classified as an alcohol-induced mood disorder in the Fourth Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-IV); severe anxiety in 10 to 30% of alcoholics, often beginning during alcohol withdrawal and which can persist for many months after cessation of drinking (alcohol-induced anxiety disorder); and auditory hallucinations and/or paranoid delusions in the absence of any obvious signs of withdrawal -- a state now called alcohol-induced psychotic disorder -- and reported at sometime in 1 to 10% of alcoholics. Treatment of all forms of alcohol-induced psychopathology includes abstinence and supportive care, with the likelihood of full recovery within several days to 6 weeks. A history of alcohol intake is an important consideration in any patient with one of these psychiatric symptoms.
THE GASTROINTESTINAL SYSTEM Esophagus and Stomach Acute alcohol intake can result in inflammation of the esophagus (possibly secondary to reflux of gastric contents) and stomach (resulting from both an increase in acid production and damage to the gastric mucosal barrier). Esophagitis can cause epigastric distress, and gastritis, the most frequent cause of gastrointestinal bleeding in heavy drinkers, can present as anorexia and/or abdominal pain. Chronic heavy drinking, if associated with violent vomiting, can produce a longitudinal tear in the mucosa at the gastroesophageal junction -- a Mallory-Weiss lesion. Although many gastrointestinal problems are reversible, two complications of chronic alcoholism, esophageal varices secondary to cirrhosis-induced portal hypertension and atrophy of the gastric mucosa, may be irreversible. Pancreas The incidence of acute pancreatitis in alcoholics (about 25 per 1000 per year) is almost threefold higher than in the general population, accounting for an estimated 10% or more of the cases of this disorder (Chap. 304). Liver Ethanol absorbed from the small bowel is carried directly to the liver, where it becomes the preferred fuel; NADH accumulates and oxygen utilization escalates; gluconeogenesis is impaired (with a resulting fall in the amount of glucose produced from glycogen); lactate production increases; and there is a decreased oxidation of fatty acids in the citric acid cycle with an increase in fat accumulation within liver cells. In the healthy individual taking no medications, these changes are reversible, but with repeated exposure to ethanol, more severe changes in liver functioning are likely to occur. These include, in overlapping stages, fatty accumulation, alcohol-induced hepatitis, perivenular sclerosis, and cirrhosis, with the latter observed in an estimated 15 to 20% of alcoholics (Chap. 298). CANCER As discussed briefly below, the leading cause of death in alcoholics is cardiovascular disease, but cancer occupies a solid second place. Women drinking as few as 1.5 drinks per day increase their risk of breast cancer 1.4-fold. For both genders, four drinks per day increases the risk for oral and esophageal cancers by approximately threefold and rectal cancers by a factor of 1.5, whereas seven to eight or more drinks per day enhances the risks for many of these cancers by a factor of five. Overall, it has been estimated that alcoholics have a rate of carcinoma 10 times higher than the general population. HEMATOPOIETIC SYSTEM Ethanol exerts multiple reversible acute and chronic effects on all blood cells. The impact on red blood cells (RBC) is an increase in size (mean corpuscular volume, MCV), usually without anemia. This change appears to reflect the effect of alcohol on stem cells. If heavy drinking is accompanied by folic acid deficiency, there can also be hypersegmented neutrophils, reticulocytopenia, and hyperplastic bone marrow; if malnutrition is present, sideroblastic changes can also be observed. Chronic heavy drinking can also decrease production of most white blood cells (WBCs), decrease granulocyte mobility and adherence, and impair the delayed-hypersensitivity response
to new antigens (with a possible false-negative tuberculin skin test). Finally, many alcoholics present with mild thrombocytopenia. When due to repeated intoxication, the low platelet count usually resolves within a week of abstinence. Thrombocytopenia can also occur secondary to hepatic cirrhosis and congestive splenomegaly (increased destruction) or to folic acid deficiency (decreased production). Ethanol itself might not have a major effect on platelet function, but polyphenols and other constituents of some alcoholic beverages, particularly wine, may interfere with platelet aggregation. CARDIOVASCULAR SYSTEM Acutely, ethanol decreases myocardial contractility and causes peripheral vasodilation, with a resulting mild decrease in blood pressure and a compensatory increase in cardiac output. Exercise-induced increases in cardiac oxygen consumption are higher after alcohol intake. These acute effects have little clinical importance for the average healthy drinker but can produce problems in men and women with cardiac disease. Chronic intake of even modest doses of alcohol can have both deleterious and beneficial effects. Regarding the latter, a maximum of one to two drinks per day over long periods may decrease the risk for cardiovascular death, perhaps through an increase in high-density lipoprotein (HDL) cholesterol or changes in clotting mechanisms. In one large national study, cardiovascular mortality was reduced by 30 to 40% among individuals reporting one or more drinks daily compared to nondrinkers, with overall mortality lowest among those consuming approximately one drink per day. Recent data have also corroborated the decreased risk for ischemic, but not hemorrhagic, stroke associated with regular light drinking. The consumption of three or more drinks per day results in a dose-dependent increase in blood pressure, which returns to normal within weeks of abstinence. As a result, heavy drinking is an important contributor to mild to moderate hypertension. Chronic heavy drinking can cause cardiomyopathy, with symptoms ranging from unexplained arrhythmias in the presence of left ventricular impairment to heart failure with dilation of all four heart chambers and hypocontractility of heart muscle. Perhaps one-third of cases of cardiomyopathy are alcohol-induced. Mural thrombi can form in the left atrium or ventricle, while heart enlargement exceeding 25% can cause mitral regurgitation. Atrial or ventricular arrhythmias, especially paroxysmal tachycardia, can also occur after a drinking binge in individuals showing no other evidence of heart disease -- a syndrome known as the "holiday heart." GENITOURINARY SYSTEM CHANGES, SEXUAL FUNCTIONING, AND FETAL DEVELOPMENT Acutely, modest ethanol doses (e.g., blood alcohol concentrations of 100 mg/dL or even less) can both increase sexual drive and decrease erectile capacity in men. Even in the absence of liver impairment, a significant minority of chronic alcoholic men may show irreversible testicular atrophy with concomitant shrinkage of the seminiferous tubules, decreases in ejaculate volume, and a lower sperm count (Chap. 335). The repeated ingestion of high doses of ethanol by women can result in amenorrhea, a decrease in ovarian size, absence of corpora lutea with associated infertility, and
spontaneous abortions. Heavy drinking during pregnancy results in the rapid placental transfer of both ethanol and acetaldehyde, which may have serious consequences for fetal development. The fetal alcohol syndrome can include any of the following: facial changes with epicanthal eye folds, poorly formed concha, and small teeth with faulty enamel; cardiac atrial or ventricular septal defects; an aberrant palmar crease and limitation in joint movement; and microcephaly with mental retardation. The specific amount of ethanol and/or specific time of vulnerability during pregnancy have not been defined, making it advisable for pregnant women to abstain completely. OTHER EFFECTS OF ETHANOL Between one-half and two-thirds of alcoholics have evidence of decreased skeletal muscle strength caused by acute alcoholic myopathy, a condition that improves but which might not disappear with abstinence. Effects of repeated heavy drinking on the skeletal system include alterations in calcium metabolism, lower bone density, and less growth in the epiphyses, with an increased risk for fractures and osteonecrosis of the femoral head. Hormonal changes include an increase in cortisol levels, which can remain elevated during heavy drinking; inhibition of vasopressin secretion at rising blood alcohol concentrations and the opposite effect at falling blood alcohol concentrations (with the final result that most alcoholics are likely to be slightly overhydrated); a modest and reversible decrease in serum thyroxine (T4); and a more marked decrease in serum triiodothyronine (T3). ALCOHOLISM (ALCOHOL ABUSE OR DEPENDENCE) Because many drinkers occasionally imbibe to excess, temporary alcohol-related pathology is common in nonalcoholics. The period of heaviest drinking is usually the late teens to the late twenties. This is also a time of high risk for temporary alcohol-related social, occupational, or driving difficulties. These phenomena are often isolated events or self-limited, but when repeated problems in multiple life areas develop, the person is likely to meet criteria for alcohol abuse or dependence. DEFINITIONS AND EPIDEMIOLOGY DSM-IVdefines alcohol dependence as repeated alcohol-related difficulties in at least three of seven areas of functioning that cluster together over any 12-month period. These problems include any combination of tolerance, withdrawal, taking larger amounts of alcohol over longer periods than intended, an inability to control use, spending a great deal of time associated with alcohol use, giving up important activities to drink, and continued use of alcohol despite physical or psychological consequences. In this diagnosis a special emphasis is placed on evidence of tolerance and/or withdrawal, a condition referred to as "dependence with a physiological component" and which is associated wtih a more severe clinical course. Dependence occurs in both men and women, in individuals from all socioeconomic strata, and in people of all racial backgrounds. The diagnosis predicts a course of recurrent problems with the use of alcohol and the consequent shortening of the life span by a decade or more. In the absence of alcohol dependence, an individual can be given a diagnosis of alcohol abuse if he or she demonstrates repetitive problems with alcohol in any one of four life areas: an inability to fulfill major obligations, use in hazardous situations such as driving,
legal problems, or use despite social or interpersonal difficulties. The clinical diagnosis of alcohol abuse or dependence rests on the documentation of a pattern of difficulties associated with alcohol use; the definition is not based on the quantity and frequency of alcohol consumption. Thus, in screening for alcohol abuse or dependence, it is important to probe for life problems and then attempt to tie in use of alcohol or another substance. Information regarding marital or job problems, legal difficulties, histories of accidents, medical problems, evidence of tolerance, etc., is an important component of all evaluations and yields data that are of use even for nonalcoholic individuals. The lifetime risk for alcohol dependence in most western countries is about 10 to 15% for men and 5% for women. When alcohol abuse is also considered, the rates are even higher. The typical alcoholic is a blue- or white-collar worker or homemaker and thus does not fit the common stereotype. GENETICS OF ALCOHOLISM Alcoholism is a multifactorial disorder in which both environmental and biologic factors contribute. The importance of genetic influences in alcoholism is supported by the higher risk for this disorder in the identical versus fraternal twin of an alcoholic and the fourfold increased risk for children of alcoholics even if adopted at birth and raised without knowledge of the problems of their biologic parents. The evidence supporting genetic influences in alcoholism has stimulated a search for trait markers of a vulnerability toward the disorder. A 15-year follow-up of 453 men originally studied at age 20 has shown that subjects with alcoholic fathers demonstrated relatively lower levels of response to alcohol, including less intense subjective feelings of intoxication, less alcohol-related impairment in cognitive and psychomotor tests, and less intense alcohol-related changes in prolactin and cortisol secretion. This low level of response to alcohol at around age 20 was a powerful predictor of later alcoholism, explaining most of the relationship between a family history of this disorder and later alcohol problems. Additional genetically influenced characteristics that contribute to the risk of alcoholism appear to include some personality traits such as higher levels of impulsivity and sensation seeking, and several electrophysiologic measures such as the P300 wave of the event-related potential (Chap. 357), which might relate to cognitive styles or evidence ofCNSdisinhibition. All the genetic factors combined appear to explain up to 60% of the risk, with environmental influences contributing at least 40%. NATURAL HISTORY For the "average" alcoholic, the age of first drink and first minor problems (e.g., an argument with a friend while drunk or an alcoholic blackout) are similar to those in the general population. However, by the early to mid-twenties, most men and women moderate their drinking (perhaps learning from minor problems), whereas difficulties for alcoholics are likely to escalate, with the first major life problem from alcohol appearing in the mid-twenties to early forties. Once established, the course of alcoholism is likely to be one of exacerbations and remissions. As a rule, there is remarkably little difficulty in stopping alcohol use when problems develop, and this step is often followed by days
to months of carefully controlled drinking. Unfortunately, these periods are almost inevitably followed by escalations in alcohol intake and subsequent problems. The course is not hopeless, because between half and two-thirds of alcoholics maintain abstinence for extended periods after treatment. Even without formal treatment or self-help groups there is at least a 20% chance of long-term abstinence. However, should the alcoholic continue to drink, the life span is shortened by an average of 15 years, with the leading causes of death, in decreasing order, being heart disease, cancer, accidents, and suicide. IDENTIFICATION OF THE ALCOHOLIC AND INTERVENTION Physicians even in affluent areas should recognize that approximately 20% of patients have alcoholism. Therefore, it is important to pay attention to the alcohol-related symptoms and signs described above as well as laboratory tests that are likely to be abnormal in the context of regular consumption of 6 to 8 or more drinks per day. These include a high-normal or slightly elevatedMCV(e.g., ³91 fL),g-glutamyl transferase (GGT) (³30 units), serum uric acid [>416 umol/L (7 mg/dL)], carbohydrate-deficient transferrin (CDT) (³20 g/L), and triglycerides [³2.0 mmol/L (180 mg/dL)]. Mild and fluctuating hypertension (e.g., 140/95), repeated infections such as pneumonia, and otherwise unexplained cardiac arrhythmias should also raise the possibility that the patient is an alcoholic. Other disorders suggestive of alcoholism include cancer of the head and neck, esophagus, or stomach as well as cirrhosis, unexplained hepatitis, pancreatitis, bilateral parotid gland swelling, and peripheral neuropathy. Once the likelihood of alcoholism is established, only a few moments are needed to gather the history of alcohol-related life problems. The patient and the spouse or another close family member should be asked about patterns of accidents, relationship difficulties, problems on the job, and driving-related difficulties, after which the role played by alcohol should be identified. All physicians should be able to take the time needed to gather such information. In addition, a simple 25-item form to be answered by the patient, the Michigan Alcohol Screening Test (MAST), is available to aid in identifying alcoholics. However, this is only a screening tool, and a careful face-to-face interview is still required for a meaningful diagnosis. The CAGE, which consists of asking about alcohol-related trouble cutting down on intake, being annoyed by criticisms, guilt, or use of an "eye-opener," can also be helpful as an initial screen. After alcoholism is identified, the diagnosis must be shared with the patient. The presenting complaint can be used as an entree to the alcohol problem. For instance, the patient complaining of insomnia or hypertension could be told that these are clinically important symptoms and that physical findings and laboratory tests indicate that alcohol appears to have contributed to the complaints and is increasing the risk for further medical and psychological problems. The physician should share information about the course of alcoholism and explore possible avenues of attacking the problem. Some patients and family members will benefit from the opportunity to read additional material (see "Bibliography"). The process of intervention is rarely accomplished in one session. For the person who refuses to stop drinking at the first intervention, a logical step is to "keep the door open," establishing future meetings so that help is available as problems escalate. In the
meantime the family may benefit from counseling or referral to self-help groups such as Al-Anon (the Alcoholics Anonymous group for family members) and Alateen (for teenage children of alcoholics). The patient should be reminded that driving while intoxicated is dangerous and illegal. THE ALCOHOL WITHDRAWAL SYNDROME Once the brain has been repeatedly exposed to high doses of alcohol, any sudden decrease in intake can produce symptoms of withdrawal. As with allCNSdepressants, the symptoms are generally the opposite of those produced by intoxication. Features include tremor of the hands (shakes or jitters); agitation and anxiety; autonomic nervous system overactivity such as an increase in pulse, respiratory rate, and body temperature; insomnia, possibly accompanied by bad dreams; and gastrointestinal upset. These withdrawal symptoms generally begin within 5 to 10 h of decreasing ethanol intake, peak in intensity on day 2 or 3, and improve by day 4 or 5. Anxiety, insomnia, and mild levels of autonomic dysfunction may persist at decreasing levels for 6 months or more as a protracted abstinence syndrome, which may contribute to the tendency to return to drinking. At some point in their lives, between 2 and 5% of alcoholics experience withdrawal seizures ("rum fits"), usually within 48 h of stopping drinking. These are usually generalized (unless there is an underlying focal lesion), and any electroencephalographic abnormalities are mild and generally return to normal within several days. The term delirium tremens (DTs) refers to delirium (mental confusion with fluctuating levels of consciousness) along with a tremor, severe agitation, and autonomic overactivity (e.g., marked increases in pulse, blood pressure, and respirations). Fortunately, this serious and potentially life-threatening complication of alcohol withdrawal is rare. Only 5 to 10% of alcohol-dependent individuals ever experience DTs; the chance of DTs during any single withdrawal is less than 1% but is higher if there has been a withdrawal seizure. DTs are most likely to develop in patients with concomitant severe medical disorders or evidence of underlying brain damage, and thus can usually be avoided if the underlying medical problems can be identified and treated. TREATMENT Acute Intoxication The first priority is to be certain that the vital signs are relatively stable without evidence of respiratory depression, cardiac arrhythmia, or potentially dangerous changes in blood pressure. Life-threatening problems require appropriate emergency care and hospitalization. The clinician must recognize that a variety of causes may produce obtundation or coma in the alcoholic patient. The possibility of intoxication with other drugs should be considered, and a blood or urine sample is indicated to screen for opioids or otherCNSdepressants such as benzodiazepines or barbiturates. A coexisting seizure disorder, head injury, meningitis, brain abscess, or other potentially life-threatening neurologic disorder may be present. Other medical conditions that must be considered include hypoglycemia, hepatic failure, or diabetic ketoacidosis.
Patients who are medically stable should be placed in a quiet environment and asked to lie on their side if fatigued in order to minimize the risk of aspiration. When the behavior indicates an increased likelihood of violence, hospital procedures should be followed, including planning for the possibility of a show of force with an intervention team. In the context of aggressiveness, patients should be clearly reminded in a nonthreatening way that it is the goal of the staff to help them to feel better and to avoid problems. If the aggressive behavior continues, relatively low doses of a short-acting benzodiazepine such as lorazepam (e.g., 1 mg by mouth) may be used and can be repeated as needed, but care must be taken so that the addition of this secondCNSdepressant does not destabilize vital signs or worsen confusion. An alternative approach is to use an antipsychotic medication (e.g., 5 mg of haloperidol liquid), but this has the potential danger of lowering the seizure threshold. If aggression escalates, the patient might require a short-term admission to a locked ward, where medications can be used more safely and vital signs more closely monitored. Withdrawal The first, and most important, step is to perform a thorough physical examination in all alcoholics who are considering stopping drinking. It is necessary to evaluate organ systems likely to be impaired, including a search for evidence of liver failure, gastrointestinal bleeding, cardiac arrhythmia, and glucose or electrolyte imbalance. The second step in treating withdrawal for even the typical well-nourished alcoholic is to give patients adequate nutrition and rest. All patients should be given oral multiple B vitamins, including 50 to 100 mg of thiamine daily for a week or more. Most patients enter withdrawal with normal levels of body water or mild overhydration, and intravenous fluids should be avoided unless there is evidence of significant recent bleeding, vomiting, or diarrhea. Medications can usually be administered orally. The third step in treatment is to recognize that most withdrawal symptoms are caused by the rapid removal of aCNSdepressant. Therefore, patients can be weaned by administering any drug of this class and gradually decreasing the levels over 3 to 5 days. While many CNS depressants are effective, the benzodiazepines have the highest margin of safety and are, therefore, the preferred class of drugs in the treatment of alcohol withdrawal. Benzodiazepines with short half-lives (Chap. 385) are especially useful for patients with serious liver impairment or evidence of preexisting encephalopathy or brain damage. On the other hand, short-half-life benzodiazepines, e.g., oxazepam or lorazepam, result in rapidly changing drug blood levels and must be given every 4 h to avoid abrupt fluctuations in blood levels that may increase the risk for seizures. Therefore, most clinicians use drugs with longer half-lives, such as diazepam or chlordiazepoxide. The goal is to administer enough drug on day 1 to alleviate most of the symptoms of withdrawal (e.g., the tremor and elevated pulse), and then to decrease the dose by 20% on successive days over a period of 3 to 5 days. The approach is flexible; the dose is increased if signs of withdrawal escalate, and the medication is withheld if the patient is sleeping or shows signs of increasing orthostatic hypotension. The average patient requires 25 to 50 mg of chlordiazepoxide or 10 mg of diazepam given orally every 4 to 6 h on the first day. For the patient withDTs, treatment can be difficult and the condition is likely to run a course of 3 to 5 days regardless of the therapy employed. The focus of care is to
identify medical problems and correct them and to control behavior and prevent injuries. Many clinicians recommend the use of high doses of benzodiazepine (doses as high as 800 mg/day of chlordiazepoxide have been reported), a treatment that will decrease the agitation and raise the seizure threshold but probably does little to improve the confusion. Other clinicians recommend the use of antipsychotic medications, such as 20 mg or more per day of haloperidol, an approach less likely to exacerbate confusion but which may increase the risk of seizures. Antipsychotic drugs have no place in the treatment of mild withdrawal symptoms. Generalized wtihdrawal seizures rarely require aggressive pharmacologic intervention beyond that given to the usual patient undergoing withdrawal, i.e., adequate doses of benzodiazepines. There is little evidence that anticonvulsants such as phenytoin are effective in drug-withdrawal seizures, and the risk of seizures has usually passed by the time effective drug levels are reached.*The rare patient with status epilepticus must be treated aggressively, as outlined in Chap. 361, initially with intravenous lorazepam. While alcohol withdrawal is often treated in a hospital, efforts at reducing costs have resulted in the development of outpatient detoxification for relatively mild abstinence syndromes. This is appropriate for patients in good physical condition who demonstrate mild signs of withdrawal despite low blood alcohol concentrations and for those without prior history ofDTs or withdrawal seizures. Such individuals still require a careful physical examination, evaluation of blood tests, and vitamin supplementation. Benzodiazepines can be given in a 1- to 2-day supply to be administered to the patient by a spouse or other family member four times a day. Patients are asked to return daily for evaluation of vital signs and to come to the emergency room if signs and symptoms of withdrawal escalate. Rehabilitation of Alcoholics After completing alcoholic rehabilitation, 60% or more of middle-class alcoholics maintain abstinence for at least a year, and many for a lifetime. As is true for any long-term disorder for which treatment requires changes in life-style (e.g., diabetes or hypertension), therapeutic approaches include general supports that meet commonsense guidelines. Considering the lack of evidence for the superiority of any specific treatment type, it is best to keep interventions simple. Maneuvers in rehabilitation fall into two general categories. First are attempts to help the alcoholic achieve and maintain a high level of motivation toward abstinence. These include education about alcoholism and instructing family and/or friends to stop protecting the person from the problems caused by alcohol. The second step is to help the patient to readjust to life without alcohol and to reestablish a functional lifestyle through counseling, vocational rehabilitation, and self-help groups such as Alcoholics Anonymous. The third component, called relapse prevention, helps the person to identify situations in which a return to drinking is likely, formulate ways of managing these risks, and develop coping strategies that increase the chances of a return to abstinence if a slip occurs. There is no convincing evidence that inpatient rehabilitation is always more effective for the average alcoholic than is outpatient care. However, more intense interventions work better than those that are less intensive, and some alcoholics do not respond to outpatient care. The decision to hospitalize can be made if (1) the patient has medical
problems that are difficult to treat outside a hospital; (2) depression, confusion, or psychosis interferes with outpatient care; (3) the patient has such a severe life crisis that it is difficult to get his or her attention as an outpatient; (4) outpatient treatment has failed; or (5) the patient lives far from the treatment center. In any setting, the best predictors of continued abstinence include evidence of higher levels of life stability (e.g., supportive family and friends) and higher levels of functioning (e.g., job skills, higher levels of education, and absence of crimes unrelated to alcohol). Whether the treatment begins in an inpatient or an outpatient setting, subsequent outpatient contact should be maintained for a minimum of 6 months and preferably a full year after abstinence is achieved. Counseling with an individual physician or through groups focuses on day-to-day living -- emphasizing areas of improved functioning in the absence of alcohol (i.e., why it is a good idea to continue to abstain) and helping the patient to manage free time without alcohol, develop a nondrinking peer group, and handle stresses on the job without alcohol. The physician serves an important role in identifying the alcoholic, treating associated medical or psychiatric syndromes, overseeing detoxification, referring the patient to rehabilitation programs, and providing counseling. The physician is also responsible for selecting which (if any) medication might be appropriate during alcoholism rehabilitation. Patients often complain of continuing sleep problems or anxiety when acute withdrawal treatment is over, problems that may be a component of protracted withdrawal. Unfortunately, there is no place for hypnotics or antianxiety drugs in the treatment of most alcoholics after acute withdrawal has been completed. Regarding insomnia, patients should be reassured that the trouble in sleeping is normal after alcohol withdrawal and will improve over the subsequent weeks and months. They should then follow a rigid bedtime and awakening schedule and avoid any naps or the use of caffeine in the evenings. The sleep pattern will improve rapidly. Anxiety can be approached by helping the person to gain insight into the temporary nature of the symptoms and to develop strategies to achieve relaxation as well as using forms of cognitive therapy. In addition, while the mainstay of alcoholic rehabilitation involves counseling, education, and cognitive techniques, several interesting medications are under active evaluation and might prove to be useful. The first is the opioid-antagonist drug naltrexone, which has been reported in several small-scale, short-term studies to decrease the probability of a return to drinking and to shorten periods of relapse. While this medication looks promising, longer-term large-scale trials in more diverse clinical settings will be required before the cost-effectiveness of naltrexone can be established. A second medication, acamprosate, has been tested in over 5000 patients in Europe, with results that appear similar to those reported for naltrexone. Currently, acamprosate is not available in the United States, although a long-term, trial of naltrexone, acamprosate, and their combination is in progress. A third medication, which has historically been used in the treatment of alcoholism, is the ALDH inhibitor disulfiram. Taken in doses of 250 mg/day, this drug produces an unpleasant (and potentially dangerous) reaction in the presence of alcohol, a phenomenon related to rapidly rising blood levels of the first metabolite of alcohol, acetaldehyde. However, few adequate double-blind controlled trials have demonstrated the superiority of disulfiram over placebo. Disulfiram has many side effects, and the reaction with alcohol can be dangerous, especially for patients with
heart disease, stroke, diabetes mellitus, and hypertension. Thus, most clinicians reserve this medication for patients who have a clear history of longer-term abstinence associated with prior use of disulfiram and for those who might take the drug under the supervision of another individual (such as a spouse), especially during discrete periods that they have identified as representing high-risk drinking situations for them (such as the Christmas holiday). More data are required before any medication can be recommended for routine use in alcohol rehabilitation. However, additional support for alcoholics is available through Alcoholics Anonymous in almost every community. Alcoholics Anonymous is a self-help group of recovering alcoholics (men and women who have stopped drinking, perhaps many years ago) that offers an effective model of abstinence, provides a sober peer group, and makes crisis intervention available when the urge to drink escalates. No matter what type of rehabilitation program is planned, the alcoholic should be offered the option of joining Alcoholics Anonymous. (Bibliography omitted in Palm version) Back to Table of Contents
388. OPIOID DRUG ABUSE AND DEPENDENCE - Marc A. Schuckit, David S. Segal The principal effects of the opioids (opiate-like drugs) are a damping of pain perception along with modest levels of sedation and euphoria. Drugs in this category include heroin, morphine, and codeine as well as many prescription analgesics and antitussive agents. Opioid drugs are widely used in medical practice, and, thus, dependence and abuse are not limited to the classic opiod-dependent person on the street. Tolerance to any one opioid is likely to extend to the others (i.e., cross-tolerance is likely), and all opioids are associated with a similar pattern of drug-related problems. Each is capable of producing dependence as defined in the Fourth Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-IV), including evidence of physical dependence, a diagnosis made in the context of a history of tolerance and/or withdrawal. The abstinence syndrome from any of the substances can be treated with administration of any of the others. PHARMACOLOGY The prototypic opiates, morphine and codeine (3-methoxymorphine), are taken directly from the milky juice of the poppy Papaver somniferum. The semisynthetic drugs produced from the morphine or thebane molecules include hydromorphone, diacetylmorphine (heroin), and oxycodone. The purely synthetic opioids, sharing many of the basic properties of opium and morphine, include meperidine, propoxyphene, diphenoxylate, fentanyl, buprenorphine, methadone, and pentazocine. Despite claims to the contrary, all these substances (including almost all prescription analgesics) are capable of producing euphoria as well as psychological and physical dependence when taken in high enough doses over prolonged periods. The opioids produce their effects by binding to different types of opioid receptors throughout the body, including the central nervous system (CNS). Endogenous opioid peptides (i.e., enkephalins, endorphins, dynorphins, and others) have been identified that appear to be natural ligands for opioid receptors. These peptides have a distinct distribution in the CNS. The receptors with which opioid peptides interact are differentially engaged in production of the various opiate effects, such as analgesia, respiratory depression, constipation, and euphoria. Substances capable of antagonizing one or more of these actions include nalorphine, levallorphan, cyclazocine, butorphanol, buprenorphine, and pentazocine, each of which has mixed agonist and antagonist properties, as well as naloxone, nalmefene, and naltrexone, which are pure opiate antagonists. All antagonist drugs (including those with mixed agonist properties), can precipitate withdrawal symptoms if administered to a patient who is physically dependent on other opioids. The availability of relatively specific antagonists has helped identify different receptor subtypes, including u receptors, which influence some of the more classic opioid actions such as pain control, reinforcement, constipation, hormone levels, and respiration;k receptors, with possible similar functions along with sedation and effects on hormones; and dreceptors, thought to relate mostly to analgesia, mood, reinforcement, and breathing. The major features of tolerance, dependence, and withdrawal are thought to be mediated primarily by u receptors. All opioid receptors are coupled to inhibitory G proteins, which mediate their actions within cells (Chap. 386).
Opioid drugs are absorbed from the gastrointestinal system, the lungs, and/or the muscles. The most rapid and pronounced effects occur following intravenous administration, with only slightly less efficient absorption after smoking or inhaling the vapor ("chasing the dragon"), and the least intense actions are seen after absorption from the digestive tract. Most of the metabolism of opioids occurs in the liver, primarily through conjugation with glucuronic acid, and only small amounts are excreted directly in the urine or feces. The plasma half-lives of these drugs range from 2.5 to 3 h for morphine to more than 22 h for methadone and even longer for levomethadyl acetate (LAAM). Street heroin is typically only 5 to 10% pure. The remainder consists of materials such as lactose and fruit sugars, quinine, powdered milk, phenacetin, caffeine, antipyrine, and strychnine, which are used to "cut" the drug and increase the profit margin. Any marked, unexpected increase in the purity of street drugs is likely to cause unintentional lethal overdoses in users expecting less effect from a "hit." THE ACUTE AND CHRONIC EFFECTS OF OPIOID DRUGS With the exception of overdose and physical dependence, most opioid effects are relatively benign and rapidly reversible. A major danger, however, comes through the use of contaminated needles by intravenous users, which increases the risk of hepatitis B and C, bacterial endocarditis, and infection with HIV (Chap. 309). Effects on Body Systems Acute changes in the gastrointestinal system are the result of decreased motility with resulting constipation and anorexia. Chronic gastrointestinal problems in opioid-dependent individuals typically occur as a consequence of hepatitis in injection drug users. Effects of opiods in the CNS include intoxication-induced nausea and vomiting (medulla), decreased pain perception (spinal cord, thalamus, and periaqueductal gray region), euphoria (limbic system), and sedation (reticular activating system). The adulterants added to street drugs may contribute to some of the more permanent nervous system damage, including peripheral neuropathy, amblyopia, myelopathy, and leukoencephalopathy. One study revealed abnormalities in cognitive function and brain computed tomography scans of opioid-dependent subjects; whether these abnormalities are due to the opioid itself, the adulterants, the consequences of dirty needles, or an unhealthy life-style is unknown. Acute opioid administration decreases levels of luteinizing hormone, with a subsequent reduction in testosterone, which might contribute to the decreased sex drive reported by most opioid-dependent people. Other hormonal changes include a decrease in the release of thyrotropin and increases in prolactin and possibly growth hormone (Chap. 328). Acute changes in the respiratory system include respiratory depression, which results from a decreased response of the brainstem to carbon dioxide tension, a component of the drug overdose syndrome described below. At even low drug doses, this effect can be clinically significant in individuals with underlying pulmonary disease. Cardiovascular changes tend to be relatively mild, with no direct opiate effect on heart rhythm or myocardial contractility, but there is a potential problem from orthostatic hypotension, probably secondary to dilation of peripheral vessels. Bacterial infections of the lungs
and heart valves can occur from contaminated needles; the latter can result in emboli and thus an increased risk for stroke. The Toxic Reaction or Overdose Syndrome High doses of opioids can result in a potentially lethal toxic reaction or overdose syndrome. While toxic reactions are seen with all opioids, the more potent drugs such as fentanyl (80 to 100 times more powerful than morphine) are especially dangerous. The typical syndrome, which occurs immediately with intravenous overdose, includes shallow respirations at a rate of two to four per minute, pupillary miosis (with mydriasis once brain anoxia develops), bradycardia, a decrease in body temperature, and a general absence of responsiveness to external stimulation. If this medical emergency is not treated rapidly, respiratory depression, cyanosis, cardiorespiratory arrest, and death can ensue. Postmortem examination reveals few specific changes except for diffuse cerebral edema. An "allergic-like" reaction to intravenous heroin, perhaps in part related to adulterants, can also occur and is characterized by decreased alertness, a frothy pulmonary edema, and an elevation in the blood eosinophil count. The first step in managing overdose is to provide any needed respiratory or cardiovascular support including intubation for airway protection if needed. Definitive treatment for the typical opioid overdose is the administration of a narcotic antagonist such as naloxone in an initial dose of 0.4 mg to 2 mg intravenously, expecting a response in 1 to 2 min. This dose can be repeated every 2 to 3 min up to a dose of 10 mg. With the exception of overdoses with buprenorphene, if no response is seen after 10 mg, it is unlikely that an opioid overdose is responsible for the respiratory depression or coma. If an intravenous line is not available, the drug can be given intramuscularly. It is important to titrate the dose relative to the patient's symptoms. The goal is to ameliorate the respiratory depression but not provoke a severe withdrawal state. Because the effects of this drug diminish within 2 to 3 h, the individual must be monitored for at least 24 h after a heroin overdose and 72 h after an overdose of a longer-acting drug such as methadone. If there is little response to naloxone alone, the possibility of a concomitant overdose with a benzodiazepine should be considered and a challenge with intravenous flumazenil, 0.2 mg/min up to a maximum of 3 mg in an hour, might be used. Patients who are physically dependent on an opioid may experience a precipitous onset of an abstinence syndrome after administration of the opioid antagonist, but aggressive treatment of this syndrome is not appropriate until all vital signs are relatively stable. As with any drug overdose, treatment of either the typical or the "allergic" type of opioid toxic reaction often requires continued supportive care until the drug effect subsides. Patients may require respiratory support (often with oxygen supplementation and positive-pressure breathing for the "allergic" type of overdose), intravenous fluids perhaps accompanied by pressor agents to support blood pressure, and gastric lavage to remove any remaining drug with care taken to use a cuffed endotracheal tube to prevent aspiration if the patient is not alert. It is important to evaluate and treat any possible anaphylactic reactions. Cardiac arrhythmias and/or convulsions, especially likely to be seen with codeine, propoxyphene, or meperidine, also need to be treated. OPIOID ABUSE AND DEPENDENCE
Definition and Epidemiology Repeated opioid use to the point of developing multiple problems is a good indicator that future abuse and dependence are likely.DSM-IVcriteria for opioid dependence are the same as those for alcohol dependence (Chap. 387). An individual is dependent if within a 12-month period repeated difficulties occur in any three areas of functioning, including tolerance, withdrawal, use of greater amounts of opiates than intended, and use despite consequences. Patients who do not have dependence but demonstrate repeated difficulties with the law, impaired ability to meet obligations, use in hazardous situations, or continued use despite problems can be labeled as having abuse. The use of opioids for intoxication is less prevalent than the use of alcohol, marijuana, and several other drugs. A 1997 national survey reported that almost 5% of men and women age 12 or above in the United States had used an opioid for intoxication, including almost 2% in the prior year and slightly less than 1% in the prior month. Focusing specifically on heroin, the lifetime prevalence was approximately 1%, with 0.3% having taken the drug in the prior year. Use patterns of these drugs were almost twice as high in another 1997 survey sampling 12th graders in high school. In all studies, prevalence rates were higher in males than females. None of the national surveys offered data regarding the prevalence of dependence. Genetics While most data on the importance of genetic influences in substance use disorders apply to alcoholism, there are interesting findings regarding other drugs. One large study of over 3000 male twin pairs reported that there are genetic influences that relate uniquely to heroin dependence and also noted additional genetic factors related to an overall vulnerability toward substance-related problems. The genetic influences operate in the context of additional environmental factors that are likely to relate both to the family of upbringing and the general environment. Genetic factors might influence personality characteristics such as impulsivity and sensation-seeking or susceptibility to develop antisocial personality disorder. Genes relating to the actions of the drug on specific neurochemical systems such as dopamine are also potential candidates for an enhanced vulnerability toward developing opioid dependence. Natural History Dependence on or abuse of opioids can be seen in at least three types of patients. First, a minority of people with nonfatal chronic pain syndromes (e.g., back, joint, and muscle disorders) misuse their prescribed drugs. If physical dependence is established, abstinence syndromes can then intensify the pain, promoting continued drug intake. Physicians can avoid contributing to physical dependence by helping the patient to accept the goal of minimization rather than disappearance of the pain and to recognize that discomfort may not be completely eliminated (Chap. 12). Analgesic medication should be only one component of treatment and limited to the oral administration of the least potent analgesic that is able to "take the edge off" the pain (e.g., ibuprofen or, if needed, propoxyphene). Behavior modification techniques, such as muscle relaxation and meditation, and carefully selected exercises should be used as appropriate to help increase function and decrease pain. Finally, nonmedicinal approaches, including electrical transcutaneous neurostimulation for muscle and joint disease, may be useful. The second group at high risk are physicians, nurses, and pharmacists, primarily because of their easy access to substances of abuse. Physicians may begin to use
opioids to help them sleep or to reduce stress or physical aches and pains. This group appears to be at especially high risk for developing dependence on the highly potent drugs such as fentanyl. Because of the growing awareness of these problems, impaired-physician programs have been established in many hospitals and by most state medical societies. Such groups attempt to identify and aid substance-impaired physicians, giving them peer support and education to help them achieve abstinence before problems escalate to the point of licensure revocation. All doctors are advised never to prescribe opioids for themselves or for members of their family -- physicians deserve the same level of care and protection from future problems as their patients. The third and most obvious group are those who buy street drugs to get high. While some of these men and women have prior histories of severe antisocial problems, most have a relatively high level of premorbid functioning. The typical person begins using opioids occasionally, often after experimenting with tobacco, then alcohol, then marijuana, and then brain depressants or stimulants. Occasional opiate use, or "chipping," might continue for some time, and some individuals never escalate their intake to the point of developing dependence. Of course, opiate-dependent individuals are likely to continue to have experience with many other drugs. At least three of these often remain as problems during the course of opioid dependence. First, alcohol is typically used to moderate withdrawal problems, to enhance the opioid high, and as a substitute when the preferred drug is not readily available, including during methadone and other treatments. This pattern of problematic drinking, often meeting criteria for alcohol dependence, is present at some time in approximately half of opioid-dependent persons. The second drug, cocaine, appears to be taken for many of the same reasons as alcohol, and is often administered intravenously with the opioid in a mixture known as a "speedball." The third class of drugs misused in combination with opioids consists of the benzodiazepines, especially among people in methadone maintenance. Once persistent opioid use is established, severe problems are likely to develop. At least 25% die within 10 to 20 years from suicide, homicide, accidents, or infectious diseases such as tuberculosis, hepatitis, or AIDS. The mortality rate has escalated in recent years in response to the AIDS epidemic among injection drug users, with an estimated 60% of these men and women carrying HIV (Chap. 309). At the same time, while the majority of opioid-dependent persons show frequent exacerbations and remissions, it is important to remember that approximately 35% achieve long-term, often permanent, abstinence. This remission is probably most often seen after the age of 40 but can occur at any point in the clinical course. While this favorable outcome can be observed in any opioid-dependent person, as is true with most drugs of abuse a better prognosis is associated with prior histories of marital and employment stability and fewer criminal activities unrelated to drugs. TREATMENT The key to diagnosis is to discard the erroneous stereotype that opioid-dependent men and women are always unemployed and homeless. Abuse or dependence is possible in any patient who demonstrates symptoms of what might be opioid withdrawal; anyone who has a chronic pain syndrome; physicians, nurses, and pharmacists or others with
easy access to opioids; and all patients who repeatedly seek out prescription analgesics. Therefore, it is important to take the time with every patient, especially those with complaints of pain, to gather a history that includes the patterns of opioid use and the list of doctors and clinics from which they have received prescriptions. If the chronic use of opioids is suspected, gathering further data from an additional informant such as a relative or close friend can be essential. Another indicator of an enhanced risk for opioid dependence is a history of pervasive antisocial problems beginning in the preteen years. Blood and urine screens can be used to identify opioids in patients in whom misuse is suspected, and clinicians should search for physical stigmata of misuse (e.g., needle marks). After identifying opioid dependence, the next step is intervention. The need for active treatment of the abstinence syndrome can be presented, and the availability of help in establishing a drug-free life-style can be emphasized. The final decision, of course, rests with the patient. This approach to intervention is presented in relation to alcoholism in Chap. 387. The Symptoms of Withdrawal Withdrawal symptoms, usually the opposite of the acute effects of the drug, include nausea and diarrhea, coughing, lacrimation, mydriasis, rhinorrhea, profuse sweating, twitching muscles, piloerection (or "goose bumps") as well as mild elevations in body temperature, respiratory rate, and blood pressure. In addition, diffuse body pain, insomnia, and yawning occur, along with intense drug craving. Drugs with a short half-life, such as morphine or heroin, cause symptoms typically within 8 to 16 h of the last dose (thus, many dependent individuals awake in mild withdrawal every morning); symptom intensity peaks within 36 to 72 h after discontinuation of the drug, and the acute syndrome disappears within 5 to 8 days. However, a protracted abstinence phase of mild symptoms (e.g., moodiness, slight changes in pupillary size, autonomic dysfunction, changes in sleep pattern) may persist for 6 or more months. These lingering symptoms, which can be relieved by administering an opioid, probably contribute to relapse. Treatment of the Withdrawal Syndrome A thorough physical examination, including an assessment of neurologic function and a search for local and systemic infections, especially abscesses, is mandatory. Laboratory testing generally includes assessment of liver function and, in intravenous users, HIV status. Proper nutrition and rest must be initiated as soon as possible. Optimal treatment of withdrawal requires administration of sufficient opioid medication on day 1 to decrease symptoms, followed by a more gradual withdrawal of the drug, usually over 5 to 10 days. Any opioid will work (all have some level of cross-tolerance), but for ease of administration, many physicians prefer to use a long-acting drug such as methadone. To estimate the first day's dose from the patient's history, 1 to 2 mg of methadone can be considered approximately equivalent to 3 mg of morphine, 1 mg of heroin, or 20 mg of meperidine. Most patients require between 10 and 25 mg of methadone orally twice on day 1, with higher doses given if prominent symptoms of withdrawal are not dampened. After several days of a stabilized drug dose, the opioid is then decreased by 10 to 20% of the original day's dose each day. However, most states restrict the prescription of opioids to dependent persons, and, in
the absence of special permits, detoxification with opioids is often proscribed or limited. Thus, pharmacologic treatments often center on relief of symptoms of diarrhea with Imodium or a nonopioid drug, of "sniffles" with decongestants, and pain with nonopioid analgesics (e.g., ibuprofen). Comfort can be enhanced with thea2-adrenergic agonist clonidine to decrease sympathetic nervous system overactivity. Given at doses of approximately 5 ug/kg (up to 0.3 mg given two to four times a day), clonidine decreases autonomic nervous system dysfunction and produces sedation. Blood pressure should be monitored closely. Some clinicians augment this regimen with low to moderate doses of benzodiazepines for 2 to 5 days to decrease agitation. A special case of opioid withdrawal is seen in the newborn made passively dependent through the mother's drug abuse during pregnancy. Some level of withdrawal develops in 50 to 90% of children of heroin-dependent mothers. As few as 25% of infants of methadone-maintenance mothers show clinically relevant withdrawal symptoms, probably because of the longer half-life of this drug. The syndrome consists of irritability, crying, a tremor (in 80%), increased reflexes, increased respiratory rate, diarrhea, hyperactivity (in 60%), vomiting (40%), and sneezing/yawning/hiccuping (in 30%). The child usually has a low birth weight but may be otherwise unremarkable until the second day, when symptoms are likely to begin. The treatment follows the same general steps used in the treatment of the physically dependent adult. The child must be carefully evaluated to rule out medical problems such as hypoglycemia, hypocalcemia, infections, and trauma; general support in a warm, quiet environment and regulation of electrolytes and glucose are also required. The infant with moderate to severe symptoms can be treated with any of the following: paregoric (0.2 mL orally every 3 to 4 h), methadone (0.1 to 0.5 mg/kg per day), phenobarbital (8 mg/kg per day), or diazepam (1 to 2 mg/kg every 8 h). Medication should be given in decreasing levels for 10 to 20 days. Dependent infants of mothers on methadone maintenance also benefit by breast feeding while the mother continues to take methadone. Rehabilitation of Opioid-Dependent Persons Despite some differences in demographics, the same general rules for rehabilitation apply to opioid-dependent persons as to alcoholics. The basic strategy includes detoxification and family support, and the process can benefit from the use of reading materials or referral to self-help groups. It is also important to establish realistic patient goals and a program of counseling and education to increase motivation toward abstinence. A long-term commitment to rebuilding a life-style without the substance is essential for preventing recidivism. Most rehabilitation approaches have common elements, regardless of the drug involved. Patients are educated about their responsibility for improving their lives, and motivation for abstinence is increased by providing information about the medical and psychological problems that can be expected if dependence continues. Patients and families are encouraged to establish an opioid-free life-style by learning to cope with chronic pain and develop realistic vocational planning (e.g., for pharmacists, physicians, and nurses). The dependent person is also encouraged to establish a drug-free peer group and to participate in self-help groups such as Narcotics Anonymous. Another important treatment component is relapse prevention aimed at identifying triggers for a
return to drugs and developing appropriate coping strategies. Much of this advice and counseling can be given by the physician, but many clinicians refer patients to more formal drug programs, including methadone maintenance clinics, programs using narcotic antagonists, and therapeutic communities. Long-term follow-up of treated patients indicates that approximately one-third were completely drug free in the previous year; 60% were no longer using opioids, although some were misusing other substances. Individuals who stay in methadone maintenance or in therapeutic communities show significant improvement in antisocial behavior and employment status. In general, the best prognosis is for those individuals who are employed, who have higher levels of education, and who remain in treatment for at least 2 months. Dependence among health care providers, such as physicians, is treated similarly, but in addition a closely supervised "diversion" procedure is usually instituted and carried out for 1 to 2 years or more. Methadone Maintenance Maintenance programs with methadone and the even longer-acting agentLAAMshould only be used in combination with education and counseling. It is important to note that drug maintenance is not aimed at "curing" opioid dependence; rather, it provides a substitute drug that is legally accessible, safer, can be taken orally, and has a long half-life so that it can be taken once a day. The goal is to help persons who have repeatedly failed in drug-free programs to improve functioning within the family and job, to decrease legal problems, and to improve health. Methadone is a long-acting opioid that possesses almost all the physiologic properties of heroin. The recipient, who has been carefully screened to rule out prior psychiatric disorders, may be maintained on a relatively low dose (e.g., 30 to 40 mg/d); a better approach is to use a higher dose (80 to 120 mg/d), because it may be more effective in blocking heroin-induced euphoria and decreasing craving. There is some evidence that the higher methadone doses result in greater retention in treatment and consequently in lower levels of arrest and relapse to street drugs. Three-quarters or more of patients, especially those receiving the higher doses, are likely to remain heroin-free for 6 months or longer. Methadone is administered as an oral liquid given once a day at the program, with weekend doses taken by the patient at home. The longer-acting analogues, such asLAAM, can be given two or three times a week, with the dose of LAAM increased to as high as 80 mg three times a week if needed. After a period of maintenance (usually 6 months to 1 year or longer), the clinician can work closely with the patient to regulate the rate of drug decrease (by about 5% per week) if possible. In the past, the British have used heroin maintenance with goals and guidelines similar to those of current methadone programs. There is no evidence that heroin maintenance has any advantages over methadone maintenance, but the heroin approach does add the risk that the drug will be easily sold on the streets. Treatment with mixed agonists-antagonists such as buprenorphine also appears beneficial, although results are not as good as with methadone. Opioid Antagonists The opiate antagonists (e.g., naloxone) compete with heroin and other opioids for receptors, reducing the effects of the opioid agonists. Administered over long periods with the intention of blocking the "high" produced if the patient takes opioids, these drugs can be useful as part of an overall treatment approach that
includes counseling and support. The most widely used antagonist in rehabilitation is naltrexone; 50 mg per day antagonizes 15 mg of heroin for 24 h, and higher doses (125 to 150 mg) block the effects of 25 mg of intravenous heroin for up to 3 days. Naltrexone is free of agonist properties, produces no known withdrawal symptoms when stopped, and its side effects tend to be mild. To avoid precipitating a withdrawal syndrome, patients should be free of opioids for a minimum of 5 days before beginning treatment with this medication. In addition, they should first be challenged with 0.4 or 0.8 mg of the shorter-acting agent naloxone to be certain they are able to tolerate the long-acting antagonist. Following this procedure, a test dose of 10 mg of naltrexone can be given, with the expectation that any withdrawal symptoms will be seen in 0.5 to 2 h. Several variations of this approach can be used with detoxification from methadone maintenance, including a fairly rapid, medically supervised plan. Over a 10-day period, the daily dose should be increased to about 100 mg on Mondays and Wednesdays and 150 mg on Fridays. Unfortunately, despite the apparent advantages of this treatment approach, some patients are resistant to continuing care. In one study, only about 60% of the patients completed 6 days of naltrexone induction, and only 10% remained in the program at the end of 6 months. However, another study reported much higher rates of compliance, with almost a third achieving continuous abstinence for at least a year. Drug-Free Programs Most existing halfway houses and recovery centers for opioid-dependent persons use some variant of the therapeutic community approach. This is an exception to the general preference for short-term residential (as opposed to outpatient) rehabilitation, since care can last a year or more while the person is taken out of the street culture and given a new life within the group. In this structure, members, including leaders who are themselves in the process of recovery, help participants gain insights into more successful strategies for coping with problems. As is true for treatments of all substance-use disorders, it is likely that counseling, behavioral treatments, and relatively simple approaches to psychotherapy add significantly to a positive outcome. Most approaches focus on teaching participants to cope with stress, enhancing their understanding of personality attributes, teaching better cognitive styles, and, through the process of relapse prevention, addressing issues that might contribute to increased craving, easy access to drugs, or periods of decreased motivation. A combination of these therapies with the approaches described above appears to give the best results. Finally, it is important to discuss prevention. Except for the terminally ill, physicians should carefully monitor opioid drug use in their patients, keeping doses as low as is practical and administering them over as short a period as the level of pain would warrant in the average person. Physicians must be vigilant regarding their own risk for opioid abuse and dependence, never prescribing these drugs for themselves. For the nonmedical intravenous drug-dependent person, all possible efforts must be made to prevent AIDS, hepatitis, bacterial endocarditis, and other consequences of contaminated needles both through methadone maintenance and by considering needle-exchange programs. (Bibliography omitted in Palm version)
Back to Table of Contents
389. COCAINE AND OTHER COMMONLY ABUSED DRUGS - Jack H. Mendelson, Nancy K. Mello The abuse of cocaine and other psychostimulant drugs appears to be increasing in many metropolitan and rural areas throughout the world, according to a year 2000 report by the National Institute on Drug Abuse (NIDA). The number of deaths associated with these drugs has also increased. In several urban areas of the United States, use of these drugs has increased more sharply among women than men. Although enhanced legal enforcement as well as educational prevention procedures have attenuated, in part, the increase in psychostimulant abuse among youths in the United States, there appears to be an enhanced worldwide risk for psychostimulant abuse and dependence. The initiation and continuation of drug abuse are determined by a complex interaction of the pharmacologic properties and relative availability of each drug, the personality and expectations of the user, and the environmental context in which the drug is used. Polydrug abuse, the concurrent use of several drugs with different pharmacologic effects, is increasingly common among individuals from all socioeconomic strata. There has been an alarming increase in particularly dangerous forms of polydrug abuse, such as the combined use of heroin and cocaine intravenously. There is no simple explanation for this change in polydrug use patterns. Drug abusers may attempt to attenuate one drug effect with another, as when heroin or alcohol is used to modulate the cocaine high. Sometimes one drug is used to enhance the effects of another, as with benzodiazepines and methadone, or cocaine plus heroin in methadone-maintained patients. Chronic cocaine and psychostimulant abuse may cause a number of adverse health consequences, ranging from pulmonary disease to reproductive dysfunction. Preexisting disorders such as hypertension and cardiac disease may be exacerbated by drug abuse, and the combined use of two or more drugs may accentuate medical complications associated with abuse of one of them. The adverse health consequences of drug abuse are further complicated by AIDS. Drug abuse increases the risk of exposure to HIV. Cocaine and psychostimulant abuse contribute to the risk for HIV infection in part by the adverse immunomodulatory effects of these drugs. In addition, concurrent use of cocaine and opiates (the "speedball") is frequently associated with needle-sharing by intravenous drug users. These individuals continue to represent the largest single group of persons with HIV infection in several major metropolitan areas in the United States as well as in urban areas in Scotland, Italy, Spain, Thailand, and China. COCAINE Cocaine is a stimulant and local anesthetic with potent vasoconstrictor properties. The leaves of the coca plant (Erythroxylon coca) contain approximately 0.5 to l% cocaine. The drug produces physiologic and behavioral effects when administered orally, intranasally (snorting), intravenously, or via inhalation following pyrolysis (smoking). Cocaine increases synaptic concentrations of the monoamine neurotransmitters dopamine, norepinephrine, and serotonin by binding to transporter proteins in presynaptic neurons and blocking reuptake. The reinforcing effects of cocaine are
related to effects on dopaminergic neurons in the mesolimbic system (Chap. 386). Prevalence of Cocaine Use Cocaine has become widely available throughout the United States, and cocaine abuse occurs in virtually all social and economic strata of society. The prevalence of cocaine abuse in the general population has been accompanied by an increase in cocaine abuse by heroin-dependent persons, including those in methadone maintenance programs. Intravenous cocaine is often used concurrently with intravenous heroin -- a combination that purportedly attenuates the postcocaine "crash" and substitutes a cocaine "high" for the heroin "high" blocked by methadone. Acute and Chronic Cocaine Intoxication There has been an increase in both intravenous administration and inhalation of pyrolyzed cocaine via smoking. Following intranasal administration, changes in mood and sensation are perceived within 3 to 5 min, and peak effects occur at l0 to 20 min. The effects rarely last>1 h. Inhalation of pyrolyzed materials includes inhaling crack/cocaine or smoking coca paste, a product made by extracting cocaine preparations with flammable solvents, and cocaine free-base smoking. Free-base cocaine, including the free base prepared with sodium bicarbonate (crack), is becoming increasingly popular because of the relative high potency of the compound and its rapid onset of action (8 to 10 s following smoking). Cocaine produces a brief, dose-related stimulation and enhancement of mood and an increase in cardiac rate and blood pressure. Body temperature usually increases, and high doses of cocaine may induce lethal pyrexia or hypertension. Because cocaine inhibits reuptake of catecholamines at adrenergic nerve endings, the drug potentiates sympathetic nervous system activity. Cocaine has a short plasma half-life of approximately 45 to 60 min. Cocaine is metabolized primarily by plasma esterases, and cocaine metabolites are excreted in urine. The very short duration of euphorigenic effects of cocaine observed in chronic abusers is probably due to both acute and chronic tolerance. Frequent self-administration of the drug (two to three times per hour) is often reported by chronic cocaine abusers. Alcohol is used to modulate both the cocaine high and the dysphoria associated with the abrupt disappearance of cocaine's effects. A metabolite of cocaine, cocaethylene, has been detected in blood and urine of persons who concurrently abuse alcohol and cocaine. Cocaethylene induces changes in cardiovascular function similar to those of cocaine alone, and the pathophysiologic consequences of alcohol abuse plus cocaine abuse may be additive when both are used together. The prevalent assumption that cocaine inhalation or intravenous administration is relatively safe is contradicted by reports of death from respiratory depression, cardiac arrhythmias, and convulsions associated with cocaine use. In addition to generalized seizures, neurologic complications may include headache, ischemic or hemorrhagic stroke, or subarachnoid hemorrhage. Disorders of cerebral blood flow and perfusion in cocaine-dependent persons have been detected with magnetic resonance spectroscopy (MRS) studies. Severe pulmonary disease may develop in individuals who inhale crack cocaine; this effect is attributed both to the direct effects of cocaine and to residual contaminants in the smoked material. Hepatic necrosis has been reported to occur following crack cocaine use.
Although men and women who abuse cocaine may report that the drug enhances libidinal drive, chronic cocaine use causes significant loss of libido and adversely affects reproductive function. Impotence and gynecomastia have been observed in male cocaine abusers, and these abnormalities often persist for long periods following cessation of drug use. Women who abuse cocaine have reported major derangements in menstrual cycle function including galactorrhea, amenorrhea, and infertility. Chronic cocaine abuse may cause persistent hyperprolactinemia as a consequence of disordered dopaminergic inhibition of prolactin secretion by the pituitary. Cocaine abuse by pregnant women, particularly the smoking of crack, has been associated with both an increased risk of congenital malformations in the fetus and perinatal cardiovascular and cerebrovascular disease in the mother. However, cocaine abuse per se is probably not the sole cause of these perinatal disorders, since many problems associated with maternal cocaine abuse, including poor nutrition and health care status as well as polydrug abuse, also contribute to risk for perinatal disease. Protracted cocaine abuse may cause paranoid ideation and visual and auditory hallucinations, a state that resembles alcoholic hallucinosis. Psychological dependence on cocaine, as manifested by inability to abstain from frequent compulsive use, has also been reported. Although the occurrence of withdrawal syndromes involving psychomotor agitation and autonomic hyperactivity remains controversial, severe depression ("crashing") following cocaine intoxication may accompany drug withdrawal. TREATMENT Treatment of cocaine overdose is a medical emergency that is often best managed in an intensive care unit. Cocaine toxicity produces a hyperadrenergic state characterized by hypertension, tachycardia, tonic-clonic seizures, dyspnea, and ventricular arrhythmias. Intravenous diazepam in doses up to 0.5 mg/kg administered over an 8-h period has been shown to be effective for control of seizures. Ventricular arrhythmias have been managed successfully by administration of 0.5 to 1.0 mg of propranolol intravenously. Since many instances of cocaine-related mortality have been associated with concurrent use of other illicit drugs (particularly heroin), the physician must be prepared to institute effective emergency treatment for multiple drug toxicities. Treatment of chronic cocaine abuse requires combined efforts by primary care physicians, psychiatrists, and psychosocial care providers. Early abstinence from cocaine use is often complicated by symptoms of depression and guilt, insomnia, and anorexia, which may be as severe as those observed in major affective disorders. Individual and group psychotherapy, family therapy, and peer group assistance programs are often useful for inducing prolonged remission from drug use. A number of medications used for the treatment of various psychiatric disorders have been administered to reduce the duration and severity of cocaine abuse and dependence. However, no available medication is both safe and highly effective for either cocaine detoxification or maintenance of abstinence. Some psychotherapeutic interventions are occasionally effective; however, no specific form of psychotherapy or behavioral modification is uniquely beneficial. MARIJUANA AND CANNABIS COMPOUNDS
Cannabis sativa contains>400 compounds in addition to the psychoactive substance, delta-9-tetrahydrocannabinol (THC). Marijuana cigarettes are prepared from the leaves and flowering tops of the plant, and a typical marijuana cigarette contains 0.5 to 1 g of plant material. Although the usual THC concentration varies between 10 and 40 mg, concentrations>100 mg per cigarette have been detected. Hashish is prepared from concentrated resin of C. sativa and contains a THC concentration of between 8 to 12% by weight. "Hash oil," a lipid-soluble plant extract, may contain a THC concentration of 25 to 60% and may be added to marijuana or hashish to enhance its THC concentration. Smoking is the most common mode of marijuana or hashish use. During pyrolysis,>150 compounds in addition to THC are released in the smoke. Although most of these compounds do not have psychoactive properties, they do have potential physiologic effects. THCis quickly absorbed from the lungs into blood and is then rapidly sequestered in tissues. It is metabolized primarily in the liver, where it is converted to 11-hydroxy-THC, a psychoactive compound, and >20 other metabolites. Many THC metabolites are excreted through the feces at a rate of clearance that is relatively slow in comparison to that of most other psychoactive drugs. Specific cannabinoid receptors (CB1 and CB2) have been identified in the central nervous system, including the spinal cord, and in the peripheral nervous system. High densities of these receptors have been found in the cerebral cortex, basal ganglia, and hippocampus. B lymphocytes also appear to have cannabinoid receptors. A naturally occurringTHC-like ligand has been identified in the nervous system, where it is widely distributed. Prevalence of Marijuana Use Marijuana is the most commonly used illegal drug in the United States. Use is particularly prevalent among adolescents; studies suggest that ~40% of high school students in the United States have used marijuana. Marijuana is relatively inexpensive and is considered by many persons to be less hazardous than the use of other controlled drugs and substances. Very potent forms of marijuana (sinsemilla) are now available in many communities, and concurrent use of marijuana with crack/cocaine and phencyclidine is increasing. Marijuana abuse by individuals from all social strata has been increasing. Acute and Chronic Marijuana Intoxication Acute intoxication from marijuana and cannabis compounds is related to both the dose ofTHC and the route of administration. THC is absorbed more rapidly from marijuana smoking than from orally ingested cannabis compounds. Acute marijuana intoxication usually consists of a subjective perception of relaxation and mild euphoria resembling mild to moderate alcohol intoxication. This condition is usually accompanied by some impairment in thinking, concentration, and perceptual and psychomotor function. Higher doses of cannabis may produce behavioral effects analogous to severe alcohol intoxication. Although the effects of acute marijuana intoxication are relatively benign in normal users, the drug can precipitate severe emotional disorders in individuals who have antecedent psychotic or neurotic problems. As with other psychoactive compounds, both set (user's expectations) and setting (environmental context) are important determinants of the type and severity of behavioral intoxication.
As is true of alcoholics, chronic marijuana abusers may lose interest in common socially desirable goals and steadily devote more time to drug acquisition and use. However,THC does not cause a specific and unique "amotivational syndrome." The range of symptoms sometimes attributed to marijuana use is difficult to distinguish from mild to moderate depression and the maturational dysfunctions often associated with protracted adolescence. Chronic marijuana use has also been reported to increase the risk of psychotic symptoms in individuals with a past history of schizophrenia. Physical Effects of Marijuana Conjunctival injection and tachycardia are the most frequent immediate physical concomitants of smoking marijuana. Tolerance for marijuana-induced tachycardia develops rapidly among regular users; angina may be precipitated by marijuana smoking in persons with a history of coronary insufficiency. Exercise-induced angina may be increased after marijuana use to a greater extent than after tobacco cigarette smoking. Patients with cardiac disease should be strongly advised not to use cannabis compounds. Significant decrements in pulmonary vital capacity have been found in regular daily marijuana smokers. Because marijuana smoking typically involves deep inhalation and prolonged retention of marijuana smoke, marijuana smokers may develop chronic bronchial irritation. Impairment of single-breath carbon monoxide diffusion capacity (DLco) is greater in persons who smoke both marijuana and tobacco than in tobacco smokers. Despite the well-documented association between tobacco smoking and lung cancer, at present there is no direct evidence that marijuana smoking induces lung cancer. However, heavy marijuana use among Americans may be too recent to permit detection of this problem. Although marijuana has also been associated with adverse effects on a number of other systems, many of these studies await replication and confirmation. A reported correlation between marijuana use and decreased testosterone levels in males has not been confirmed. Decreased sperm count and sperm motility and morphologic abnormalities of spermatozoa following marijuana use have also been reported. Administration of high doses of marijuana to female rhesus monkeys suppresses pituitary gonadotropins and gonadal steroids. Prospective studies demonstrated a correlation between impaired fetal growth and development and heavy marijuana use during pregnancy. Marijuana has also been implicated in derangements of the immune system; in chromosomal abnormalities; and in inhibition of DNA, RNA, and protein synthesis; however, these findings have not been confirmed or related to any specific physiologic effect in humans. Tolerance and Physical Dependence Habitual marijuana users rapidly develop tolerance to the psychoactive effects of marijuana and often smoke more frequently and try to secure more potent cannabis compounds. Tolerance for the physiologic effects of marijuana develops at different rates; e.g., tolerance develops rapidly for marijuana-induced tachycardia but more slowly for marijuana-induced conjunctival injection. Tolerance to both behavioral and physiologic effects of marijuana decreases rapidly upon cessation of marijuana use. Withdrawal signs and symptoms have been reported in chronic cannabis users, with the severity of symptoms related to dosage and duration of use. These include tremor,
nystagmus, sweating, nausea, vomiting, diarrhea, irritability, anorexia, and sleep disturbances. Withdrawal signs and symptoms observed in chronic marijuana users are usually relatively mild in comparison to those observed in heavy opiate or alcohol users and rarely require medical or pharmacologic intervention. More severe and protracted abstinence syndromes may occur after sustained use of high potency cannabis compounds. Therapeutic Use of Marijuana Marijuana, administered as cigarettes or as a synthetic oral cannabinoid (dronabinol), has been proposed to have a number of properties that may be clinically useful in some situations. These include antiemetic effects in chemotherapy recipients, appetite-promoting effects in AIDS, reduction of intraocular pressure in glaucoma, and reduction of spasticity in multiple sclerosis and other neurologic disorders. With the possible exception of AIDS-related cachexia, none of these attributes of marijuana compounds is clearly superior to other readily available therapies. Furthermore, any therapeutic benefit of marijuana must be balanced against the many unhealthy psychoactive effects associated with its use. METHAMPHETAMINE The abuse of methamphetamine, also referred to as "meth," "speed," "crank," "chalk," "ice," "glass," or "crystal," has been declining in many metropolitan areas and communities throughout the United States. This decrease is attributed in part to drug seizures and the closures of clandestine laboratories that produce methamphetamine illegally. Prevention programs focusing upon methamphetamine abuse have also increased. Most persons who abuse amphetamine self-administer the drug orally, although there have been reports of methamphetamine administration by inhalation and intravenous injection. Individuals who abuse or become dependent upon methamphetamine state that use of this drug induces feelings of euphoria and decreases fatigue associated with aversive life situations. Adverse physiologic effects observed as a consequence of methamphetamine abuse include headache, difficulty concentrating, diminished appetite, abdominal pain, vomiting or diarrhea, disordered sleep, paranoid or aggressive behavior, and psychosis. Severe, life-threatening toxicity may present as hypertension, cardiac arrythmia or failure, subarachnoid hemorrhage, ischemic stroke, intracerebral hemorrhage, convulsions, or coma. Amphetamines increase the release of monoamine neurotransmitters (dopamine, norepinephrine, and serotonin) from presynaptic neurons. It is thought that the euphoric and reinforcing effects of this class of drugs are mediated through dopamine and the mesolimbic system, whereas the cardiovascular effects are related to norepinephrine. Magnetic resonance spectroscopy studies suggest that chronic abuse may injure the frontal areas and basal ganglia of the brain. Therapy of acute methamphetamine overdose is largely symptomatic. Ammonium chloride may be useful to acidify the urine and enhance clearance of the drug. Hypertension may respond to sodium nitroprusside or a-adrenergic antagonists. Sedatives may reduce agitation and other signs of central nervous system overactivity. Treatment of chronic methamphetamine dependence may be accomplished in either an inpatient or outpatient setting using strategies similar to those described above for cocaine abuse.
MDMA (3,4-methylenedioxymethamphetamine), or Ecstasy, is a derivative of methamphetamine. Ecstasy is usually taken orally but may be injected or inhaled. In addition to amphetamine-like effects, MDMA can induce vivid hallucinations and other perceptual distortions. These toxicities are similar to those of lysergic acid diethylamide (LSD) and may be mediated through the release of serotonin. LYSERGIC ACID DIETHYLAMIDE The discovery of the psychedelic effects ofLSD in 1947 led to an epidemic of LSD abuse during the 1960s. Imposition of stringent constraints on the manufacture and distribution of LSD (classified as a Schedule I substance by the U.S. Food and Drug Administration), as well as public recognition that psychedelic experiences induced by LSD were a health hazard, have resulted in a reduction in LSD abuse. The drug still retains some popularity among adolescents and young adults, however, and there are indications that LSD use among young persons has been increasing in some communities in the United States. LSDis a very potent drug; oral doses as low as 20 ug may induce profound psychological and physiologic effects. Tachycardia, hypertension, pupillary dilation, tremor, and hyperpyrexia occur within minutes following oral administration of 0.5 to 2 ug/kg. A variety of bizarre and often conflicting perceptual and mood changes, including visual illusions, synesthesias, and extreme lability of mood, usually occur within 30 min after LSD intake. The action of LSD may persist for 12 to 18 h, even though the half-life of the drug is only 3 h. Tolerance develops rapidly forLSD-induced changes in psychological function when the drug is used one or more times per day for 4 or more days. Abrupt abstinence following continued use does not produce withdrawal signs or symptoms. There have been no clinical reports of death caused by the direct effects of LSD. The most frequent medical emergency associated withLSD use is panic episode (the "bad trip"), which may persist up to 24 h. Management of this problem is best accomplished by supportive reassurance ("talking down") and, if necessary, administration of small doses of anxiolytic drugs. Adverse consequences of chronic LSD use include risk for schizophreniform psychosis and derangements in memory function, problem solving, and abstract thinking. Treatment of these disorders is best carried out in specialized psychiatric facilities. PHENCYCLIDINE Phencyclidine (PCP), a cyclohexylamine derivative, is widely used in veterinary medicine to briefly immobilize large animals and is sometimes described as a dissociative anesthetic. PCP binds to ionotropic n-methyl-d-aspartate (NMDA) receptors in the nervous system, blocking ion current through these channels. PCP is easily synthesized; its abusers are primarily young people and polydrug users. It is used orally, by smoking, or by intravenous injection. It is also used as an adulterant inTHC,LSD, amphetamine, or cocaine. The most common street preparation, angel dust, is a white granular powder that contains 50 to 100% of the drug. Low doses (5 mg) produce
agitation, excitement, impaired motor coordination, dysarthria, and analgesia. Users may have horizontal or vertical nystagmus, flushing, diaphoresis, and hyperacusis. Behavioral changes include distortions of body image, disorganization of thinking, and feelings of estrangement. Higher doses of PCP (5 to 10 mg) may produce hypersalivation, vomiting, myoclonus, fever, stupor, or coma. PCP doses of ³10 mg cause convulsions, opisthotonus, and decerebrate posturing, which may be followed by prolonged coma. The diagnosis ofPCPoverdose is difficult because the patient's initial symptoms may suggest an acute schizophrenic reaction. Confirmation of PCP use is possible by determination of PCP levels in serum or urine; PCP assays are available at most toxicologic centers. PCP remains in urine for 1 to 5 days following high-dose intake. PCPoverdose requires life-support measures, including treatment of coma, convulsions, and respiratory depression in an intensive care unit. There is no specific antidote or antagonist for PCP. PCP excretion from the body can be enhanced by gastric lavage and acidification of urine. Death from PCP overdose may occur as a consequence of some combination of pharyngeal hypersecretion, hyperthermia, respiratory depression, severe hypertension, seizures, hypertensive encephalopathy, and intracerebral hemorrhage. Acute psychosis associated withPCP use should be considered a psychiatric emergency since patients may be at high risk for suicide or extreme violence toward others. Phenothiazines should not be used for treatment because these drugs potentiate PCP's anticholinergic effects. Haloperidol (5 mg intramuscularly) has been administered on an hourly basis to induce suppression of psychotic behavior. PCP, likeLSD and mescaline, produces vasospasm of cerebral arteries at relatively low doses. Chronic PCP use has been shown to induce insomnia, anorexia, severe social and behavioral changes, and, in some cases, chronic schizophrenia. POLYDRUG ABUSE Although drug abusers often report a preference for a particular drug, such as alcohol or opiates, the concurrent use of other drugs is common. Multiple drug use often involves substances that may have different pharmacologic effects from the preferred drug. Concurrent use of dissimilar compounds such as stimulants and opiates or stimulants and alcohol is not unusual. The diversity of reported drug use combinations suggests that achieving some perceptible change in state, rather than any particular direction of change (stimulation or sedation), may be the primary reinforcer in polydrug use and abuse. There is also evidence that intoxication with alcohol or opiates is associated with increased tobacco smoking. There is relatively little systematic information available about multiple drug abuse interactions. However, the combined use of cocaine, heroin, and alcohol increases the risk for toxic effects and adverse medical consequences over risks associated with use of a single drug. One determinant of polydrug use patterns is the relative availability and cost of the drugs. There are many examples of situationally determined drug-use patterns. For example, alcohol abuse, with its attendant medical complications, is one of the most serious problems encountered in former heroin addicts participating in methadone maintenance programs.
The physician must recognize that perpetuation of polydrug abuse and drug dependence is not necessarily a symptom of an underlying emotional disorder. Neither alleviation of anxiety nor reduction of depression accounts for initiation and perpetuation of polydrug abuse. Severe depression and anxiety are as frequently the consequences of polydrug abuse as they are the antecedents. There is also evidence that some of the most adverse consequences of drug use may be reinforcing and contributing to the continuation of polydrug abuse. TREATMENT Adequate treatment of polydrug abuse, as well as other forms of drug abuse, requires innovative programs of intervention. The first step in successful treatment is detoxification, a process that may be difficult because of the abuse of several drugs with different pharmacologic actions (e.g., alcohol, opiates, and cocaine). Since patients may not recall or may deny simultaneous multiple drug use, diagnostic evaluation should always include urinalysis for qualitative detection of psychoactive substances and their metabolites. Treatment of polydrug abuse often requires hospitalization or inpatient residential care during detoxification and the initial phase of drug abstinence. When possible, specialized facilities for the care and treatment of chemically dependent persons should be used. Outpatient detoxification of polydrug abuse patients is likely to be ineffective and may be dangerous. As in the treatment of alcohol abuse, no single therapeutic modality has been shown to be uniquely effective in inducing remission. Polydrug abuse is a chronic disorder with an unpredictable pattern of remission and recrudescence. Even temporary remissions with attendant physical, social, and psychological improvements are preferable to the continuation or progressive acceleration of polydrug abuse and its related adverse medical and interpersonal consequences. In polydrug abuse, as in many chronic disorders, definitive "cures" rarely occur. The concerned physician should continue to assist polydrug abuse patients throughout the cyclic oscillations of this complex behavior disorder, recognizing that resumption of drug use may be the rule rather than the exception. (Bibliography omitted in Palm version) Back to Table of Contents
390. NICOTINE ADDICTION - David M. Burns The use of tobacco leaf to create and satisfy nicotine addiction was introduced to Columbus by Native Americans and spread rapidly to Europe. The use of tobacco as cigarettes, however, is predominantly a twentieth century phenomenon, as is the epidemic of disease caused by this form of tobacco. Nicotine is the principal constituent of tobacco responsible for its addictive character. Addicted smokers regulate their nicotine intake and blood levels by adjusting the frequency and intensity of their tobacco use both to obtain the desired psychoactive effects and avoid withdrawal. Unburned cured tobacco contains nicotine, carcinogens, and other toxins capable of causing gum disease and oral cancer. When tobacco is burned, the resultant smoke contains, in addition to nicotine, carbon monoxide and >4000 other compounds that result from volatilization, pyrolysis, and pyrosynthesis of tobacco and various chemical additives used in making different tobacco products. The smoke is composed of a fine aerosol, with a particle size distribution predominantly in the range to deposit in the airways and alveolar surfaces of the lungs, and a vapor phase. The bulk of the toxicity and carcinogenicity of the smoke resides in the aerosolized particulate phase, which contains a large number of toxic constituents and >40 carcinogenic compounds. The aggregate of particulate matter, after subtracting nicotine and moisture, is referred to as tar. The vapor phase contains carbon monoxide, respiratory irritants, and ciliotoxins as well as many of the volatile compounds responsible for the distinctive smell of cigarette smoke. The alkaline pH of smoke from blends of tobacco utilized for pipes and cigars allows sufficient absorption of nicotine across the oral mucosa to satisfy the smoker's need for this drug. Therefore, smokers of pipes and cigars tend not to inhale the smoke into the lung, confining the toxic and carcinogenic exposure (and the increased rates of disease) largely to the upper airway for most users of these products. The acidic pH of smoke generated by the tobacco used in cigarettes dramatically reduces absorption of nicotine in the mouth, necessitating inhalation of the smoke into the larger surface of the lungs in order to absorb quantities of nicotine sufficient to satisfy the smoker's addiction. The shift to using tobacco as cigarettes, with resultant increased deposition of smoke in the lung, has created the epidemic of heart disease, lung disease, and lung cancer that dominates the current disease manifestations of tobacco use. DISEASE MANIFESTATIONS OF CIGARETTE SMOKING Over 400,000 individuals die prematurely each year in the United States from cigarette use; this represents approximately one out of every five deaths in the United States. Approximately 40% of cigarette smokers will die prematurely due to cigarette smoking unless they are able to quit. The major diseases caused by cigarette smoking are listed in Table 390-1, with the relative risks for each disease listed for male and female current smokers. The incidence of smoking-related diseases is proportionately greater in younger than in older smokers, particularly for coronary artery disease and stroke. At older ages, the
background rate of disease in nonsmokers increases, diminishing the fractional contribution of smoking and the relative risk; however, absolute excess rates of disease mortality found in smokers compared to nonsmokers increase with increasing age. The organ damage caused by smoking and the number of smokers who die from smoking are both greater among the elderly, as one would expect from a process of cumulative injury. Cardiovascular Diseases Cigarette smokers are more likely than nonsmokers to develop large vessel atherosclerosis as well as small vessel disease. Approximately 90% of peripheral vascular disease in the nondiabetic population can be attributed to cigarette smoking, as can approximately 50% of aortic aneurysms. In contrast, 20 to 30% of coronary artery disease and approximately 10% of occlusive cerebrovascular disease are caused by cigarette smoking. There is a multiplicative interaction between cigarette smoking and other cardiac risk factors such that the increment in risk produced by smoking among individuals with hypertension or elevated serum lipids is substantially greater than the increment in risk produced by smoking for individuals without these risk factors. In addition to its role in promoting atherosclerosis, cigarette smoking also increases the likelihood of myocardial infarction and sudden cardiac death by promoting platelet aggregation and vascular occlusion. Reversal of these effects may explain the rapid benefit of smoking cessation for a new coronary event demonstrable among those who have survived a first myocardial infarction. This effect may also explain the substantially higher rates of graft occlusion among continuing smokers following vascular bypass surgery for cardiac or peripheral vascular disease, as well as the high failure rate of angioplasty procedures among continuing smokers. Cessation of cigarette smoking reduces the risk of a second coronary event within 6 to 12 months after quitting, and rates of first myocardial infarction or death from coronary heart disease also decline within the first few years following cessation. After 15 years of cessation, the risk of a new myocardial infarction or death from coronary heart disease in former smokers is similar to that in those who have never smoked. Cancer Cancers of the lung, larynx, oral cavity, esophagus, pancreas, kidney, and urinary bladder are caused by cigarette smoking. In addition, there is evidence suggesting that cigarette smoking may play a role in increasing the risk of cervical and stomach cancer. There is conflicting evidence on the relationship of cigarette smoking and cancer of the breast, but overall there does not appear to be a causal link. There is a lower risk of uterine cancer among postmenopausal women who smoke. The risks of cancer increase with the increasing number of cigarettes smoked per day and the duration of smoking, and there are synergistic interactions between cigarette smoking and alcohol use for cancer of the oral cavity, esophagus, and possibly lung. Several occupational exposures also synergistically increase lung cancer risk among cigarette smokers, most notably occupational asbestos and radon exposure. Cessation of cigarette smoking reduces the risk of developing cancer relative to continuing smoking, but even 20 years after cessation there is a modest persistent increased risk of developing lung cancer.
Respiratory Disease Cigarette smoking is responsible for>90% of chronic obstructive pulmonary disease. Within 1 to 2 years of beginning to smoke regularly, many young smokers will develop inflammatory changes in their small airways, although lung function measures of these changes do not predict development of chronic airflow obstruction. After³20 years of smoking, pathophysiologic changes in the lungs develop and progress proportional to smoking intensity and duration. Chronic mucous hyperplasia of the larger airways results in a chronic productive cough in as many as 80% of smokers over age 60. Chronic inflammation and narrowing of the small airways and/or enzymatic digestion of alveolar walls resulting in pulmonary emphysema can result in reduced expiratory airflow sufficient to produce clinical symptoms of respiratory limitation in approximately 15% of smokers. Changes in the small airways of young smokers will reverse after 1 to 2 years of cessation. There may also be a small increase in measures of expiratory airflow following cessation among individuals who have developed chronic airflow obstruction, but the major change following cessation is a slowing of the rate of decline in lung function with advancing age rather than a return of lung function toward normal. Pregnancy Cigarette smoking is associated with several maternal complications of pregnancy: premature rupture of membranes, abruptio placentae, and placenta previa; there is also a small increase in the risk of spontaneous abortion among smokers. Infants of smoking mothers are more likely to experience preterm delivery, have a higher perinatal mortality, are small for their gestational age, are more likely to die of sudden infant death syndrome, and appear to have a developmental lag for at least the first several years of life. Other Conditions Smoking delays healing of peptic ulcers; increases the risk of osteoporosis, senile cataracts, and macular degeneration; and results in premature menopause, wrinkling of the skin, gallstones and cholecystitis in women, and male impotence. Environmental Tobacco Smoke Long-term exposure to environmental tobacco smoke increases the risk of lung cancer and coronary artery disease among nonsmokers. It also increases the incidence of respiratory infections, chronic otitis media, and asthma in children as well as causing exacerbation of asthma in children. PHARMACOLOGIC INTERACTIONS Cigarette smoking may interact with a variety of other drugs in ways that may have clinically significant implications (Table 390-2). Cigarette smoking induces the cytochrome P450 system, which may alter the metabolic clearance of drugs such as theophylline. This effect may result in more drug toxicity among nonsmokers on fixed drug dosage schedules and in inadequate serum levels in smokers as outpatients when the dosage is established in the hospital under nonsmoking conditions. Correspondingly, serum levels may rise when smokers are hospitalized and not allowed to smoke. Smokers may also have higher first-pass clearance for drugs such as lidocaine, and the stimulant effects of nicotine may reduce the effect of benzodiazepines or beta blockers.
OTHER FORMS OF TOBACCO USE Other major forms of tobacco use are moist snuff deposited between the cheek and gum, chewing tobacco, pipes and cigars, and recently bidi (tobacco wrapped in tendu or temburni leaf and commonly used in India) and clove cigarettes. Oral tobacco use leads to gum disease and can result in oral cancer. All forms of burned tobacco generate toxic and carcinogenic smoke similar to that of cigarette smoke. The differences in disease consequences of use relate to frequency of use and depth of inhalation. The risk of upper airway cancers is similar among cigarette and cigar smokers, while those who have smoked only cigars have a much lower risk of lung cancer, heart disease, and chronic obstructive pulmonary disease. However, cigarette smokers who switch to pipes or cigars do tend to inhale the smoke, increasing their risk; and it is likely that comparable inhalation and frequency of exposure to tobacco smoke from any of these forms of tobacco use will lead to comparable disease outcomes. Recent prevalence-of-use data have suggested a resurgence of cigar and bidi use among adolescents of both genders, raising concerns that these older forms of tobacco use are once again causing a public health concern. LOWER TAR AND NICOTINE CIGARETTES Since the bulk of the toxicity of cigarette smoke is contained in the tar, and since nicotine is the principal addictive agent in cigarettes, it has been suggested that cigarettes that deliver less tar and nicotine to the smoker might be safer. Studies of smokers of low-yield cigarettes suggest that there may be a 10 to 20% reduction in the risk of developing lung cancer among those who reduce the nominal tar yield of their cigarettes by³50%. However, this benefit is only evident if smokers do not compensate for the lower nicotine delivery with an increased intensity of smoking, and most studies show that smokers of low-yield cigarettes do compensate. Because of their addiction to nicotine, most smokers tend to preserve their intake of nicotine, and correspondingly their tar intake, when they shift to lower nicotine cigarettes. Newer, very low yield cigarettes commonly use vents in the filters or other engineering designs to reduce the tar and nicotine when the cigarette is smoked by machine. However, the delivery of tar and nicotine is much higher when these cigarettes are smoked by actual smokers. Current evidence suggests that if there is any disease-reduction benefit for smokers of low-yield cigarettes, it is too small to be clinically meaningful, and individuals should be discouraged from thinking of low-yield cigarettes as a substitute for cessation. CESSATION The process of stopping smoking is often a cyclical one, with the smoker sometimes making multiple attempts to quit and failing before finally being successful. Approximately 70 to 80% of smokers would like to quit smoking, approximately one-third of current smokers attempt to quit each year, and ³90% of these unassisted quit attempts fail. Smokers have been categorized into those who are not thinking about quitting (precontemplation), those who are thinking about quitting (contemplation), and
those who are in the action phase of quitting. A useful conceptualization of the cessation process is one where smokers cycle through the stages of cessation; each time smokers go around the cycle, a few more smokers become successful in their cessation efforts. One goal of clinician-based smoking interventions then becomes moving smokers from one stage of the cessation cycle to another, and efforts can be focused on moving the smoker to the next stage rather than focusing exclusively on immediate cessation. The move from thinking about quitting to making a quit attempt is often triggered by a variety of environmental stimuli independent of physician control. The cost of cigarettes can be a powerful trigger for cessation attempts. Media campaigns, particularly when coupled with cessation events, are also able to trigger cessation attempts in large numbers of smokers. Changes in workplace rules to restrict smoking in the workplace have been associated with quit attempts in substantial numbers of workers. However, physician advice to quit, particularly around an acute illness, is also a powerful trigger for cessation activity, with up to half of patients who are advised to quit making a cessation effort. Telephone counseling and nicotine-replacement therapy are all useful enhancers of long-term cessation success. Clinic-based cessation programs have a substantial benefit for long-term cessation for those who can be recruited to participate, and physician recommendation can double the fraction of smokers who are willing to participate in these programs. PREVENTION Approximately 90% of individuals who will become cigarette smokers initiate the behavior during adolescence. Factors that promote adolescent initiation are parental or older generation cigarette smoking, tobacco advertising and promotional activities, the availability of cigarettes, and the social acceptability of smoking. The need for an enhanced self-image and to imitate adult behavior is greatest for those adolescents who have the least external validation of their self-worth, which may explain in part the enormous differences in adolescent smoking prevalence by socioeconomic and school performance strata. Prevention of smoking initiation must begin early, preferably in the elementary school years. Physicians who deal with adolescents should be sensitive to the prevalence of this problem in their patient population. Effective physician-based interventions for adolescent smokers remain to be developed, but current clinical guidelines suggest that physicians should ask all adolescents whether they have experimented with tobacco or currently use tobacco, reinforce the facts that most adolescents and adults do not smoke, and explain that all forms of tobacco are both addictive and harmful. GENETIC CONSIDERATIONS Several genes have been associated with nicotine addiction. Some reduce the clearance of nicotine, and others have been associated with an increased likelihood of becoming dependent on tobacco and other drugs as well as a higher incidence of depression. Genetic alterations that involve the neurotransmitter dopamine, and
possibly the serotoninergic and cholinergic neuroregulatory pathways, are being explored for their contribution to development of addiction to tobacco and other substances. The precise role these genetic differences play in development and maintenance of nicotine addiction remains to be determined, but it is unlikely that genetic factors are the principal determinants of addiction. Rates of smoking initiation among males, and corresponding rates of nicotine addiction, have dropped by almost 50% since the mid-1950s, suggesting that factors other than genetics are the principal determinants of whether individuals will become addicted. It is more likely that genetic polymorphism represents a range of biologic susceptibility conditioning the intensity of cigarette use and the probability that experimentation with tobacco as an adolescent leads to addiction as an adult. PHYSICIAN INTERVENTION Physicians can make a clear difference in promoting successful cessation among their smoking patients, and the Agency for Health Care Policy and Research (AHCPR) has developed clinical guidelines for health care system-based smoking cessation (Table 390-3). All patients should be asked whether they smoke, their past experience with quitting, and whether they are currently interested in quitting. Those who are not interested in quitting should be encouraged and motivated to quit; provided a clear, strong, and personalized physician message that smoking is an important health concern; and offered assistance if they become interested in quitting in the future. There is a relationship between the amount of assistance a patient is willing to accept, and the success of the cessation attempt. A quit date should be negotiated, usually not the day of the visit but within the next few weeks, and a follow-up contact by office staff around the time of the quit date should be provided. There are a variety of nicotine-replacement products, including over the counter nicotine patch and gum, as well as nicotine nasal and oral inhalers available by prescription. Clonidine and, more recently, antidepressants such as bupropion have also been shown to be effective; some evidence supports the combined use of nicotine-replacement therapy and antidepressants. Nicotine-replacement therapy is provided in different dosages for use with smokers of different numbers of cigarettes per day. Antidepressants are more effective in those with a history of depression symptoms. At this time there are few clear indications favoring the use of one agent over another as initial therapy. Current recommendations are to offer pharmacologic treatment to all who will accept it and to provide counseling and other support to the patient as a part of the cessation attempt. Cessation advice alone is likely to increase success by 50% compared with no intervention; a more comprehensive approach with advice, pharmacologic assistance, and counseling can increase cessation success by almost threefold. In order for physicians to incorporate cessation assistance into their practice successfully, it is essential to change the infrastructure in which the physician practices. The following are simple changes: (1) including questions on smoking and interest in cessation on patient-intake questionnaires, (2) asking patients whether they smoke as part of the initial vital sign measurements made by office staff, (3) listing smoking as a problem in the medical record, and (4) automating follow-up contact with the patient on their quit date. These changes are essential to institutionalizing smoking intervention
within the practice setting; without this institutionalization, the best intentions of physicians to intervene with their patients who smoke are often lost in the time crush of a busy practice. (Bibliography omitted in Palm version) Back to Table of Contents
PART FIFTEEN -ENVIROMENTAL AND OCCUPATIONAL HAZARDS SECTION 1 -SPECIFIC ENVIRONMENTAL AND OCCUPATIONAL HAZARDS 391. SPECIFIC ENVIRONMENTAL AND OCCUPATIONAL HAZARDS - Howard Hu, Frank E. Speizer It cannot be overemphasized that an appropriate environmental/occupational history is an essential part of the medical workup of many chronic diseases. The general approach to the patient whose illness may have been caused or exacerbated by environmental or occupational hazards is detailed inChap. 5. The term hazards in this context is generally synonymous with toxins and toxic exposures and encompasses chemical factors as well as other risks posed by the physical environment and by selected natural phenomena. These hazards may exist in the general environment or in the workplace. Strictly speaking, smoking, alcohol ingestion, nutritional factors, and infectious agents can also be considered chemical or environmental hazards. Once a specific hazard has been identified as a factor in the pathogenesis of an illness or as an imminent threat, the clinical approach must include the development of a strategy for preventing further exposure and for treating the specific manifestations of the illness, using antidotes and supportive measures. In the following chapters, specific hazards are considered, including acute poisoning and drug overdose; heavy metal poisoning; disorders caused by venoms, bites, and stings; drowning and near-drowning; electrical injuries; and radiation injury. The health effects of ambient air pollution, occupational respiratory exposures, passive smoking, and assorted toxic air pollutants are discussed briefly inChap. 254. Space does not allow specific discussion in this text of many other important categories of hazards, such as organic solvents; chemicals used in the plastics, synthetic textiles, and rubber industries; and pesticides. The reader should consult other detailed texts or electronic information sources for clinical data on these topics. In this volume, however, brief attention is focused on several selected issues in light of recent developments in research that have enhanced our understanding of the way these hazards may interact with human behavior and consequently pose increased risks to both individuals and society. HAZARDOUS WASTE AND GROUNDWATER CONTAMINATION The term hazardous waste embodies toxic chemicals, radioactive materials, and biologic or infectious wastes. In many communities, hazardous waste has emerged as a major public health concern. In the United States, some 50,000 sites (defined by specific criteria) have been estimated to contain hazardous chemicals; 1000 or so of these have been included as "Superfund sites" on a National Priority List drawn up by the Environmental Protection Agency (EPA). New or unrecognized sites are likely to exist as well. These sites may require long-term remedial action. The spectrum of substances contained at the sites is wide and theoretically may include any of some 30,000 chemicals that are commonly used in commerce. However, the EPA keeps fewer than 200 chemicals on a special hazardous substance list in light of their toxicity, the frequency with which they are encountered, and other factors. One difficulty in
anticipating risks associated with hazardous waste sites is that the substances are usually present in mixtures whose composition is seldom fully known. In addition, with respect to toxicity, chemicals may interact with one another in an additive, protective, or synergistic fashion, and little knowledge exists on which to base predictions regarding the interactions of these complex mixtures. Waste-site employees and the surrounding community can incur hazardous exposures through the inhalation of toxic vapors or dusts emanating directly from a waste site or an on-site incinerator; the ingestion of water contaminated by surface runoff or by material leaching through soil into surface water or groundwater; the ingestion of contaminated plants, fish, or other wildlife; or direct contact. This last risk is particularly likely for children, who may enter a poorly secured site. Perhaps the exposure of greatest concern to community residents has been the contamination of groundwater by volatile organic compounds or solvents (VOCs); together, the widespread detection of low levels of VOCs in groundwater and the several studies suggesting an association between heavy VOC contamination of drinking water and cancer probably account for the high priority given in public opinion polls to avoiding cancer risks. A 1983 study found that 11 of the 20 chemicals most commonly detected at National Priority List waste sites were VOCs (Table 391-1). Current regulatory policy rests on the assumption that there is no threshold below which a carcinogen exerts no effect or risk. Thus, once a substance is identified as a probable carcinogen (see below), it is regulated to a concentration that is believed to be accompanied by an acceptable level of risk. Clearly, great uncertainty exists regarding methods used to classify drinking-water carcinogens and to extrapolate the risks related to exposure to these substances. Regardless,VOCcontamination in groundwater is likely to continue to be a high-priority issue in the public arena. ENVIRONMENTAL CARCINOGENS Based on studies and reviews of the literature by the International Agency for Research on Cancer, enough evidence exists to classify around 60 substances and processes as probably or definitely carcinogenic in humans (Table 391-2). Some processes are deemed carcinogenic on the basis of epidemiologic evidence, even though the specific causative agent cannot always be clearly identified. Tumor promoters are not distinguished from tumor initiators in this listing, and the chemical structures and modes of action are diverse. Around 150 additional agents and processes have been designated as possibly carcinogenic on the basis of studies of bacteria and animals as well as human epidemiologic studies. The extent to which inferences can be made from nonhuman studies is controversial but certainly depends on minimal standards in the execution of such studies. For example, the Interagency Regulatory Liaison Group recommends that for a carcinogen assay to be considered positive, the test must have been performed on at least 50 animals of each sex in two different species with at least three dose groups (control and two dose levels) over the lifetime of the animals. BUILDING-RELATED ILLNESSES Reports of discomfort and symptoms in relation to office environments began in the United States in the 1970s. Research has led to the recognition that some
building-related illnesses have a clear etiology; these illnesses include hypersensitivity diseases, infections, and exacerbations of asthma due to airborne irritants. However, the majority of such complaints, particularly those of mucous membrane irritation, fatigue, and headache, have no clear etiology. Terms such as sick-building syndrome (SBS; also called tight-building syndrome) and nonspecific building-related illnesses have been used to designate this constellation of symptoms, which have been found in most investigations to occur most often in sealed buildings with centrally controlled mechanical ventilation. Early characterizations of SBS as mass psychogenic illness have not been borne out in the majority of cases by subsequent epidemiologic investigations. Since indoor air-exchange rates were sharply reduced in the 1970s to conserve energy, current hypotheses focus on inadequate dilution of irritants arising from building materials (such as formaldehyde-containing particle board), office supplies (such as carbonless copy paper and photocopy developer solution), toxins from mold and bacterial endotoxin, and personal care products used by occupants as risk factors for SBS. Confirmation of these hypotheses and further characterization of SBS await additional research. MULTIPLE-CHEMICAL SENSITIVITY The multiple-chemical sensitivity (MCS) syndrome is a diagnosis that has increasingly been given to patients with a wide variety of symptoms that they attribute to exposure at very low levels to a number of commonly encountered chemicals. The syndrome usually begins after a well-defined environmental event, such as a reaction to a more clearly toxic dose of an organic solvent, pesticide, or respiratory irritant. Some cases of MCS begin asSBS. Affected persons commonly report symptoms such as fatigue, malaise, headache, dizziness, lack of concentration, memory loss, and "spaciness" -- symptoms that overlap somewhat with those of other diagnoses of uncertain etiology, such as chronic fatigue syndrome. The pathogenesis of MCS is obscure, and no proven methods exist for its diagnosis, evaluation, and treatment. Case series suggesting a high prevalence of affective disorders indicate that psychological factors may play a role in causing MCS and/or in determining its severity; however, evidence does not support MCS as a purely psychogenic illness. A few studies of MCS patients suggest that the biologic mechanism of MCS may involve neurogenic inflammation of the nasal mucosa (as indicated by abnormal rhinolaryngoscopic findings) linked to central nervous system dysfunction (as indicated by alterations seen on single photon emission computed tomography); however, well-controlled research remains sparse, and no firm conclusions can be drawn. Other than the ruling out of other treatable conditions and the avoidance of exacerbating exposures, no specific recommendations for the management of MCS patients can yet be made. A panel of European scientists convened by the World Health Organization recommended that the designation MCS be replaced by the term idiopathic environmental illness (IEI). PERSISTENT ORGANIC POLLUTANTS Persistent organic pollutants (POPs) are a class of chemical compounds that tend to travel thousands of miles if released into the atmosphere, to accumulate in the food chain, and to persist in the environment as well as in human tissues (principally fat cells). Although the list of POPs is long, 12 have been identified as particularly important: nine pesticides (aldrin, chlordane, DDT, dieldrin, endrin, heptachlor,
hexachlorobenzene, mirex, and toxaphene), dioxins and furans (byproducts of incineration), and polychlorinated biphenyls (PCBs, fluids used mainly as dielectrics in transformers). The persistence and lipid solubility of these compounds allow them to bioconcentrate several thousand-fold as they are passed up the food chain to humans. High levels of exposure to a number of POPs have been shown to contribute to birth defects, infertility, immunosuppression, impaired cognitive development, and some types of cancers. These effects have been linked to the potential of POPs to act as endocrine disruptors -- i.e., hormonal mimics. The further production and use of most POPs have been banned, but concern remains over the possible low-level effects of POPs that persist in the environment and in human tissues. Concern has been raised, for example, that population exposures to POPs are contributing to worldwide declines in sperm density and increased rates of congenital hypospadias and testicular and breast cancer; epidemiologic studies testing these theories have yielded mixed results, however, and more research is needed. GLOBAL CLIMATIC CHANGES An increasing body of evidence indicates that human activities are responsible for global climatic changes, which, in turn, may be directly or indirectly increasing human exposure to environmental hazards. The depletion of stratospheric ozone by chlorinated fluorocarbons, with a consequent increase in ultraviolet radiation exposure, has been firmly established. Increased risks of skin cancers and cataracts are accepted as results of this phenomenon. Less clear is whether the immunosuppressive effects of ultraviolet radiation detected in animals and in vitro have significant clinical impacts on human resistance to infection. Although uncertainties in climate modeling persist, an increasing if not overwhelming amount of evidence indicates that anthropogenic greenhouse gases are fostering global warming. A prominent concern is that global warming can abet the introduction and dissemination of serious infectious diseases, such as mosquito-borne infections (malaria, dengue, and viral encephalitis) and waterborne infectious and toxin-related illnesses (cholera, shellfish poisoning). The World Health Organization has identified global warming as one of the largest public health challenges facing the twenty-first century. (Bibliography omitted in Palm version) Back to Table of Contents
392. DROWNING AND NEAR-DROWNING - Jerome H. Modell It is an unexpected tragedy when a previously healthy person dies or is exposed to severe cerebral hypoxia and suffers permanent brain damage. For many years, drowning was considered a "fight for survival": Arms flailing and screaming for help, a person who could not swim struggled to remain on the surface of the water to reach safety. This situation, however, is rarely reported by persons at the scene of aquatic emergencies. Furthermore, no single set of circumstances comprises drowning or near-drowning. It may be a secondary event following such precursors as head or spinal trauma; hypoxia-induced unconsciousness; or unconsciousness due to preexisting cardiovascular disease, sudden cardiac death, or myocardial infarction. The initiating event is usually unknown, so the drowned or near-drowned victim must be treated based on probable physiologic effects of the near-drowning itself. If survival with normal brain function is to occur, a thorough understanding of the pathophysiology of drowning and an organized approach to therapy are imperative. PATHOPHYSIOLOGY OF DROWNING Approximately 90% of near-drowning victims aspirate fluid into their lungs. In those who do not aspirate fluid, hypoxemia results simply from breath holding, laryngospasm, or apnea. In those who do aspirate, the volume and the composition of the fluid determine the physiologic basis of the hypoxemia. Freshwater aspiration alters the surface tension properties of pulmonary surfactant and makes alveoli unstable, which causes a decreased ventilation/perfusion ratio. Some alveoli collapse and become atelectatic, which produces a true or absolute intrapulmonary shunt, while others are poorly ventilated and produce a relative shunt; in either case, significant pulmonary venous admixture occurs. Fresh water in the alveoli is hypotonic and is rapidly absorbed and redistributed throughout the body. While some have proposed that water continues to enter the lungs after death, at autopsy the lungs of victims who died in the water frequently contain little water. Also, it has been shown experimentally that if a dead body is submerged in tagged or colored water, water is not found in the lungs at autopsy. These findings support the premise that active respiration determines the volume of water aspirated. Hypertonic seawater pulls additional fluid from the plasma into the lungs, and thus the alveoli are fluid-filled but perfused, which causes substantial pulmonary venous admixture. With both types of water, pulmonary edema may occur secondary to events such as fluid shifts, a change in capillary permeability, or cerebral hypoxia, which causes neurogenic pulmonary edema. Regardless of the cause, pulmonary edema adds to the ventilation/perfusion abnormality. Water that is grossly contaminated with bacteria or that contains particulate matter may complicate the picture. Particulate matter can obstruct the smaller bronchi and respiratory bronchioles. Grossly contaminated water increases the risk of severe pulmonary infection. Neither problem is sufficiently common, however, to justify recommending specific therapy routinely for all victims. At least 85% of near-drowned victims are thought to aspirate 22 mL/kg of water or less, which does not result in a clinically significant alteration of blood volume or serum
electrolyte concentrations. After resuscitation, by the time blood is analyzed, serum electrolyte concentrations are usually normal or close to normal. Significant changes are documented in only approximately 15% of those who cannot be resuscitated and only rarely in those who are resuscitated. These findings suggest that either a small amount of water was aspirated, fluid was rapidly redistributed, or both. Therefore, electrolyte disturbance rarely needs treatment. When a large quantity of water is aspirated, seawater causes hypovolemia, which concentrates extracellular electrolytes, and fresh water causes acute hypervolemia. If enough water is aspirated that plasma becomes severely hypotonic and the patient is hypoxemic, red cell membranes can rupture, and plasma hemoglobin and serum potassium concentrations increase significantly. However, this development has been reported only rarely. With rapid redistribution of fluid and development of pulmonary edema, even freshwater victims frequently demonstrate hypovolemia by the time they reach the hospital. Hypercarbia, which is associated with apnea and/or hypoventilation, is less often documented by blood gas analysis than is hypoxemia. While hypoxemia due to pulmonary venous admixture persists in all near-drowned victims who aspirate water, hypercarbia is usually corrected sooner with artificial mechanical ventilation and improved minute ventilation and, thus, is reported in only a small percentage of victims evaluated at the hospital. Besides hypoxemia, metabolic acidosis also persists in most patients. Abnormal cardiovascular function, usually ascribed to hypoxemia, is brief with effective, timely therapy. Abnormality in renal function is uncommon, but when it does occur, it too is secondary to hypoxemia, altered renal perfusion, or, in extremely rare circumstances, significant hemoglobinuria. TREATMENT The first step is retrieving the victim from the water, and, if necessary, performing artificial ventilation and circulation. The American Heart Association recommends that an abdominal thrust not be used routinely in victims of submersion. This recommendation was upheld by a special committee of the Institute of Medicine convened in 1994 specifically to evaluate the efficacy of an abdominal thrust in the treatment of near-drowned victims. In these patients, an abdominal thrust may lead to regurgitation of gastric contents and, thus, to aspiration of the vomitus. Further, an abdominal thrust may delay ventilatory or circulatory resuscitation. Therefore, an abdominal thrust should be used only when the airway is obstructed with a foreign body or when the victim fails to respond to mouth-to-mouth ventilation. Because emergency services and intensive pulmonary and cardiovascular care have improved during the past 25 years, central nervous system depression now presents the major therapeutic challenge in near-drowning. The rate of survival with normal cerebral function varies considerably in retrospective studies. Some factors that adversely influence survival are prolonged submersion, delay in initiation of effective cardiopulmonary resuscitation, severe metabolic acidosis (pH< 7.1), asystole upon arrival at a medical facility, fixed dilated pupils, and a low Glasgow coma score (10 MHz) electrical power can couple electrical energy across an air gap into the body without charge transport across the skin surface (capacitive coupling). Low-frequency electricity causes tissue injury primarily by permeabilizing cell membranes, electroconformational denaturation of cell membrane proteins, and thermal denaturation of tissue proteins. Factors that determine the anatomic pattern, the extent of tissue injury, and the relative contribution of heat versus direct electrical damage include the amount of current, anatomic location, and the contact duration. The type of clothing, the use of protective gear, and the power capability of the electrical source also contribute to the wide range of clinical manifestations in victims of electrical shock. In addition, a very high-energy electrical arc can produce a strong thermoacoustic blast force leading to barotrauma. Associated falls and skin burns are frequent, exacerbating the injury. Cataracts characteristically occur after rapid and brief exposure of the eyes to hot gases and arc-mediated electrical current. The latency period for development of cataracts averages approximately 6 months. Peripheral nerve and skeletal muscle tissues are most vulnerable to membrane damage by applied electrical currents. The "no-let-go" phenomenon results from the passage of more than 14 to 16 mA longitudinally through the forearm that induces tetanic contractions of muscles controlling handgrip. The resulting involuntary muscle spasm may lead to joint dislocations and spine fractures. When current of>50 mA is passed hand-to-hand or hand-to-foot, there is enough induced depolarization of myocardial membranes to cause cardiac arrhythmias, particularly if the induced depolarization occurs during early myocardial repolarization. Disruption of extremity skeletal muscle and nerve cell membranes by the process of electroporation results when more than 0.5 to 1 A is passed through the extremity. Electroporation damage accumulates on the time scale of milliseconds, leading to lethal cellular injury. With more prolonged contacts in the range of seconds, thermal damage in the subcutaneous tissues becomes substantial. Because the vulnerability to supraphysiologic temperature exposure is similar regardless of tissue type, all tissues in the current path are burned when pathologic levels of heating occur. Extensive disruption of cell membranes leads to release of myoglobin and hemoglobin, which enter the circulation. Acute renal failure can result from intrarenal crystalization of these molecules. Acute renal failure superimposed on extensive tissue injury has a very high mortality rate. DIAGNOSIS For the more common low-frequency electrical injuries, at least two skin contact wounds are present. Differences in wound size and topography are largely determined by the surface contact area, the shape of the objects that conducted the current through the victim, and the duration of contact. Cardiac arrhythmias and most respiratory disturbances must be rapidly detected by examination of the pulse, chest, and electrocardiogram. The next priority is to determine the location and extent of tissue damage. Injured skeletal muscle and nerves are often
found beneath undamaged skin. Lateral spine x-rays or computed tomography (CT) are needed to rule out unstable spine fracture patterns. X-ray images of the extremities involved are also important to rule out skeletal fractures or joint dislocations. Blood chemistries should be immediately evaluated and monitored. Metabolic acidosis and elevated serum potassium levels may exist as consequences of extensive skeletal muscle injury. Serum CPK levels will rise over several hours if there is significant rhabdomyolysis. Tissue edema begins to form because of increased vascular permeability and the release of intracellular contents into the extravascular space. Muscle compartment syndrome and compression neuropathies are common manifestations. If available, magnetic resonance imaging (MRI) scans can rapidly localize tissue edema. Where severe heating has coagulated the blood vessels, tissue injury may exist in the absence of edema. Muscle compartment fluid pressures should be measured where edema is present. If MRI is not available, then the muscle compartments in the current path between contact points should be monitored for elevated interstitial fluid pressure. Elevated compartment pressures may evolve during resuscitation. Muscle compartment fluid pressures >30 cmH20 are indications for fasciotomy. It may be necessary to check the pressures every 8 h for 24 h. Radionucleotide scanning with99mTc-pyrophosphate can also be useful to detect tissue damage. These scans, however, take 4 to 6 h to complete and are mostly useful in the less severe injuries. If there is a history of loss of consciousness,CT of the head is indicated. TREATMENT The first priority is to disconnect the patient from the electrical power source. When high-capacity circuits are involved, disconnection must not be attempted before the circuit is deenergized. Cervical spine fracture should be assumed until proven otherwise. Critical initial considerations are evaluation and support of vital organ function and, secondarily, assessment of the extent of injury. After very-high voltage trauma, prolonged cardiopulmonary resuscitation (CPR) may be necessary before the stunned myocardium regains the ability to sustain a coordinated rhythm. Patients with significant wounds and tissue injury as well as those with vital organ injury require hospital admission. It is unlikely for cardiac arrhythmias to develop if cardiac injury is not detectable on initial presentation. Peripheral nerve injury invariably occurs even in minor shocks and usually resolves over several days. If symptoms persist, however, they may be controlled with cyclooxygenase inhibitors alone or with antioxidants. Small wounds can be managed by cleaning and applying topical antibiotics. Major neuropsychological and stress disorders often follow a terrifying "no-let-go" experience. Management often involves psychiatric consultation. For more substantial trauma, a Foley catheter and large bore peripheral intravenous lines delivering normal saline at a rate sufficient to generate a 30 to 50 mL/h urine output are essential. If the urine is visibly pigmented with myoglobin or hemoglobin, the output should be doubled and alkalinized to a pH >6 by adding bicarbonate to the intravenous solutions until the urine has cleared. In the most severe injuries, hyperpermeability of peripheral capillaries may result in rapid interstitial (third space) fluid accumulation. In such cases, it may be necessary to increase blood oxygen levels
and add dextran to resuscitation fluids. Cardiac arrhythmias must be immediately controlled by antiarrhythmic drugs simultaneously with the correction of serum pH and electrolyte abnormalities. Brain injury-related seizures must be controlled with antiepileptic agents. Patients who have lost central nervous system (CNS) control of respiration or airways should be intubated and mechanically ventilated. A paralyzed ventilated patient may need monitoring by electroencephalogram (EEG) to assess seizure control. Appropriate management of corneal burns or abrasions, tympanic membrane rupture, and closed head injury should be instituted. Large skin burn wounds are often present because of arc-mediated contacts and clothing ignition. Care should be taken to prevent rapid loss of body heat through open wounds. Tetanus prophylaxis should be administered. Perfusion of devascularized tissue must be quickly restored. Diminished pulses or decreased tissue oxygen by transcutaneous pulse oximetry are indications for escharotomy releases. Fasciotomy is often required. The classic clinical signs of pain in acute compartment syndrome cannot be relied on because of associated nerve injury. In addition to decompression of extremity muscle compartments, decompression of nerve within edematous fibrous and osseous conduits (e.g., carpal tunnel, Guyon's canal, and tarsal tunnel) should be carried out to help prevent compression neuropathy. Care to avoid tissue drying or desiccation is important. Debridement of nonviable subcutaneous tissue should be performed as soon as a general anesthetic can be safely administered. Anaerobic bacterial infection of devascularized skeletal muscle is a common complication. Intravenous penicillin G and/or hyperbaric oxygen as prophylactic antibiotics are sometimes utilized but have unproven value. Radiographs taken to rule out fractures may reveal air bubbles in the subcutaneous tissues. This gas results from tissue boiling when prolonged Joule heating has occurred. Rehabilitation into society and gainful employment are the ultimate objectives. For severely injured victims these goals require functional muscle and nerve reconstruction as well as correction of scar contractures. Psychological and neurocognitive problems are expected and require treatment by experts. Persistent peripheral neurologic problems are also common and often require detailed evaluation and therapeutic intervention. LIGHTNING INJURIES Lightning injury is a powerful manifestation of arc-mediated electrical contact. Arcing occurs when the voltage gradient in air exceeds 2 million V/m. The arc consists of a hot ionized gas of subatomic particles that is highly conductive. Peak lightning currents reach into the range of 30,000 to 50,000 A for a duration of 5 to 10 us. Lightning arc temperatures reach up to 30,000°K, which generates thermoacoustic blast waves, commonly called thunder. Peak blast pressures reach 4 or 5 atm in the immediate vicinity of a lightning strike, and up to 1 or 2 atm 1 m away. Substantial barotrauma can result.
Like radiofrequency current, lightning current flows along the surfaces of conducting objects. Initially, the flow of an enormous current through the lightning strike generates a very large surrounding magnetic field pulse. This magnetic field pulse can induce current flow in the body, enough to disrupt cardiac andCNSfunction. When lightning current enters the ground, it spreads out radially, which sets a large current traveling along the surfaces of the ground. A substantial voltage drop can occur between the feet of a nearby individual. The voltage drops between widely separated feet can reach 1500 to 2000 V and can induce a 2 to 3 A current flow in the legs for a 10-us period. Victims of direct lightning strikes experience a multimodal injury. Superficial burns on the skin represent the current path along the skin surface. The intense brief shock pulse seems to arrest all electrophysiologic processes. The victim appears lifeless. ProlongedCPR may be necessary. Muscle and nerve necrosis is rare in survivors. Deeper injury results when the victim is in contact with a large conducting object such as a truck or fence that has been struck by lightning, which then discharges over several milliseconds through the victim. Delay in resuscitation is the most common cause of death. Bystanders are usually afraid to touch the victim while precious minutes pass. However, unless the victim is on an insulating platform, there is no residual electric charge on the body afer several milliseconds. When needed,CPRshould be given without hesitation. Victims should be cared for in an intensive care unit until life-threateningCNS and cardiac injuries are ruled out. Late neurologic and ophthalmologic sequelae often develop. (Bibliography omitted in Palm version) Back to Table of Contents
394. RADIATION INJURY - Stephen M. Hahn, Eli Glatstein All human beings are constantly exposed to ionizing radiation. Environmental sources include the cosmic radiation from space and radiation from the ground and from inhaled and ingested materials. Airline travel and mining both increase exposure to the background radiation. For example, air travel at 30,000 ft exposes individuals to a dose equivalent of 0.5 mrem/h. Radiation originating in the body comes mainly from radioactive potassium, which emits beta and gamma rays. Lungs are exposed to irradiation from inhaled air, which contains small amounts of radioactive radon. The cosmic exposure contributes approximately 28 mrem per year. The ground and internal sources contribute approximately 26 and 27 mrem per year, respectively. The most prominent man-made sources of radiation include x-ray equipment, nuclear weapons, and radioactive medications. TERMINOLOGY AND DEFINITIONS The first major unit of radiation exposure was the roentgen (R), defined as an amount of x-rays or gamma rays that produces a specific amount of ionization in a unit of air under standard temperature and pressure (Table 394-1); this quantity can be measured directly in an ionization chamber. The rad, or radiation absorbed dose, is defined as 100 ergs/g of tissue. Thus, the rad represents a net deposition of energy in a three-dimensional volume, because x-rays attenuate as they traverse tissue. The rad has been replaced by the Systeme Internationale (SI) unit of the gray (Gy), which represents 100 rad. Roentgens and rads can be converted by means of various tables; the relation between them depends on photon energy. The above definitions reflect physical variables. The unit that reflects the biologic response and that can be used to compare the effects of various types of radiation is the unit of dose equivalence, the rem (roentgen equivalent in man). The rem has been replaced by theSI unit, the sievert (Sv), which equals 100 rem. These units reflect the exposure or absorption dose multiplied by a biologic factor that represents the biologic effectiveness of the specific type of radiation (see below). TYPES OF IONIZING RADIATION The absorption of energy from radiation in tissue often leads to excitation or ionization. Excitation involves elevation of an electron in an atom or molecule to a higher energy state without actual ejection of the electron. Ionization involves actual ejection of one or more electrons from the atom. Ionizing radiation is subclassified as electromagnetic (photon) or particulate radiation (Table 394-2). X-rays and gamma rays are examples of electromagnetic photon radiation. They differ only in their source: X-rays are produced mechanically, by making electrons strike a target, which causes the electrons to give up their kinetic energy as x-rays, while gamma rays are produced by nuclear disintegration of radioactive isotopes. X-rays can be thought of as packets of energy, or photons. X-rays have no mass or charge, travel in straight lines, and attenuate continuously as they traverse tissue. Gamma rays have similar properties. Each photon contains an amount of energy equal to hn, where h is Planck's constant. The critical difference between nonionizing and
ionizing radiation is the energy of individual photons, not the energy of the total dose. Types of particulate radiation include electrons, protons, alpha particles, neutrons, negative pi-mesons, and heavy charged ions; these have discrete mass and charge (except for neutrons, which lack charge; Table 394-2). Electrons, or beta particles, are small and negatively charged and can be accelerated to close to the speed of light. They decelerate fairly rapidly in tissue and penetrate it to only a limited depth. Thus, electron beams are often used to treat superficial problems. Protons are positively charged and have a mass about 2000 times that of an electron. Protons stop abruptly, depending on their energy; in the process of sudden deceleration, most of their energy is given up, which tends to cause ionization just before the proton stops. This region of enhanced ionization, sometimes called the Bragg peak, means that proton beams exert their effects in a relatively compact region. Alpha particles are helium nuclei, consisting of two protons and two neutrons. The mass and charge are great enough that these particles do not penetrate far through matter unless they have tremendous energy; even a piece of paper is enough to protect against most alpha particles. Because these particles are charged, they can be accelerated in electrical fields. Neutrons are similar in mass to protons (having an atomic mass of 1), but they are not charged and therefore cannot be accelerated in an electrical field. Neutron beams are produced by colliding charged particles into a suitable target or are emitted as a fission product of heavy radioactive atoms. Heavy charged ions are nuclei of heavier elements that have a positive charge owing to the stripping away of some or all of the orbiting electrons. Equal doses of different types of radiation do not necessarily produce equal biologic effects; thus 1 Gy of neutrons produces a greater biologic effect than 1 Gy of x-rays. The biologic effects produced by a given dose of radiation can be quantified by the relative biologic effectiveness (RBE) value, which relates them to the effects produced by 250-kV photon radiation as a standard. In general, the greater the RBE value for a given type of radiation, the greater the biologic effect. The RBE value will be greater for more densely ionizing radiation, such as neutrons. The RBE value depends on the linear energy transfer (see below), the dose, the dose rate, and the nature of the biologic system. The linear energy transfer (LET) is the amount of ionization occurring per unit length of the radiation track. It is usually expressed as kilovolts per micron and increases with the square of the charge of the incident particle. High-LET radiation is biologically different from low-LET (i.e., conventional) radiation: Hypoxic and oxygenated cells respond similarly to high-LET irradiation, whereas it takes about three times as much low-LET radiation to produce a given killing effect in hypoxic cells as in oxygenated cells. It is thought that low-LET radiation must produce multiple hits on DNA to destroy a cell, whereas high-LET radiation need produce only a single hit on DNA to kill a cell. Representative values of LET andRBE are given in Table 394-3. Radiation, especially x-rays, is absorbed and causes ionization in three major ways: the photoelectric effect, the Compton effect, and pair production. At low energies (30 to 100 keV), as in diagnostic radiology, the photoelectric effect is important. In this process, the incident photon interacts with an electron in one of the outer shells of an atom (typically
K, L, or M). If the energy of the photon is greater than the binding energy of the electron, then the electron is expelled from the orbit with a kinetic energy that is equal to the energy of the incident photon minus the binding energy of the electron. The photoelectric effect varies as a function of the cube of the atomic number of the material exposed (Z3); this fact explains why bone is visualized much better than soft tissue on radiographs. At higher energies, as used in therapeutic radiology, the Compton effect dominates. In this process, the incident photon interacts with an electron in an orbital shell. Part of the incident photon energy appears as kinetic energy of electrons, and the residual energy continues as a less energetic deflected photon. At energy levels above 1.02 MeV, the photons may be absorbed through pair production. In this process, both a positron and an electron are produced in the absorbing material. A positron has the same mass as an electron but has a positive instead of a negative charge. The positron travels a very short distance in the absorbing medium before it interacts with another electron. When that happens, the entire mass of both particles is converted to energy, with the emission of two photons in exactly opposite directions. BIOLOGIC EFFECTS OF RADIATION Radiation must produce double-strand breaks in DNA to kill a cell, owing partly to the high capacity of mammalian cells for repairing single-strand damage. Radiation can also produce effects indirectly by interacting with water (which makes up approximately 80% of a cell's volume) to generate free radicals, which can damage the cell. Free radicals are highly reactive chemical entities that lack a stable number of outer-shell electrons. A free radical is not stable and has a life span of a fraction of a second. It is estimated that most x-ray-induced cell damage is due to the formation of hydroxyl radicals, as follows:
The result of radiation damage is cell death. The biologic effects on epithelial cell reproduction are typically expressed only when the damaged cells attempt to divide. Another biologic effect is the induction of cancerous growth by mutation many years after radiation exposure. Patients who receive radiation have a significant risk of neoplasm two to three decades after their exposure; this risk is significantly higher than that of the population as a whole. RADIATION-INDUCED CHROMOSOME ABERRATIONS Chromosome breaks can occur when cells are irradiated. The broken ends of chromosomes can combine with broken ends of different chromosomes. These abnormal combinations are most readily seen during mitosis. Chromosome abnormalities typically occur in cells irradiated in the G1 phase of the cell cycle, before the doubling of genetic material. If cells are irradiated in the G2 phase, chromatid aberrations may result. The frequency of chromosomal aberrations in peripheral circulating lymphocytes correlates with the dose received. The dose can be estimated by comparing the chromosomal changes to in vitro cultures exposed to controlled doses
of irradiation. The minimum dose that can be detected by peripheral lymphocyte analysis is about 0.1 to 0.2 Sv (10 to 20 rem). Lymphocyte analysis may provide evidence of recent total-body exposure. CELL SURVIVAL CURVE The dose-response curve for all mammalian cells appears to have a linear-quadratic relationship. In simple terms, the mathematical model that explains the relationship between the dose and the fraction of surviving cells has both linear and exponential components. The linear component results from double-stranded chromosomal breaks produced by single hits. The exponential component represents breaks produced by multiple hits. Figure 394-1 shows the shape of a typical survival curve for mammalian cells exposed to radiation. The fraction of cells surviving is plotted on a semilogarithmic scale. For x-rays or gamma rays, the dose-response curve has a shoulder that is followed by a straight line curve as the dose is increased. The shoulder represents the cell's ability to repair sublethal injury. For alpha particles or lower energy neutrons, the dose-response curve is a straight line from the origin. Thus, the survival rate is an exponential function of the dose. In all mammalian cell lines studied, increases in the radiation dose decrease the survival rate of cells. However, a number of factors may contribute to a relative resistance to radiation in human tumors in vivo, including hypoxia and expression of particular oncogenes, such as ras. The biologic basis for radiation resistance has not been fully defined. Four important processes that occur after radiation exposure can be summarized as the "four R's" of radiobiology. The first is repair. Repair is temperature dependent and is thought to represent the enzymatic mechanisms for healing intracellular injury. The second R is reoxygenation, a process whereby oxygen (and other nutrients) are actually better distributed to viable cells following radiation injury and cell killing. The third R is repopulation, the ability of the cell population to continue to divide and to replace dying and dead cells. The fourth R is redistribution, which reflects the variability of a cell's radiosensitivity over the cell cycle. Radiosensitivity can vary through the cell cycle by as much as a factor of 3. The G1 phase has the most variable length of all the phases of the cell cycle. For most cell lines, cells that have a short G1 period are most sensitive at the G2/mitosis interface, less sensitive in G1, and most resistant toward the end of the synthesis (S) period. Radiation therapy is effective in cancer treatment when it exerts greater cytotoxic effects on tumor cells than on normal tissues. A major determinant of the therapeutic index is exploiting differences in the four R's between tumor cells and normal tissues by delivering the radiation in dose fractions. CLINICAL FINDINGS ON FRACTIONATION The clinical radiation response may be related to the interactions of various growth factors and cytokines. For example, radiation can induce growth factors and cytokines such as tumor necrosis factor (TNF), interleukin (IL) 1. TNF can induce proliferation of fibroblasts and enhance the inflammatory response. TNF and IL-1 have been shown to
radioprotect hematopoietic cells in vitro by increasing the D0of the cell survival curve. TNF also enhances killing of a human tumor cell line by irradiation. TNF may produce radioprotection or radiosensitization depending on the cell type. Efforts to modulate radiation effects with TNF remain experimental. Other factors implicated in the radiation response are basic fibroblast growth factor and platelet-derived growth factorb, which may be associated with late effects of radiation on vessels. The degree and the duration of functional recovery of normal tissues are related to the number of stem cells surviving after irradiation. If the stem cells are destroyed in the irradiated volume and replacement from adjacent tissues is inadequate, radiation injury will persist. True late effects develop independent of early reactions; they occur despite recovery from acute radiation injury. Table 394-4 shows the frequency of radiation tolerance seen with fractionated radiotherapy at 5 years of follow-up. These numbers are rough estimates at best. The clinical manifestations of irradiation will depend on the volume of the organ irradiated, the total dose, the dose per fraction, and the length of time taken to deliver the dose. Dose per fraction is the most important factor determining normal tissue effects. In addition, the cellular consequences of treatment can be progressive over time. Thus, length of follow-up is also crucial in judging clinical sequelae. Central Nervous System Traditionally, the central nervous system (CNS) has been described as relatively resistant to radiation-induced changes. When the human brain is treated with standard fractionation (1.8 to 2.0 Gy/d), acute reactions are seldom observed. SubacuteCNSreactions to radiation treatment are more common. The clinical manifestations may include Lhermitte's sign, which is a self-limited paresthesia occurring with flexion of the neck. It is believed to be due to transient demyelination of the spinal cord following significant radiation exposure. It can be seen 1 to 3 months after completion of radiation treatment to the spinal cord. The frequency of Lhermitte's sign varies according to the type of radiation therapy and can be as high as 15% after mantle-field radiation. Mild encephalopathy and focal neurologic changes can occur after irradiation limited to the cranium. If radiation treatments to the brain are given at the same time that chemotherapeutic agents are administered, the effects can be more severe, presumably reflecting altered permeability to the drugs. The effect of cranial irradiation is believed to be secondary to radiation effects on the replicating oligodendrocytes and possibly on the microvasculature. Both clinical and radiologic changes may simulate tumor progression and can often pose diagnostic and treatment dilemmas. Postirradiation pathology and associated clinical symptoms typically begin 6 to 36 months after radiation therapy and are related to the total dose and volume treated. Fraction size appears to be the most important variable affecting the rate of postirradiation brain necrosis. Neurocognitive changes can also be seen in children after cranial irradiation. The important pretreatment factors that predict the degree of lateCNSeffects include the age at which cranial irradiation was given and neurocognitive functional level at the time of treatment.
A unique late effect of cranial irradiation combined with chemotherapy, known as leukoencephalopathy, has been described in some patients. Leukoencephalopathy is a necrotizing reaction usually noted 4 to 12 months after combined treatment with methotrexate and cranial irradiation. Dementia and dysarthria may progress to seizures, ataxia, or death. Transverse myelitis after radiation treatment is a spinal cord reaction similar to cerebral necrosis. This syndrome consists of progressive and irreversible leg weakness and loss of bladder function and sensation referrable to a single spinal cord level. Flaccid paralysis eventually occurs. Symptoms can occur as early as 6 months after radiation treatment, but the usual time to onset is 12 to 24 months. Lhermitte's sign does not correlate with transverse myelitis. Skin Skin reaction can be seen within 2 weeks of fractionated radiotherapy, a delay that correlates with the time required for cells to move from the basal to the keratinized layer of skin. The severity of the reaction depends on the skin dose per fraction and the total dose delivered to an area of skin. Erythema is observed, soon followed by dry desquamation. The skin at this time can be erythematous, warm, and sometimes edematous. The vessels in the upper dermis are dilated, and inflammatory infiltration with granulocytes, macrophages, eosinophils, plasma cells, and lymphocytes is noted. When a severe skin reaction occurs, it is usually located where the beam strikes the skin tangentially. Moist desquamation consists of eruption of the epidermal layer. Healing is through reepithelialization from cells of less affected basal layers. When skin reactions are severe, treatment interruptions are needed to permit healing. Dry desquamation is treated conservatively. Symptoms of dryness can be alleviated by advising the patient to wear only cotton fabric next to the affected skin and to refrain from the use of irritants of any kind. If treatment becomes necessary, hydrophilic agents that do not contain heavy metals are recommended. Petroleum jellies should not be used, as they may trap bacteria and increase the chance of infection. Moist desquamation is best managed by leaving the affected area dry and open to air. A chronic reaction to radiation can be seen starting 6 to 12 months after irradiation. The epidermis is usually atrophic and may be more easily injured than normal skin. Interstitial fibrosis may also be increased. Hyperpigmentation of irradiated skin outlining the treatment field can be seen within a couple of months after completion of irradiation. This will fade gradually. The skin becomes thin, and hair loss may be permanent. Radiation therapy can induce second malignancies, which tend to be more aggressive than cancers arising in patients without significant radiation exposure. Heart and Blood Vessels When cardiac disease appears after radiation treatment, it is often difficult to tell to what extent the radiation treatment was causative. The pathogenesis of atherosclerotic heart disease is multifactorial. Exposure of a large heart volume to high-dose radiation therapy accelerates the development of coronary artery disease. Acute "pericarditis" may result from cardiac irradiation. The symptoms may include chest pain and fever, with or without pericardial effusion. This syndrome is usually self-limited and typically manifests itself a few months after treatment. Asymptomatic pericardial effusion may be the most common manifestation of
radiation-induced heart disease. It is usually detected by chest x-ray and confirmed by an echocardiogram. Most patients with symptomatic radiation-induced constrictive pericarditis will have received more than 40 Gy to a large portion of the heart. The risk increases significantly with cardiac doses greater than 50 Gy. Chronic cardiac changes may have their onset from 6 months to several years after irradiation. The clinical symptoms may indicate chronic constrictive disease due to pericardial, myocardial, and endocardial fibrosis -- a pancarditis. The clinical signs may include dyspnea, chest pain, venous distention, pleural effusion, and paradoxical pulse. Lung The clinical symptoms of radiation pneumonitis can be separated into early and late phases. During the early phase, clinical manifestations may include dyspnea, cough, and fever. Shortness of breath is relatively infrequent. It is more common to observe only the radiologic changes on a chest x-ray, without clinical symptoms. The clinical signs and symptoms of radiation pneumonitis may appear in 3 to 6 weeks if a large region of lung is irradiated to a dose above 25 Gy. An infiltrate outlining the treatment field may become evident on the chest x-ray. Radiation changes should not occur outside the treated field. Computed tomography can often help in distinguishing radiation pneumonitis from other causes of the infiltrate. The incidence of radiation pneumonitis can be reduced with careful treatment planning designed to lower the total dose given to the treated lung volume. Permanent scarring that results in respiratory compromise may develop if the dose and the volume of lung irradiated are excessive. Dyspnea and cough may be severe and debilitating. Patients with symptoms of radiation pneumonitis may respond rapidly to glucocorticoids, but the medication has little effect on fibrotic changes. Glucocorticoids must be tapered very slowly to avoid rebound exacerbation of symptoms, which can prove lethal for some patients. Prophylactic administration of glucocorticoids is of no proven merit. Supportive care includes bronchodilators and oxygen at the lowest possible FIO2. Digestive Tract Pathologic changes of the epithelial layer occur early during radiation treatments. The underlying submucosa may become edematous, with dilation of capillaries. Recovery from radiation damage can be expected within a few weeks after completion of radiation therapy, provided that sufficient numbers of stem cells are left. The radioresponsiveness of the aerodigestive tract, like that of other structures, is not uniform but varies according to the location. Patients often have symptoms from radiation exposure that are similar to other forms of acute gastritis. The clinical signs include epigastric pain, loss of appetite, nausea, and vomiting. Decreased gastric acidity is observed after 15 to 20 Gy of fractionated radiation therapy. The tolerance of the stomach to radiation is also aggravated by addition of systemic chemotherapy, such as 5-fluorouracil. The germinal centers of the bowel mucosa are in the crypts of Lieberkuhn. Newly formed cells move upward along the walls of the crypts as transitional cells, undergoing maturation. The epithelial lining of the small bowel is the most rapidly renewed system in the human body and is completely renewed in 3 to 6 days. Within 12 to 24 h after the
first dose of radiation therapy, pathologic evidence of dead cells are seen in the mucosal lining. Complete denudation of the mucosal surface rarely occurs during a regular course of radiation treatment because of the high capacity of the mucosa for regeneration. However, a focal area of erosion may be seen. The histologic appearance may be nearly normal within 2 to 3 weeks after radiation therapy. Clinical manifestations of acute radiation enteropathy are nausea and vomiting, diarrhea, and cramping pain. Relevant factors contributing to the pathogenesis of diarrhea include malabsorption and alterations in the intestinal bacterial flora. The severity of symptoms, as in other anatomic areas, is proportional to the irradiated volume and the total dose. Symptoms of chronic radiation enteropathy include diarrhea, abdominal cramping, nausea, malabsorption, vomiting, and obstruction. Progressive fibrosis, perforation, fistula formation, and stenosis of the irradiated portion of the bowel can occur during the chronic phase of radiation enteropathy. Most clinical manifestations of chronic changes occur between 6 months and 5 years after radiation therapy. Conservative noninvasive treatment can frequently control gastrointestinal symptoms. A low-residue or elemental diet may be beneficial. When nonsurgical treatment fails to relieve severe symptoms, surgical intervention is often indicated. Bladder Radiation injury to the bladder generally becomes symptomatic 3 to 6 weeks after the start of treatment, and symptoms usually subside 3 to 4 weeks after completion of radiation therapy. Patients often complain of increased frequency and dysuria. Cystoscopy often shows diffuse mucosal changes similar to those of acute cystitis. Sometimes desquamation and ulceration can be seen. Without infection, urinary symptoms are managed symptomatically. Concurrent chemotherapy with cytotoxic agents such as cyclophosphamide increases the severity of the acute bladder reaction. The late effects of high radiation doses to the bladder may include interstitial fibrosis, telangiectasia, and ulceration. The blood vessels may be dilated and prone to rupture, resulting in painless hematuria. These changes are often difficult to distinguish from tumor recurrence and progression. A contracted bladder may result from doses in excess of 60 Gy. Testes and Ovaries In general, type B spermatogonia are exquisitely sensitive to the effects of radiation. The type A spermatogonia are thought to be more resistant because their longer cell cycle time allows considerable variation in radiosensitivity among different phases of the cell cycle. Sertoli cells and Leydig cells are less radiosensitive than the spermatogonia. Elevated levels of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) have been observed after as little as 75 cGy. Doses as low as 10 cGy to the testicles may result in injury to the type B spermatogonia. The single dose required for permanent sterilization on normal human males is believed to be between 6 and 10 Gy. In normal human males, sperm count recovery requires 9 to 18 months after a fractionated dose of 8 to 100 cGy. The radiation dose necessary to induce ovarian failure is age-dependent. A single dose of 3 to 4 Gy can induce amenorrhea in almost all women over 40 years of age. In young
women, oogenesis is much less sensitive to radiation than is spermatogenesis in men. ACUTE TOTAL-BODY IRRADIATION The data regarding the acute effects of total-body irradiation on humans come primarily from Japanese survivors of the atomic bomb, Marshallese exposed to radioactive fall-out in 1954, and persons exposed to radiation from the Chernobyl nuclear accident. Early symptoms of acute total-body irradiation, known as the prodromal radiation syndrome, last for a limited time. Clinical manifestations depend on the total-body dose. At doses >100 Gy, death usually occurs 24 to 48 h later from neurologic and cardiovascular failure. This is known as the cerebrovascular syndrome. Because cerebrovascular damage causes death very quickly, the failures of other systems do not have time to develop. At doses between 5 and 12 Gy, death may occur in a matter of days as a result of the gastrointestinal syndrome. The symptoms during this period may include nausea, vomiting, and prolonged diarrhea for several days leading to dehydration, sepsis, and death. A total-body dose >10 Gy is uniformly fatal unless supportive therapy (fluid, electrolytes, blood products, and antibiotics) is given. The process of intestinal denudation depends on the dose and may take between 3 and 120 days. Death from intestinal denudation usually occurs before the full effects of radiation on the blood-forming elements are seen. At total-body doses between 2 and 8 Gy, death may occur 2 to 4 weeks after exposure from bone marrow failure, the hematopoietic syndrome. The full effect of radiation is not apparent until the mature hematopoietic cells are depleted. Clinical symptoms during this period may include chills, fatigue, and petechial hemorrhage. Peripheral blood lymphopenia develops during the first 12 to 48 h after any significant exposure. Beyond 5 to 6 Gy, the rate and magnitude of the drop are not well correlated to radiation exposure. Some stem cells may survive acute exposure to³10 Gy. Death is from infection or bleeding and usually occurs before anemia can develop (red blood cell half-life is 100 to 120 days). The LD50/60(the dose at which 50% of the population is dead by 60 days) is around 3.25 Gy if support is not given. There is considerable variability in the total-body dose tolerated. The very young and the old are more radiosensitive than middle-aged and young adult individuals. Females in general appear to be more tolerant of radiation than males. Persons exposed to25 years. Thus, blood lead levels may decline significantly while the body's total burden of lead remains heavy. The toxicity of lead is probably related to its affinity for cell membranes and mitochondria, as a result of which it interferes with mitochondrial oxidative phosphorylation and sodium, potassium, and calcium ATPases. Lead impairs the activity of calcium-dependent intracellular messengers and of brain protein kinase C. In addition, lead stimulates the formation of inclusion bodies that may translocate the metal into cell nuclei and alter gene expression. CLINICAL TOXICOLOGY Symptomatic lead poisoning in childhood generally develops at blood lead levels>3.9 umol/L (80 ug/dL) and is characterized by abdominal pain and irritability followed by lethargy, anorexia, pallor (resulting from anemia), ataxia, and slurred speech. Convulsions, coma, and death due to generalized cerebral edema and renal failure
occur in the most severe cases. Subclinical lead poisoning [blood lead level>1.4 umol/L (> 30 ug/dL)] can cause mental retardation and selective deficits in language, cognitive function, balance, behavior, and school performance despite the lack of discernible symptoms. Epidemiologic studies and meta-analyses of studies regarding lead's effect on the intellectual function of children indicate that cognition is probably impaired in a dose-related fashion at blood lead levels well below 1.4 umol/L (30 ug/dL) and that no threshold for this effect is likely to exist above the lowest measurable blood lead level of 0.05 umol/L (1 ug/dL). The impact is greatest when the exposure is of long duration and has been most apparent when it takes place around the age of 2 years; however, the impact of fetal lead exposure remains to be clarified, particularly in view of the observation that maternal bone lead stores can be mobilized to a significant degree during pregnancy, with consequent exposure of the fetus. In adults, symptomatic lead poisoning, which usually develops when blood lead levels exceed 3.9 umol/L (80 ug/dL) for a period of weeks, is characterized by abdominal pain, headache, irritability, joint pain, fatigue, anemia, peripheral motor neuropathy, and deficits in short-term memory and the ability to concentrate. Encephalopathy is rare. A "lead line" sometimes appears at the gingiva-tooth border after prolonged high-level exposure. Some individuals develop these symptoms and signs at lower blood lead levels [1.9 to 3.9 umol/L (40 to 80 ug/dL)] and/or with briefer periods of exposure. Chronic subclinical lead exposure is associated with interstitial nephritis, tubular damage (with tubular inclusion bodies), hyperuricemia (with an increased risk of gout), and a decline in glomerular filtration rate and chronic renal failure. Epidemiologic evidence also suggests that blood lead levels in the range of 0.34 to 1.7 umol/L (7 to 35 ug/dL) are associated with increases in blood pressure, decreases in creatinine clearance, and decrements in cognitive performance that are too small to be detected as a lead effect in individual cases but nevertheless may contribute significantly to the causation of chronic disease. An additional issue for both children and adults is whether lead that has accumulated in bone and lain dormant for years can pose a threat later in life, particularly at times of increased bone resorption such as pregnancy, lactation, and senile osteoporosis. Elevation of the bone lead level appears to be a risk factor for anemia, hypertension, cardiac conduction delays, and impairment of cognitive function. Hyperthyroidism has been reported to cause lead toxicity in adults by mobilizing stores of bone lead acquired during childhood. Genetic polymorphisms, such as variants of the gene that codes for aminolevulinic acid dehydratase (a critical enzyme in the production of heme) or the C282Y hemochromatosis gene, may confer differences in susceptibility to lead retention and toxicity; ~15% of Caucasians have a variant form of one of these genes. This issue is the focus of continued research. LABORATORY FINDINGS In 1991, the Centers for Disease Control and Prevention designated 0.48 umol/L (10 ug/dL) as the blood lead level of concern in children. A specific set of interventions is recommended when the level exceeds this value. OSHA requires the regular measurement of blood lead in lead-exposed workers and the maintenance of blood lead
levels10 mmol/L -- suggests the presence of a low-molecular-weight solute such as an alcohol, glycol, or ketone or an unmeasured electrolyte or sugar. The osmolal gap can also provide an estimate of the amount of anion present (Table 396-3). Ketosis suggests acetone, isopropyl alcohol, or salicylate poisoning. Hypoglycemia may be due to poisoning withb-adrenergic blockers, ethanol, insulin, oral hypoglycemic agents, quinine, and salicylates, whereas hyperglycemia can occur in poisoning with acetone, b-adrenergic agonist, calcium channel blockers, iron, theophylline, or Vacor. Hypokalemia can be caused by barium, a b-adrenergic agonist, a diuretic, theophylline, or toluene; hyperkalemia suggests poisoning with ana-adrenergic agonist, ab-adrenergic blocker, cardiac glycosides, or fluoride. Radiologic studies may also be useful for diagnostic purposes. Pulmonary edema (adult respiratory distress syndrome, or ARDS) can be caused by poisoning with carbon monoxide, cyanide, an opioid, paraquat, phencyclidine, a sedative-hypnotic, or salicylate; by inhalation of irritant gases, fumes, or vapors (ammonia, metal oxides, mercury); or by prolonged anoxia, hyperthermia, or shock. Aspiration pneumonia is common in patients with coma, seizures, and petroleum distillate ingestion. Radiopaque densities may be visible on abdominal x-rays following the ingestion of calcium salts, chloral hydrate, chlorinated hydrocarbons, heavy metals, illicit drug packets, iodinated compounds, potassium salts, psychotherapeutic agents, lithium, phenothiazines, enteric-coated tablets, or salicylates. Theelectrocardiogram (ECG) can be useful to assist with the differential diagnosis and to guide treatment. Bradycardia and atrioventricular (AV) block may occur in patients poisoned bya-adrenergic agonists, antiarrhythmic agents, beta blockers, calcium channel blockers, cholinergic agents (carbamate and organophosphate insecticides), cardiac glycosides, lithium, magnesium, or tricyclic antidepressants. QRS- and QT-interval prolongation may be caused by hyperkalemia and by membrane-active drugs (Table 396-1). Ventricular tachyarrhythmias may be seen in poisoning with cardiac glycosides, fluorides, membrane-active drugs, sympathomimetics, or agents that cause hyperkalemia or potentiate the effects of endogenous catecholamines (e.g.,
chloral hydrate, aliphatic and halogenated hydrocarbons). Analysis of urine and blood (and occasionally of gastric contents and chemical samples) may be useful to confirm or rule out suspected poisoning. Interpretation of laboratory data requires knowledge of the tests used for screening and confirmation (thin-layer, gas-liquid, or high-performance liquid chromatography; colorimetric and fluorometric assays; enzyme-multiplied and radioimmunoassays; gas chromatography; mass spectrometry), their sensitivity (limit of detection) and specificity, the preferred biologic specimen for analysis, and the optimal time of specimen sampling. Personal communication with the laboratory is essential. A negative result on a screen may mean the substance is not detectable by the test used or that its concentration is too low for detection at the time of sampling. In the latter case, repeating the test at a later time may yield a positive result. Although some rapid screening tests for a limited number of drugs of abuse are available, comprehensive screening tests require 2 to 6 h for completion, and immediate management must be based on the history, physical examination, and routine ancillary tests. In addition, when the patient is asymptomatic, or when the clinical picture is consistent with the reported history, qualitative screening is neither clinically useful nor cost-effective. It is of greatest value in patients with severe or unexplained toxicity who have coma, seizures, cardiovascular instability, metabolic or respiratory acidosis, and nonsinus cardiac rhythms. Quantitative analysis is useful for poisoning with acetaminophen, acetone, alcohol (including ethylene glycol), antiarrhythmics, anticonvulsants, barbiturates, digoxin, heavy metals, lithium, paraquat, salicylate, and theophylline, as well as for carboxyhemoglobin and methemoglobin. Results can often be available within an hour. The response to antidotes may be useful for diagnostic purposes. Resolution of altered mental status and abnormal vital signs within minutes of intravenous administration of dextrose, naloxone, or flumazenil is virtually diagnostic of hypoglycemia, narcotic poisoning, and benzodiazepine intoxication, respectively. The prompt reversal of acute dystonic (extrapyramidal) reactions following an intravenous dose of benztropine or diphenhydramine confirms a drug etiology. Although the reversal of both central and peripheral manifestations of anticholinergic poisoning by physostigmine is diagnostic, this antidote may cause arousal in patients withCNSdepression of any etiology. TREATMENT General Principles Treatment goals include support of vital signs, prevention of further poison absorption, enhancement of poison elimination, administration of specific antidotes, and prevention of reexposure (Table 396-4). Specific treatment depends on the identity of the poison, the route and amount of exposure, the time of presentation relative to the time of exposure, and the severity of poisoning. Knowledge of the offending agents' pharmacokinetics and pharmacodynamics is essential. During the pretoxic phase, prior to the onset of poisoning, decontamination is the highest priority, and treatment is based solely on the history. The maximum potential toxicity based on the greatest possible exposure should be assumed. Since decontamination is more effective when accomplished soon after exposure, the initial
history and physical examination should be focused and brief. It is also advisable to establish intravenous access and initiate cardiac monitoring, particularly in patients with potentially serious ingestions or unclear histories. When an accurate history is not obtainable, and a poison causing delayed toxicity or irreversible damage is suspected, blood and urine should be sent for toxicologic screening and, if indicated, for quantitative analysis. During absorption and distribution, blood levels may be greater than those in tissue and may not correlate with toxicity. However, high blood levels of agents whose metabolites are more toxic than the parent compound (acetaminophen, ethylene glycol, or methanol) may indicate the need for additional interventions (antidotes, dialysis). Most patients who remain or become asymptomatic 4 to 6 h after ingestion will not develop subsequent toxicity and can be discharged safely. Longer observation may be necessary for patients who have ingested agents that slow gastric emptying and intestinal motility as this will delay dissolution, absorption, and distribution characteristics. Extended observation may also be indicated for agents that are converted in the body to toxic metabolites (Table 396-1). During the toxic phase, the time between the onset of poisoning and the peak effects, management is based primarily on clinical and laboratory findings. Effects after an overdose begin sooner, peak later, and last longer than they do after a therapeutic dose. Resuscitation and stabilization are the first priority. All symptomatic patients should have an intravenous line, oxygen saturation determination, cardiac monitoring, and continuous observation. Baseline laboratory, ECG, and x-ray evaluation may also be appropriate. Intravenous glucose (unless documented to be normal), naloxone, and thiamine should be considered in patients with altered mental status, particularly those with soma or seizures. Decontamiation may also be appropriate. Measures that enhance poison elimination may shorten the duration of toxicity and lessen its severity. However, the risks must be weighed against the benefits. Diagnostic certainty (usually via laboratory confirmation) is generally a prerequisite. Intestinal dialysis with repetitive doses of activated charcoal is usually safe and can enhance the elimination of many poisons. Diuresis and chelation therapy enhance the elimination of a relatively small number of poisons, and their use is associated with potential complications. Extracorporal methods are effective in removing many poisons, but their expense and risk make their use reasonable only in patients who would otherwise have an unfavorable outcome. During the resolution phase of poisoning, supportive care and monitoring should continue until clinical, laboratory, andECGabnormalities have resolved. Since chemicals are eliminated from the blood before tissues, blood levels are usually lower than tissue levels during this phase and may not correlate with toxicity. This is particularly true when extracorporeal elimination procedures are used. Redistribution from tissues may cause a rebound increase in the blood level after termination of these procedures. When a metabolite is responsible for toxic effects, continued treatment of an asymptomatic patient might be necessary because of a potentially toxic blood level (acetaminophen, ethylene glycol, and methanol).
Supportive Care The goal of supportive therapy is to maintain physiologic homeostasis until detoxification is accomplished and to prevent and treat secondary complications such as aspiration, bedsores, cerebral and pulmonary edema, pneumonia, rhabdomyolysis, renal failure, sepsis, thromboembolic disease, and generalized organ dysfunction due to prolonged hypoxia or shock. Admission to an intensive care unit is indicated for the following: patients with severe poisoning (coma, respiratory depression, hypotension, cardiac conduction abnormalities, cardiac arrhythmias, hypothermia or hyperthermia, seizures); those needing close monitoring, antidotes, or enhanced elimination therapy; those showing progressive clinical deterioration; and those with significant underlying medical problems. Patients with mild to moderate toxicity can be managed on a general medical service, intermediate care unit, or emergency department observation area, depending on the anticipated duration and level of monitoring needed (intermittent clinical observation versus continuous clinical, cardiac, and respiratory monitoring). Patients who have attempted suicide require continuous observation and measures to prevent self-injury until they are thought unlikely to make further attempts. Respiratory care Endotracheal intubation for protection against the aspiration of gastrointestinal contents is of paramount importance in patients withCNSdepression or seizures as this complication can increase morbidity and mortality. Mechanical ventilation may be necessary for patients with respiratory depression or hypoxia and to facilitate therapeutic sedation or paralysis in order to prevent hyperthermia, acidosis, and rhabdomyolysis associated with neuromuscular hyperactivity. Since clinical assessment of respiratory function is often inaccurate, the need for oxygenation and ventilation is best determined by oximetry or arterial blood gas analysis. The gag reflex is not a reliable indicator of the need for intubation. A patient may maintain airway patency while being stimulated but not if left unattended. Those who cannot respond to voice or who are unable to sit and drink fluids without assistance are best managed by prophylactic intubation. Drug-induced pulmonary edema is usually noncardiac rather than cardiac in origin. ProfoundCNSdepression and cardiac conduction abnormalities suggest the latter etiology. Measurement of pulmonary artery pressure may be necessary to establish etiology and direct appropriate therapy. Extracorporeal measures (membrane oxygenation, venoarterial perfusion, cardiopulmonary bypass), partial liquid (perfluorocarbon) ventilation, and hyperbaric oxygen therapy may be appropriate for severe but reversible respiratory failure. Cardiovascular therapy Maintenance of normal tissue perfusion is critical to allow for complete recovery once the offending agent has been eliminated. If hypotension is unresponsive to volume expansion, treatment with norepinephrine, epinephrine, or high-dose dopamine may be necessary (Chap. 38). Intraaortic balloon pump counterpulsation and venoarterial or cardiopulmonary perfusion techniques should be considered for severe but reversible cardiac failure. Bardyarrhythmias associated with hypotension generally should be treated as described inChap. 229. Glucagon and calcium may be effective in both beta blocker and calcium channel blocker poisoning. Antibody therapy may be indicated for cardiac glycoside poisoning.
Supraventricular tachycardia associated with hypertension andCNSexcitation is almost always due to agents that cause generalized physiologic excitation (Table 396-1). Most cases are mild or moderate in severity and require only observation or nonspecific sedation with a benzodiazepine. For cases that are severe or associated with hemodynamic instability, chest pain, orECGevidence of ischemia, specific therapy is indicated. For patients with sympathetic hyperactivity, treatment with a combined alpha and beta blocker (labetalol), a calcium channel blocker (verapamil or diltiazem), or a combination of a beta blocker and a vasodilator (esmolol and nitroprusside) is preferred. For those with anticholinergic poisoning, physostigmine is the treatment of choice. Supraventricular tachycardia without hypertension is generally secondary to vasodilation or hypovolemia and responds to fluid administration. Lidocaine and phenytoin are generally safe for ventricular tachyarrhythmias, but beta blockers can be hazardous unless the arrhythmia is clearly due to sympathetic hyperactivity. For ventricular tachyarrhythmias due to tricyclic antidepressants and probably other membrane-active agents (Table 396-1), class IA, IC, and III antiarrhythmic agents are contraindicated (because of similar electrophysiologic effects), but sodium bicarbonate may be helpful. Magnesium sulfate and overdrive pacing (by isoproterenol or a pacemaker) may be useful in patients with torsade de pointes and prolonged QT intervals. Magnesium and antidigoxin antibodies should be considered in patients with severe cardiac glycoside poisoning. Invasive (esophageal or intracardiac)ECGrecording may be necessary to determine the origin (ventricular or supraventricular) of wide-complex tachycardias (Chap. 230). If the patient is hemodynamically stable, however, it may be prudent to observe rather than to treat with another potentially proarrhythmic agent. Arrhythmias may be resistant to drug therapy until underlying acid-base, electrolyte, oxygenation, and temperature derangements are corrected. Central nervous system therapies Neuromuscular hyperactivity and seizures can lead to hyperthermia, lactic acidosis, and rhabdomyolysis, with their attendant complications, and should be treated aggressively. Seizures caused by excessive stimulation of catecholamine receptors (sympathomimetic or hallucinogen poisoning and drug withdrawal), or decreased activity ofg-aminobutyric acid (GABA) (isoniazid poisoning) or glycine (strychnine poisoning) receptors are best treated with enhancers of GABA effects such as benzodiazepines or barbiturates. Since benzodiazepines and barbiturates act by slightly different mechanisms (the former increases the frequency and the latter increases the duration of chloride channel opening in response to GABA), therapy with both may be effective when neither is effective alone. Seizures caused by isoniazid, which inhibits the synthesis of GABA, may require high doses of pyridoxine (which facilitates the synthesis of GABA). Seizures resulting from membrane destabilization (beta blocker or cyclic antidepressant poisoning) may require a membrane-active anticonvulsant such as phenytoin as well as GABA enhancers. For poisons with central dopaminergic effects (phencyclidine), an agent with opposing activity, such as haloperidol, may be useful. In anticholinergic and cyanide poisoning, specific antidotal therapy may be necessary. The treatment of seizures secondary to ischemia, edema, or metabolic abnormalities should include correction of the underlying cause. Neuromuscular paralysis is indicated in refractory cases. Electroencephalographic (EEG) monitoring and continuing treatment of seizures are necessary to prevent permanent neurologic damage.
Other measures Temperature extremes, metabolic abnormalities, hepatic and renal dysfunction, and secondary complications should be treated by standard therapies. Prevention of Poison Absorption Gastrointestinal Decontamination Whether or not to perform gastrointestinal decontamination, and which procedure to use, depends on the time since ingestion; the existing and predicted toxicity of the ingestant; the availability, efficacy, and contraindications of the procedure; and the nature, severity, and risk of complications. In animal and human volunteer studies, the efficacy of activated charcoal, gastric lavage, and syrup of ipecac decreases with time, and there are insufficient data to support or exclude a beneficial effect when they are used more than 1 h after ingestion. Due to the lack of clinical studies using control groups without treatment, the efficacy of these procedures for improving the outcome of overdose patients has not been established. The average time from ingestion to presentation for treatment is over 1 h for children and over 3 h for adults. Most patients will recover from poisoning uneventfully with good supportive care alone, but complications of gastrointestinal decontamination, particularly aspiration, can prolong this process. Hence, gastrointestinal contamination should be performed selectively, not routinely, in the management of overdose patients. It is clearly unnecessary when predicted toxicity is minimal or the time of expected maximal toxicity has passed without significant effect. Activated charcoal has comparable or greater efficacy, fewer contraindications and complications, is less aversive and invasive than ipecac or gastric lavage, and is the preferred method of gastrointestinal decontamination in most situations. Activated charcoal is prepared as a suspension in water, either alone or with a cathartic. It is given orally via a nippled bottle (for infants), or via a cup, straw, or small-bore nasogastric tube. The recommended dose is 1 g/kg body weight, using 8 mL of diluent per gram of charcoal if a premixed formulation is not available. Palatability may be increased by adding a sweetener (sorbitol) or a flavoring agent (cherry, chocolate, or cola syrup) to the suspension. Charcoal adsorbs ingested poisons within the gut lumen, allowing the charcoal-toxin complex to be evacuated with stool. The complex can also be removed from the stomach by induced emesis or lavage. In vitro, charcoal adsorbs³90% of most substances when given in an amount equal to 10 times the weight of the substance. Charged (ionized) chemicals such as mineral acids, alkalis, and highly dissociated salts of cyanide, fluoride, iron, lithium, and other inorganic compounds are not well adsorbed by charcoal. In animal and human volunteer studies, charcoal decreases the absorption of ingestants by an average of 73% when given within 5 min of ingestant administration, 51% when given at 30 min, and 36% at 60 min. Charcoal is at least equally as effective as ipecac syrup or gastric lavage. Experimentally, lavage followed by charcoal is more effective than charcoal alone, and charcoal before and after lavage is more effective than charcoal alone or charcoal after lavage. In the treatment of poisoned patients, however, charcoal alone generally results in a better clinical outcome than either treatment with ipecac followed by charcoal or lavage followed by charcoal. Side effects of charcoal include nausea, vomiting, and diarrhea or constipation. Charcoal may also prevent the absorption of orally administered
therapeutic agents. Complications include mechanical obstruction of the airway, aspiration, vomiting, and bowel obstruction and infection caused by inspissated charcoal. Charcoal is not recommended for patients who have ingested corrosives because it obscures endoscopy. Gastric lavage is performed by sequentially administering and aspirating about 5 mL fluid per kilogram of body weight through a no. 28 French orogastric tube in children and a no. 40 French tube in adults. Except for infants, tap water is acceptable. The patient should be placed in Trendelenburg and left lateral decubitus positions to prevent aspiration (even if an endotracheal tube is in place). Lavage decreases ingestant absorption by an average of 52% if performed within 5 min of ingestion administration, 26% if performed at 30 min, and 16% if performed at 60 min. Its efficacy is similar to that of ipecac. Significant amounts of ingested drug are recovered in one-tenth of patients. Aspiration is a common complication (occurring in up to 10% of patients), especially when lavage is perfomed improperly. Serious complications (tracheal lavage, esophageal and gastric perforation) occur in approximately 1% of patients. For this reason, the physician should personally insert the lavage tube and confirm its placement, and the patient must be cooperative or adequately restrained (with pharmacologic sedation if necessary) during the procedure. Gastric lavage is contraindicated in corrosive or petroleum distillate ingestions because of the respective risks of gastroesophageal perforation and aspiration-induced hydrocarbon pneumonitis. Syrup of ipecac can be used for the home management of patients with accidental ingestions, reliable histories, and mild predicted toxicity. It may delay the administration and decrease the effectiveness of activated charcoal, oral antidotes, and whole-bowel irrigation and is very rarely appropriate for patients treated at a health care facility. It is administered orally in a dose of 30 mL for adults, 15 mL for children, and 10 mL for small infants. Clear liquids should also be given. Ipecac irritates the stomach and stimulates the central chemoreceptor trigger zone. Vomiting usually occurs about 20 min after administration. The dose may be repeated if vomiting does not occur. In animal and human volunteer studies, ipecac decreases ingestant absorption by an average of 60% if given within 5 min of ingestant administration, 32% if given at 30 min, and 30% if given at 60 min. Side effects include lethargy in children (12%) and protracted vomiting (8 to 17%). Chronic ipecac use (by patients with anorexia nervosa or bulimia) may cause electrolyte and fluid abnormalities, cardiac toxicity, and myopathy. Except for aspiration, serious complications are rare. Gastric or esophageal tears and perforations and stroke have been reported. Ipecac is contraindicated in patients with recent gastrointestinal surgery,CNSdepression, or seizures, and in those who have ingested corrosives or rapidly acting CNS poisons (camphor, cyanide, tricyclic antidepressants, propoxyphene, strychnine). Whole-bowel irrigation is performed by administering a bowel-cleansing solution containing electrolytes and polyethylene glycol (Golytely, Colyte) orally or by gastric tube at a rate of up 0.5 L/h in children and 2.0 L/h in adults until rectal effluent is clear. The patient must be in a sitting position. Although data are limited, whole-bowel irrigation may be at least equally as effective as other decontamination procedures. It may be appropriate for those who have ingested foreign bodies, packets of illicit drugs, slow-release or enteric-coated medications, and agents that are poorly adsorbed by charcoal (e.g., heavy metals). It is contraindicated in patients with bowel obstruction,
ileus, hemodynamic instability, and compromised unprotected airways. Cathartic salts (disodium phosphate, magnesium citrate and sulfate, sodium sulfate) or saccharides (mannitol, sorbitol) promote the rectal evacuation of gastrointestinal contents. The most effective cathartic is sorbitol in a dose of 1 to 2 g/kg of body weight. Alone, cathartics do not prevent ingestant absorption and should not be used as a method of gut decontamination. Their primary use is to prevent constipation following charcoal administration. Abdominal cramps, nausea, and occasional vomiting are side effects. Complications of repeated dosing include hypermagnesemia and excessive diarrhea. Cathartics are contraindicated in patients who have ingested corrosives and in those with preexisting diarrhea. Magnesium-containing cathartics should not be used in patients with renal failure. Dilution (i.e., drinking 5 mL/kg of body weight of water or another clear liquid) should be accomplished as soon as possible after the ingestion of corrosives (acids, alkali). However, it may increase the dissolution rate (and hence absorption) of capsules, tablets, and other solid ingestants and should not be used in these circumstances. Endoscopic or surgical removal of poisons may be useful in rare situations, such as ingestion of a potentially toxic foreign body that fails to transit the gastrointestinal tract, a potentially lethal amount of a heavy metal (arsenic, iron, mercury, thallium), or agents that have coalesced into gastric concretions or bezoars (barbiturates, gluthethimide, heavy metals, lithium, meprobamate, sustained-release preparations). Patients who become toxic from cocaine due to its leakage from multiple ingested drug packets require immediate surgical intervention. Decontamination of Other Sites Immediate, copious flushing with water, saline, or another available clear, drinkable liquid is the initial treatment for topical exposures (exceptions include alkali metals, calcium oxide, phosphorus). Saline is preferred for eye irrigation. A triple wash (water, soap, water) may be best for dermal decontamination. Inhalational exposures should be treated initially with fresh air or oxygen. The removal of liquids from body cavities such as the vagina or rectum is best accomplished by irrigation. Solids (drug packets, pills) should be removed manually, preferably with visual guidance. Enhancement of Poison Elimination Although the elimination of most poisons can be accelerated by therapeutic interventions, the pharmacokinetic efficacy (removal of drug at a rate greater than that accomplished by intrinsic elimination) and clinical benefit (in terms of a shortened duration of toxicity or improved outcome) of such interventions are often more theoretical than proven. Hence, the decision to use such measures should be based on the actual or predicted toxicity and the potential efficacy, cost, and risks of therapy. Multiple-Dose Activated Charcoal Repetitive oral dosing with charcoal can enhance the elimination of previously absorbed substances by binding them within the gut as they are excreted in the bile, secreted by gastrointestinal cells, or passively diffuse into the gut lumen (reverse absorption or enterocapillary exsorption). Doses of 0.5 to 1 g/kg body weight every 2 to 4 h, adjusted downward to avoid regurgitation in patients with decreased gastrointestinal motility, are generally recommended. Experimentally, this
treatment enhances the elimination of nearly all substances tested. Pharmacokinetic efficacy approaches that of hemodialysis for some agents (e.g., phenobarbital, theophylline). Multiple-dose therapy is not effective in accelerating elimination of chlorpropamide, tobramycin, or agents that adsorb poorly to charcoal. Complications include intestinal obstruction, pseudoobstruction, and nonocclusive intestinal infarction in patients with decreased gut motility. Forced Diuresis and Alteration of Urinary pH Diuresis and ion trapping via alteration of urine pH may prevent the renal reabsorption of poisons that undergo excretion by glomerular filtration and active tubular secretion. Since membranes are more permeable to nonionized molecules than to their ionized counterparts, acidic (low-pKa) poisons are ionized and trapped in an alkaline urine, and basic poisons are ionized and trapped in an acid urine. Saline diuresis can enhance the renal excretion of alcohols, bromide, calcium, fluoride, lithium, meprobamate, potassium, and isoniazid. Alkaline diuresis (a urine pH³7.5 and a urine output of 3 to 6 mL/kg body weight per hour) enhances the elimination of chlorphenoxyacetic acid herbicides, chlorpropamide, diflunisal, fluoride, methotrexate, phenobarbital, sulfonamides, and salicylates. Contraindications include congestive heart failure, renal failure, and cerebral edema. Acid-base, fluid, and electrolyte parameters should be monitored carefully. Acid diuresis enhances the renal elimination of amphetamines, chloroquine, cocaine, local anesthetics, phencyclidine, quinidine, quinine, strychnine, sympathomimetics, tricyclic antidepressants, and tocainide. Its use, however, has been largely abandoned because of potential complications and lack of clinical efficacy. Extracorporeal Removal Peritoneal dialysis, hemodialysis, charcoal or resin hemoperfusion, hemofiltration, plasmapheresis, and exchange transfusion are capable of removing any toxin from the bloodstream. Agents most amenable to enhanced elimination by dialysis have low molecular mass (60 min from medical care, a proximal lympho-occlusive constriction band may limit the spread of venom when applied within 30 min. To avoid worsening tissue damage, however, the band should not interrupt arterial blood flow. The bitten extremity should be splinted if possible and kept at approximately heart level. Measures to be avoided include incising or cooling the bite site, giving the victim an alcoholic beverage, or applying electric shocks. For elapid or sea snake bites, the Australian pressure-immobilization technique, in which the entire bitten extremity is wrapped with an elastic or crepe bandage and then splinted, is highly effective. The bandage is applied with the same snugness used for a sprained ankle. This technique greatly restricts absorption and circulation of venom. The utility of this method in viperid poisoning requires further research, as it theoretically could compound local tissue damage by restricting venom to the local tissues. Hospital Management In the hospital, the victim should be closely monitored (vital signs, cardiac rhythm, and oxygen saturation) while a history is quickly obtained and a
brief but thorough physical examination is performed. The level of erythema and/or swelling in a bitten extremity should be marked and limb circumferences measured in several locations every 15 min until swelling has stabilized. Large-bore intravenous access in unaffected extremities should be established. Early hypotension is due to pooling of blood in the pulmonary and splanchnic vascular beds. Hours later, hemolysis and loss of intravascular volume into soft tissues may play important roles. Fluid resuscitation with normal saline or Ringer's lactate should be initiated for clinical shock. If the blood pressure response is inadequate after administration of 20 to 40 mL/kg of body weight, then a trial of 5% albumin (10 to 20 mL/kg) is in order. If tissue perfusion fails to respond to volume resuscitation and antivenom infusion (see below), vasopressors (e.g., dopamine) should be administered. Invasive hemodynamic monitoring (central venous and/or pulmonary arterial pressures) can be helpful in such cases, although obtaining access is riskier if coagulopathy is present. Blood should be drawn for laboratory evaluation as soon as possible. Blood typing and cross-matching procedures can be affected over time by circulating venom. Also important are a complete blood count to evaluate the degree of hemorrhage or hemolysis, studies of renal and hepatic function, coagulation studies to identify signs of consumptive coagulopathy, and testing of urine for blood or myoglobin. In severe cases or in the face of significant comorbidity, arterial blood gas studies, electrocardiography, and chest radiography are indicated. Attempts to locate a source of appropriate antivenom should begin early in all cases of known venomous snakebite, regardless of symptoms. In the event that signs and symptoms progress rapidly, any delay in the administration of antivenom is dangerous. Antivenoms rarely offer cross-protection against snake species other than those used in their production unless the species are closely related. An example of good cross-protection is that of Australian tiger snake (Notechis scutatus) antivenom for sea snake bites (see below). The package insert accompanying a particular antivenom should be consulted for information regarding the spectrum of coverage. In the United States, assistance in finding antivenom can be obtained 24 hours a day from the University of Arizona Poison and Drug Information Center (telephone: 520-626-6016). Rapidly progressive or severe local findings (soft tissue swelling, ecchymosis, petechiae, etc.) or manifestations of systemic toxicity (signs and symptoms or laboratory abnormalities) are indications for the administration of intravenous antivenom. The package insert outlines techniques for reconstitution of antivenom (when necessary), skin-testing procedures (for potential allergy), and appropriate starting doses. Most commercial antivenoms are of equine origin and carry a risk of anaphylactic, anaphylactoid, and delayed-hypersensitivity reactions. Skin testing does not reliably predict which patients will have an allergic reaction to equine antivenom; false-negative and false-positive results are common. Before antivenom infusion, the patient should receive appropriate loading doses of intravenous antihistamines (e.g., diphenhydramine, 1 mg/kg to a maximum of 100 mg; and cimetidine, 5 to 10 mg/kg to a maximum of 300 mg) in an effort to limit acute reactions. Modest expansion of the patient's intravascular volume with crystalloids may also be beneficial in this regard. Epinephrine should be immediately available, and the antivenom dose to be administered should be diluted (e.g., in 1000 mL of normal saline for adults or in 20 mL/kg for children). This volume can be decreased if necessary (e.g., if the victim has a history of congestive heart
failure). The antivenom should be started slowly, with the physician at the bedside to intervene in the event of an acute reaction. The rate of infusion can be increased gradually in the absence of allergic phenomena until the total starting dose has been administered (over a total period of 1 to 4 h). Further antivenom may be necessary if the patient's clinical condition worsens. Laboratory values should be rechecked hourly, particularly if abnormal, until stability is apparent. The management of a life-threatening envenomation in a victim with an apparent allergy to antivenom requires significant expertise. Consultation with a poison specialist, an intensive care specialist, and/or an allergist is recommended. Often antivenom can still be administered in these situations under closely controlled conditions and with intensive premedication (e.g., with epinephrine, antihistamines, and steroids). Care of the bite wound should include application of a dry sterile dressing and splinting of the extremity with padding between the digits. Once the administration of an indicated antivenom has been initiated, the extremity should be elevated above heart level to relieve edema. Tetanus immunization should be updated as appropriate. The use of prophylactic antibiotics is controversial, as the incidence of secondary infection following venomous snakebite appears to be low. Many authorities, however, prescribe a broad-spectrum antibiotic (such as ampicillin or a cephalosporin) for the first few days. If swelling in the bitten extremity raises concern that subfascial muscle edema may be impeding tissue perfusion (muscle-compartment syndrome), intracompartmental pressures should be checked by any minimally invasive technique (e.g., the wick catheter). If pressures are elevated and remain so despite antivenom administration, prompt surgical consultation for possible fasciotomy should be obtained. This complication, fortunately, is rare after snakebites. Whether or not antivenom is given, any patient with signs of venom poisoning should be observed in the hospital for at least 24 h. A patient with an apparently "dry" bite should be watched for at least 6 to 8 h before discharge, as significant toxicity occasionally develops after a delay of several hours. The onset of systemic symptoms is commonly delayed for a number of hours after bites by several of the elapids (including the coral snakes) and sea snakes. Patients bitten by these reptiles should be observed in the hospital for 24 h. Significant work is being done in several regions of the world to produce safer, more effective antivenoms. Much of this work involves production of ovine-based antivenoms that are further purified and enzymatically cleaved to yield functional F(ab) fragments of the immunoglobulin molecules. These antivenoms are currently in clinical use in many countries, and trials of one product are under way in the United States. MORBIDITY AND MORTALITY The overall mortality rates for venomous snakebite are low in areas of the world with rapid access to medical care and appropriate antivenom. In the United States, for example, the mortality rate is 75 ug/100 g of foodstuff is considered hazardous to humans. In the 1972 New England "red tide," the concentration of saxitoxin in blue mussels exceeded 9000 ug/100 g of foodstuff. Saxitoxin appears to block sodium conductance, inhibiting neuromuscular transmission at the axonal and muscle membrane levels. Within minutes to a few hours after ingestion of contaminated shellfish, there is the onset of intraoral and perioral paresthesias, notably of the lips, tongue, and gums, that progress rapidly to involve the neck and distal extremities. The tingling or burning sensation later changes to numbness. Other symptoms rapidly develop and include lightheadedness, disequilibrium, incoordination, weakness, hyperreflexia, incoherence, dysarthria, sialorrhea, dysphagia, thirst, diarrhea, abdominal pain, nausea, vomiting, nystagmus, dysmetria, headache, diaphoresis, loss of vision, chest pain, and tachycardia. Flaccid paralysis and respiratory insufficiency may follow 2 to 12 h after ingestion. In the absence of hypoxia, the victim often remains alert but paralyzed. Treatment is supportive and based on symptoms. If the victim comes to medical attention within the first few hours after poison ingestion, the stomach should be emptied by gastric lavage and then irrigated with 2 L (in 200-mL aliquots) of a solution of 2% sodium bicarbonate. The administration of activated charcoal (50 to 100 g) and a cathartic (sorbitol, 20 to 50 g) makes empirical sense but has not been proved effective. Some authors advise against administration of magnesium-based solutions, such as certain cathartics, cautioning that hypermagnesemia may contribute to suppression of nerve conduction. The most serious problem is respiratory paralysis. The victim should be closely observed in a hospital for at least 24 h for respiratory distress. With prompt recognition of ventilatory failure, endotracheal intubation and assisted ventilation prevent anoxic myocardial and brain injury. DOMOIC ACID INTOXICATION In late 1987 in eastern Canada, an outbreak of gastrointestinal and neurologic symptoms (amnestic shellfish poisoning) occurred after consumption of mussels found to be contaminated with domoic acid. A heat-stable neuroexcitatory amino acid whose biochemical analogs are kainic acid and glutamic acid, domoic acid binds to the kainate type of glutamate receptor with three times the affinity of kainic acid and is 20 times as
powerful a toxin. Mussels can be tested for domoic acid by mouse bioassay and high-performance liquid chromatography. The regulatory limit for domoic acid in shellfish is 20 parts per million. The abnormalities noted within 24 h of ingesting contaminated mussels (Mytilus edulis) include arousal, confusion, disorientation, and memory loss. The median time of onset is 5.5 h. Other prominent symptoms include severe headache, nausea, vomiting, diarrhea, abdominal cramps, hiccoughs, arrhythmias, hypotension, seizures, ophthalmoplegia, hemiparesis, mutism, grimacing, agitation, emotional lability, coma, copious bronchial secretions, and pulmonary edema. Histologic study of brain tissue taken at autopsy has shown neuronal necrosis or cell loss and astrocytosis, most prominently in the hippocampus and the amygdaloid nucleus -- findings similar to those in animals poisoned with kainic acid. Several months after the primary intoxication, victims still demonstrate chronic residual memory deficits and motor neuronopathy or axonopathy. Nonneurologic illness does not persist. Therapy is supportive and based on symptoms. Since kainic acid neuropathology seems to be nearly entirely seizure mediated, an emphasis should be placed on anticonvulsive therapy, for which diazepam appears to be as effective as any other drug. SCOMBROID Scombroid (mackerel-like) fish include the albacore, bluefin, and yellowfin tuna; mackerel; saury; needlefish; wahoo; skipjack; and bonito. Nonscombroid fish that produce scombroid poisoning include the dolphinfish (mahimahi, Coryphaena), kahawai, sardine, black marlin, pilchard, anchovy, herring, amberjack, and Australian ocean salmon. In the northeastern and mid-Atlantic United States, bluefish has been linked to scombroid poisoning. Because greater numbers of nonscombroid fish are being recognized as scombrotoxic, the syndrome may more appropriately be called pseudoallergic fish poisoning. Under conditions of inadequate preservation or refrigeration, the musculature of these dark- or red-fleshed fish undergoes bacterial decomposition, which includes the decarboxylation of the amino acid L-histidine to histamine, histamine phosphate, and histamine hydrochloride. Histamine levels of >20 to 50 mg/100 g are noted in toxic fish, with levels in excess of 400 mg/100 g on occasion. The toxin is heat stable and is not destroyed by domestic or commercial cooking. Affected fish typically have a sharply metallic or peppery taste; however, they may be normal in appearance, color, and flavor. Symptoms occur within 15 to 90 min of ingestion and include flushing (sharply demarcated; exacerbated by ultraviolet exposure; particularly pronounced on the face, neck, or upper trunk), a sensation of warmth without elevated core temperature, conjunctival hyperemia, pruritus, urticaria, angioneurotic edema, bronchospasm, nausea, vomiting, diarrhea, epigastric pain, abdominal cramps, dysphagia, headache, thirst, pharyngitis, burning of the gingiva, palpitations, tachycardia, dizziness, and hypotension. Without treatment, the symptoms generally resolve within 8 to 12 h. The reaction may be more severe in a person who is concurrently ingesting isoniazid
because of blockade of gastrointestinal tract histaminase. Therapy is directed at reversing the histamine effect with antihistamines, either H-1 or H-2. If bronchospasm is severe, an inhaled bronchodilator -- or in rare, extremely severe circumstances, injected epinephrine -- may be used. Glucocorticoids are of no proven benefit. Protracted nausea and vomiting, which may empty the stomach of toxin, may be controlled with a specific antiemetic, such as prochlorperazine. The persistent headache of scombroid poisoning may respond to cimetidine or a similar antihistamine if standard analgesics are not effective. PFIESTERIA In the summer of 1997, reports of adverse reactions after casual exposure to Maryland waters infested with the fish-eating dinoflagellate Pfiesteria prompted the Centers for Disease Control and Prevention (CDC) to undertake multistate surveillance and to establish a case definition. As defined by the CDC, the human disease syndrome associated with Pfiesteria is characterized by either of two groups of signs and symptoms: (1) memory loss, confusion, or acute skin burning on direct contact with infested water; or (2) at least three of the following: headache, rash (flat red sores), eye irritation, upper respiratory irritation, muscle cramps, and gastrointestinal symptoms. Since the initial reports from Maryland, many such cases have followed both casual exposure to infested water and laboratory work with Pfiesteria (which is currently conducted in biohazard III facilities). Research on Pfiesteria has been complicated by a variety of factors, including the lack of a test for detection of its toxins, which have yet to be purified, and the organism's complex life cycle, which includes at least two dozen stages. In nature, the proximity of a school of fish elicits Pfiesteria's transformation into a flagellated zoospore that releases at least two toxins: a water-soluble, neuroactive toxin that kills fish within minutes, and a fat-soluble toxin that causes epidermal delamination. Polluted environments appear to favor Pfiesteria. For the treatment of Pfiesteria-associated syndromes, one teaspoon of milk of magnesia followed by one scoop of cholestyramine in 8 ounces of water and 70% sorbitol solution is administered daily for 2 weeks. (Bibliography omitted in Palm version) Back to Table of Contents
398. ECTOPARASITE INFESTATIONS, ARTHROPOD BITES AND STINGS - James H. Maguire, Andrew Spielman Ectoparasites are arthropods or helminths that infest the skin of other animals from which they derive sustenance. They may penetrate beneath the surface of the host or attach superficially by their mouthparts. These organisms damage their hosts by inflicting direct injury, by eliciting a hypersensitivity reaction, or by inoculating toxins or pathogens. The main medically important ectoparasites are arachnids (including mites and ticks), insects (including lice, fleas, and flies), pentastomes (tongue worms), and leeches. Arthropods may also harm humans through brief encounters in which they take a blood meal or attempt to defend themselves by biting, stinging, or inoculating venoms. Various arachnids (spiders, scorpions), insects (including bees, hornets, wasps, ants, flies, bugs, caterpillars, and beetles), millipedes, and centipedes produce ill effects in this manner, as do certain ectoparasites of animals, including ticks, biting mites, and fleas (discussed in this chapter as biting arthropods). More people in the United States die each year as a consequence of arthropod stings than from poisonous snake bites. ECTOPARASITE INFESTATIONS SCABIES The human itch mite, Sarcoptes scabiei (Fig. 398-CD1), which infests some 300 million persons each year, is one of the most common causes of itching dermatoses throughout the world. Gravid female mites measuring 0.3 to 0.4 mm in length burrow superficially beneath the stratum corneum for a month, depositing two or three eggs a day. Nymphs that hatch from these eggs mature in about 2 weeks through a series of molts and then emerge as adults to the surface of the skin, where they mate and subsequently reinvade the skin of the same or another host. Transfer of newly fertilized female mites from person to person occurs by intimate personal contact and is facilitated by crowding, uncleanliness, and multiple sexual partners. Medical practitioners are at particular risk of infestation. Transmission via sharing of contaminated bedding or clothing is infrequent because these mites cannot survive much more than a day without host contact. In the United States, scabies may account for 2 to 5% of visits to dermatologists; involved particularly often are children, immigrants from developing countries, and close household contacts. Outbreaks occur in nursing homes, mental institutions, and hospitals. The itching and rash associated with scabies derive from a sensitization reaction directed against the excreta that the mite deposits in its burrow (Plate IID-52). For this reason, an initial infestation remains asymptomatic for 4 to 6 weeks, and a reinfestation produces a hypersensitivity reaction without delay. Scratching generally destroys the burrowing mite, but symptoms remain even in its absence. Burrows become surrounded by infiltrates of eosinophils, lymphocytes, and histiocytes, and a generalized hypersensitivity rash later develops in remote sites. By destroying these pathogens, immunity and associated scratching limit most infestations to fewer than 15 mites per person. Hyperinfestation with thousands or millions of mites, a condition known as crusted scabies or Norwegian scabies, may result from glucocorticoid use, immunodeficiency diseases (including AIDS and infection with human T-lymphotropic virus type I), and neurologic and psychiatric illnesses that interfere with itching and
scratching. Patients with scabies report intense itching that worsens at night and after a hot shower. Typical burrows may be difficult to find because they are few in number and may be obscured by excoriations. Burrows appear as dark wavy lines in the epidermis, measure 3 to 15 mm, and end in a small pearly bleb that contains the female mite. Such lesions generally develop on the volar wrists, between the fingers, on the elbows, and on the penis. Small papules and vesicles, often accompanied by eczematous plaques, pustules, or nodules, are symmetrically distributed in these sites (Fig. 398-CD2) and in skin folds under the breasts and around the navel, axillae, belt line, buttocks, upper thighs, and scrotum. Except in infants, the face, scalp, neck, palms, and soles are spared. Burrows and other typical lesions may be sparse in persons who wash frequently, and topical glucocorticoid treatment and bacterial superinfection may alter the appearance of the rash. Atypical presentations of scabies include bullous lesions, which resemble those of bullous pemphigoid, and vesicular lesions, which resemble those of dermatitis herpetiformis. Superinfection with nephritogenic strains of streptococci has led to acute glomerulonephritis. Crusted scabies resembles psoriasis in its typical widespread erythema, thick keratotic crusts, scaling, and dystrophic nails. Characteristic burrows are not seen in crusted scabies, and patients usually do not itch, although their infestations are highly contagious and have been responsible for outbreaks of classic scabies in hospitals. Bacteremia occurs frequently in AIDS patients with crusted scabies and prominent fissures. Persons with massive infestations occasionally present with diffuse pruritus and generalized papules or with minimal or no cutaneous signs. A diagnosis of scabies should be considered in patients with pruritus and symmetric polymorphic skin lesions in characteristic locations, particularly if there is a history of household contact with a case. Burrows should be sought and unroofed with a sterile needle or scalpel blade, and the scrapings should be examined microscopically for the mite, its eggs, and its fecal pellets. A drop of mineral oil facilitates removal of the sample. Biopsies or scrapings of papulovesicular lesions may also be diagnostic. In the absence of identifiable mites or mite products, the diagnosis is based on clinical presentation and history. The possibility of other sexually transmitted diseases should be excluded in adults with scabies. TREATMENT For the treatment of scabies, 5% permethrin cream is less toxic than the once commonly used 1% lindane preparations and is effective against lindane-tolerant infestations. Both scabicides are applied thinly but thoroughly behind the ears and from the neck down after bathing and are removed 8 h later with soap and water. Lindane is absorbed through the skin, and its overuse has led to seizures and aplastic anemia. It should not be applied to pregnant women or infants. Alternatives include topical crotamiton cream, benzyl benzoate, and sulfur ointments. Successful treatment of crusted scabies requires the application first of a keratolytic agent such as 6% salicylic acid (to improve the penetration of scabicides) and then of scabicides to the scalp, face, and ears (with care to avoid the eyes). Repeated treatments or the sequential use of several agents may be necessary. A single oral dose of ivermectin (200 ug/kg) effectively treats scabies in otherwise healthy persons. Patients with crusted scabies
may require two or more doses of ivermectin. Although ivermectin may become the agent of choice for treating crusted scabies, it has not yet received approval by the U.S. Food and Drug Administration (FDA) for any form of scabies. Its use should be reserved for persons who fail to respond to topical scabicides, the elderly, persons with generalized eczema, and other persons who may not tolerate topical therapy. Although effectively treated scabies infestations become noninfectious within a day, itching and rash due to hypersensitivity frequently persist for weeks or months. Unnecessary re-treatment of the affected patients may provoke contact dermatitis. Antihistamines, salicylates, and calamine lotion relieve itching during treatment, and topical glucocorticoids are useful for the pruritus that lingers after effective treatment. An oral antibiotic may be necessary for bacterial superinfections that fail to resolve with antiscabietic therapy. Relapses of scabies may be due to infestations of the scalp when topical therapy is applied only from the neck down. To prevent reinfestations, bedding and clothing should be washed in hot water, and close contacts, even if asymptomatic, should be treated simultaneously. OTHER MITE INFESTATIONS Species of Demodex, the follicle mite, live in hair follicles and sebaceous glands of the face and ears. The wormlike mites measure up to 0.4 mm in length and, if carefully sought, can be found on almost all persons. They appear not to cause disease, although their density is high in persons with rosacea. House dust mites of the genus Dermatophagoides infest houses throughout the world, living on furniture and rugs and feeding on shed human dander. Exposure to their allergens causes asthma, rhinitis, conjunctivitis, and eczema in persons with house dust allergies. Management includes immunotherapy with mite extracts and environmental interventions such as frequent vacuuming and removal of rugs from bedrooms to reduce mite density. PEDICULOSIS (LOUSE INFESTATIONS) All three species of human louse feed at least once a day on human blood. Pediculus humanus var. capitis (Fig. 398-CD3) infests the head (Fig. 398-CD4), P. humanus var. corporis the clothing (Fig. 398-CD5), and Phthirus pubis (Fig. 398-CD6) mainly the hair of the pubis (Fig. 398-CD7). Females cement their eggs (nits) firmly to hair or clothing. The saliva of lice produces an intensely irritating maculopapular or urticarial rash in sensitized persons. Head lice, which infest an estimated 6 to 12 million people in the United States, are transmitted directly from person to person and occasionally by shared headgear and grooming implements. The prevalence is highest among school-aged girls who wear long hair; in the United States, black children are less frequently infested than other children. Excoriations of pruritic lesions on the scalp, neck, and shoulders lead to oozing, crusting, matting of hair, bacterial infections, and regional lymphadenopathy. Adult lice are frequently seen crawling in the hair with a velocity that approaches 25 mm/min. Body lice remain in clothing except when feeding and cannot survive more than a few hours away from the human host. It follows, therefore, that P. humanus var. corporis
mainly infests disaster victims or indigent persons who do not change their clothes. Transmission by direct contact or by sharing of clothing and beds is enhanced under crowded conditions. The fact that the body louse leaves febrile persons or corpses as they become cold facilitates the transmission of typhus, louse-borne relapsing fever, and trench fever (Chap. 177). Trench fever and endocarditis due to Bartonella quintana have emerged as diseases of homeless persons living in large cities of the United States and Europe. Pruritic lesions are particularly common around the neckline. Chronic infestations result in the postinflammatory hyperpigmentation and thickening of skin known as vagabonds' disease. The cosmopolitan crab or pubic louse is transmitted mainly by sexual contact but can infest eyelashes, axillary hair, and hair in other sites as well as pubic hair. Children with pubic lice generally acquire their infestations from parents rather than via sexual transmission. Polymerase chain reaction (PCR) analysis of the blood meal of lice permits identification of host DNA in cases of child abuse or rape. Intensely pruritic lesions and 2- to 3-mm blue macules (maculae ceruleae) develop at the site of bites. Blepharitis commonly accompanies infestations of the eyelashes. A suspected diagnosis of pediculosis is confirmed by the finding of nits or adult lice on hairs or in clothing. The dorsoventrally flattened adult lice measure 2 to 4 mm in length and have three pairs of legs ending in claws that enable them to grasp hair shafts or clothing. Oval nits measure 0.8 mm in length and are opaque white or cream-colored (body and head lice) or dark brown (pubic lice). TREATMENT The preferred treatment is a 10-min application of 1% permethrin creme rinse, which kills both lice and eggs and is available without prescription. An alternative, 0.5% malathion, requires a prescription and must be left in place for 8 to 12 h. Other agents, such as the more toxic 1% lindane and pyrethrins with piperonyl butoxide, are not ovicidal and require a second application 1 week after the first to kill hatching nymphs. Dead or hatched nits, which remain attached to hair sheaths and become translucent or opalescent, may falsely suggest an active infection. Resistance of head lice to permethrin, malathion, and lindane has been reported. When a properly applied treatment fails, a higher concentration (5%) of permethrin may be tried or the class of pediculicide changed (e.g., by switching from permethrin to malathion). Ivermectin may be useful in cases of resistance to both malathion and permethrin but has not been approved for this purpose by theFDA. After louse infestations have been treated with insecticide, the hair should be combed with a fine-toothed nit comb to remove nits. Combs and brushes should be disinfected in hot water at 65°C for 5 min or soaked in insecticide for 1 h. Body lice can be eliminated by bathing and application of topical pediculicides from head to foot. Clothes and bedding are deloused by heat sterilization in a dryer at 65°C for 30 min or by fumigation. Infestations with pubic lice are treated with topical pediculicides except for eyelid infestations (phthiriasis palpebrum), which respond to a coating of petroleum applied for 3 to 4 days or 1% yellow oxide of mercury ointment applied four times daily for 2 weeks. TUNGIASIS
Tunga penetrans, like other fleas, is a wingless, laterally flattened insect measuring 2 to 4 mm in length that feeds on blood. Also known as the chigoe flea, sand flea, or jigger, it occurs in tropical regions of Africa and the Americas. Adults live in sandy soil and burrow under the skin between toes, under nails, or on the soles of bare feet. The fleas engorge on blood and grow from pinpoint to pea size over a 2-week period. The lesions resemble a white pustule with a central black depression and may be pruritic or painful. Occasional complications include tetanus, bacterial infections, and autoamputation of toes. Tungiasis is treated by removal of the intact flea with a sterile needle or scalpel, tetanus vaccination, and topical antibiotics. MYIASIS Myiasis refers to infestations by maggots, mainly due to the larvae of metallic-colored screw-worm flies or botflies. Maggots invade living or necrotic tissue or body cavities and produce different clinical syndromes depending on the species of fly. Furuncular Myiasis (Fig. 398-CD8) In forested parts of Central and South America, larvae of Dermatobia hominis (the human botfly) produce boil-like subcutaneous nodules 2 to 3 cm in diameter. The adult female captures a mosquito or other bloodsucking insect and deposits her eggs beneath its abdomen. When the carrier insect attacks a human or bovine host several days later, the warmth and moisture of the host's surface stimulate the larvae to hatch and penetrate the skin. After 6 to 12 weeks, the larvae mature and drop to the ground, where they pupate. The African tumbu fly, Cordylobia anthropophaga, produces similar lesions. Dozens of eggs are deposited on sand or drying laundry that is contaminated with urine or sweat. Larvae hatch on contact with the body, penetrate the skin, and produce boils from which they emerge 8 or 9 days later. A diagnosis of furuncular myiasis is suggested by uncomfortable lesions with a central breathing pore that emits bubbles when submerged in water. There is often a sensation of movement under the skin that may lead to severe emotional distress. Tumbu fly larvae can be removed by manual expression after the air pore is coated with petroleum to suffocate the larvae and induce them to emerge. Removal of Dermatobia larvae is facilitated by injection of a local anesthetic into the surrounding tissue, but surgical excision is often necessary because up-pointing spines hold the larva firmly in place. Creeping Dermal Myiasis Maggots of the horse botfly, Gasterophilus intestinalis, do not mature after penetrating human skin but migrate for weeks in the epidermis. The resulting pruritic and serpiginous eruption resembles cutaneous larva migrans caused by Ancylostoma braziliense. Horseback riders become infested when eggs deposited on the flank of the horse hatch against their bare legs. The black spines of the larvae can be identified after mineral oil is smeared over the lesion. Larva are removed with a needle. The larvae of the cattle botfly (Hypoderma species) invade more deeply and produce boil-like swellings. Wound (Fig. 398-CD9) and Body Cavity Myiasis Certain flies are attracted to blood and pus, and their newly hatched larvae enter wounds or diseased skin. Larvae of species such as Phaenicia sericata, the green-bottle fly, remain superficial and confined to necrotic tissue and were used in the past to debride purulent wounds. Other species,
including the screw-worms (Chrysomyia bezziana in Asia and Africa and Cochliomyia hominivorax in Latin America) and the flesh fly (Wohlfahrtia vigil in northern North America), invade more deeply into viable tissue and produce large suppurating lesions. Larvae that infest wounds also may infest body cavities such as the mouth, nose, ears, sinuses, anus, vagina, and lower urinary tract, particularly in unconscious or otherwise debilitated patients. The consequences range from harmless colonization to destruction of the nose, meningitis, and deafness. Treatment involves removal of maggots and debridement of tissue. Other Forms of Myiasis The maggots responsible for furuncular and wound myiasis may also cause ophthalmomyiasis. Sequelae include nodules in the eyelid, retinal detachment, and destruction of the globe. In addition, the adult sheep botfly, Oestrus ovis, may deposit larvae in the eyes of persons tending sheep and goats, and the larvae may produce a conjunctival infestation and acute conjunctivitis. True intestinal myiasis occurs when eggs or larvae of the drone fly (Eristalis tenax) are ingested with contaminated food, mature in the gut, and cause enteritis. Most instances in which maggots are found in human feces are the result of larviposition by flesh flies on recently passed stools. PENTASTOMIASIS Pentastomids, or tongue worms, are parasites with characteristics of both helminths and arthropods and are classified in a separate phylum. The wormlike adults inhabit the respiratory passages of reptiles and carnivorous mammals. Human infestation with Linguatula serrata is common in the Middle East and occurs in the Sudan following ingestion of encysted larval stages in raw liver or lymph nodes of sheep and goats, the intermediate hosts. The larvae migrate to the nasopharynx and produce an acute self-limiting syndrome known as halzoun (Marrara in the Sudan), which is characterized by pain and itching of the throat and ears, coughing, hoarseness, dysphagia, and dyspnea. Severe edema may cause obstruction and necessitate tracheostomy, and ocular invasion has been described. Diagnostic larvae measuring 5 to 10 mm in length are found in the copious nasal discharge or vomitus. Human beings become infected with Armillifer armillatus by ingesting eggs in contaminated food or drink or after handling the definitive host, the African python. Larvae encyst in various organs but rarely cause symptoms unless they compress vital structures or perforate an organ during migration. Cysts occasionally require surgical removal as they enlarge during molting, but they are usually encountered as an incidental finding at autopsy. There are reports of the cutaneous larva migrans syndrome due to other pentastomes (Reighardia and Sebekia species) in Southeast Asia and Central America. LEECH INFESTATIONS Medically important leeches are annelid worms that attach to their hosts with chitinous cutting jaws and draw blood with muscular suckers. The medicinal leech, Hirudo medicinalis, is still used occasionally to reduce venous congestion in surgical flaps or replanted body parts. This practice has been complicated by wound infections, myonecrosis, and sepsis due to Aeromonas hydrophila, which colonizes the gullets of commercially available leeches.
Ubiquitous aquatic leeches that parasitize fish, frogs, and turtles readily attach to the skin of human beings and avidly suck blood. More notorious are the land leeches (Haemadipsa) that live in moist vegetation of tropical rain forests. Attachment is usually painless. Hirudinin, a powerful anticoagulant secreted by the leech, causes continued bleeding after the leech has detached. Healing of the wound is slow, and bacterial infections are not uncommon. Several species of aquatic leeches in Africa, Asia, and southern Europe can enter through the mouth, nose, and genitourinary tract and attach to mucosal surfaces at sites as deep as the esophagus and trachea. Bleeding may be intense. Externally attached leeches are removed by steady gentle traction. Removal is hastened by application of alcohol, salt, vinegar, or a flame to the leech. Internally attached leeches may detach on exposure to gargled saline or may be removed by forceps. DELUSIONAL INFESTATIONS The groundless conviction that one is infested with arthropods or other parasites is an extremely difficult disorder to treat and unfortunately is not rare. Patients report infestations of their skin, clothing, or homes and describe sensations of something moving in or on their skin. Excoriations often accompany complaints of pruritus or insect bites. Patients bring in as evidence of infestation specimens that are identified microscopically as plant-feeding or peridomestic arthropods, pieces of skin, vegetable matter, or inanimate objects. In suspected cases, it is imperative to rule out true infestations and neuropathies, environmental irritants such as fragments of fiberglass, and other causes of tingling or prickling sensations. Pharmacotherapy with pimozide, which blocks dopamine receptors, has been more helpful than psychotherapy in treating this disorder. ARTHROPOD BITES AND STINGS SPIDER BITES Of the>30,000 recognized species of spider, only about 100 defend themselves aggressively and have fangs sufficiently long to penetrate human skin. The venom that spiders use to immobilize and digest their prey can cause necrosis of skin and systemic toxicity. While the bites of most spiders are painful but not harmful, envenomations of the brown or fiddle spiders (Loxosceles species), widow spiders (Latrodectus species), and other species may be life-threatening. Identification of the offending spider should be attempted, since specific treatments exist for bites of widow and brown recluse spiders and since injuries attributed to spiders are frequently due to other causes. Recluse Spider Bites and Necrotic Arachnidism Severe necrosis of skin and subcutaneous tissue follows envenomation by Loxosceles reclusa, the brown recluse spider, and by at least four other species of Loxosceles in the southern and midwestern United States. Other spiders that produce necrotic ulceration include the hobo spider (Tegenaria agrestis) in the Pacific Northwest, the sac spiders (Chiracanthium species) throughout the United States and abroad, the South American brown spider Loxosceles laeta in Central and South America, and other Loxosceles species in Africa and the Middle East. All these spiders measure 7 to 15 mm in body length and 2 to 4 cm in leg span. Recluse spiders are brown and have a dark violin-shaped spot on their dorsal
surface; hobo spiders are brown with gray markings; and sac spiders may be pale yellow, green, or brown. These spiders are not aggressive toward human beings and bite only if threatened or pressed against the skin. They hide under rocks and logs or in caves and animal burrows, and they emerge at night to hunt other spiders and insects. They invade homes, particularly in the fall, and seek dark and undisturbed hiding spots in closets, in folds of clothing, or under furniture and rubbish in storage rooms, garages, and attics. Bites often occur while the victim is dressing and are sustained primarily to the arms, neck, and lower abdomen. The clear viscous venoms of these spiders contain an esterase, alkaline phosphatase, protease, and other enzymes that produce tissue necrosis and hemolysis. Sphingomyelinase B, the most important dermonecrotic factor, binds cell membranes and promotes chemotaxis of neutrophils, leading to vascular thrombosis and an Arthus-like reaction. Initially, the bite is painless or produces a stinging sensation. Within the next few hours, the site becomes painful and pruritic, with central induration surrounded by a pale zone of ischemia and a zone of erythema. In most cases, the lesion resolves without treatment over 2 to 3 days. In severe cases, the erythema spreads, and the center of the lesion becomes hemorrhagic and necrotic with an overlying bulla. A black eschar forms and sloughs several weeks later, leaving an ulcer that may be ³25 cm in diameter and eventually a depressed scar. Healing usually takes place within 3 to 6 months but may take as long as 3 years if adipose tissue is involved. Local complications include injury to nerves and secondary infection. Fever, chills, weakness, headache, nausea, vomiting, myalgia, arthralgia, maculopapular rash, and leukocytosis may develop within 72 h of the bite. In rare instances, acute complications such as hemolytic anemia, hemoglobinuria, and renal failure are fatal. TREATMENT Initial management includes local cleansing, application of sterile dressings and cold compresses, and elevation and loose immobilization of the affected limb. Analgesics, antihistamines, antibiotics, and tetanus prophylaxis should be administered if indicated. Within the first 48 to 72 h, the administration of dapsone, a leukocyte inhibitor, may halt the progression of lesions that are becoming necrotic. Dapsone is given in oral doses of 50 to 100 mg twice daily after glucose-6-phosphate dehydrogenase deficiency has been ruled out. The efficacy of locally or systemically administered glucocorticoids has not been demonstrated, and a potentially useful Loxosceles-specific antivenin has not been approved for use in the United States. Debridement and later skin grafting may be necessary after signs of acute inflammation have subsided, but immediate surgical excision of the wound is detrimental. Patients should be monitored closely for signs of hemolysis, renal failure, and other systemic complications. Widow Spider Bites The bite of the female widow spider is notorious for the effect of its potent neurotoxin. Latrodectus mactans, the black widow, has been found in every state of the United States except Alaska and is most abundant in the southeast. It measures up to 1 cm in body length and 5 cm in leg span, is shiny black, and has a red hourglass marking on the ventral abdomen. Other dangerous North American Latrodectus species include L. geometricus (the brown widow), L. bishopi (the red widow), L. variolus, and L.
hesperus, and there are related species in other temperate and subtropical parts of the world. Widow spiders spin their webs under stones, logs, plants, or rock piles or in dark spaces in barns, garages, and outhouses. Bites are most common in the summer and early autumn and occur when the web is disturbed or when the spider is trapped or provoked. The buttocks or genitals are sites of bites incurred by humans while sitting in an outdoor privy. The initial bite goes unnoticed or is perceived as a sharp pinprick. Two small red marks, mild erythema, and edema develop at the fang entrance site. The oily yellow venom that is injected does not produce local necrosis, and some persons experience no other symptoms. However, a-latrotoxin, the most active component of the venom, binds irreversibly to nerves and causes release and eventual depletion of acetylcholine, norepinephrine, and other neurotransmitters from presynaptic terminals. Within 30 to 60 min, painful cramps spread from the bite site to large muscles of the extremities and the trunk. Extreme rigidity of the abdominal muscles and excruciating pain may suggest peritonitis, but the abdomen is not tender on palpation. Other features include salivation, diaphoresis, vomiting, hypertension, tachycardia, labored breathing, anxiety, headache, weakness, fasciculations, paresthesia, hyperreflexia, urinary retention, uterine contractions, and premature labor. Rhabdomyolysis and renal failure have been reported, and respiratory arrest, cerebral hemorrhage, or cardiac failure may end fatally, especially in very young, elderly, or debilitated persons. The pain begins to subside during the first 12 h but may recur during several days or weeks before resolving spontaneously. TREATMENT Treatment consists of local cleansing, application of ice packs, and tetanus prophylaxis. Hypertension that does not respond to analgesics and antispasmodics, such as benzodiazepines or methocarbamol, requires specific antihypertensive medication. Intravenous administration of one or two vials of a widely available equine antivenin rapidly relieves pain and can be life-saving. Because of the risk of anaphylaxis and serum sickness, antivenin should be reserved for severe cases involving respiratory arrest, uncontrollable hypertension, seizures, or pregnancy. Envenomations by Tarantulas and Other Spiders Tarantulas are long-lived, hairy spiders of which 30 species are found in the United States, primarily in the southwest. The tarantulas that have become popular household pets are usually imported species with bright colors and a leg span of up to 25 cm. Tarantulas bite only when threatened and cause no more harm than a bee sting, but the venom occasionally provokes deep pain and swelling. Several species are covered with urticating hairs that are launched in the thousands when a threatened spider rubs its hind legs across the dorsal abdomen. These hairs penetrate human skin and produce pruritic papules that last for weeks. Failure to wear gloves or to wash the hands after handling the Chilean Rose tarantula, the most popular pet spider, has resulted in transfer of hairs to the eye and devastating ocular inflammation. Treatment of bites includes local washing and elevation of the bitten area, tetanus prophylaxis, and analgesic administration. Antihistamines and topical or systemic glucocorticoids are given for exposure to urticating hairs.
Atrax robustus, the Sydney funnel-web spider of Australia, and Phoneutria species, the South American banana spiders, are among the most dangerous spiders in the world because of their aggressive behavior and potent neurotoxins. Envenomation by A. robustus causes a rapidly progressive neuromotor syndrome that can be fatal within 2 h. The bite of the banana spiders causes severe local pain followed by profound systemic symptoms and respiratory paralysis that can lead to death within 2 to 6 h. Specific antivenins for envenomation by each of these spiders are available. Lycosa species (wolf spiders) are found throughout the world and may produce painful bites and transient local inflammation. SCORPION STINGS Scorpions are crablike arachnids that feed on ground-dwelling arthropods and small lizards, which they grasp with a pair of frontal pinchers and paralyze by injecting venom from a stinger on the tip of the tail. Painful but relatively harmless scorpion stings need to be distinguished from the potentially lethal envenomations that are produced by about 30 of the approximately 1000 known species and cause more than 5000 deaths worldwide each year. Scorpions feed at night and remain hidden during the day in crevices or burrows or under wood, loose bark, or rocks on the ground. They seek cool spots under buildings and often enter houses, where they get into shoes, clothing, or bedding or enter bathtubs and sinks in search of water. Scorpions sting human beings only when disturbed. Scorpions of the United States Of the 40 or so scorpion species in the United States, only the bark scorpion (Centruroides sculpturatus or C. exilicauda) produces a venom that can be lethal. Stings of the other species, such as the common striped scorpion C. vittatus and the large Hadrurus arizonensis, cause immediate sharp local pain followed by edema, ecchymosis, and a burning sensation. Symptoms typically resolve within a few hours, and skin does not slough. Allergic reactions to the venom sometimes develop. The deadly C. sculpturatus of the southwestern United States and northern Mexico measures about 7 cm in length and is yellow-brown in color. Its venom contains neurotoxins that cause sodium channels to remain open and neurons to fire repetitively. In contrast to the stings of nonlethal species, C. sculpturatus envenomations are usually associated with little swelling, but prominent pain, paresthesia, and hyperesthesia can be accentuated by tapping on the affected area (the tap test). These symptoms soon spread to other locations; dysfunction of cranial nerves and hyperexcitability of skeletal muscles develop within hours. Patients present with restlessness, blurred vision, abnormal eye movements, profuse salivation, lacrimation, rhinorrhea, slurred speech, difficulty in handling secretions, diaphoresis, nausea, and vomiting. Muscle twitching, jerking, and shaking may be mistaken for a seizure. Complications include tachycardia, arrhythmias, hypertension, hyperthermia, rhabdomyolysis, and acidosis. Symptoms progress to maximal severity in about 5 h and subside within a day or two, although pain and paresthesia can last for weeks. Fatal respiratory arrest is most common among young children and the elderly. Other Dangerous Scorpions Envenomations by Leiurus quinquestriatus in the Middle
East and North Africa, by Mesobuthus tamulus in India, by Androctonus species along the Mediterranean littoral and in North Africa and the Middle East, and by Tityus serrulatus in Brazil cause massive release of endogenous catecholamines with hypertensive crises, arrhythmias, pulmonary edema, and myocardial damage. Acute pancreatitis occurs with stings of Tityus trinitatis in Trinidad, and central nervous toxicity complicates stings of Parabuthus and Buthotus scorpions of South Africa. Tissue necrosis and hemolysis may follow stings of the Iranian Hemiscorpius lepturus. TREATMENT Identification of the offending scorpion aids in planning therapy. Stings of nonlethal species require at most ice packs, analgesics, or antihistamines. Because most victims of dangerous envenomations (such as those produced by C. sculpturatus) experience only local discomfort, they can be managed at home with instructions to return to the emergency department if signs of cranial-nerve or neuromuscular dysfunction develop. Aggressive supportive care and judicious use of antivenin can reduce or eliminate mortality from more severe envenomations. Keeping the patient calm and applying pressure dressings and cold packs to the sting site decrease the absorption of venom. A continuous intravenous infusion of midazolam controls the agitation, flailing, and involuntary muscle movements produced by scorpion stings. Close monitoring during treatment with this drug and other sedatives or narcotics is necessary for persons with neuromuscular symptoms because of the risk of respiratory arrest. Hypertension and pulmonary edema respond to nifedipine, nitroprusside, hydralazine, or prazosin, and bradyarrhythmias can be controlled with atropine. Commercially prepared antivenins are available in several countries for some of the most dangerous species. A caprine C. sculpturatus antivenin (not yetFDAapproved) is available as an investigational drug from the Arizona State University for use only in Arizona. Because of the risk of anaphylaxis or serum sickness following administration of goat serum, use of the antivenin is controversial. Intravenous administration of antivenin rapidly reverses cranial-nerve dysfunction and muscular symptoms but does not affect pain and paresthesia. The benefit of scorpion antivenin has not been established in controlled trials. Prevention In scorpion-infested areas, shoes, clothing, bedding, and towels should be shaken and inspected before being used. Removal of wood, stones, and debris from yards and campsites eliminates hiding places for scorpions, and household spraying of insecticides can deplete their source of food. CHIGGERS AND OTHER BITING MITES Chiggers are the larvae of trombiculid (harvest) mites that normally feed on mice in grassy or brush-covered sites in the tropics and subtropics and (less frequently) in temperate areas during warm months. They wait for hosts on low vegetation and attach themselves to passing animals or to people. The larva then pierces the skin of its host and deposits a tubelike structure in the dermis through which it imbibes lymph and tissue juices. This highly antigenic "stylostome" serves as the focus of an exceptionally pruritic papular, papulovesicular, or papulourticarial lesion that may be 2 cm in diameter and that develops within hours of attachment in persons previously sensitized to mite
antigen. Feeding mites appear as tiny red vesicles in hair follicles. Scratching invariably destroys the body of a mite attached to a person. These lesions generally vesiculate and develop a hemorrhagic base. Itching and burning last for weeks. The rash is most common on the ankles or near tight-fitting clothes that obstruct the mites' movements. Chiggers are the vectors of scrub typhus in tropical and subtropical parts of Asia. Repellents are useful for preventing chigger bites. Certain mesostigmatid mites that infest the nests of mice or birds feed on human beings when their usual hosts have been displaced. For example, intense episodes of itching dermatitis in humans may follow the removal of trash from a human residence or the departure of pigeons that have been nesting on a window air-conditioner. Other mites that infest grain, straw, cheese, or other animal products occasionally produce similar episodes. Persons who have close contact with dogs -- and, to a lesser extent, cats -may develop a self-limited pruritic papulovesicular rash from bites of cheyletiellid mites that cause a mangelike condition in these animals. Mouse mites are the vectors of rickettsialpox in cities of the northeastern United States. Fowl and chicken mites transmit the viruses of St. Louis encephalitis and western equine encephalitis. Although sanitary measures effectively prevent rickettsialpox, removal of accumulated refuse may result in a transient period of elevated risk. Diagnosis of mite-induced dermatitides (including those caused by chiggers) relies heavily on a history of exposure to the source of the mite, since the tiny mite may escape notice or may already have fallen off or been scratched off the lesions. Antihistamines or topical steroids effectively reduce mite-induced pruritus. HYMENOPTERA STINGS Insects that sting to defend their colonies or subdue their prey belong to the order Hymenoptera, which includes apids (bees and bumblebees), vespids (wasps, hornets, and yellow jackets), and ants. Their venoms contain a wide array of amines, peptides, and enzymes that are responsible for local and systemic reactions. Although the toxic effect of multiple stings can be fatal, nearly all of the 50 or more deaths due to hymenopteran stings in the United States each year are the result of allergic reactions. Bee and Wasp Stings Bees lose their venom apparatus in the act of stinging and subsequently die, while vespids can sting numerous times in succession. The familiar honeybees (Apis mellifera) and bumblebees (Bombus and other genera) attack when a colony is disturbed, but the extremely aggressive Africanized honeybees respond to minimal intrusions rapidly and in large numbers. Since their introduction into Brazil in 1957, these "killer bees" have spread through South and Central America to the southern and western United States. The common vespids in the United States include the yellow jacket, notable for the yellow and black bands on its abdomen; the bald-faced hornet, with a black body and a white face; the brown hornet, measuring 2.5 to 3.5 cm in length; and the paper wasps, which have variously colored elongate bodies. Vespids sting in defense of their nests, which they often build near human dwellings and suspend from eaves or shubbery, plaster onto walls, or burrow into wood or soil. Yellow jackets feed on sugary substances and decaying meat and are annoyingly abundant at recreation sites and
around garbage, particularly in the late summer and fall. Venom is produced in glands at the posterior end of the abdomen and is expelled rapidly by contraction of muscles of the venom sac, which has a capacity of up to 0.1 mL in large insects. The venoms of different species of hymenopterans are biochemically and immunologically distinct. Direct toxic effects are mediated by mixtures of low-molecular-weight compounds such as serotonin, histamine, and acetylcholine and several kinins. Polypeptide toxins in honeybee venom include mellitin, which damages cell membranes; mast cell-degranulating protein, which causes histamine release; apamin, a neurotoxin; and adolapin, which has anti-inflammatory action. Enzymes in venom include hyaluronidase, which allows the spread of other venom components, and phospholipases, which may be among the major venom allergens. There appears to be little cross-sensitization between honeybee and wasp venoms. Uncomplicated stings cause immediate pain, a wheal-and-flare reaction, and local edema and swelling that subside in a few hours. Stings from accidentally swallowed insects may induce life-threatening edema of the upper airways. Multiple stings can lead to vomiting, diarrhea, generalized edema, dyspnea, hypotension, and collapse. Rhabdomyolysis and intravascular hemolysis may cause renal failure. Death from the direct effects of venom has followed 300 to 500 honeybee stings. Large local reactions that spread³10 cm around the sting site over 24 to 48 h are not uncommon. These reactions may resemble cellulitis but are caused by hypersensitivity rather than secondary infection. Such reactions tend to recur on subsequent exposure but are seldom accompanied by anaphylaxis and are not prevented by venom immunotherapy. An estimated 0.4 to 4.0% of the U.S. population exhibits clinical immediate-type hypersensitivity to insect stings, and 15% may have asymptomatic sensitization manifested by positive skin tests. Persons who experience severe allergic reactions are likely to have similar reactions after subsequent stings; occasionally, adults who have had mild reactions later experience serious reactions. Mild anaphylactic reactions from insect stings, as from other causes, consist of nausea, abdominal cramping, generalized urticaria, flushing, and angioedema. Serious reactions, including upper airway edema, bronchospasm, hypotension, and shock, may be rapidly fatal. Severe reactions usually begin within 10 min of the sting and only rarely develop after 5 h. Unusual complications, including serum sickness, vasculitis, neuritis, and encephalitis, develop several days or weeks after a sting. TREATMENT Stingers embedded in the skin should be scraped or brushed off with a blade or a fingernail but not removed with forceps, which may squeeze more venom out of the venom sac. The site should be cleansed and disinfected and ice packs used to slow the spread of venom. Elevation of the affected site and administration of analgesics, oral antihistamines, and topical calamine lotion relieve symptoms; application of meat tenderizer containing papain is of no proven value. Large local reactions may require a short course of oral therapy with glucocorticoids. Patients with numerous stings should be monitored for 24 h for evidence of renal failure or coagulopathy.
Anaphylaxis is treated with subcutaneous injection of 0.3 to 0.5 mL of epinephrine hydrochloride in a 1:1000 dilution; treatment is repeated every 20 to 30 min if necessary. Intravenous epinephrine (2 to 5 mL of a 1:10,000 solution administered by slow push) is indicated for profound shock. A tourniquet may slow the spread of venom. Parenteral antihistamines, fluid resuscitation, bronchodilators, oxygen, intubation, and vasopressors may be required. Patients should be observed for 24 h for recurrent anaphylaxis. Prevention Persons with a history of allergy to insect stings should carry a sting kit with a preloaded syringe containing epinephrine for self-administration in case of a sting. These patients should seek medical attention immediately after using the kit. To avoid stings when outdoors, individuals can wear shoes and protective clothing and avoid attracting insects with sweet foods, bright-colored clothes, perfumes, or cosmetics. Venom Immunotherapy Repeated injections of purified venom produce a blocking IgG antibody response to venom and reduce the incidence of recurrent anaphylaxis from between 50 and 60% to