Pediatric Surgery and Urology: Long-Term Outcomes

Citation preview

This page intentionally left blank

Pediatric Surgery and Urology

This highly regarded textbook provides a unique clinical reference for all pediatric surgeons. This new edition analyzes and updates what is known about long-term outcomes in pediatric surgery and urology. The editors have succeeded in bringing together critical reviews written by leading international experts in pediatric surgery and urology. The second edition of this successful and popular textbook has been completely revised and updated with new chapters on urolithiasis, small bowel transplantation, pancreatitis, breast disorders, and a completely new section on trauma. An understanding of long-term outcomes is critical if individual surgeons and health policy-makers are to achieve optimum results in current clinical practice. This is an essential reference source for pediatric surgeons and urologists, pediatricians, adult specialists, and others dealing with the sequelae of childhood surgical problems. Reviews of the First Edition “It is an excellent book, well written by a host of experienced surgeons reviewing the results of the surgery of their particular subspecialties and interests . . . This book should not only be bought by departments of pediatric surgery and pediatric urology, but should also be read by all surgeons and trainees who operate on infants and children.” British Journal of Urology “ . . . institutional experience and expert clinical review of the literature make the text current and applicable to daily practice . . . A book that emphasizes outcome rather than diagnosis and procedures has become essential, as health care moves to measures of efficacy, safety, acceptability, and cost effectiveness . . . I heartily endorse this work to my colleagues.” Robert J. Touloukian, M. D., review in JAMA.

Pediatric Surgery and Urology Long-term Outcomes

Second Edition

Edited by

Mark D. Stringer, M.S., F.R.C.S., F.R.C.P.C.H. Professor of Paediatric Surgery, Department of Paediatric Surgery, St James’s University Hospital Leeds, UK

Keith T. Oldham, M.D., F.A.C.S. Professor and Chief, Division of Pediatric Surgery, Medical College of Wisconsin Marie Z Uihlein Chair and Surgeon-in-Chief, Children’s Hospital of Wisconsin, Milwaukee, Wisconsin, USA

Pierre D. E. Mouriquand, M.D., F.R.C.S. (Eng.), F.A.P.U. Professor of Pediatric Urology and Head of Pediatric Surgery, Department of Pediatric Urology, Debrousse Hospital, Claude Bernard University Lyon I, Lyon, France

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge cb2 2ru, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521839020 © Cambridge University Press 2006 This publication is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2006 isbn-13 isbn-10

978-0-511-24930-3 eBook (EBL) 0-511-24930-6 eBook (EBL)

isbn-13 isbn-10

978-0-521-83902-0 hardback 0-521-83902-5 hardback

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Every effort has been made in preparing this publication to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn fromactual cases, every effort has been made to disguise the identities of the individuals involved. Nevertheless, the authors, editors and publishers can make no warranties that the information contained herein is totally free fromerror, not least because clinical standards are constantly changing through research and regulation. The authors, editors and publishers therefore disclaimall liability for direct or consequential damages resulting fromthe use of material contained in this publication. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

We dedicate this book to our patients and their families.

Contents

List of contributors Acknowledgments Preface

page xi xxi xxiii

Part I General issues 1 Introduction and historical overview (a) North American perspective

3 3

W. Hardy Hendren

(b)European perspective

13

Barry O’Donnell

2 Principles of outcomes analysis

17

Brigid K. Killelea, Eric L. Lazar, and Michael G. Vitale

3 Outcomes analysis and systems of children’s surgical care

29

Keith T. Oldham, Mark D. Stringer, and Pierre D. E. Mouriquand

4 Perinatal mortality and morbidity: outcome of neonatal intensive care

39

Janet M. Rennie

5 Psychological aspects of pediatric surgery

54

Lorraine Ludman

Part II Head and neck 6 Cleft lip and palate

67

Alistair G. Smyth

7 Lymphangiomas

79

Salam Yazbeck and Maria Di Lorenzo† †

deceased.

vii

viii

Contents

8 Thyroid and parathyroid (including thyroglossal disorders)

24 Small bowel disorders 90

Jaimie D. Nathan and Michael A. Skinner

9 Salivary glands disorders

25 Cystic fibrosis 104

John D. Langdon

10 Head and neck tumors Robert Ord and Rui Fernandes

135

Donald Nuss

150

158

184

192 192

208

Philip A. J. Chetcuti and John W. L. Puntis

16 Antireflux procedures

217

Adam Goldin

17 Esophageal replacement

227

Steven W. Bruch and Arnold G. Coran

18 Esophageal achalasia

232

Thomas T. Sato

19 Congenital malformations of the breast

242

Younghoon R. Cho and Arun K. Gosain

32 Gastrointestinal motility disorders

255

Mark D. Stringer

33 The Malone antegrade continence enema (MACE) procedure

270

416

429

Ashok Rajimwale and Padraig S. J. Malone

34 Splenectomy

435

Jennifer H. Aldrint and Henry E. Rice

35 Biliary atresia

446

Edward R. Howard

36 Choledochal cyst

465

Takeshi Miyano

37 Biliary stone disease

480

Faisal G. Qureshi, Evan P. Nadler, and Henri R. Ford

491 491

´ eric ´ Gauthier, Daniele ` Pariente, and Fred Sophie Branchereau

(b) Endoscopic treatment 21 Abdominal wall defects

401

Manu R. Sood and Paul E. Hyman

38 Portal hypertension (a) Surgery and interventional radiology

Part IV Abdomen 20 Abdominal surgery: general aspects

31 Anorectal malformations: experience with the posterior sagittal approach ˜ and Marc A. Levitt Alberto Pena

Spencer W. Beasley

(b) Nutrition, growth, and respiratory function

385

Risto J. Rintala and Mikko Pakarinen

´ erique ´ Fred Sauvat and Yann Revillon

15 Esophageal atresia (a) Surgical aspects

374

Prem Puri and Alan Mortell

30 Hirschsprung’s disease

Dana Mara Thompson and Robin T. Cotton

14 Pulmonary resection and thoracotomy

362

Jacqueline Saito

29 Appendicitis

Nikki Almendinger, Son Lee West and Jay Wilson

13 Surgical management of airway obstruction

351

Frederick Alexander

28 Intestinal failure

12 Congenital diaphragmatic hernia

329

Marion C. W. Henry and R. Lawrence Moss

27 Inflammatory bowel disease in children

Part III Thorax 11 Chest wall deformities

315

Mark Davenport and Hilary Wyatt

26 Necrotizing enterocolitis 117

305

Julie R. Fuchs and Jacob C. Langer

502

Mark D. Stringer

David K. Magnuson

22 Inguinal and umbilical hernias

286

Emma J. Parkinson and Agostino Pierro

23 Infantile hypertrophic pyloric stenosis David C. G. Crabbe

39 Persistent hyperinsulinemic hypoglycemia in infancy

510

Pascale de Lonlay and Jean-Jacques Robert

296

40 Acute and chronic pancreatitis in children ` Pierre Tissieres and Claude Le Coultre

519

Contents

Part V Urology

56 Gonadal tumors

707

J. S. Valla

41 Introduction

531

Pierre D. E. Mouriquand

42 Upper tract dilation

533

Venkata R. Jayanthi and Stephen A. Koff

43 Posterior urethral valves

540

Peter M. Cuckow

44 Vesicoureteric reflux (a) Definition and conservative management

† David

555 555

571

583

595

759 759

Michael L. Ritchey and Nadeem N. Dhanani

60 Rhabdomyosarcoma

769

782

Dave R. Lal, Charles A. Sklar, and Michael P. LaQuaglia

799

Jean de Ville de Goyet and Jean-Bernard Otte

62 Extragonadal germ cell tumors 621

Thomas E. Novak and John P. Gearhart

49 Surgery for neuropathic bladder and incontinence

59 Wilms’ tumor (a) Wilms’ tumor: N. American experience

61 Liver tumors and resections 611

Laurence S. Baskin

48 Bladder exstrophy

745

Patrick G. Duffy, Gill A. Levitt, and Anthony J. Michalski

Justine M. Schober

47 Hypospadias

58 Neuroblastoma

(b) Wilms’ tumor: European experience

Christopher R. J. Woodhouse

46 Feminization (surgical aspects)

727

Joel Shilyansky

Gough

45 Genitoplasty in exstrophy and epispadias

57 Late effects of the treatment of childhood cancer Mark F. H. Brougham and W. Hamish B. Wallace

Judith H. van der Voort and Kate Verrier Jones

(b) Surgical treatment

Part VI Oncology

815

Christian J. Streck and Andrew M. Davidoff

63 Hemangiomas and vascular malformations

826

Adam M. Vogel and Steven J. Fishman

631

Prasad Godbole and Duncan T. Wilcox

50 Non-neuropathic bladder–sphincter dysfunction

Part VII Transplantation 643

¨ ¨ ´ Goran Lackgren and Tryggve Neveus

51 Undescended testes

652

John M. Hutson

52 Circumcision

Adrian S. Woolf

54 Multicystic kidney

884

Jorge Reyes and Geoffrey Bond

67 Heart and lung transplantation 683

858

Paolo Muiesan and Nigel D. Heaton

66 Intestinal and multivisceral transplantation 675

845

Maria H. Alonso, Greg Tiao, and Frederick C. Ryckman

65 Liver transplantation 664

Peter M. Cuckow

53 The single kidney

64 Renal failure and transplantation

897

Thomas Spray and Stephanie M. P. Fuller

Gianantonio M. Manzoni and Anthony A. Caldamone

55 Urolithiasis Bartley G. Cilento, Jr., Gerald C. Mingin, and Hiep T. Nguyen †

deceased.

695

Part VIII Trauma 68 Introduction scoring and trauma management systems Peter F. Ehrlich

915

ix

x

Contents

69 Prognosis and recovery of pediatric head injury

72 Myelomeningocele and hydrocephalus 926

P. David Adelson, Neil Buxton, and Richard Appleton

70 Truncal trauma

958

Bruce A. Kaufman

73 Outcomes after maternal–fetal surgery

966

Preeti Malladi, Karl Sylvester, and Craig T. Albanese

936

Steven Stylianos and Michael G. Vitale

74 Conjoined twins

1005

Lewis Spitz

Part IX Miscellaneous Index 71 Vascular access A. Martin Barrett and Roly Squire

947 Colour plates between pages 648 and 649

1015

Contributors

P. David Adelson Department of Pediatric Neurosurgery Children’s Hospital of Pittsburgh and The University of Pittsburgh 3705 Fifth Avenue Pittsburgh, PA 15213-2583, USA Craig T. Albanese Lucile Packard Children’s Hospital Stanford Medical Center 780 Welch Road, Suite 206 Stanford, CA 94305-5733, USA Jennifer H. Aldrint Division of Pediatric Surgery PO Box 3815 Duke University Medical Center Durham, NC 27710, USA Frederick Alexander Department of Pediatric Surgery Cleveland Clinic Children’s Hospital 9500 Euclid Avenue M14 Cleveland, OH 44195, USA Nikki Almendinger Department of Surgery Children’s Hospital 300 Longwood Avenue Fegan 3 Boston, MA 02115, USA

xi

xii

Contributors

Maria H. Alonso Department of Pediatric General and Thoracic Surgery Transplantation Division Cincinnati Children’s Hospital Medical Center 3333 Burnet Avenue Cincinnati, OH 45229, USA

Steven W. Bruch University of Michigan Medical School C. S. Mott Children’s Hospital Section of Pediatric Surgery F3970 1500 E. Medical Center Drive Ann Arbor, Mi 48109-0245, USA

Richard Appleton The Roald Dahl EEG Unit Department of Paediatric Neurology Royal Liverpool Children’s Hospital Alder Hey, Eaton Road Liverpool L12 2AP, UK

Neil Buxton Department of Paediatric Neurosurgery Royal Liverpool Children’s Hospital Alder Hey, Eaton Road Liverpool L12 2AP, UK

A. Martin Barrett Department of Paediatric Surgery Royal Victoria Infirmary Queen Victoria Road Newcastle upon Tyne NE1 4LP, UK

Anthony A. Caldamone Division of Pediatric Urology Hasbro Children’s Hospital Brown University School of Medicine Providence, RI, USA

Laurence S. Baskin Department of Urology University of California 400 Parnassus Avenue, A640, Box 0738 San Francisco, CA 94143, USA

Philip A. J. Chetcuti Clarendon Wing Leeds General Infirmary Leeds LS2 9NS, UK

Spencer W. Beasley Department of Paediatric Surgery Christchurch Hospital Private Bag 4710 Christchurch, New Zealand

Younghoon R. Cho Department of Plastic Surgery Medical College of Wisconsin 9200 West Wisconsin Avenue Milwaukee, WI 53226, USA

Geoffrey Bond Department of Pediatric Neurosurgery Children’s Hospital of Pittsburgh and The University of Pittsburgh 3705 Fifth Avenue Pittsburgh, PA 15213-2583, USA

Bartley G. Cilento, Jr. Department of Pediatric Urology The Children’s Hospital 300 Longwood Avenue Boston, MA 02115-5737, USA

Sophie Branchereau Division of Surgery Federation of Paediatrics 78 rue du G´en´eral Leclerc Centre Hospitalier Universitaire Bicˆetre Le Kremlin Bicˆetre, France Mark F. H. Brougham Department of Paediatric Haematology and Oncology Royal Hospital for Sick Children 17 Millerfield Place Edinburgh EH9 1LW, UK

Arnold G. Coran University of Michigan Medical School C. S. Mott Children’s Hospital Section of Pediatric Surgery F3970 1500 E. Medical Center Drive Ann Arbor, MI 48109-0245, USA Robin T. Cotton Department of Pediatric Otolaryngology Cincinnati Children’s Hospital Medical Center 3333 Burnet Avenue Cincinnati, OH 45229, USA

Contributors

David C. G. Crabbe Department of Paediatric Surgery Clarendon Wing Leeds General Infirmary Leeds LS2 9NS, UK Peter M. Cuckow Department of Paediatric Urology Institute for Child Health and Great Ormond Street Hospital for Children NHS Trust 30 Guilford Street London WC1N 3JH, UK

Mark Davenport Department of Paediatric Surgery King’s College Hospital Denmark Hill London SE5 9RS, UK

Andrew M. Davidoff St. Jude Children’s Research Hospital 332 N Lauderdale Memphis TN 38105, USA

Pascale de Lonlay D´epartement de P´ediatrie Hˆopital Necker-Enfants Malades 149 rue de S`evres 75743 Paris Cedex 15, France

Jean de Ville de Goyet Transplant and Paediatric Surgery St. Luc University Hospital 10 Avenue Hippocrate B-1200 Brussels, Belgium

Nadeem N. Dhanani Department of Urology University of Texas Houston Medical School 6431 Fannin Street, Suite 6018 Houston Texas 77030, USA

Maria Di Lorenzo† (formerly) Department of Surgery University of Montreal Hˆopital Sainte-Justine 3175 Cˆote Ste-Catherine Montreal Quebec Canada H3T 1C5 Patrick G. Duffy Department of Paediatric Urology Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust 30 Guilford Street London WC1N 3JH, UK Peter F. Ehrlich University of Michigan Medical School C. S. Mott Children’s Hospital Section of Pediatric Surgery F3970 1500 E. Medical Center Drive Ann Arbor, MI 48104, USA Rui Fernandes Department of Oral and Maxillofacial Surgery University of Maryland 666 W Baltimore Street Baltimore, MD 21201, USA Steven J. Fishman Department of Surgery Children’s Hospital 300 Longwood Avenue Fegan 3 Boston, MA 02115, USA Henri R. Ford Children’s Hospital of Los Angeles University of Southern California Los Angeles, CA, USA Julie R. Fuchs Department of Pediatric Surgery J. W. Riley Hospital for Children Indianapolis, IN, USA †

deceased.

xiii

xiv

Contributors

Stephanie M. P. Fuller Division of Cardiothoracic Surgery Heart, Lung, and Heart Transplant Services Children’s Hospital of Philadelphia 34th Street and Civic Center Blvd. Suite 8527, Main Building Philadelphia, PA 19104, USA ´ eric ´ Gauthier Fred Division of Surgery Federation of Paediatrics Centre Hospitalier Universitaire Bicˆetre 78 rue du G´en´eral Leclerc Le Kremlin Bicˆetre, France John P. Gearhart Brady Urological Institute The Johns Hopkins Hospital 600 North Wolfe Street Baltimore, MD 21287-2101, USA Prasad Godbole Department of Paediatric Urology Sheffield Children’s NHS Trust, Sheffield, UK Adam Goldin Division of Pediatric Surgery Department of Surgery University of Washington Seattle, WA, USA Arun K. Gosain Department of Plastic Surgery Medical College of Wisconsin 9200 W. Wisconsin Avenue Milwaukee, WI 53226, USA

W. Hardy Hendren Children’s Hospital 300 Longwood Avenue Boston, MA 02115, USA Marion C. W. Henry Section of Pediatric Surgery Yale University School of Medicine 333 Cedar Street, FMB 132 PO Box 208062 New Haven, CT 06520-8062, USA Edward R. Howard (formerly) Department of Surgery King’s College Hospital Denmark Hill London SE5 9RS, UK John M. Hutson Department of General Surgery Royal Children’s Hospital Flemington Road Parkville, VIC 3052, Australia Paul E. Hyman Department of Pediatrics University of Kansas Hospital 3901 Rainbow Boulevard Kansas City, KS 66160, USA Venkata R. Jayanthi Division of Urology Children’s Hospital Room ED343, Ed Building 700 Children’s Drive Columbus, OH 43205-2696, USA

†

David Gough Department of Paediatric Urology Royal Manchester Children’s Hospital Pendlebury Manchester M27 4HA, UK Nigel D. Heaton Liver Transplant Surgical Service King’s College Hospital Denmark Hill London SE5 9RS, UK †

deceased.

Bruce A. Kaufman Department of Neurosurgery Children’s Hospital of Wisconsin 9000 W Wisconsin Avenue MS 405 Milwaukee, WI 53226, USA Brigid K Killelea In-CHOIR (International Center for Health Outcomes and Innovation Research) 600W. 168th Street, 7th Floor, New York, NY10032, USA

Contributors

Stephen A. Koff Division of Urology Children’s Hospital Room ED343, Ed Building 700 Children’s Drive Columbus, OH 43205-2696, USA

Gill A. Levitt Department of Paediatric Urology Great Ormond Street Hospital for Children NHS Trust 30 Guilford Street London WC1N 3JH, UK

¨ ¨ Goran Lackgren Department of Paediatric Urology Uppsala University Akademiska Sjukhuset Ing 95 S-75185 Uppsala, Sweden

Marc A. Levitt Colorectal Center for Children Cincinnati Children’s Hospital for Pediatric Surgery 3333 Bunet Avenue, ML 2023 Cincinnati, OH 46229, USA

Dave R. Lal (formerly) Department of Surgery and Pediatrics Memorial Sloan-Kettering Cancer Center C-1176, 1275 York Avenue New York New York 10021, USA John D. Langdon (formerly) Department of Oral and Maxillofacial Surgery King’s College London London, UK Jacob C. Langer Room 1526 Department of Surgery Hospital for Sick Children University of Toronto 555 University Avenue Toronto ON M5G 1X8, Canada Michael P. LaQuaglia Department of Surgery and Pediatrics Memorial Sloan-Kettering Cancer Center C-1176, 1275 York Avenue New York NY 10021, USA Eric L. Lazar Children’s Hospital of New York 3959 Broadway Room 201 South New York NY 10032, USA Claude Le Coultre Paediatric Surgery Department Children’s Hospital 6 rue Willy Donze-HUG CH-1211 Geneva 14, Switzerland

Lorraine Ludman Institute of Child Health Great Ormond Street Hospital for Children NHS Trust 30 Guilford Street 46229, USA London WCIN 3JH, UK David K. Magnuson University Hospital Health System Rainbow Babies & Childrens Hospital 1100 Euclid Ave, Room RBC 122 Cleveland, OH 44106, USA Preeti Malladi Stanford University Pediatric Surgery 780 Welch Rd., Ste. 206 Stanford, CA 94305-5733, USA Padraig S. J. Malone Southampton University Hospitals Tremona Road Southampton SO16 6YD, UK Gianantonio M. Manzoni Sezione Urologia Pediatrica, Divisione di Urologia Ospedale di Circolo e Fondazione Macchi Varese Italy Anthony J. Michalski Department of Paediatric Urology Great Ormond Street Hospital London WC1N 3JH, UK Gerald C. Mingin Department of Urology Denver Children’s Hospital Colorado, CO, USA

xv

xvi

Contributors

Takeshi Miyano Department of Pediatric General and Urogenital Surgery Juntendo University School of Medicine 2-1-1 Hongo Bunkyo-ku Tokyo 113-8421, Japan Alan Mortell Children’s Research Centre Our Lady’s Hospital for Sick Children Crumlin Dublin 12, Ireland R. Lawrence Moss Section of Pediatric Surgery Yale University School of Medicine 333 Cedar Street, FMB 132 PO Box 208062 New Haven, CT 06520-8062, USA Pierre D. E. Mouriquand Department of Paediatric Urology/Surgery Debrousse Hospital 29 rue Soeur Bouvier 69322 Lyon Cedex 05, France Paolo Muiesan Liver Transplant Surgical Services King’s College Hospital, Denmark Hill London SE5 9RS, UK

Hiep T. Nguyen Department of Pediatric Urology The Children’s Hospital 300 Longwood Avenue Boston, MA 02115-5737, USA Thomas E. Novak Brady Urological Institute The Johns Hopkins Hospital 600 North Wolfe Street Baltimore, MD 21287-2101, USA Donald Nuss Pediatric Surgery Children’s Hospital of the King’s Daughters 601 Children’s Lane, Suite 5A Norfolk, VA 23507, USA Barry O’Donnell 28 Merlyn Road, Ballsbridge Dublin 4, Ireland Keith T. Oldham Children’s Hospital of Wisconsin 9000 W Wisconsin Avenue PO Box 1997 Milwaukee, WI 53201, USA

Evan P. Nadler New York University School of Medicine 530 First Avenue, Suite 10 W New York, NY 10016, USA

Robert Ord Department of Oral and Maxillofacial Surgery University of Maryland 666 W Baltimore Street Baltimore, MD 21201, USA

Jaimie D. Nathan Division of Pediatric Surgery Duke University Medical Center PO Box 3815 Durham, NC, 27710, USA

Jean-Bernard Otte Transplant and Paediatric Surgery St. Luc University Hospital 10 Avenue Hippocrate B-1200 Brussels, Belgium

´ Tryggve Neveus Department of Paediatrics Uppsala University Akademiska Sjukhuset Ing 95 S-75185 Uppsala, Sweden

Mikko Pakarinen Children’s Hospital University of Helsinki PO Box 281 FIN-00029 HUS, Finland

Contributors

` Pariente Daniele Division of Radiology Federation of Paediatrics Centre Hospitalier Universitaire Bicˆetre 78 rue du G´en´eral Leclerc Le Kremlin Bicˆetre, France Emma J. Parkinson Department of Paediatric Surgery Institute of Child Health 30 Guilford Street London WC1N 1EH, UK ˜ Alberto Pena Colorectal Center for Children Cincinnati Children’s Hospital for Pediatric Surgery 3333 Bunet Avenue ML 2023 Cincinnati, OH 46229, USA Agostino Pierro Department of Paediatric Surgery Institute of Child Health and Great Ormond Street Hospital NHS Trust 30 Guilford Street London WC1N 1EH, UK John W. L. Puntis Clarendon Wing Leeds General Infirmary Leeds LS2 9NS, UK Prem Puri Children’s Research Centre Our Lady’s Hospital for Sick Children Crumlin Dublin 12, Ireland Faisal G. Qureshi University of Pittsburgh Medical Center 300 Halket Street, Suite 5500 Pittsburgh, PA 15213, USA Ashok Rajimwale Southampton University Hospitals Tremona Road Southampton SO16 6YD, UK

Janet M. Rennie Department of Neonatal Medicine Elizabeth Garrett Anderson Obstetric Hospital University College London Hospitals Huntley Street London WCIE 6 DH, UK Yann Revillon Service de Chirurgie Pediatrique 149 rue de Sevres 75743 Paris Cedex 15, France Jorge Reyes Transplant Surgery University of Washington Seattle, USA Henry E. Rice Division of Pediatric Surgery PO Box 3815 Duke University Medical Center Durham, NC 27710, USA Risto J. Rintala Department of Pediatric Surgery Children’s Hospital University of Helsinki PO Box 281 FIN-00029 HUS, Finland Michael L. Ritchey Department of Surgery and Pediatrics Division of Urology University of Texas Houston Medical School 6431 Fannin Street, Suite 6018 Houston Texas 77030, USA Jean-Jacques Robert D´epartement de P´ediatrie Hˆopital Necker – Enfants Malades 149 rue de S`evres 75743 Paris Cedex 15, France Frederick C. Ryckman Department of Pediatric General and Thoracic Surgery Transplantation Division Cincinnati Children’s Hospital Medical Center 3333 Burnet Avenue Cincinnati, OH 45229, USA

xvii

xviii

Contributors

Jacqueline Saito Division of Pediatric Surgery University of Alabama Birmingham, AL, USA Thomas T. Sato Children’s Hospital of Wisconsin 9000 W Wisconsin Avenue C320 Milwaukee, WI 53226, USA ´ erique ´ Fred Sauvat Service de Chirurgie Pediatrique 149 rue de Sevres 75743 Paris Cedex 15, France Justine M. Schober 333 State Street, Suite 201 Erie, PA 16506 USA and Department of Urology Hamot Medical Center Erie, PA, USA Joel Shilyansky Children’s Hospital of Wisconsin 9000 W. Wisconsin Avenue C320 Milwaukee, WI 53226, USA Michael A. Skinner Division of Pediatric Surgery Duke University Medical Center PO Box 3815 Durham, NC 27710, USA Charles A. Sklar Department of Surgery and Pediatrics Memorial Sloan-Kettering Cancer Center C-1176, 1275 York Avenue New York New York 10021, USA Alistair G. Smyth Northern and Yorkshire Cleft Lip and Palate Service Clarendon Wing The General Infirmary at Leeds Gt. George Street Leeds LS2 9NS, UK

Manu R. Sood Department of Pediatric Gastroent erology Children’s Hospital of Wisconsin 9000 W. Wisconsin Avenue Milwaukee, WI 53226, USA Lewis Spitz Department of Paediatric Surgery Institute of Child Health and Great Ormond Street Hospital NHS Trust 30 Guilford Street London WC1N 1EH, UK Thomas Spray Division of Cardiothoracic Surgery Heart, Lung, and Heart Transplant Services Children’s Hospital of Philadelphia 34th Street and Civic Center Blvd. Suite 8527, Main Building Philadelphia, PA 19104 USA Roly Squire Department of Paediatric Surgery Gledhow Wing St. James’s University Hospital Beckett Street Leeds LS9 7TH, UK Christian J. Streck St. Jude Children’s Research Hospital 332 N Lauderdale Memphis TN 38105, USA Mark D. Stringer Children’s Liver and GI Unit Gledhow Wing St. James’s University Hospital Leeds, LS9 7TF, UK Steven Stylianos Department of Pediatric Surgery Miami Children’s Hospital 3200 SW 60th CT-Suite 201 Miami, FL 33155, USA Karl Sylvester Stanford University Pediatric Surgery 780 Welch Rd., Ste. 206 Stanford, CA 94305-5733 USA

Contributors

Dana Mara Thompson Department of Pediatric Otolaryngology and the Aerodigestive and Sleep Center Cincinnati Children’s Hospital Medical Center 3333 Burnet Avenue Cincinnati, OH 45229, USA

W. Hamish B. Wallace Department of Paediatric Haematology and Oncology Royal Hospital for Sick Children 17 Millerfield Place Edinburgh EH9 1LW, UK

Greg Tiao Department of Pediatric General and Thoracic Surgery Transplantation Division Cincinnati Children’s Hospital Medical Center 3333 Burnet Avenue Cincinnati, OH 45229, USA

Son Lee West Department of Surgery Children’s Hospital 300 Longwood Avenue, Fegan 3 Boston, MA 02115, USA

` Pierre Tissieres Multidisciplinary Paediatric Intensive Care Unit Bicˆetre Hospital 78 rue de Genberal Leclerc 92475 Le Kremlin-Bicˆetre, France J. S. Valla Service de Chirurgie Pediatrique Hˆopital Lenval 56 avenue de Californie 06200 Nice, France Judith van der Voort KRUF Children’s Kidney Centre for Wales University Hospital of Wales Heath Park Cardiff CF14 4XW, UK Kate Verrier Jones KRUF Children’s Kidney Centre for Wales University Hospital of Wales Heath Park Cardiff CF14 4XW, UK Michael G. Vitale Children’s Hospital of New York 3959 Broadway, 8 North New York, NY 10032, USA Adam M. Vogel Department of Surgery Children’s Hospital 300 Longwood Avenue Fegan 3 Boston, MA 02115, USA

Duncan T. Wilcox Department of Pediatric Urology The University of Texas Southwestern Medical Center at Dallas Dallas, Texas, USA Jay Wilson Department of Surgery Children’s Hospital 300 Longwood Avenue Fegan 3 Boston, MA 02115, USA Christopher R. J. Woodhouse Institute of Urology Gower Street Campus 48 Riding House Street London W1W 7EY, UK Adrian S. Woolf Nephro-Urology Unit, UCL Institute of Child Health 30 Guilford Street London WC1N 1EH, UK Hilary Wyatt Department of Child Health King’s College Hospital Denmark Hill London SE5 9RS, UK Salam Yazbeck Department of Surgery University of Montreal Hˆopital Sainte-Justine 3175 Cˆote Ste-Catherine Montreal Quebec H3T 1C5, Canada

xix

Acknowledgments

We are indebted to our partners, Emma, Karen, and Jessica, and to our children, Paul, Stephen, Catherine, Christian, Brian, David, and Caroline. Without their support and encouragement, our endeavors as pediatric surgeons would not be possible. We also wish to thank Peter Silver, Senior Editor at Cambridge University Press, for his steadfast support and encouragement throughout the project and Joseph Bottrill, Production Editor, and Mary Sanders (Copyeditor), for their expertise in collating the material for this second edition.

xxi

Preface

The second edition of this book attempts to bring together and analyze what we currently know about the long-term effects of conditions and operative procedures in pediatric surgery and urology. The subject of long-term outcomes has been relatively neglected in the past. However, the realization that there is an ever-expanding cohort of children reaching maturity with a variable legacy from childhood surgical problems has prompted a more critical and detailed study of outcomes with the aim of optimizing current surgical practice. Encouragingly, all major international pediatric surgical meetings now include data on long-term outcomes. Furthermore, in the decade since the first edition of this text was conceived, the quality and relevance of the data are substantially more robust. Much more is now known about long-term function and quality of life. Data on outcomes provide a barometer for healthcare, indicating its efficacy, safety, acceptability, and sometimes its cost-effectiveness. Such data will be used increasingly to shape public policy and guide surgical practice. Perhaps, more than in any other field of surgery, it is necessary for pediatric surgeons and urologists to look critically at long-term outcomes and the effect these have on the quality of life of their patients and their families. By inviting contributions from leading experts around the world including the USA, Europe, Australia, and Japan, we have collected together critical analyses of the literature, framed within the context of a wealth of institutional and personal experience. We are grateful to our many authors for their outstanding efforts. We hope that the second edition of this text is reasonably comprehensive, but it is not intended to be encyclopedic. A completely new section on trauma has been added, together with new chapters on the organization and delivery of pediatric surgical care, urolithiasis, pancreatitis, intestinal motility

xxiii

xxiv

Preface

disorders, small bowel transplantation, and pediatric breast disorders. Since the first edition was published, Professor Ted Howard has retired and stepped down from the editorial team. He was a major influence in the initiation of this project and a keen advocate of the importance of longterm follow-up. We are much indebted to him and wish him a long and happy retirement.

This book is intended to be a unique reference work for pediatric surgeons and urologists, but it should also be of value to pediatricians, adult specialists, and others who are involved in the long-term follow-up of patients with congenital malformations. We hope that clinicians will continue to be stimulated to look critically and honestly at their long-term results so that we can better inform our patients and their families in the future.

Part I General issues

1a Introduction and historical overview: North American perspective W. Hardy Hendren Children’s Hospital, Boston, MA, USA

William E. Ladd (1880–1967) is considered to be the father of pediatric surgery in North America. However, delving into records from the Massachusetts General Hospital (MGH), founded in 1821, discloses pediatric surgical cases cared for by surgeons 60 years before Ladd’s time. A book published in 1839 by John C. Warren,1 who performed the first publicly demonstrated operation under general anesthesia at the MGH in 1846, included pediatric cases. A generation later, his son, J. Mason Warren, wrote a book after anesthesia was well established in which many more pediatric cases were described.2 When the Children’s Hospital in Boston was founded in 1869 (Philadelphia Children’s Hospital predated it in 1855), Benjamin Shaw, resident physician at the MGH wrote to the press, “Our existing institutions public and private provide adequately for hospital treatment of children.” He stated that 190 of 1264 (14%) admissions to the MGH in 1868 were children. A recent book, The Children’s Hospital of Boston, Built Better Than They Knew, by Clement A. Smith,3 describes a century of history at Children’s. From 1882 to 1914 the most frequent surgical admission was for bone and joint tuberculosis. Early on there were no recognized cases of appendicitis, because Reginald Fitz at the MGH had not yet described that entity. In contrast there were 113 admissions for appendicitis from 1911 to 1913. When the present Children’s Hospital was built in 1914, next to Harvard Medical School, a herd of cows was maintained across the street to provide tuberculosis-free milk for the patients. Ladd graduated from the Harvard Medical School in 1906. He trained as a general surgeon at the Boston City Hospital. He held an appointment as a Volunteer Assistant Surgeon at Children’s in 1910, but maintained a private practice of general surgery and gynecology. On December 6, 1917, during World War

I, a munitions ship exploded in the harbor at Halifax, Nova Scotia. There were hundreds of deaths and injuries. A plea for help was made to physicians from Boston. Ladd was among those who responded. Many of the patients were children. His career path changed soon thereafter to concentrate on pediatric surgery, succeeding James Stone as Chief of Surgery in 1927. He became the first geographic full-time surgeon for children and soon established a link with the Peter Bent Brigham Hospital, recognizing that pediatric surgeons should have strong ties with general surgery. Another surgeon who responded to the call from Nova Scotia was Ernest Armory Codman (1869–1940). This textbook, which stresses long-term outcomes of pediatric surgery, would be incomplete without mention of Codman. He was a classmate of Harvey Cushing. Codman developed the anesthesia chart, with name, diagnosis, operation, vital signs, and remarks. When he wrote a paper on ether anesthesia and presented it for review, a senior surgeon at MGH described it as “too frank for the good of the hospital, for it described in detail the cases which I lost.” When X-ray diagnosis was introduced in 1896, Codman became interested in that new tool. Although he was a surgeon, he became the first radiologist at the Children’s Hospital in 1899. It was established as the first Pediatric Radiology Department in America. Although contemporary literature for the past decade has emphasized the importance of long-term follow-up and quality improvement, Codman preached that philosophy almost a century ago. He introduced the “end result idea.” His approach was methodical, complete, and precise. When the Clinical Congress of Surgeons of North America met in 1912, Codman was made chairman of a committee on hospital standardization to

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


3

4

W. H. Hendren

Fig. 1a.1. Cartoon shown by Ernest A. Codman on January 8, 1915, to a meeting of surgeons at the Boston Medical Library. It created a storm of resentment. The ostrich is kicking golden eggs into the outstretched hands of MGH staff who are labeled: Surgical, Gynecologic, Obstetric, and Pediatric (and humbug) Teams, and a Death Bed Team. President Lowell of Harvard stands above, straddling a bridge across the Charles River and saying, “I wonder if clinical truth is incompatible with medical science? Could my clinical professors make a living without humbug?” The ostrich is eating humbugs, and muses “If I only dared look and see, I might find a doctor who could cure my own ills.” At the Clinical Truth Table are the Board of Trustees of the MGH. They are commenting, “If we let her know the truth about our patients do you suppose she would still be willing to lay?” At the top right is a Bill Head crediting the Massachusetts General Hospital with the first demonstration of Anesthesia, Practical Social Service, and Emancipation from Humbug by the End Result System. On the left, above the clinical teams, stands Harvard Medical School. Harvard College, Massachusetts Institute of Technology and Bunker Hill Monument are seen in the background across the river.

improve the quality of patient care. His zeal, together with his overt criticism of surgical practices of that time, created great animosity from other surgeons. In 1914, at a meeting in Philadelphia, he said, “Hospitals are responsible for the care given by their staff and should carefully note the results of each surgeon, and all of that should be made public.” His most famous attack on the establishment occurred on January 8, 1915, at a meeting of surgeons in the Boston Medical Library. He unveiled a large cartoon, shown in Fig. 1a.1. He was virtually ostracized by colleagues in Boston Surgery. Indeed, Codman emphasized the need for longterm outcome research almost a century ago. It is a fascinating chapter in surgery. A recent biography of Codman, by William J. Mallon4 described that saga. Codman’s contributions were enormous. His book on the shoulder is a classic. He started the bone sarcoma registry, which still exists. He wrote many scholarly papers on a wide range of general surgical and orthopedic subjects. The cartoon resides today in the Boston Medical Library, Achieve Room, in the Countway Library of Medicine at Harvard Medical School. Returning to Ladd, he was a founding member of the Board of Surgery and the American Association for

Fig. 1a.2. Department of Surgery under William E. Ladd, 1932. In back row, Ladd is the tallest figure fifth from the left; Thomas Lanman is the seventh. In the front row, Robert Gross, an intern, is seated second from the left. Note cottage-type buildings for patients.

Introduction: North American perspective

Plastic Surgery (Fig. 1a.2). He wrote about many subjects, including pyloric stenosis, intussusception, biliary atresia, cleft lip, exstrophy of the bladder, Wilms’ tumor, and malrotation of the intestine. He devised the treatment for malrotation of the intestine with midgut volvulus. The operation today is still termed “The Ladd’s Procedure.” In 1939 he saved a newborn with esophageal atresia,5 a day after the same event occurred in Minneapolis by Dr. Logan Leven.6 The early cases had division of the tracheoesophageal fistula, marsupialization of the upper pouch in the neck, and insertion of a gastrostomy. Later an antethoracic esophagus was constructed; the lower two-thirds was a Roux-en-Y loop of jejunum placed beneath the skin anterior to the sternum; the gap between the cervical esophagus and the jejunal loop was constructed from a skin tube, performed in stages. His associate Dr. Thomas Lanman, published in 1940 a series of 32 esophageal atresia failures,7 predicting that, “Given a suitable case in which the patient is seen early I feel that with greater experience, improved technique, and good luck, the successful outcome of a direct anastomosis can and will be reported in the near future”.7 That came true in 1941, when Cameron Haight of Ann Arbor, Michigan reported such a case.8 A published genealogy of North American Pediatric Surgery9 documents a direct line of descent from Ladd to 66% of all pediatric surgeons and 73% of training directors in North America. Reminiscences about Ladd by Orvar Swenson,9 William Clatworthy,9 and Alexander Bill,10 describe Ladd’s eminent position in surgery and the great respect that he enjoyed from those associated with him. The William E. Ladd Chair in Surgery was established at Harvard Medical School in 1941. Ladd was the first incumbent, foIlowed by Robert Gross, Aldo Castaneda, and Richard Jonas. It was my privilege as a medical student in 1951 to meet Dr. Ladd. He gave us an informal talk about scrofula. Ten years later, Dr. Edward D. Churchill, Chief of Surgery at the MGH, appointed me to the staff of the MGH, with the charge to develop a Pediatric Surgical Division. An early successful case was a 3 lb infant with esophageal atresia. Dr. Ladd, accompanied by Dr. Lanman, although both long retired, came to the MGH to discuss the case at Surgical Grand Rounds. Imagine what that meant to a young surgeon just getting started! Churchill gave me a cartoon showing a family of birds on a tree limb, in descending order of size, the big birds next to the trunk. At the end of the branch was a tiny bird with one foot on the limb flapping furiously to stay there! That cartoon might well have applied also to other of my surgical colleagues who were introducing pediatric surgery as a specialty in various places. Uniformly there was apprehension and a cool reception by established surgeons. In 1967, while walking

down the operating room corridor one evening I chanced to look into an induction room. There was Dr. Ladd about to be anesthetized for surgery on a hip. It was a privilege to hold his hand as he went to sleep. He died that year, at age 87, ending a distinguished career in surgery. Edward D. Churchill (1894–1972), Chief of Surgery at MGH, also contributed importantly to North American Pediatric Surgery. He developed the “rectangular” Surgical Residency,11 in which smaller numbers of residents start but most finish a full 5-year program. This contrasted with the steep pyramid system extant at Johns Hopkins, Yale, Peter Bent Brigham Hospital, and Duke. The pyramid system was similar to surgical training in Germany. It graduated a superb and experienced surgeon, but many fell at the wayside who did not reach the top of the pyramid. A rectangular system is most common today. It assures full training for most who start such a program, although there may be an additional year or more for those destined for an academic career or a specialty like pediatric surgery. Churchill developed segmental resection of the lung;12 some of his patients were children with bronchiectasis. Churchill’s establishing a division of pediatric surgery at MGH resulted eventually in entry to pediatric surgery by 41 members of the MGH surgical residency staff. Some became program heads in North America: Scott Adzick, Philadelphia; Jay Vacanti, Boston, MGH; Robert Shamberger, Boston, Children’s; Michael Harrison, San Francisco; Michael LaQuaglia, Sloan-Kettering Hospital, New York; Dennis Lund, Madison, Wisconsin; Lucien Leape, Boston, Floating Hospital; Dale Johnson, Salt Lake City; Judson Randolph, Washington D.C.; Willis Williams, Atlanta; Timothy Canty, Louisville; Michael Mitchell, Seattle; Terry Hensle, New York; Kenneth Crooks, Columbus; Jens Rosenkrantz, Los Angeles; and Judah Folkman and Hardy Hendren, each now Emeritus Chiefs in Boston. Robert E. Gross (1905–1988) (Fig. 1a.3) was the most outstanding of those who trained under Ladd. He graduated from Harvard Medical School in 1931. Charles F. McKhann, Professor of Pediatrics, wrote a letter of recommendation about Gross to Ladd. Quoting from this letter given to me by his son, Charles McKhann Jr.: “Mr. R. E. Gross, a member of the fourth year class of Harvard Medical School, has asked me to write you concerning his qualifications for the position of House Officer in the Children’s Hospital. Mr. Gross is an interested, eager and accurate student, somewhat above average, has a pleasant personality, and a good appearance. He should make a satisfactory House Officer.” This understated letter did not portend what Gross would accomplish. His training included first a residency in Pathology under S. Burt Wolbach. He trained in general surgery under Elliot Cutler at the Peter Bent Brigham, and in Pediatric Surgery

5

6

W. H. Hendren

Fig. 1a.3. Department of Surgery in 1959 under Robert E. Gross. Front row, left to right: Ernest Barsamian, Samuel Schuster, Thomas Holder, Robert Gross, Luther Longino, Donald MacCollum, Robert Smith, and Hardy Hendren. Second row, left to right: Lawrence Hill, Arnold Colodny, Donald Brief, John Crowe, Judson Randolph, David Collins, Morton Wooley, Lon Curtis, and Mayo Johnson.

under Ladd. Seven years after McKhann’s letter, while he was Chief Surgical Resident, Gross successfully divided a patent ductus arteriosus.13 The operation was performed when Ladd was out of town. Ladd never forgave Gross for that professional sleight. Gross later confided that he did not think Ladd would have allowed the operation if he had been present. The patient, Lorraine Sweeny, then 8 years old, is alive and well today, 66 years later. This underscores how pediatric surgeons strive to attain a long and productive life for their patients, often not possible in adults. Ladd retired in 1945. Franc Ingraham, Chief of Neurosurgery, was appointed as interim Chief while an ad hoc committee deliberated for 2 years (not uncommon at Harvard). Gross was appointed Chief in 1947. Churchill was chairman of the ad hoc committee! Visitors who came to Boston to observe or work with Ladd and Gross included several who became very distinguished pediatric surgeons and mentors. C. Everett Koop, sent by his Chief, Isadore Ravdin, returned to Philadelphia to be the first Surgeon in Chief at that Children’s Hospital.14 Willis J. Potts returned to the Children’s Memorial Hospital in Chicago. Robert Zachary returned to Sheffield, England and trained many registrars in pediatric surgery. Jesus Lozoya-Solis returned to Mexico City to become the dom˜ was inant figure in pediatric surgery there. Alberto Pena one of his many pupils. Lozoya believed pediatric surgeons should be first a pediatrician and secondarily a surgeon, differing from the opinion of Ladd and Gross. That pattern

became prevalent in many Spanish-speaking countries of Central and South America. Clarence Crafoord of Stockholm, Sweden reported successful resection of coarctation of the aorta in 1944.15 Gross soon followed with his own important contributions to coarctation.16 From his laboratory came also the use of human cadaver aortic grafts to replace the narrow aortic segment in coarctation cases which are too long for excision and primary anastomosis.17 The grafts were freeze-dried and radiated for sterilization. This was a landmark contribution to the field of vascular surgery. Perchance I was recently contacted by an 88-year-old man who is Dr. Gross’s first homograft coarctation repair. He is in good health, and had just returned from competitive bowling! Ladd and Gross published a book in 1941 on Abdominal Surgery of Childhood.18 It presented experience and statistics of follow-up in pediatric surgery at Children’s, from 1915 to 1941. The book is a classic. Recently a copy was given to me which was given to Wolbach in 1941 by Gross. An inscription acknowledges Gross’ gratitude for Wolbach’s mentoring and guidance, expressing his high regard for Wolbach as his primary teacher and advocate. In 1953 Gross published his own single author book, The Surgery of Infancy and Childhood.19 It was dedicated to Wolbach. The copy given to Wolbach by Gross, which was recently given to me, bears the inscription: “Dear Uncle Burt, With this book come my deepest thanks for all you’ve done in so many ways. Devotedly, Bob.” The book remains a classic today. It should be read by all pediatric surgeons. Many thousands of copies were published in multiple printings, in several different languages. Therefore, it can usually be found and purchased (for 15–20 times its original cost!). Gross described repair of anomalies of the great vessels, which constrict the trachea, esophagus or both.20 Little has been added concerning vascular rings in the past 50 years. He described treating omphalocele by wide undermining of the skin, and temporary closure over the viscera, leaving a huge abdominal wall hernia to repair later.21 An interim technique used by some for several years was painting the sac with mercurochrome, awaiting its gradual epithelialization and contracture. His pupil, Samuel R. Schuster, later introduced temporary silastic covering of the protruding viscera, with staged closure soon thereafter.22 This has endured. Willis Potts authored a unique book in 195923 describing some of the more common entities in pediatric surgery. He wrote, “I want to dedicate this book to the infant who has the great misfortune of being born with a serious deformity. All life is before him and what is done during the first

Introduction: North American perspective

few days may decide whether life will be a joy or a burden. If this infant could speak it would beg imploringly of the surgeon, ‘Please exercise the greatest gentleness with my miniature tissues and try to correct the deformity in the first operation. Give me blood and the proper amount of fluids and electrolytes; add plenty of oxygen to the anesthesia and I will show you that I can tolerate a terrific amount of surgery. You will be surprised at the speed of my recovery and I shall always be grateful to you’.” Regarding imperforate anus, Potts wrote, “In general, atresia of the rectum is more poorly handled than any other congenital anomaly of the newborn. A properly functioning rectum is an unappreciated gift of greatest price. The child who is so unfortunate as to be born with an imperforate anus may be saved a lifetime of misery and social seclusion by the surgeon who with skill, diligence and judgment performs the first operation on the malformed rectum”. Pediatric surgeons continue to strive to correct the difficult entity of imperforate anus, most espousing the posterior sagittal approach ˜ 24 introduced by DeVries and Pena. Although Gross opened the field of congenital heart surgery, others soon followed him. Alfred Blalock at Johns Hopkins advanced treatment of blue babies by introducing the Blalock–Taussig subclavian–pulmonary artery shunt.25 Gross regretted that he had not paid heed to Helen Taussig who had visited him earlier from Hopkins and suggested “making a ductus” as treatment for blue babies. Willis Potts devised a direct aorta to pulmonary artery shunt.26 Other surgeons worked to perfect a heart–lung machine. First was John Gibbon, of Jefferson Medical College in Philadelphia. He successfully closed an atrial septal defect for a child using a cardiopulmonary bypass but did not venture further into more complex defects. His research had started in the laboratory of Edward Churchill in Boston, who did not think there was much merit in that idea! In 1952 John Lewis showed that simple atrial defects could be closed with brief inflow occlusion and hypothermia, but this did not give enough time to repair complex malformations. Gross employed a rubber well sewn to the right atrium to access and close atrial septal defects.27 Results of that were imperfect. It was C. Walton Lillehei in Minneapolis who focused the spotlight on congenital heart surgery. On April 25, 1954, he closed a large interventricular septal defect in a 4-yearold girl, with the father providing cardiopulmonary support by cross-circulation between parent and child, despite vigorous outcry by members of the medical department. His surgical chief, Owen Wangansteen, stood by him. He then demonstrated repair of complex anomalies, such as Tetralogy of Fallot.28 This interesting surgical history was recently reported in riveting detail in a book entitled, King of Hearts by G. Wayne Miller, in 2000.29 Thereafter a practical car-

diopulmonary bypass machine was introduced by Lillehei and soon adopted by many surgeons, a key development in bringing congenital heart surgery to its current advanced state. Gross stepped down as Surgeon in Chief in 1967, to concentrate on cardiac surgery. He retired in 1972. Gross revealed later that, during his entire career, he had operated with good vision in only one eye. This was told to one of his former residents who was facing loss of vision in one eye from a melanoma. Gross in retirement underwent surgery for his congenital cataract. That gave him binocular vision he had lacked during his brilliant surgical career. A Harvard Chair was named in his honor in 1985, on the occasion of his 80th birthday. The author was its first incumbent. Gross died with Alzheimer’s disease on October 11, 1988, closing the career of a giant in American Surgery. Ladd and Gross were both accomplished and imaginative surgeons. Their department of surgery trained a multitude of pediatric surgeons in North America. This heritage has been self-perpetuating. Recognition of pediatric surgery as a specialty did not come easily. Randolph outlined the genesis of other early programs in the United States:30 Willis Potts in Chicago, 1947; Everett Koop in Philadelphia, 1950; Ovar Swenson at Boston Floating Hospital, 1952; William Clatworthy in Columbus, 1955; William Kiesewetter in Pittsburgh, 1958; Thomas Santulli, in New York. There were also Canadian programs in Montreal, Toronto, Winnipeg, Ottawa, and Vancouver. The first full-time pediatric surgeon in North America was Herbert Coe in Seattle; he had visited Children’s Hospital in Boston when Ladd had not yet limited his work to children. Coe was the driving force in establishing in 1948 the Surgical Section of the American Academy of Pediatrics. The Journal of Pediatric Surgery began in 1965 through the efforts of Everett Koop and Stephen Gans. The American Pediatric Surgical Association was founded in 1970. Recognition of special competence in pediatric surgery by the American Board of Surgery was finally accomplished in 1975 through the efforts of Harvey Beardmore and others. The British Association of Pediatric Surgeons which formed in 1953, has promoted interchange of knowledge among surgeons the world over. It has been fascinating to be privy to all of these developments spanning more than half a century. The field of pediatric surgery has been virtually transformed in the past 50 years. Space does not permit mentioning all of the contributions and all of the contributors, but some deserve to be highlighted. Douglas Stephens of Melbourne, Australia, and later Chicago, taught us much about the embryology and classification of anorectal and genitourinary malformation.31 Jonathan Rhodes of

7

8

W. H. Hendren

Philadelphia reported the first primary pullthrough in the neonate with imperforate anus.32 Cloacal malformations, once a no-man’s land of pediatric surgery, have become reparable.33,34 Infants with cloacal exstrophy always died, until a survivor was reported in 1960 by Peter Rickham, then in Liverpool.35 Today, most of these infants can be repaired in a satisfactory fashion to lead useful lives. Key to development of much major pediatric surgery was the emergence of pediatric anesthesiology, which can support even small babies through a prolonged surgery as well as help surgeons care for them postoperatively. It is now appreciated that fluid loss and necessary replacement are much greater than we practiced 50 years ago to maintain metabolic hemostasis in the pediatric surgical patient. Orvar Swenson, another prot´eg´e of Ladd, pioneered the enormous contribution of unlocking the mystery of Hirschsprung’s disease. With Alexander Bill, he proved the pathology to be the aganglionic distal segment and described an effective operation for it.36 Originally, the clinical picture of Hirschsprung’s disease was an older child with lifelong constipation. Soon it was recognized that the most common presentation is a newborn with intestinal obstruction. It accounts for one-third of neonatal cases of bowel obstruction. To be sure, Duhamel, Soave, Rehbein, and others made modifications, but Swenson’s carefully documented clinical research remains the basis for treatment of this problem half a century later. Swenson recently described the long-term result of his experience with Hirschsprung’s.37 Martin described a practical surgical solution for the infant with total colonic aganglionosis.38 Judah Folkman succeeded Gross as Surgeon in Chief at Children’s Hospital in 1968. Twenty-three years earlier Koop had learned pediatric surgery from Ladd and Gross. In 1968 Folkman went to Philadelphia for a “cram course” in pediatric surgery by Koop, a successful quid pro quo for both institutions! For 14 years Folkman carried the Herculean load of that office plus overseeing a large surgical research laboratory, mentoring many postdoctoral research fellows. He chose in 1982 to concentrate his enormous productivity on the ever-expanding research laboratory program, devoted in large part to the field of angiogenesis research. This had begun with Folkman’s astute observation that tumors grow by attracting new blood vessels. He reasoned that control of angiogenesis might control growth of tumors, as well as other non-malignant conditions characterized by vascular ingrowth, such as diabetic retinopathy. Alpha interferon, an angiogenic inhibitor, was shown to reduce dramatically the mortality of lifethreatening giant hemangiomas in infants.39 Like many spectacular advances, Folkman’s work provoked derisive

commentary and skepticism initially. However, his vision was amply vindicated by many other scientists who joined in this investigative effort around the world. Many clinical trials are in progress today at many major hospitals. The author succeeded Folkman as Chief of the Department of Surgery in 1982. Joseph Murray opened the field of organ transplantation in Dec. 1954 when he transplanted a kidney from one identical twin to his brother in end-stage renal failure.40 Later in 1962, he demonstrated successful cadaver renal transplantation using immunosuppression. For the seminal advance of transplantation, Murray was awarded The Nobel Prize in 1990.41 Soon Starzl opened the field of liver transplantation.42 Now half a century later heart, lung, pancreas, intestine, and even multivisceral organ transplantation have been added to the surgeons’ repertoire. Organ availability has been a problem since transplantation began. Joseph Vacanti started the field of tissue engineering to solve this dilemma.43 Dudrick’s introduction of total parenteral nutrition44 has saved countless lives of children, and adults, where nutritional needs cannot be met otherwise. This began with a long struggle to save a baby with gangrenous bowel from malrotation and midgut volvulus at Philadelphia Children’s Hospital. Another boon to mankind! Neonatal physiology and nutrition45 advanced greatly as pediatric surgery grew in scope and stature.45 Patricia Donahoe made great strides in developmental biology, starting with investigation of the Mullerian Inhibiting Substance.46 Her investigations broadened to study mechanisms of fetal growth and differentiation.47 Robert Bartlett conceived the idea of extracorporeal membrane oxygenation while working as a resident under Robert Gross.48 This saved countless lives in children and adults with cardiopulmonary failure. Ancillary to that was vast improvement in treatment of neonates with severe diaphragmatic hernia. Michael Harrison and Alfred deLormier imaginatively introduced the field of fetal surgery.49 Harrison’s pupils, Adzick, Jennings, Flake, and Longaker and others carried this new field further forward, making it possible to salvage infants with once fatal problems. Burn treatment was pushed to the forefront by the Coconut Grove Nightclub fire on Nov. 28, 1942 in Boston.50 Much was learned about burns in caring for victims of that well known tragedy. Ultimately, expeditious excision and grafting of extensive deep burns was developed,51,52 and the use of artificial skin to close large burn wounds after excision.53,54 Burn mortality dropped remarkably through early excision and grafting.55

Introduction: North American perspective

Limb replantation began in 1963 when Ronald Malt at Massachusetts General Hospital reunited the severed arm of a young boy after traumatic amputation,56 sparking worldwide salvage of many digits and limbs. Tracheal resection in children was made possible by the pioneering work of Hermes Grillo at MGH, who developed segmental tracheal and bronchial resection for benign and malignant conditions in adults. This soon became equally useful in children.57,58 Cancer chemotherapy was another seminal advance for pediatric surgery, greatly improving the cure rate of many malignancies. It was my privilege as a surgical resident to witness introduction of this in 1955 in Wilms’ tumor cases by Sidney Farber, Chief of Pathology at Children’s. Like Gross, Farber was a pupil of S. Burt Wolbach. Farber began with the premise that some types of cancer might be controlled by chemical agents. At the Fourth Annual Internal Cancer Research Conference in St. Louis, MO, in September, 1947, a case was presented of a patient who had a remarkable clinical response to a new experimental chemotherapeutic agent, Teropterin (teryl-triglutamic acid). This was the result of the keen observation by Brian L. Hutchings of the Lederle Laboratories. In 1942 Hutchings was trying to produce folic acid in large quantities and noted that one of the vats contained a filtrate that stimulated growth of a certain microorganism, whereas other vats produced no such growth. This filtrate, initially thought to be folic acid turned out to be a folic acid analog, Teropterin. It was shown subsequently that sarcomatous tumors transplanted into mice disappeared with the addition of this drug. The anonymous patient presented at the conference was “Babe” Ruth, the celebrated slugger of the New York Yankees. In 1946 he presented with hoarseness, left retroorbital pain, and a neck mass. The neck mass was partially excised. The primary tumor was not laryngeal in origin as generally believed but a primary nasopharyngeal cancer at the base of the skull. He had a remarkable temporary relief from pain and regression of the tumor on treatment with teropterin.59 Sidney Farber investigated the role of this agent and other folic acid agonists and antagonists. Farber’s early work led to saving the lives of many children with leukemia,60 which was once uniformly fatal. He organized the laboratory and clinical facility for chemotherapy, named the Jimmy Fund, in honor of one of the early tumor successes, a patient with an intestinal lymphoma. That patient is alive today. It was the beginning of a new era. The impact of chemotherapy for pediatric malignancies is now well established after nearly half a century of close collaboration between medical oncologists, radiotherapists, and surgeons. The Jimmy

Fund Building was joined to the Charles A. Dana Cancer Center for adults, and was designated a comprehensive cancer center in 1973. Today, it is called the Dana-Farber Cancer Center. Pediatric urology as a recognized field was virtually nonexistent 50 years ago. On my arrival at Children’s Hospital, the resident just finishing gave the news that one duty would be to conduct the Urology Clinic every Monday afternoon. When professing to have no knowledge about pediatric urology, he was reassuring by saying “Don’t worry, nobody else does either!” The clinic had many children with various tubes to be changed: nephrostomy, cystostomy, or urethral. There was no senior supervision of the clinic. That sparked an interest to learn something about urologic problems. Looking back through half a century, pediatric urology became a robust and recognized surgical specialty.61 It began with a few general urologists who founded the Society for Pediatric Urology in 1951. On the 50th Anniversary, there were 437 members and 9 honorary members. Dr. Meredith Campbell of New York was the standard bearer. He was President for 5 years. His two volume Pediatric Urology62 described the anomalies and diseases of the child’s urinary tract. It was a scholarly work and illustrated routine operations of that era. All present pediatric urologists should read the book to appreciate the great advances in the past 50 years. Pediatric urology has paralleled the development of pediatric general surgery, although one generation later. As a specialty it was accepted slowly, just as occurred in pediatric surgery in 1975. The American Board of Urology recently approved certification of special competence, 30 years after that was achieved for pediatric general surgery. The Urology Section of the American Academy of Pediatrics, founded in 1971 currently has 462 members. Its annual meeting became the premier forum for those interested in this field. To illustrate progress in the specialty, several areas deserve mention. Exstrophy of the bladder was managed mainly by ureterosigmoidostomy diversion and cystectomy until Jeffs,63 and Chisholm64 and others showed that primary repair is feasible although not always successful. This concept was advanced further with bladder augmentation, continent diversion, and even total repair in the newborn (bladder closure, ureteral reimplantation, and epispadias repair simultaneously). Cloacal exstrophy was deemed insoluble until Rickham reported a survivor in 1960.35 From this beginning extensive repair became feasible, with urinary continence in most and pull through of the colon in some of those with contractile muscle of the pelvis and perineum.65

9

10

W. H. Hendren

Ureterocele was treated largely by wide unroofing; this relieved obstruction but produced massive reflux which furthered renal damage and often produced incontinence in those with ectopic ureteroceles which require repair of the cleft bladder neck and urethra. Superior results were achieved by total repair of the sometimes complex anatomy in one stage.66 Urinary tract imaging was formerly limited to intravenous pyelogram, retrograde pyelogram and static cystogram (often under anesthesia). Cine radiography taught much about both function and anatomy, but at the price of excessive radiation exposure. Development of more advanced techniques, such as “spot films,” ultrasound study, nuclear contrast, magnetic resonance technique, and computerized tomography improved enormously the accuracy of urodiagnosis. Endoscopic visualization of the urinary tract 50 years ago using battery powered incandescent light bulbs for illumination was crude to say the least, especially in infants for whom small caliber endoscopes were not available. Fiberoptic endoscopy revolutionized that modality, especially when flexible technique developed. Each training department should retain an old incandescent cystoscope to give trainees an appreciation for the blessings of new technology! Urethral valves were described accurately in 1919 by Hugh Hampton Young, the father of North American Urology, at Johns Hopkins Hospital.67 They were destroyed by blindly inserting a cold punch, which did not allow much accuracy. Open cystotomy carried down into the prostate urethra afforded better valve excision but sometimes produced incontinence.19 Accurate valve surgery resulted from voiding cystourethrography and the improved vision with fiberoptic scopes. The result was recognition that urethral valves, like most pathology, occur in a spectrum of severity68 from grade 4 with severe obstruction and hydroureteronephrosis, to grade 1 “mini valves,” which show only subtle radiographic findings and have no upper tract dilatation. Vesicoureteral reflux came to the forefront in the late 1950s and early 1960s, as dynamic urography developed. Wyland Leadbetter and his associate Victor Politano described the most often used technique of abolishing reflux by tunneling reimplantation of the ureter:69 many variations followed. This work not only focused clinical awareness on the importance of reflux, but also served as a starting wedge for much of reconstructive urology. Ileal loop introduced by Bricker in 195070 for drainage after anterior pelvic exenteration became an important method of drainage for both adults and children. It lessened greatly use of ostomy tubes of various sorts, often reducing

incidence of urinary infection which always accompanies long standing tube drainage. However, after a decade of ileal loops, it became apparent that upper tracts often deteriorated, secondary to non- sterile reflux.71 This was prevented by use of the colon conduit in which non-refluxing tunnels can be made.72 Megaureter was largely ignored in the literature, but its repair was a logical next step after ureteral reimplantation became well established. A book devoted exclusively to the ureter published in 1967 largely omitted this condition, with conclusion that it lacked importance and was not repairable.73 However, after Bischoff in Germany described shortening and tapering the ureter,74 operation was introduced in North America to shorten, taper, and tunnel reimplant the ureter with a high success rate.75 Clinical experience and urodynamic study proved that a dilated ureter does not propel urine effectively because the walls cannot coapt. Tubeless ureterostomy became popular, using end, loop, Roux-en-Y and even circle technique, as well as pyelostomy. That phase was short lived after the high complication rate of those diversions was reported,76 and it was shown that the infant urinary tract can be primarily reconstructed without prior drainage.77 Evaluation of the changes in urologic surgery described above evolved to reconstruction of previously diverted urinary tracts.78,79 This led to the conclusions that (i) most diversions can be undiverted and (ii) most diversions are not necessary in the first place! Hypospadias repair was vastly different at Boston Children’s Hospital in 195319 from what surgeons practice today. Mild cases were not repaired. Severe cases were done in three stages, and were delayed until pubertal years. At that time MacCollum described 40 patients who had undergone satisfactory chordee release, but only 18 had completed stage 3. He stated, “The results have been exceedingly satisfactory and we see little need for improvement or change in operative technique, with the possible exception that earlier repair might have been desirable in some of the boys, a consideration which we did not find to be of great importance in many cases.” From that viewpoint much changed, most recognizing that early repair is both feasible and desirable, and that there are many ways this can be accomplished, giving a penis which looks normal and has satisfactory function for both micturition and procreation!

Conclusions A retrospective view of both pediatric surgery and pediatric urology in the 20th century shows monumental changes and enormous benefits for pediatric patients. This evokes

Introduction: North American perspective

speculation about what lies ahead as the efforts of many surgeons introduce solutions to many of the problems which still exist. A symposium on Pediatric Surgery was published in 1938.80 The editor was William E. Ladd. His introduction stated, “Undoubtedly great strides have been made in this field of surgery in the last few years and I have confidence that greater advances are soon to follow.” That prophesy was correct. It will surely prove true again when we look back in the distant future to our care of children at the beginning of this century.

REFERENCES 1. Warren, J. C. Surgical Observations on Tumors with Cases and Operations. Boston: Crocker and Brewster; London: John Churchill, Princes St., Soho, 1839. 2. Warren, J. M. Surgical Observations with Cases. Boston: Ticknor and Fields, 1867. 3. Smith, C. A. The Children’s Hospital of Boston “Built Better Than They Knew.” Boston: Little Brown, 1983. 4. Mallon, W. J. Ernest Amory Codman: The End Result of a Life in Medicine. Philadelphia: W. B. Saunders, 2000. 5. Ladd, W. E. The surgical treatment of esophageal atresia and tracheoesophageal fistulas. N. Engl. J. Med. 1944; 230:625. 6. Leven, N. L. Congenital atresia of the esophagus with tracheoesophageal fistula. J. Thorac. Surg. 1941; 10:648. 7. Lanman, T. H. Congenital atresia of the esophagus: a study of thirty-two cases. Arch. Surg. 1940; 41:1060. 8. Haight, C. & Towsley, H. A. Congenital atresia of the esophagus with tracheoesophageal fistula; extrapleural ligation of fistula and end-to-end anastomosis of esophageal segments. Surg. Gynecol. Obstet. 1943; 76:672–688. 9. Glick, P. H. & Azizkhan, R. G. A Genealogy of North American Pediatric Surgery From Ladd Until Now. St. Louis, MO: Quality Medical Publishing, 1997. 10. Bill, H. William E. Ladd, M. D., Great Pioneer of North American Pediatric Surgery. Progr. Pediatr. Surg. 1986; 20:52–59. 11. Grillo, H. C. Edward D Churchill and the “rectangular” surgical residency. Surgery 2004; 136:947–952. 12. Churchill, E. D. The segmental and lobular physiology and pathology of the lung. J. Thorac. Surg. 1949; 18:279. 13. Gross, R. E. & Hubbard, J. P. Surgical ligation of a patent ductus arteriosus: report of first successful case. J. Am. Med. Assoc., 1939; 112:729. 14. Koop, C. E. A perspective on the early days of pediatric surgery. J. Pediatr. Surg. 1998; 33:953–960. 15. Crafoord, C. & Nylin, G. Congenital coarctation of the aorta and its surgical treatment. J. Thorac. Surg. 1945; 14:347. 16. Gross, R. E., Hufnagel, C. A. Coarctation of the aorta. Experimental studies regarding its surgical correction. N. Engl. J. Med. 1945; 233:287. 17. Gross, R. E. Treatment of certain coarctations by homologous grafts: a report of 19 cases. Ann. Surg. 1951; 134:753.

18. Ladd, W. E. & Gross, R. E. Abdominal Surgery of Infancy and Childhood. Philadelphia: W.B. Saunders, 1941. 19. Gross, R. E. The Surgery of Infancy and Childhood. Philadelphia: W. B. Saunders. 1953. 20. Gross, R. E. & Newhauser, E. B. D. Compression of the trachea or esophagus by vascular anomalies, surgical therapy in 40 cases. Pediatrics 1951; 7:69. 21. Gross, R. E. A new method for surgical treatment of large omphaloceles. Surgery 1948; 24:277. 22. Schuster, S. R. A new method for the staged repair of large omphaloceles. Surg. Gynecol. Obstet. 1967; 125:837–850. 23. Potts, W. J. The Surgeon and The Children. Philadelphia: WB Saunders, 1959. 24. DeVries, P. A. & Pena, A. Posterior sagittal anorectoplasty. J. Pediar. Surg. 1982; 17:638–643. 25. Taussig, H. B. Analysis of malformations of the heart amenable to a Blalock–Taussig operation. Am. Heart. J. 1943; 36:321. 26. Potts, W. J., Smith, S., & Gibson, S. Anastomosis of aorta to a pulmonary artery. J. Am. Med. Assoc. 1945; 132:627. 27. Gross, R. E., Pomeranz, A. A., Watkins, E. et al. Surgical closure of defects of the interauricular septum by use of an atrial well. N. Engl. J. Med. 1952; 247:455. 28. Lillehei, C. W., Cohen, M., Warden, H. E. et al. Direct vision intracardiac surgical correction of the tetralogy of Fallot, pentology of Fallot, and pulmonary atresia: report of first 10 cases. Ann. Surg. 1955; 142:418. 29. Miller, G. W. King of Hearts. New York: Random House 2000. 30. Randolph, J. G. The first of the best. J. Pediatr. Surg. 1985; 20:580–591. 31. Stephens, F. D. Congenital Malformations of the Rectum, Anus and Genito-urinary Tract. London: Livingstone, 1963. 32. Rhodes, J. E., Pipes, R. L., & Randall, J. P. Simultaneous abdominal and perineal approach in operations for imperforate anus with atresia of rectum and rectosigmoid. Ann. Surg. 1948, 127:552. 33. Hendren, W. H. Cloacal malformations: experience with 105 cases. J. Pediatr. Surg. 1992; 27:890–901. 34. Hendren, W. H. Cloaca, the most severe degree of imperforate anus: experience with 195 cases. Ann. Surg. 1998; 228:331– 346. 35. Rickham, P. P. Vesico-intestinal fissure. Arch. Dis. Child. 1960; 35:97–102. 36. Swenson, O. & Bill, A. H. Resection of rectum and rectosigmoid with preservation of the sphincter for benign spastic lesions producing megacolon: an experimental study. Surgery 1948; 24:212. 37. Swenson, O. Early history of the therapy of Hirschsprung’s disease: facts and personal observations over 50 years. J. Pediatr. Surg. 1996; 31:1003–1008. 38. Martin, L. Surgical management of total colonic aganglionosis. Ann. Surg. 1972; 176:343. 39. Ezekowitz, R. A., Mulliken, J. B., & Folkman, J. Interferon alpha 2A therapy of “life-threatening” hemangiomas in infancy. N. Engl. J. Med. 1992; 326:1456–1463.

11

12

W. H. Hendren

40. Merrill, J. P., Bartlett, J. E., Harrison, J. H. et al. Successful homotransplantations of the human kidney between identical twins. J. Am. Med. Assoc. 1956; 160:277–282. 41. Murray, J. E. The first successful organ transplantation in man. Scand. J. Immunol. 1994; 39:1–11. 42. Starzl, T. E. & Demetris, A. J. Liver Transplantation. Chicago: Yearbook Medical Publ., 1990. 43. Choi, R. S. & Vacanti, J. P. Preliminary studies of tissueengineered intestine using isolated epithelial organoid units on tubular synthetic biodegradable scaffolds. Transpl. Proc. 1997; 29:848–851. 44. Dudrick, S. J., Wilmore, D. W., Vars, H. M. et al. Long-term total parenteral nutrition with growth, development and positive nitrogen balance. Surgery 1968; 64:134. 45. Coran, A. G. Parenteral nutrition in infants and children. Surg. Clin. North. Am. 1981; 61:1089–1100. 46. Donahoe, P. K., Ito, Y., & Hendren, W. H. A graded organ culture assay for the detection of Mullerian inhibiting substance. J. Surg. Res. 1877; 23:141–148. 47. Teixera, J., He, W. W., Shah, P. C. et al. Developmental expression of a candidate Mullerian inhibiting substance type II receptor. Endocrinology 1996; 137:160–165. 48. Bartlett, R. H., Gazzaniga, A. B., Jeffries, M. R. et al. Extracorporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy. Trans. Am. Soc. Artif. Intern. Organs 1976; 22:80– 93. 49. Harrison, M. R., Golbus, M. S., & Filly, R. A. The Unborn Patient, 2nd edn. Philadelphia, PA, 1990. 50. Symposium on the Management of the Coconut Grove Burns at the Massachusetts General Hospital. Ann. Surg. 1943; 117:801– 965. 51. Cope, O., Langohr, J. L., Moore, F. D. et al. Expeditious care of full-thickness burn wounds by surgical excisions and grafting. Ann. Surg. 1947; 125:1–22. 52. Hendren, W. H., Constable, J. D., & Zawacki, B. E. Early partial excision of major burns in children. J. Pediatr. Surg. 1968; 3:445– 464. 53. Burke, J. F., Yannas, I. V., Quinby, W. C. Jr et al. Successful use of a physiologically acceptable artificial skin in the treatment of extensive burn injury. Ann. Surg. 1981; 194:413–428. 54. Burke, J. F. From Desperation to skin regeneration: progress in burn treatment. J. Trauma 1990; 30:S36–S40 (suppl 12). 55. Tompkins, R. G., Remensyder, J. P., Burke, J. F. et al. Significant reductions in mortality for children with burn injuries through the use of prompt eschar excision. Ann. Surg. 1988; 208:33– 41. 56. Malt, R. A. & McKhann, C. F. Replantation of severed arms. J. Am. Med. Assoc. 1964; 189:716–722. 57. Grillo, H. C. & Zannini, P. Management of obstructive tracheal disease in children. J. Pediatr. Surg. 1984; 19:414–416.

58. Grillo, H. C. Pediatric tracheal problems. Chest Surg. Clin. North. Am. 1996; 6:693–700. 59. Bikhazi, N. B., Kramer, A. M., Spiegel, J. H. et al. “Babe” Ruth’s illness and its impact on medical history. Laryngoscope 1999; 109:1–2. 60. Frei, E. 3rd, Jaffe, N., & Farber, S. Treatment of acute leukemia. N. Engl. J. Med. 1971; 287:1357. 61. Hendren, W. H. From an acorn to an oak. J. Pediatr. Surg. 1999; 34:46–58. 62. Campbell, M. F. Pediatric Urology, 2 vol. New York: The MacMillan Co. 1937. 63. Jeffs, R. D. Exstrophy of the urinary bladder. In Pediatric Surgery, Chicago, Il: Year Book Medical Publ., 1986. 64. Chisholm, T. C. & McPharland, F. A. Exstrophy of the urinary bladder. In Pediatric Surgery, Chicago, Il: Year Book Medical Publ, 1982. 65. Lund, D. P. & Hendren, W. H. Cloacal exstrophy: a 25 year experience with 50 cases. J. Pediatr. Surg. 2001; 36:68–75. 66. Hendren, W. H. & Michael, M. E. Surgical correction of ureteroceles. J. Urol. 1979; 121:590–597. 67. Young, H. H., Frontz, W. A., & Baldwin, J. C. Congenital obstruction of the posterior urethra. J. Urol. 1919; 3:289. 68. Hendren, W. H. Posterior urethral valves in boys: a broad clinical spectrum. J. Urol. 1971; 106:298–307. 69. Politano, V. A. & Leadbetter, W. F. An operative technique for the correction of vesicoureteral reflux. J. Urol. 1958; 79:932–941. 70. Bricker, E. M. Bladder substitution after pelvic evisceration. Surg. Clin. North Am. 1950; 30:1511. 71. Middleton, A. W. & Hendren, W. H. Ileal conduits in children at Massachusetts General Hospital from 1955 to 1970. J. Urol. 1976; 115:591–595. 72. Althausen, A. F., Hagen-Cook, K., & Hendren, W. H. Nonrefluxing colon conduit: experience with 70 cases. J. Urol. 1978; 120:35–39. 73. Bergman, H. (ed) The Ureter. New York: Harper & Row, 1967. 74. Bischoff, P. Megaureter. Br. J. Urol. 1957; 29:416. 75. Hendren, W. H. Operative repair of megaureter in children. J. Urol. 1969; 101:491–507. 76. Hendren, W. H. Complications of ureterostomy. J. Urol. 1978; 120:269–281. 77. Hendren, W. H. A new approach to infants with severe obstructive uropathy: early complete reconstruction. J. Pediatr. Surg. 1970; 5:184–199. 78. Hendren, W. H. Reconstruction of previously diverted urinary tracts in children. J. Pediatr. Surg. 1973; 8:135–150. 79. Hendren, W. H. Urinary tract refunctionalization after longterm diversion: a 20 year experience with 177 patients. Ann. Surg. 1990; 212:478–495. 80. Ladd, W. E. (Ed) Surgery in Children. Symposium. Am. J. Surg. 1938; 39:227–475.

1b Introduction and historical overview: European perspective Barry O’Donnell Ballsbridge, Dublin, Ireland

‘One man can make a difference’ John Fitzgerald Kennedy (1917–1963)

Each country has its own history of pediatric surgery. Each tale is usually short because, although many children’s hospitals are more than 100 years old, the growth of pediatric surgery as a specialty in Europe dates from after the Second World War. Before that, surgical services were provided by general surgeons, many of whom saw their appointment to the children’s hospital as an exotic and seldom-visited outpost of a far-flung empire. One man changed that. His name was “D.B.” – Denis John Walko (Aboriginal for “big man”) Browne (1892–1967) – surgeon at The Hospital for Sick Children, Great Ormond Street, London, from 1928 to 1957. An Australian who had served at Gallipoli, he was in the trenches of the First World War. The upright military majestic figure (6ft 3in) with, as he put it, the “long, thin, Irish upper lip,” was truly the mainspring of the subject in western Europe. His original mind and wide knowledge of logic and philosophy enabled him to produce “surprise” – the touchstone of genius – solutions to such varied problems as club-foot, cleft lip, and hypospadias. For all his ingenious techniques and specially designed instruments (“the combined mousetrap and can-opener is rarely a success as either”), his most important contribution was to encourage others, not just in the United Kingdom but all over Europe, to concentrate their efforts on the surgery of the child. He promoted his loyal lieutenants, David Waterston (1910–1985) and Harold (Nicky) Nixon (1917–1990), as well as encouraging David Innes Williams (born 1919) the founder of pediatric urology. He performed every form of surgery from hydrocephalus to club-foot, and in-between he would repair an esophageal atresia. Everything he did

and touched was irresistibly logical. Most important of all, he was a warm, generous spirit who always put the child first. In Britain, Hunterian Professorships are awarded for innovation. It is a comment on Denis Browne that he was given this award four times. Another apostle of his, Peter Rickham (1918–2003), went to Liverpool where, with the encouragement of the much-loved Isobella Forshall (1900– 1989), he built Liverpool into England’s second biggest unit. One overlooks Scotland at one’s peril. Glasgow had the Royal Hospital for Sick Children since 1883, and a dedicated surgeon, Matthew White, was a near-contemporary of D.B.’s – and he in turn encouraged his young people such as Wallace Dennison and Sam Davidson. The founding of the British Association of Paediatric Surgeons (BAPS) in 1953 was a further catalyst to the growth of the specialty in Europe. Although run from the UK, members were enrolled from all over the world and it became the organization outside the Americas. D.B. was its first president, from 1953 to 1957. But from the continent of Europe came David Vervat (Rotterdam), Fritz Rehbein (Bremen), Franco Soave (Genoa), G. Winkel Smith (Copenhagen), Bernard Duhamel (Paris), John Shanley (Dublin), Mattai Sulumma (Helsinki), Max Grob (Zurich) and Theodor Ehrenpreis (Stockholm). These were the principal national representatives in those early years. Bob Zachary of Sheffield was at these early meetings, as was the indefatigable Douglas Stephens of Melbourne. Most early units were established in the teeth of opposition from professors of surgery and the system specialists who, with some notable exceptions (and this reservation is put in only for legal purposes and has no basis in fact), flatly denied the necessity for the new discipline. “What will you people do?” was the question. At that time the main

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


13

14

B. O’Donnell

arguments in favor of the neophytes were neonatal surgery and Hirschsprung’s disease, the cure and cause of which had been discovered by Orvar Swenson in Boston. Superior results and shorter bed stay soon became apparent in common problems such as pyloric stenosis (which still had a 3% mortality in the 1950s), daycare of infant hernia, the abandonment of drainage in appendicitis, and orchidopexy based on Denis Browne’s anatomical classification. There were some false dawns. The natural, rare enigmatic regression of some neuroblastomas in patients under 1 year was not appreciated at that time, and its “cure” was mistakenly attributed to repeated injections of vitamin B12 ; but there were many advances based on reclassification. Perhaps the most striking of these was in Wilms’ tumor which was then thought by the pathologists to be a single entity without any concept of an anaplastic, uniformly lethal group. It is chastening to realize now that there are at least 13 subgroups of Wilms’ tumor, and this figure seems to be added to every year or so. The concept of the controlled trial, one of the biggest medical advances of the past 50 years, has never had much allure for surgeons. This has led us into a number of misconceptions. Vesicoureteric reflux was a growth industry in the 1960s; it was looked on then as a progressive rather then regressive disease, and its association with kidney damage was ill-understood. It was assumed that the cause of the reflux was the radiologically impressive but unmeasurable bladder-neck obstruction. This meant that the valuable operation of antireflux ureteric reimplant was often combined with a rarely useful revision of the bladder neck. We were slow learners; we frequently disbelieve our own research. Open spina bifida, which was for many decades an epidemic in western Europe, seemed an appropriate subject for pediatric surgeons, particularly as they would treat the whole child rather than a collection of systems. Indeed, at one stage in the 1970s it almost looked as if the issue of spina bifida was set to dominate the entire subject of pediatric surgery. Evangelical reports of really good outcomes following emergency newborn closure quickly appeared, and controlled trials were felt by the prophets and the promoters to be unnecessary. However, the results were not really influenced by early operation. When there was a report that 60% of people with spina bifida in a certain area “led normal lives,” H. H. Nixon of London was moved to murmur “This is what I have long suspected, normal life in Xville is bloody awful.”

Changes in disease Neonatal surgery has been the flagship of the specialty, never more so than in the early years. But the pattern of

disease has changed. Intestinal malrotation and duodenal obstruction have been replaced as the commonest problems by two conditions that did not exist in the 1950s. Neonatal necrotizing enterocolitis surfaced as a problem for surgeons in the 1960s as smaller and smaller premature infants needing more and more care survived to develop this “survivor’s” disease. Most were under 1000 g and many were under 750 g. As with so many new conditions, the early reports were received with bafflement until they began turning up in everyone’s practice. The other entirely “new” condition was gastroschisis. This was thought of in the late 1960s to be a variant of “ruptured exomphalos,” but aggregate figures showed many differences, particularly the rarity of associated malformations in the “new” condition. The youth of the mothers – few are over 23 years – remains perplexing in this late-occurring malformation. Forty years ago, Hirschsprung’s disease (HD) was rarely if ever diagnosed in the newborn, but it has become one of the commonest causes of neonatal intestinal obstruction. It took years to discover that, unless there was rapid decompression of the distended bowel, the dreaded “enterocolitis” supervened, increasing the mortality and long-term morbidity of a condition that was thought to be eminently curable. I asked Dr Orvar Swenson in 1965: “What is the biggest problem in Hirschprung’s disease?” He said: “enterocolitis.” I had never heard of it until then. Hirschsprung’s disease was on every scientific program, and still is. Mattai Sulumma of Helsinki said in 1977: “In the year 2000 you will still be talking about HD.” We felt we had an ingenious but basically simple solution to this fascinating problem. “Skip segments” were totally discredited and intestinal neuronal dysplasia was seen as a great rarity of little practical importance. We now know otherwise. Then we were more concerned with the often heated debates about the “best” operation. The rectosigmoidectomy of Swenson, the retrorectal pull-through of Duhamel, and the endorectal pull-through of Soave, all gave “excellent” results in expert hands. If you were getting less than perfect results, then you were not doing the operation properly. We believed this and it took some time to show that there were other factors. That’s what long-term results are all about. Up to 1970, almost all intussusceptions were reduced by surgery before it was shown first that a barium enema and then an air enema was the treatment of choice. The introduction of metronidazole into the management of appendicitis, and with it the great reduction in anaerobic infection, reduced morbidity and reduced the incidence of wound infection in this common condition from over 10% to less than 2%.

Introduction: European perspective

Fig. 1b.1. The British Association of Paediatric Surgeons (BAPS) Annual Meeting, Rotterdam, the Netherlands, July 1964. (Left to right) Front, standing: Queen Juliana of the Netherlands (Patron of the meeting), Peter Rickham (Liverpool); First step: M. E. Muller (Hamburg), Isabella Forshall (Liverpool), Carlo Montagnani (Rome), V. Kafka (Prague). Second step: Theodor Ehrenpreis (Stockholm), C.E. Koop (Philadelphia), David Vervat (Rotterdam). Third step: Bob Zachary (Sheffield). Back row: Fritz Rehbein (Bremen), David Waterston (London), A.G. Brom (St Gallen, Switzerland), Sir Denis Browne (London). This selected group who were presented to the Queen typifies the international nature of the BAPS.

The really big advances In my view, two of the most important developments in that 40 years were, on the one hand, the advances in anesthesia and intensive care (including intravenous nutrition), and on the other hand the technological advances, particularly radiologic imaging. Intensive care began by keeping the sick postoperative patients in the hopefully named “recovery rooms” overnight, and the high nursing ratios made the difference between life and death to many. Some patients who had complicated surgery lasting more than 6 hours never came round in those days. In or about 1952, Peter Rickham of Liverpool – whose energy could be measured only on the Richter scale – moved a child with a repaired esophageal atresia into a side-room off the main ward and slept beside his patient for 10 days, feeding and giving him one-to-one care for those vital early days and nights. Was this the beginning of neonatal intensive care? Long intravenous (IV) lines were unknown, and at the beginning of IV feeding in the 1970s peripheral lines were resited daily to reduce venous thrombosis. The concept of empowering the nurses was slow to develop and “intravenous teams” were unknown. There is now a world shortage of intensive care nurses

owing to the “burnout” of caring for large numbers of marginal “life or death” patients. Intensive care is said to consume 20% of the entire US medical expenditure. Who could have predicted this? Technology drives progress, and although the heart–lung machine captured the public imagination, it was the spectrum of advances in what was “radiology” and is now “imaging” that has benefited a far greater number of patients. Ultrasound came into use in the early 1970s, and babies – before and after birth – were studied as never before. The fact that ultrasound interpretation was highly personal (the most important part of any X-ray report is the initials at the bottom) put a premium on old-fashioned talent, and recruitment into the specialty flourished. Computerized axial tomography (CT) in the early 1980s and magnetic resonance imaging (MRI) in the mid-decade opened our eyes still further. Nuclear medical imaging made a big impact, particularly in pediatric urology, where it made us revise our views in many areas (especially kidney function). Scarred kidneys that looked “not too bad” on the intravenous urogram were often shown to have poor function and to be not worth preserving. But there were so many subtle changes. Improved cannulae meant no more “cutdowns.” As a resident my first 1000 IVs were cutdowns. Surgeons got better instruments, especially vascular clamps, needle-holders, and even scissors. Better sutures and needles accelerated the trend towards finer and finer materials, which in turn encouraged us to use magnification for at least part of many operations. Better lighting came with fiberoptic lenses in the early 1970s, and for those who remember this change from the filament bulb it was like the jet engine succeeding the propeller, particularly in all forms of endoscopy. The cardiac surgeons led the way with the use of headlights, although Denis Browne had always worn one from the 1940s on and so had many of his followers. One disappointment was that the laser had a much more limited role than we had expected; and another was the failure of glues to replace sutured anastomoses. Are surgeons any better? Well, they have produced better results. Good craftsmanship takes time and surgeons no longer have to operate against the clock. The new breed are better taught and many have honed their skills in animal laboratories. I have a lasting and affectionate memory of Dr. Orvar Swenson standing on the other side of the table from “skin to skin” as I divided my first patent ductus arteriosus. He sweated the most, and not without reason.

Time is the touchstone It has been said that up to 1900 a patient had only a 50% chance of benefiting from an encounter with a doctor, and it is only since about 1950 that we have been able to prove

15

16

B. O’Donnell

that this ratio has dramatically improved. Statistically validated drug trials are a vital part of modern medicine, but surgery has trodden a more pragmatic and less defensible path. This is one reason why long-term follow-up is so valuable. It often allows a reclassification of a disease possibly into a group that does well and a group that does badly. The differences in outcome between high and low imperforate anus is a classic example. Even classification by birthweight and associated anomalies allowed us better to predict the outcome. Large numbers are important to get reliable reproducible data, but many of the most acceptable figures have come from smaller countries where stable populations cared for under comparable conditions give valuable information. Ilmo Louhimo’s long-term follow-up of a variety of conditions in Finland is a good example for us all.

Hospital notes are now preserved because the law and the lawyers insist, but many invaluable records have been destroyed in the past as a false economy because there was “no space” for storage. Computerized records and the paperless revolution have been slow to come to hospitals. To anyone interested in a condition, I would still say get two boxes of 6 by 4 (15 cm by 10 cm) record cards and put the diagnoses and patients’ names in one box and the patients’ details and diagnoses in the other, and enter the follow-up data into each. The ability to put symbols in sequence has been one of man’s greatest achievements, and this simple system is not subject to meltdown or viruses. Reliable long-term results are the cornerstones of our professional achievements.

2 Principles of outcomes analysis Brigid K. Killelea, Eric L. Lazar, and Michael G. Vitale Children’s Hospital of New York, NY, USA

Outcomes research differs in many ways from more traditionally performed clinical research. Whereas conventional “bench laboratory” experiments are designed to quantify short-term biological parameters for a given number of subjects, outcomes research employs strategies and methods to determine how disease and clinical interventions affect patient populations over time. Often, these studies are carried out in hopes of affecting public policy. In order to do so, study design must be rigorous, satisfying several criteria to ensure successful execution and valid results. For example, groups of equally distributed patients must be carefully selected, ethical and disease appropriate interventions must be implemented, and endpoints must be identified to accurately reflect how these study subjects are affected. At the same time, critical interpretation of results requires that potential biases be acknowledged and accounted for. Furthermore, as we will see, the assessment of children undergoing surgical treatment carries with it certain challenges unique to this population.

A brief history of outcomes research From the earliest days of contemporary medical practice, observers have questioned the effects of various medical therapies. In fact, the ability to judge and assess the effects of clinical intervention is a necessary prerequisite to gauge the evolution of medical treatment. Codman was among the first physicians to stress the importance of formally assessing the results of clinical intervention. As a result, he is considered the father of Outcomes Research. An orthopedic surgeon in Boston during the 1920s and 1930s, Codman espoused his “End Result Idea.”1

Thus in the year, 1910, at the age of forty, began the great and still unsuccessful interest of my life, over which I have toiled harder and suppressed more regrets than over any other star gazing period of my career . . . I had become interested in what I have called the END RESULT IDEA, which was merely the common sense notion that every hospital should follow every patient it treats, long enough to determine whether or not the treatment has been successful, and then to inquire “if not, why not?” with a view to preventing similar failures in the future.’ (AE Codman in The Shoulder, 1934)

Neither Boston nor the broader healthcare community was ready for Codman’s idea in 1934, but change was not far off. The federal government markedly expanded after World War II, leading to unprecedented growth in the US healthcare system and in federal funding for biomedical research and education. The research budget of the National Institutes of Health (NIH) grew from approximately $26 million in 1945 to almost $7 billion in 1990 (both in 1988 inflation-adjusted dollars).2,3 Subspecialization, academic medicine, and medical innovation flourished in this expansionist environment. Under this system, hospital bills were generated based on tests and services rendered. Charts were reviewed after patients were discharged, and payments were made retrospectively. Likewise, physicians were reimbursed by insurance companies and individuals on a fee-for-service basis; doctors were free to order diagnostic tests, medications, etc. as they deemed appropriate, without present-day worries of cost containment. Not surprisingly, this growth was followed by an era of increasing concern about cost. Beginning in the 1970s, the diffusion of technology into medical practice was suddenly a target. Big ticket items like CT scanners, renal dialysis machines (and a federal mandate for universal coverage

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


17

18

B. K. Killelea, E. L. Lazar, and M. G. Vitale

for end-stage renal disease patients) were commonly cited examples of technologies that far exceeded initial estimates of cost.4 The government responded with the creation of several organizations dedicated to the study and regulation of this boom in technological advancement: in 1976 the Medical Device Amendments to the Food, Drug, and Cosmetic Act were introduced to assess the impact of applied biomedical advances and emerging medical interventions; the NIH created the consensus development conferences in 1977, to organize evidence-based assessments of medical practice and controversial issues for both providers and the public;5 and the National Information Center on Health Services Research and Health Care Technology (NICHSR) was created in 1978 to coordinate the federal government’s assessment activities.6 Not long after, in the 1980s payers began to take a serious interest in controlling costs. For the first time, payments became fixed. The prospective payment system (PPS), based on diagnosis-related groups (DRGs), was introduced for Medicare recipients. Under this system, hospitals received a fixed amount of reimbursement based not on actual services rendered, but rather on predetermined payment schedules based on diagnoses. The PPS system encouraged fewer diagnostic tests, shorter lengths of hospital stay, greater emphasis on outpatient care, and decreased technology expenditures. During this same period, third-party payers like health maintenance organizations (HMOs) and preferred provider organizations (PPOs), saw an explosion in membership. In the current healthcare environment, HMO coverage has been extended to both Medicaid and Medicare recipients. By the end of 2003, approximately 4.6 million people, 11% of Medicare beneficiaries, were covered by managed health care plans. By employing utilization controls, HMOs have kept costs down. Hospitals and individual practitioners have started competing with other providers for managed care contracts in order to sustain adequate patient loads. But do they offer the same quality of care; are the plans accessible to all who need them, and what is the overall cost to the federal government? To answer these questions there has been an increase in demand for better information about both the quality and cost of current health care. This demand transcends national borders and health system design. Utilization review, done either retrospectively or prospectively, has also helped shape the development of outcomes research by ensuring that the delivery of health care and related services is clinically appropriate and medically necessary.7 Beginning in the 1980s, utilization review has been increasingly performed on a prospective basis. Based on physicians’ requests for treatment, designated organizations are charged with determining whether or not treatment is medically necessary. After explicit disease-

specific criteria are reviewed for each case, treatment is either approved or denied. Utilization review is used to control costs for everything from diagnostic tests and imaging, length of hospital stay, to requests for ambulatory surgery. Data from this activity allows nationwide trends in health care utilization to be identified. For example, recent evidence has suggested a geographic disparity in the utilization of surgical procedures across the country.8 Significant differences in rates of prostatectomy, lower back surgery, and carotid endarterectomy have all been well documented for different regions in the United States.9,10 These regional differences (often up to 20-fold) have not been explained by parallel differences in disease prevalence. Rather, they appear to be based largely on major variations in practice styles among physicians11 and surgeon availability and experience. In light of these differences, the importance of a defined standard of care becomes evident. But this standard of care must be applicable under real life situations, not idealized settings. Furthermore, this definition ought to be based not on individual experience, but on carefully designed research generalizable to large groups of patients. In 1962 the Food and Drug Administration (FDA) expanded its mission to include the assessment not only of drug safety but efficacy. Defined as “. . . the effect of a health care intervention on the outcome of care under “ideal” or experimental conditions,” it was not long until measurements of efficacy were expected in outcomes research as well. In the past, carefully controlled outcomes studies were carried out by highly skilled specialists under ideal conditions to determine an intervention’s efficacy. But in everyday practice, few patients and physicians are able to abide by such controlled environments. Furthermore, the practical application of these therapeutic interventions under real life conditions may also reveal unanticipated risks and benefits. Therefore, researchers and practitioners have become increasingly interested in defining outcomes under general or routine conditions to define its effectiveness. These developments highlight the introduction of what was commonly referred to by Relman as the era of assessment and accountability.13 In 1988 he proposed that the time had come for the medical community to “. . . provide a basis for decisions on the future funding and organization of health care, . . .” and “. . . the relative costs, safety, and effectiveness of all the things physicians do or employ in the diagnosis, treatment, and prevention of disease.”12 It appears that we are now well within this era, and that the science of clinical outcomes research, its design, and related methodologies have emerged as the tools of navigation. The spectrum of what constitutes clinical research has broadened; we are asking questions that relate not only to short-term biological endpoints like

Principles of outcomes analysis

mortality, but we are also interested in measuring physiological and socio-economic parameters that affect morbidity, functional status, and access to care. Advances in psychometrics have established a range of reliable and valid instruments for measuring quality of life.13–15 Hence, we now have the ability to quantify these outcomes and make comparisons across different patient populations. As emphasis has shifted toward less dramatic and often non-lifesaving interventions, we are faced with new ethical considerations; can we afford the sometimes marginal improvements in health status? The inclusion of cost data and cost-effectiveness analyses are of the utmost importance in answering these questions. At the same time that technological and pharmaceutical advances have broadened our capacity to treat, so must our ability to measure their effectiveness and costs.

Foundations of clinical research and clinical evidence Thus, the merger of a number of cultural, technological and economic forces led to the development of a more rigorous “systematic assessment of clinical intervention” now referred to as “Outcomes Research.” New methodological techniques of study design and evaluation have evolved to address questions about the demographics of disease and the applicability of interventions to larger populations. In order to derive clinical evidence to answer such questions, new ideas and treatments need to be tested, and patients observed. To avoid ambiguity in the interpretation of results, study design must control for confounding variables that may affect patient groups. In an effort to accommodate these concerns, various types of studies have evolved, including randomized controlled clinical trials, registries, cohort studies, observational studies, and meta-analyses, to name a few.

Surgical innovation Despite methodological advances of clinical investigation made in the second half of the twentieth century, evaluation of surgical procedures and devices has lagged. In 2000, the FDA approved 30 new molecular therapies and 90 pharmaceuticals. Prior to this approval, stringent FDA regulations require that new drugs undergo three phases of clinical trials to monitor for adverse and potential side effects. No such regulations exist for surgical procedures. Since there is no formal governmental regulatory system for the development and evaluation of clinical procedures, clinical evaluation has traditionally been far less formal.9

Surgical innovation poses unique challenges to traditional models of clinical evaluation. Acceptance of a new surgical technique begins with performance of radical procedures, usually upon animals, followed by technical refinement along a learning curve over time. These procedures are often pioneered by a handful of specialists, usually at large academic centers and then undergo a decentralized period of incremental modification in everyday clinical practice until they become established.9 Following early clinical experience in a series of patients, many procedures do not undergo randomized controlled clinical trials before their diffusion into more widespread use. However, it is important to keep in mind that there exist several ethical and practical considerations which preclude RCTs from being used in the development of new surgical procedures. For example, double blinding is practically impossible in the operating room, and controls may include standard accepted surgery or alternative treatments involving drugs or devices, which may themselves be in the process of improvement and development. Furthermore, technical expertise may still be evolving over the course of the study, rendering comparisons between procedures performed several months apart fraught with bias. In addition, patient and family bias to obtain particular procedure contribute to difficulty in organizing surgical RCTs. Faced with these issues, observational studies may be particularly useful in this arena, as they permit monitoring of clinical practice and changes in health outcomes. In addition, they are well suited to document incremental changes that typically take place as a procedure finds its way into the surgical armamentarium.

The challenge in pediatric assessment We have made dramatic strides in our ability to provide treatment for children with a wide range of health problems. However, our understanding of the true effects of our interventions has not evolved at an equal pace. New pressures of accountability brought on by a rapidly evolving system of healthcare financing have underscored the need for standardized, valid measures of outcome, which are appropriately designed to reflect changes in health status for specific areas of clinical intervention. Responding to this need, there has been a burgeoning interest in the area of outcomes assessment and measuring quality of life following surgical intervention in adults. However, much less attention has been focused on pediatric outcomes assessment. The evaluation of broadly defined outcomes, including general health status and quality of life, has several inherent difficulties in a pediatric population. First, any assessment

19

20

B. K. Killelea, E. L. Lazar, and M. G. Vitale

of functional status must be performed in a developmental context. Key aspects of quality of life such as physical, emotional, and social function develop rapidly as the child ages. Therefore, there is a need for age-adjusted normative values, which allow for the comparison of children with various problems to healthy children of the same age. Secondly, the validity of information attained directly from children is suspect; such data can only be collected on relatively older children. Therefore, parents or caretakers are often proxies for direct patient-based responses. Lastly, unlike adult populations, there is a relatively low prevalence of serious disease in children; often, long periods of follow up are needed to identify the natural history of a disease and the effect of treatment. Given the diversity and relative rarity of significant pediatric problems, the collection of adequate numbers of patients can be a challenge. Multi-center studies and research using large administrative datasets are effective strategies to pursue clinical questions in these patient groups. By pooling data from different sources, more powerful studies can be designed, but again, critical interpretation of these results necessitates acknowledgement of increased opportunity for bias as compared to RCTs. Despite some difficulties, the development of valid methods of assessing outcomes in this unique pediatric population is necessary in the evolution of surgical practice and innovation in this area. Fortunately, measures to assess functional status and quality of life in children have recently become widely available. The Child Health Questionnaire (CHQ)16 is perhaps the best validated measure for the assessment of general health status in children. Akin to the Short Form-36 (SF-36) which has been widely used in the adult literature17 the CHQ consists of a short questionnaire which can be scored, and generates multiple domains which span the spectrum of physical, psychosocial, and social health in injured children. Age-adjusted normative values are available and play an important role for the comparison of health status in children after trauma, for which pre-morbid scores are not available. The Pediatric Orthopedic Society of North America has developed another health status questionnaire, which also exhibits good validity and reliability across a range of pediatric musculoskeletal conditions.14

The governmental mandate The challenges inherent in pediatric evaluation of this type and the shortage of work in this area have not gone unnoticed. Recently, there has been a strong societal push, and subsequent governmental mandate, to stimulate pediatric

clinical research. Congress passed legislation to foster clinical research in children (the Pediatric Research Initiative Act), and Federal agencies have responded in turn. In particular, the National Institutes of Health (NIH) has put into place several new initiatives to support pediatric clinical trials. For example, the National Institute of Child Health and Human Development has created a network initiative to support more than 50 industry trials of non-psychiatric drugs, the NIMH has set up a clinical research network to support pediatric psychopharmacology, and the NHBLI is establishing a clinical trials network for testing emerging treatments for pediatric heart disease. Moreover, the FDA has created financial incentives for pharmaceutical and device companies to test their products in children (1997 FDA Modernization Act and April 1999 FDA rule requiring companies that seek approval for new drugs to run trials that include children).

Issues of design and methods The challenges facing investigators who wish to assess a new device or novel therapy in children influence the design phase of a study. Does one plan an RCT or review a database for clinical outcomes before and after a certain intervention? As we shall see, in pediatrics especially, there are many issues, both ethical and biological which must be taken into consideration. For example, one must take into account available data and resources, issues of time and cost, and ethical considerations related to forming a control or no-treatment group. We discuss several of the more common types of study design, and consider the strengths and weaknesses of each in the context of pediatric surgery.

Randomized controlled clinical trials It has been routinely asserted that the RCT is the gold standard by which evidence for any change in clinical practice must be gathered. The head-to-head competition of two therapies while controlling for important confounding factors has an irresistible intellectual appeal, well-grounded in the scientific method. In fact, graphically illustrating the hierarchy of clinical evidence, the double blind, randomized clinical trial sits atop a pyramid, towering over all other forms of evidence. The RCT, therefore, is a logical first choice as a study design when considering a new therapy. In it, the results of a treatment are evaluated by comparing a group of patients who received no treatment, i.e., a control group, to a group of patients equally matched for age, sex, ethnicity, etc. who received the therapeutic intervention

Principles of outcomes analysis

under investigation. The control group gives the expected rate of the study outcome under the null hypothesis of no effect, and the intervention group gives the observed rate of this outcome under the investigational treatment. Randomization is crucial in the design of an RCT. Groups that are well matched have less chance of introducing biases that may affect outcome; sample sizes and costs can thus be reduced. The result is a relatively homogeneous sample population. However, this very strategy makes the results only narrowly applicable to patients at large. Extrapolation to individuals or groups outside the inclusion/exclusion criteria is not, strictly speaking, valid. For example, trial results of an oral anticoagulant in males aged 20–40 may not be applicable to post-menopausal females. The designers of such a trial might have strategically excluded this group because of the peri-menopausal hormonal changes that can affect coagulation. This latter group may behave quite differently, obscuring clear results in such a trial. Moreover, physiologic flux and uncertainty in this group could make participation unsafe. Nonetheless, excluding “all-comers” limits the generalizability of the results. In essence, many RCTs demonstrate efficacy, but not effectiveness. Some trialists have addressed this concern about restrictive entry criteria by advocating large, simple RCTs with few inclusion or exclusion criteria. Typically, these studies require tens of thousands of participants to overcome the issue of variability and garner sufficient statistical power. Surgical trials, however, cannot easily be done on this scale and in children such accrual is impossible, even in multicenter trials. The key feature of RCTs, the randomization of patient groups, is designed to distribute known and unknown factors that might affect outcome. Randomization can also be difficult in surgical trials because of the perceived magnitude and finality of the decisions. Suppose, for example, we were trialing a life-sparing new cardiac procedure for a congenital cardiac anomaly. In this example, the control group might receive current medical management. Few people would be willing to be randomized to the medical arm of this trial if death was very likely, given the opportunity for survival with surgery. The case is considerably different from enrolling patients in a drug trial where differences in outcome are not so dramatic. How then, could such a trial ever be executed? Zelen proposed a strategy to account for potential participant aversion to one of the arms of a randomized trial.18 In his scheme, participants are randomized to either a consented or an unconsented arm of the study. Those randomized to the consented arm are permitted to choose which of the two therapies they wish to receive. The participants

randomized to the unconsented arm of the study receive the standard therapy and serve as controls. Despite barriers to acceptability, there is a rational basis for this allocation scheme. If, at the start of the trial, people in both groups have an equal chance of developing the outcome of interest, conclusions about the intervention can be drawn. Final comparisons in outcome are made not between treatment groups, but between consented and unconsented arms. Ultimately, the statistical power of the study depends upon the number of participants that choose the experimental therapy. Clearly, this strategy clearly has drawbacks; the two groups may not be matched with regard to demographic characteristics, and the consented groups may very well be biased for one reason or another. There is also an ethical issue in that randomization occurs without consent in the control group since they never know a priori that they are part of a trial. In the end, while it is not an altogether unsound practice, the Zelen strategy is not widely accepted; a trial designed on this basis would be carefully scrutinized by those who seek to use the intervention. Randomization also requires that the clinician have equipoise concerning the two therapies being compared. A surgeon who believes that surgery should always be attempted prior to medical therapy could hardly convince his patients to be randomized to the latter group. The proclivity to surgery may rest only on anecdote and be greater for the treatment of a skewed population (perhaps patients with severe disease), but without the clinician’s acknowledgment of the lack of confirmatory data, a trial simply will not work. A second feature of the RCT is blinding. In the example of a drug trial, patients who are blinded do not know whether they are in the treatment group or the control group. In a double-blind study, observers who administer the drug or collect patient data have no knowledge of what group patients are in either. Blinding poses a challenge in surgical trials. All would agree that a surgeon needs to see what is being done in the operating room and when comparing two surgical therapies, it is unlikely that we would ever be able to successfully blind the surgeon. Consider the previous example of laparoscopic cholecystectomy. We could not even theoretically propose a method to blind the surgeon. One could blind the patient by the use of large dressings but we would anticipate an early break to the blind in such cases. In surgery, we must usually settle for an appraisal of the functional outcomes by an individual blind to the allocation. Related to blinding is the placebo effect. Patients who are blinded don’t know whether they are receiving the pharmacologically active drug under investigation, or a placebo; these groups are used as controls. While the expectation

21

22

B. K. Killelea, E. L. Lazar, and M. G. Vitale

is that control groups will report no change, sometimes this is not the case. The placebo effect, therefore, is the measurable, observable, or felt improvement in health not attributable to treatment. In place of a placebo, in surgery we could consider a sham operation. The risks here are great, however. In drug trials, the risk of placebo is that active treatment is not being pursued. In sham surgery, the added risk of the surgery itself must be considered. This is ethically thin ice – performing sham surgery for the sole purpose of blinding and controlling for the placebo effect. Hence, these trials are rare in surgery. Sham surgery is occasionally used, however, as was demonstrated in the cranial implantation of fetal cells to treat Parkinson’s disease.19 Because of the concern that one could not accurately compare a group getting fetal cell implants to any group other than a sham craniotomy group, sham operations were performed to account for the effect of surgery itself. The results in this trial debunked the therapy and prevented widespread adaptation of an ineffective therapy, but the data could not have been convincingly obtained without the contribution of participants willing to undergo sham surgery. In general, however, sham surgery has limited practical applicability and is not often used. Another shortcoming of the RCT, particularly where surgery is concerned, relates to its time course. Wellexecuted RCTs take considerable time to plan, recruit, accrue, and analyze. The time course for most trials is about 5 years. Unfortunately, some surgical therapies, especially those that are market driven, gain wide acceptance faster than we can trial them.9 Laparoscopic cholecystectomy was never subjected to the scrutiny of a trial and within 5 years of its introduction, the practice was already widespread. Most would agree that initiating a trial now would have little impact on current practice patterns. Furthermore, accrual of patients in the laparotomy arm would be impossible, given laparoscopy’s now apparent benefits. In sum, in certain instances, RCTs risk being viewed as irrelevant. The proposal of an RCT in surgery must account for two related features not seen in drug trials. The first concerns surgical skill and technique. In a drug trial, a medication is given (or not) and efficacy is evaluated. Drug administration is a fairly straightforward event, with little room for variance in its execution. In surgery, administration of the “intervention” is not quite as straightforward – there can be profound differences among surgeons with regard to skill and technique. This increased variability adds uncertainty to the outcome of a surgical trial. A surgical approach may work when done a certain way, with attention to certain details and nuances, but less well or not at all without the scrutiny of a trial. Variability is introduced into a surgi-

cal trial because of these differences. Next, new techniques require a run-in period to approach the plateau of the learning curve prior to the start of a trial. If a participant center enrols patients without such a run-in period, the outcome will be biased toward the null as there will be early technical failures contributing to the analysis. These related features, variable skill and technique and the learning curve, require that the surgical trialist insure, as nearly as is possible, the techniques and skills of the participating surgeons. While neither impossible nor prohibitive, such requirements add considerable time and expense to preparing a surgical trial. In pediatric surgery especially, most areas of clinical inquiry involve fewer patients, many of whom would be used during standardization and run-in periods, leaving fewer cases for randomization in the trial itself. In a perfect world, all existing therapies would be subjected to RCT methodology to prove their presumed merit and new therapies would only be adopted after rigorous evaluation. Indeed, there are many clinical research authorities who feel that virtually no new therapy should be adopted prior to establishing its superiority in an RCT.20 Such a clean sweep of all of clinical medicine, however, is unlikely and the cost would be prohibitive. Even if the cost and chaos were tolerable, is the RCT the most desirable assessment tool in general? In surgery? In children? In point of fact, we have seen some of the limitations of RCTs that merit consideration before embarking on a trial destined to fail or be misinterpreted in a broader application. In sum, trialing a new surgical procedure prematurely places everyone on the upslope of the learning curve, and trialing too late risks making the results irrelevant. Accepting randomization (on the part of the clinician) requires equipoise which is often lacking despite the lack of sound evidence to support a particular approach. Blinding and controlling for the placebo effect in surgical trials add considerable expense and are ethically dubious under some circumstances. How then do we proceed and meet the goals of evidence-based medicine in an era of assessment and accountability given these challenges? How do we plan to study a new device or therapy? First, we must remember that, today, many surgical trials are not focused on crude outcomes, such as survival. Rather, we are becoming increasingly interested in metrics related either to the economic sequelae of illness (lost work days or length of stay) or non-biologic attributes of illness (health-related quality of life). The traditional trial design is ideal to study an intervention in which the medical outcome is clear and the goal is to establish the superior approach. Perhaps, however, the rigid approach of an RCT is excessive in the setting of these alternate economic or quality of life outcomes.

Principles of outcomes analysis

Observational studies It is important to recognize the strengths inherent in forms of evidence gathered elsewhere on the clinical outcomes pyramid. Observational data contained in large, administrative databases, for example, can provide a wealth of information about practice patterns evident in a variety of geographic or payer/provider specific situations. While such databases are limited by data points defined by administrative needs, usable information is nonetheless contained therein. Such databases can be used to gauge the performance of a given region, institution, or provider with regard to outcome. These “report cards” are increasingly common in our consumer driven models of health care delivery. So, while we might hesitate to conduct an RCT assigning a patient to one hospital or another to see which had better outcomes for a given surgical diagnosis, we can make such a comparison using an administrative database and surrogate outcome, such as length of stay (LOS) or the occurrence of complications, to gauge performance. Adjustments can easily be made for severity of illness, patient socioeconomic and educational status, volume/outcome relationships, or any other important confounders. These adjustments may not have the crisp appeal of an RCT, but they are no less rigorous or reliable. The key to making successful inferences lies in understanding the meaning, limitations, and reliability of the data points. By way of simple example, many admission databases code infants and children as unemployed. Any inference about socioeconomic status that is based upon occupation would be wrong unless the investigator is thoroughly familiar with such nuances. In order to avoid problems associated with surrogate endpoints, many centers are developing their own databases. These “homegrown” databases include, a priori, demographic data and endpoints of clinical relevance thereby increasing the sophistication of potential analyses. Observational studies can be done prospectively, retrospectively, or as a cross-sectional study. In a study of risk factors, for example, patient characteristics can be analyzed retrospectively to see if they are related to the disease. In prospective cohort studies the degree of exposure, risk of occurrence, and effect of treatment can all be followed months or years into the future.

controlled cohort study. A cohort is a group of individuals who share common characteristics, and are usually about the same age.21 Ideally, this group is assembled at a synchronous point in time and observed until a particular endpoint is reached, usually either the conclusion of the study period, or death. As time passes and the group ages, changes in health status are observed reflected in morbidity and mortality of the cohort.23 Cohort studies are useful in determining the risk of developing disease. To do this, we must know the incidence of a disease – simply the number of new cases that develop in a cohort over the study period. Incidence is expressed as the number of new cases per person at risk for developing the disease. Generally, epidemiologists refer to two kinds of risk: attributable risk and relative risk. Both compare the incidence of disease in two or more cohorts that have had different exposures to some possible risk factor. Attributable risk is calculated by subtracting the incidence of disease in the unexposed cohort from the incidence of disease in the exposed cohort. This measure gives an absolute increase in the incidence of disease among those exposed to the risk factor. Relative risk designates how many times more likely it is that an exposed person will become diseased than a non-exposed person. This number, expressed as a ratio, is calculated by dividing the number of people who develop the disease by those who do not for a given cohort. In addition to defining prevalence and studying risk, cohort studies are useful for observing the natural history of a disease, and patients’ response to treatment. Since the sample sizes are usually quite large, findings are credible and cause and effect relationships can be established. They are not, however appropriate for disease with long latency periods or diseases that are very uncommon. In addition, they are more vulnerable to bias than RCTs. Without randomization, in order to assemble comparable groups of patients, attempts are made to match patients in the control group with characteristics of patients in the case group except for the exposure or disease. If differences between the groups do exist that are themselves predictive of outcome, then there is a selection bias. Any observed difference in incidence between the two groups may just be reflective of the differences between the two groups.17 In addition, cohort studies are also subject to patient drop-out and cross-over between groups.

Cohort studies

Case control studies

After the RCT, the next best study design for evaluating disease and the effect of treatment is the prospective, matched,

Another type of observational study, the case-control study, uses observational data to investigate potential

23

24

B. K. Killelea, E. L. Lazar, and M. G. Vitale

relationships between exposure to one or more risk factors and the development of disease.22,23 By selecting a group of patients with a disease (cases) and a comparable group of patients without disease (controls), researchers can then look back in time to try and establish a relationship between exposure and disease in the two groups. Although they are less expensive and quicker to conduct than prospective cohort studies or RCTs, there are limitations to case control studies, and they are susceptible to bias. When choosing a control group, the investigator must be careful not to overmatch. Groups that are overmatched may be so similar that a disease’s association with another characteristic of interest becomes masked.24 For example, suppose we are studying a disease associated with over consumption of meat. Choosing controls based on level of education or income would be overmatching since these characteristics are not independently associated with development of the disease. People who are more educated and/or have higher incomes generally tend to eat more meat and may erroneously lead us to infer a relationship between socioeconomic status and disease.

Case reports and case series As we have seen, there are a variety of methodologies and strategies that produce valid inferences and permit evidence-based decision making, and the circumstances vary under which each of these various methods is most effective. For the sake of completeness, case reports and case series remain publishable and have an important role in stimulating inquiry, but we cannot infer anything about incidence, or risk from them. Furthermore, outcomes cannot be assessed nor adopted for widespread use based on these more primitive forms of evidence. They are useful ways of describing an unusual presentation or response to treatment, or a novel type of therapy or procedure.

Meta-analysis Newer to the outcomes tool kit is the meta-analysis. Metaanalysis is a useful way to combine results from multiple studies by calculating a weighted average of study-specific results.25,26 In pediatric surgery, where cases are often few and far between, meta-analyses hold special promise, but there are limitations. Multiple trials in different clinical settings, perhaps each with different inclusion/exclusion criteria, will generate different effect sizes. The more variability within a trial, the greater the tendency will be toward

the null. It may also be difficult to reconcile trial data from that yielded by observational methods. Furthermore, the reader of clinical science may not have access to trials that were not reported. It is important, therefore, to keep in mind the potential for publication bias: negative studies (those that affirm the null hypothesis) are either not reported or not published. Very recently, medical journal editors have proposed registering all trials so as to make available these studies for inclusion into the overall assessment of any given therapy. Meta-analysis can synthesize diverse results arising from many types of investigational strategies and yield a statistically powerful estimate of risk. Using clinical trials databases can broaden the pool of studies to include those published in languages other than English. Of greatest value is the ability to increase statistical power for secondary outcomes by combining many well-done studies. What is newer is the inclusion of patient-centered outcomes. In addition to a biologic endpoint, marker, or mortality, there is new interest in assessing the quality of life of our patients that reflects the impact of their disease and our treatments. Quality of life certainly has different meanings for different individuals but we would all agree that the central elements must account for the ability to complete activities of daily living in a variety of domains – functional, social, and emotional. In the next section, we will focus on some of the endpoints for analysis in outcomes research, namely health related quality of life (HRQOL) and cost.

End-points for analysis Focusing intense effort solely on cure, disease-free survival, and lowering or eliminating a biologic marker ignores other important attributes of patients’ “outcomes.” HRQOL includes global assessments of well-being and functional status; as such they are becoming the new currency in outcomes research. These additional domains add depth to our understanding of the effects of our interventions. Interpretation of the outcome of a therapeutic intervention requires more than assessment of the biologic endpoint. So, while we may fail to cure some of our patients whose disease is simply not amenable to treatment, some patients may have a better quality of life as a result of treatment – a joint replacement patient may walk with less pain or an oncology patient may beat the fatigue of anemia, for example. On the other hand, we may extend life by many months with a given intervention, but at some price – being ventilator dependent may not be an acceptable or desirable outcome for many patients. HRQOL data enhances

Principles of outcomes analysis

our understanding of the therapeutic intervention in ways not previously appreciated.

Quality of life Scientific rigor demands objective, quantitative data. The very word quality in the phrase “quality of life” (QOL) evokes a qualitative process and in the minds of some, subordinates such research. As have been briefly introduced, contemporary instruments for the measurement of quality of life are vigorously validated and highly reliable. This section examines the background to quantitative measurement of the health-related quality of life. Illness and the treatments that we offer have an impact on the life of the patient beyond the biophysical parameters that constitute the usual health assessments. Classically, QOL instruments encompass multiple aspects of life – physical, psychological, and emotional functioning. Each of these aspects, known as domains, can be sub-divided into various dimensions. The physical, for example, can be viewed as a composite of objective physical ability, restriction of activity, physical symptoms, beliefs and feelings about physical health, and specific disorders. Each of these dimensions can be quantified on an ordinal scale, generating an overall domain score. Separate domain scores can be combined for a global or overall score. The greatest utility of such an instrument lies in its sensitivity over time for a given individual. Usually, higher scores correlate with a better QOL and even small changes in QOL are reflected in corresponding changes in score. An excellent review of the general principles of measuring quality of life is offered by Testa.27 QOL instruments come in two varieties – generic and disease specific. The classic Short Form (SF-36) is one of several widely used generic instruments which assesses HRQOL. The Short Form was derived from the longer Medical Outcomes Study (MOS) questionnaire which contained 100 items. The 36-item version has been well validated and assesses seven domains. Numerous disease specific instruments have been validated for patients living with renal disease, epilepsy, asthma, heart failure, rheumatologic disease, and most recently, obesity, among many others. One of the greatest challenges in assessing HRQOL in children relates to their ever-changing developmental background, particularly in the youngest children. That is, normative values for children are really a moving target. Instruments, such as the SF-36, are simply not valid in children or adolescents. Perhaps the most widely used child health instrument in the United States is the Child Health Questionnaire (CHQ), which comes in several forms. The self-administered version is an 87-item form that has 11

dimensions and is useful in the older child and adolescent (10–18 years of age). There are three parental versions available, denoted by the suffix PF (parental form) which contain 28, 50, and 98 items. The CHQ-PF28, CHQ-PF50 and the CHQ-PF87 allow the parent to answer as a surrogate for children as young as five years of age (as well as the older children to age 18). Of course, whenever a surrogate is involved, there is always the potential for loss of signal or bias. It is important to remember, however, that the greatest value is these instruments lies in change over time, not necessarily the absolute value of an isolated score. Another US instrument is the Pediatric Quality of Life Inventory (PedsQL) which has three self-administered versions for the age groups 5–7, 8–12, and 13–18. This 23-item instrument measures function in four domains and takes about 10 minutes to complete. A unique feature of the PedsQL, is that in addition to a module of disease-specific items, it contains a core group of items that constitutes the generic health-related quality of life. This feature is particularly useful in that it allows for comparison of quality of life between groups of patients with different diseases. In addition, there are many other disease specific instruments, including those for children with orthopedic problems, asthma, obesity, and even those who have suffered traumatic injury. The science of constructing, validating, and using the various available instruments is well covered in other texts.28 It is important to remember, however, that it is a scientific process and not for the novice. One cannot sit with pen and paper and jot down a good questionnaire for use in a study and expect to have a valid, reliable instrument. Items must be tested, refined, and re-tested. This iterative process can take months to years before normative values can be obtained. Specific items must map strongly to one domain, rather than weakly to several. The response scale must be chosen and then tested. Several questions of design and age-appropriate measures must be considered. Should there be a visual analog scale, an ordinal scale, or pictures? Which items provide for internal validation and is there external validation? There is no question that, in general, one is better off using an established instrument and accepting its flaws rather than creating one de novo. Instruments that are selected for use must also be culturally appropriate. Instruments originating in the US should not be administered elsewhere until an item-by-item evaluation has been made to change whatever culturally based items may exist. Even in the US, an instrument may need to be modified for a particular cultural enclave or region under study. Simple translations from one language to another are not sufficient if the culturally based concepts remain foreign to the subject. Changes to any of the items may

25

26

B. K. Killelea, E. L. Lazar, and M. G. Vitale

change the instrument’s sensitivity, specificity, reliability, and/or validity. Hence, the vetting process often has to be done for each country or region where the instrument is intended for use. Since some pediatric illnesses and conditions are not very common, this may mean, particularly for the disease-specific instruments, exhausting the pool of available patients for study in simply validating the instrument. These admonishments aside, the endeavor of trying to measure child health outcomes is of obvious importance and worth the time and effort to do properly.29 Biologic outcomes will always remain important, but as our healthcare delivery evolves, uniformly good outcomes will be expected and even demanded. It will be in the improvement of quality of life that will differentiate one treatment option from another rather than survival.

Utility Utility is a central component to formal decision analysis and policy making. Utility is measured on a scale of zero to one and is best thought of as preference for a given health state. Clearly, it is value-based and is entirely dependent on the person who determines its value. Utility has its basis in game theory and is understood as a standard gamble. For example, suppose we were to ask a subject whether they preferred living with their disease and accepting its eventual outcome (paralysis or death), vs. trying a treatment that could improve their health, but possibly risk hastening the outcome. Clearly, the person making the choice has to understand the chance of the treatment helping vs. making things worse. Furthermore, much of this decision depends upon the time remaining until the inevitable, untreated outcome. For example, for a near event, we accept a lower probability of success to go forward. The patient with a ruptured abdominal aortic aneurysm will accept surgery despite low likelihood of success. To undergo scoliosis surgery with all of its attendant risks, however, is more difficult to rationalize since the event that is being aborted is many years in the future. The essential problem is that few people understand game theory (witness the many who play away their salary in casinos), and children require a surrogate to make the judgment calls. In essence, however, if someone is experiencing a good quality of life, that person is unlikely to risk very much in order to improve their lot. Perhaps a more accessible concept is known as time trade-off.30 Rather than giving the subject a choice of a certainty vs. a probability – time trade-off asks the subject to choose between two certainties. The choice comes down to how much time one is willing to give up to live in a desired

health state rather than an alternative health state. Hence, a subject might tell us that to live without the effects of chemotherapy might be worth 2 years of life. Utility scores may be obtained from specially designed questionnaires, not unlike those used for scoring quality of life. More often, however, a face-to-face interview is required to make sure that the subject understands exactly what is being asked by way of assigning a weighted value to their current health state. This can be an intrusive process and again, for children, it can be most difficult, particularly since concepts of end of life are involved. In the end, however, the goal is to provide a modifier to measured survival as a result of a given treatment. Perhaps liver transplant extends the life of a patient with an inborn error of metabolism. This patient, however, may be plagued with hospitalizations, taking multiple daily medications, and opportunistic infections as a result of the therapy (transplant). Further, travel is restricted because of frequent drug level testing and doctors’ appointments. This person might live an additional 10 years compared to a conventionally treated patient, an outcome which we would traditionally define as a great success. The utility of the treatment might be low, however, as a result of all of the above conditions and the patient might rate it 0.4. The quality adjusted life years (QALYs) are thus 4 (10 × 0.4). As noted, the judge can greatly affect the utility score. This individual’s spouse or child or parent might rate the utility closer to 0.9 since none of the conditions bother them and the value of having their loved one around is very high. Thus, the QALYs are 9, much higher and indicative of a true success. The endeavor to link quality of life with a weighted value can help make individual decisions where therapeutic choice is involved. Furthermore, health policy can be based on patient experience and favor treatments that not only work but enhance living.

Cost Most of the clinical science of outcomes research has its basis in cost. It is the competition for third-party payment that drives clinicians to demonstrate a superior product (better outcomes) at lower cost. Health care costs are complex and in many cases, difficult to ascertain. First, there is the charge for services and therapies. The charge is sometimes not rationally related to the cost and the actual and true cost may not be easy to specify. Second, agreements with vendors and suppliers may alter costs dramatically. To illustrate, some drug companies make suture material and other supplies. They might lower the cost of these items if a hospital agrees to purchase their drug products on an exclusive basis at a premium. The

Principles of outcomes analysis

hospital might agree to such an arrangement because, overall, they are saving money. However, the charges for a given patient who received no sutures and used few supplies but who received 20 doses of premium priced thirdgeneration drug are high. The patient taking no drug and having multiple dressing changes with special supplies and sutures is having his costs shifted to the first patient. This makes analyzing the real cost of care in a given case very difficult and variable from hospital to hospital. Hence, resource costs are difficult to obtain yet they are more accurate then using payer information (altered by contractual adjustments). Traditionally, one thinks of cost–benefit analyses in which the cost of delivering care is balanced against the benefit gained. The costs are not always dollars spent in achieving the goal, however. If a patient experiences a complication and can no longer perform a vital function (like work) then the cost must include the pain and suffering as well as the lost work life. Likewise, the benefits are not always just the immediate health benefits, but may involve future productive life gained. These extended concepts are very important but very hard to place value on, and so traditional cost–benefit analyses are actually rather limited. In pediatric surgery, the benefits are even more difficult to score – what is the value of repairing a congenital heart anomaly with a low likelihood of survival to a productive work life? From a cold balance sheet, we would never endeavor to undertake such a repair – but what of advances that have come from this surgical boldness. What is the value to our society in terms of other patients in the future? How is the cost and value of future, unknown benefits calculated? The endeavor of inquiry and investigation has value that is simply not accounted for in these traditional models. An alternative, cost–utility analysis uses the utility measures just described. Now, alternative therapies can be compared by the common QALYs measure. So, comparing a therapy that enhances the QOL but decreases survival with one that decreases QOL but increases survival is possible. By dividing the total cost in dollars of both options by the QALYs, one can get a sense of the cost effectiveness of each therapy. However one calculates cost, it is a clear mandate that cost-effectiveness is perhaps the single most important aspect of any therapy. Gone are the days when the monies needed to achieve a health-related goal were limitless. With available healthcare resources, it now becomes the burden of every clinician to spend and invest wisely in therapies that work, have value, and that lead to enhanced qualityadjusted survival.

Conclusions The science of outcomes research has come a long way since the days of Codman. Throughout this evolution, we have seen the progression of different methodologies, and the refinement of more appropriate end-points of evaluation. These developments have not only accelerated the pace of clinical research, but have improved the quality and applicability of resulting data. Looking ahead to the future, the emergence of new technologies, including the internet, will continue to shape this process and make the results accessible to a larger audience. Now more than ever, it is evident that rigorous evaluation of clinical and surgical practice is an important tool in the ongoing quest to best meet the needs of our patients.

REFERENCES 1. Codman, E. A. The Shoulder. Boston, MA: Thomas Todd; 1934. 2. Bloom, F. E. & Randolph, M. A. Funding Health Sciences Research: A Strategy to Restore Balance. Washington, DC: National Academy Press; 1990. 3. Finneray, K. (ed.) The Federal Role in Research and Development: Report of a Workshop. Washington, DC: National Academy Press; 1986. 4. Rettig, R. A. Health policy in radiology: technology assessment – an update. Invest. Radiol. 1991; 26:165. 5. NIH Consensus Development Program. About the Consensus Program. Available at: http://www.consensus.nih.gov/ about/about.htm. Accessed August 31, 2004. 6. United States National Library of Medicine, National Institutes of Health. National Information Center on Health Services Research and Health Care Technology (NICHSR). Available at: http://www.nom.nih.gov/nichsr/nichsr.html. Accessed August 31, 2004. 7. Gray, B. H. & Field, M. J. Controlling and Changing Patient Care? The Role of Utilization Management. Washington, DC: National Academy Press; 1989. 8. Gelijns, A. C. Innovation in Clinical Practice: The Dynamics of Medical Technology Development. Washington, DC: National Academy Press; 1991. p. 82. 9. Roos, L. L. Nonexperimental data systems in surgery. Int. J. Technol. Assess. Health Care 1989; 5(3):341–356. 10. Wennberg, J. E., Roos, N. P., Sola, L, Schor, A., & Jaffe, R. Use of claims data systems to evaluate health care outcomes: mortality and reoperation following prostatectomy. J. Am. Med. Assoc., 1987; 257(7):933–936. 11. Eddy, D. M. Variations in physician practice: the role of uncertainty. Health Aff. (Millwood), 1984; 3(2):74–89. 12. Relman, A. S. Assessment and accountability: the third revolution in medical care. N. Engl. J. Med. 1988; 319(18):1220–1222.

27

28

B. K. Killelea, E. L. Lazar, and M. G. Vitale

13. Doltroy, L. H. et al. The POSNA pediatric musculoskeletal functional health questionnaire: report on reliability, validity, and sensitivity to change. Pediatric Outcomes Instrument Development Group. Pediatric Orthopaedic Society of North America. J. Pediatr. Orthop. 1998; 18(5):561–571. 14. Landgraf, J. M., Canadian–French, German and UK versions of the Child Health Questionnaire: methodology and preliminary item scaling results. Quality Life Res. 1998; 7(5):433–445. 15. Bukstein, D. A., McGrath, M. M., Buchner, D. A., Landgraf, J., & Gross, T. F. Evaluation of a short form for measuring healthrelated quality of life among pediatric asthma patients. J. Allergy Clin. Immunol. 2000; 105(2 Pt 1):245–251. 16. Landgraf, J. M., Abetz, L., & Ware, J. E. The CHQ User’s Manual. 1st edn. Boston, MA: The Health Institute, New England Medical Center, 1996. 17. Hays, R. D., Sherbourne, C. D., & Mazel, R. M. The RAND 36-item health survey 1.0. Health Econ. 1993; 2(3):217–227. 18. Zelen, M. Alternatives to classic randomized trials. Surg. Clin. North. Am. 1981; 61(6):1425–1432. 19. Trott, C. T., Fahn, S., Greene, P. et al. Cognition following bilateral implants of embryonic dopamine neurons in PD: a double blind study. Neurology 2003; 60(12):1938–1943. 20. Green, S. B. Using observational data from registries to compare treatments: the fallacy of omnimetrics. Stat. Med. 1984; 3(4):361–373.

21. Timmreck, T. C. An Introduction to Epidemiology. 3rd edn. Sudbury, MA: Jones and Bartlett Publishers, 2002. 22. Schlesselman, J. J. Case-control Studies: Design, Conduct, Analysis. New York, NY: Oxford University Press, 1982. 23. Fletcher, R. H., Fletcher, S. W., & Wagner, E. H. Clinical Epidemiology: The Essentials. Baltimore, MD: Williams and Wilkins, 1988. 24. Friedman, G. D. Primer of Epidemiology. 5th edn. New York, NY: McGraw Hill; 2004. 25. Dickersin, K. & Berlin, J. A. Meta-analysis: state of the science. Epidemiol. Rev. 1986; 14:154. 26. Sacks, H. S., Berrier, J., Rietman, D., Ancona-Berk, V. A., Chalmers, T. C. Meta-analyses of randomized controlled trials. N. Engl. J. Med. 1987; 316:450. 27. Testa, M. A. & Simonson, D. C. Assessment of quality-of-life outcomes. N. Engl. J. Med. 1996; 334(13):835–840. 28. Streiner, D. L. & Norman, G. R. Health Measurement Scales: A Practical Guide to their Development and Use. 2nd edn. Oxford: Oxford University Press, 1995. 29. Dougherty, D. & Simpson, L. A. Measuring the quality of children’s health care: a prerequisite to action. Pediatrics 2004; 113(1)185–198. 30. Torrance, G. W. Measurement of health state utilities for economic appraisal. A review. J. Health. Econ. 1986; 5:1– 30.

3 Outcomes analysis and systems of children’s surgical care Keith T. Oldham1 Mark D. Stringer2 and Pierre D. E. Mouriquand3 1 Children’s Hospital of Wisconsin, Milwaukee, WI, USA Children’s Liver and GI Unit, St James’s University Hospital, Leeds, UK 3 Department of Paediatric Urology/Surgery, Debrousse Hospital, Lyon, France 2

Institutions which provide the basis for contemporary children’s specialty units and pediatric clinical care first developed in the nineteenth and early twentieth centuries. These reflected efforts to care for urban populations of children brought together by the forces of industrialization and were often the work of altruistic individuals, typically but not exclusively, well-to-do women, who were motivated by compassion for children. Their primary mission was often to provide an acceptable level of charity care to poor children in their communities. This is, in fact, still the case for many major children’s hospitals worldwide. The concept that children are unique patients with particular developmental, physiological and psychological needs, for whom a demonstrably better outcome could be obtained with specialized medical care came more recently. With particular regard to children’s surgical care, events detailed in Chapter 1 by Dr. Hendren (North America) and Dr. O’Donnell (Europe/UK) gave rise to what we recognize as contemporary pediatric surgery by the mid twentieth century. Specialization and credentialing were progressively driven by the view that the quality of care for children was improved by having practitioners fully dedicated to children, by providing focused professional training for individuals interested in childhood disease, and by creating a children’s specific environment for the delivery of that care. The wider medical community has not been free of controversy over these points. Perceived economic threat and fear of professional or institutional loss remain powerful forces even today when we discuss issues like regionalization or consolidation of certain services. Although certifying bodies in much of the world differentiate pediatric from adult practice and training, the lines of distinction between certain patient groups, particular

providers and competing institutions remain a source of controversy. As we begin the twenty-first century, many active discussions continue around questions such as: how to best design pediatric trauma systems; where to perform congenital heart surgery; who is best able to do childhood transplants or provide pediatric anesthesia; what is general (adult) as opposed to pediatric urology or surgery; and indeed, within pediatric surgery or urology, should everyone repair a cloaca or perform a portoenterostomy for biliary atresia? These types of questions are currently answered differently around the world, driven by unique local history, available resources and needs. It is the view of the authors of this chapter and the editors of this text, that policy decisions about providers and venue of care are best driven by relevant, risk-adjusted and accurate outcomes data. What are the best outcomes that can be achieved in correcting a particular anomaly or treating a childhood surgical condition? What are the institutional characteristics and what is the background and training of the individuals who achieve these best outcomes? Where are healthcare system boundaries best drawn to improve quality of care and deliver value? How can optimal outcomes be achieved within the economic constraints of a particular healthcare system and with the least disruption for children and their families? These questions are the fundamental reason to publish this text, an effort to assemble the most critical contemporary analysis of outcomes from international experts in children’s surgical care. Each individual chapter provides relevant data with regard to long-term outcomes in a specific area. Even when they are considered collectively, the reader will recognize that some of the questions posed above will remain

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


29

30

Keith T. Oldham, Mark D. Stringer, Pierre D. E. Mouriquand

unanswered. In the following pages, the authors examine three areas of pediatric surgical practice in some detail: childhood trauma, hepatobiliary surgery (specifically biliary atresia), and pediatric urology. We have chosen these particular areas to illustrate the types of data and analyses that are becoming available to children’s surgeons and policy makers in various countries. We have made no attempt to be encyclopedic, rather to use these examples in an illustrative manner. Other examples can be found in other chapters in this book (e.g., Chapter 6: Cleft Lip and Palate). Both the potential power of these data, as well as some of the obstacles impeding reform of current practices, will become apparent. The areas selected for discussion are chosen because they involve complex systems of care with multiple specialists dealing with critically ill children, as well as highly complex technical procedures where individual surgical skills are relevant. Thus issues of both individual experience and institutional structure and volume enter into the discussion. Whilst there is considerable variation among particular types of surgical procedures and in the quality of comparative data, it appears increasingly clear that there is an inverse relationship between hospital and surgeon case volume and surgical mortality and morbidity.1–5 This general relationship extends beyond the world of surgery.6 The challenge is to match particular patients with appropriate caregivers, delineating relevant procedures and volume thresholds that are both meaningful and practical. In the sections that follow, we attempt to model types of data available which enable rational policy decisions yielding optimal care for pediatric patients.

Trauma Trauma is a major pediatric health issue worldwide and it is the leading cause of childhood death in the developed world after the age of 1 year. For children and adolescents between 1 and 20 years of age in the United States in 2001, unintentional injury accounted for 12 916 deaths and intentional injury accounted for approximately 3300 additional deaths.7 Approximately two-thirds of these deaths in the United States are related to motor vehicles. In 2000, approximately 228 000 children were hospitalized and 10.9 million were treated in an emergency department for injury related problems.8 The estimated direct cost in 2004 for injury related medical care in the United States was 117 billion USD and approximately one-third of this was expended upon children and adolescents. Despite the number of pediatric patients involved (both generally and within particular subsets) and the obvious scope of the problem,

consolidation of pediatric trauma care into specific children’s hospitals or programs is not the norm in the United States; indeed, the majority of injured children receive their care in adult oriented environments. There are a number of reasons for this. One important issue is simply that current pediatric systems and centers are not staffed or equipped to handle such a large number of trauma patients. Because specific pediatric systems have not been developed in many regions, more than half of pediatric trauma patients in the United States receive their care in adult centers or general hospitals, even in large urban areas. Access to pediatric specialty care in rural environments is a particular challenge in trauma as well as in many other areas of health care. In the early 1990s, several authors suggested that outcome following pediatric trauma, as measured by survival, was equivalent in adult trauma centers which cared for children.9–11 These reports were based generally on retrospective, single institution reviews using TRISS methodology which allows comparison of individual program or institutional survival rates with national norms extrapolated from a largely adult cohort (the Major Trauma Outcomes Study). These comparisons have been criticized for both this flawed methodology (incorrect reference sampling frame) and for the fact that their reported results do not compare favorably with best practice benchmarks, that is, survival in dedicated pediatric trauma centers. The point is also made that survival is a relatively crude measure of pediatric trauma care quality12–14 since fatalities are relatively uncommon. More recently, Potoka et al.15 reported a statistically significant survival advantage for the most severely injured children who had head, spleen and liver injuries, when treated at pediatric trauma centers in the state of Pennsylvania in 2000, in comparison to those managed in adult trauma centers in the State over the same time period. These data were derived from a mandatory and uniform statewide trauma reporting system which has produced a uniquely detailed database. Because overall mortality rates for pediatric trauma are generally low, 1.7–8.9%,3,6–9,11 functional outcome and quality of life should be considered to be relevant measures of the quality of childhood trauma care. Head injury is the principal cause of both death and disability in pediatric trauma patients. In a follow-up study in 2001 using the same Pennsylvania state database, Potoka et al. analyzed functional outcome, i.e., disability, in severely injured children (Injury Severity Score (ISS) > 15 and age 0–16 years).14 The analysis was complex, taking into account different types of certification among trauma centers, and including age-specific evaluation of feeding, locomotion, expression,

Outcomes analysis and systems of children’s surgical care

Table 3.1. Pennsylvania trauma outcomes study. Patient demographics (ISS > 15)14

Number of patients treated Mean age (y) Sex (% male) Mechanism (% blunt injury) Mean ISS Mean GCS

Pediatric trauma center

Adult trauma center (with added qualifications)

Adult trauma center Level I

Adult trauma center Level II

553 8.7 ± 4.2 71.4 94.8 21.5 ± 7.2 12.0 ± 4.5

782 10.8 ± 4.8a 68.9 95.0 23.5 ± 8.4 11.3 ± 4.8

206 15.1 ± 1.8a 71.4 75.2b 25.4 ± 10.4 11.2 ± 5.0

546 12.0 ± 4.0a 71.6 94.5 23.1 ± 8.8 11.8 ± 4.6

P < 0.05 vs. PTC by Student’s test. P < 0.005 vs. PTC by x2 . ISS, Injury Severity Score. GCS, Glasgow Coma Score.

a

b

Table 3.2. Pennsylvania trauma outcome study. Functional outcome (ISS > 15)14 Number of dependent patients at discharge (%)

Functional category

Pediatric trauma center

Adult trauma center (with added qualifications)

Adult trauma center Level I

Adult trauma center Level II

Feeding Locomotion Transfer Social Interaction Expression

80/550 (14.6) 138/550 (25.1) 141/549 (25.7) 59/545 (10.8) 61/548 (11.1)

167/781 (21.4)a 247/780 (31.7)a 234/779 (30.0) 157/779 (20.2)a 148/779 (19.0)a

44/206 (21.4)a 61/206 (29.6)a 65/206 (31.6)a 28/198 (14.1) 29/199 (14.6)

84/545 (15.4) 152/546 (27.8) 144/546 (26.4) 67/541 (12.4) 73/543 (13.4)

P < 0.05 vs. PTC by x2 . ISS, Injury Severity Score.

a

mobility and social interaction. Key findings are summarized in Tables 3.1 and 3.2. The data in Table 3.2 show number (percentage) of dependent patients specifically identified within 48 hours of hospital discharge using the five noted functional domains (feeding, locomotion, transfer mobility, social interaction, and expression). Each domain was given a numerical score (1–4): complete independence (4); independent with device (3); modified dependence (2); complete dependence (1); and age < 24 months excluded. Scores of 3 or 4 were considered “independent” while 1 or 2 were considered “dependent”, and patients with age < 24 months were excluded as this analysis is not relevant to this age group. The data in Table 3.1 show that patients treated at pediatric trauma centers and adult centers were similar except for age and mechanism of injury. Adult trauma centers managed an older cohort of children and penetrating injuries were more frequent; however, ISS and GCS were not statistically different. The data in Table 3.2 show significantly fewer dependent patients at the time of hospital discharge from pediatric trauma centers in a number

of categories. This is particularly apparent when comparing pediatric trauma centers to adult trauma centers with added qualifications in the domains of feeding, locomotion, social interaction and expression. These differences are not explicable by differences in age. In a somewhat modified additional analysis, stratifying by body region of injury, patients with an ISS > 15 (severe injury) and head injury had the lowest probability of dependency in all domains when treated in a pediatric trauma center (Table 3.3). One of the authors (KTO) has taken a similar approach to examine the same general question of whether there is a differential outcome for injured children treated in different types of facilities. Like the previous reports, the reference data are derived from an increasingly available and useful tool, a very large, audited database. The KID 2000 database is a government maintained tool containing 2 516 833 hospital discharge records for patients ≤ 20 years of age from 27 of the 50 United States for the year 2000. This represents more than 70% of the entire US pediatric population at that time. The data set contains E-codes, a system of descriptive numerical codes that specifically

31

32

Keith T. Oldham, Mark D. Stringer, Pierre D. E. Mouriquand

Table 3.3. Pennsylvania trauma outcome study. Functional outcome by body region of injury (ISS > 15 and head injury) Number of dependent patients at discharge (%)

Functional category

Pediatric trauma center

Adult trauma center (with added qualifications)

Adult trauma center Level I

Adult trauma center Level II

Feeding Locomotion Transfer Social Interaction Expression

61/374 (16.3) 98/374 (26.2) 96/374 (25.7) 48/370 (13.0) 50/373 (13.4)

121/479 (25.3)a 156/479 (32.6)a 152/478 (31.8)a 121/478 (25.3)a 114/477 (23.9)a

22/82 (26.8)a 26/82 (31.7) 27/82 (32.9) 20/76 (26.3)a 21/77 (27.3)a

45/270 (16.7) 69/271 (25.5) 69/271 (25.1) 37/269 (13.8) 44/270 (16.3)

P < 0.05 vs. PTC by x2 . ISS, Injury Severity Score.

a

provide information about emergency care and injury. Analysis of this large data set using E-codes identified 210 895 cases of pediatric injury. These cases were then coded using ICDMAP-90® (Trianalytics, Inc., Baltimore, MD), a validated software tool which provides an injury severity score (ISS) for each record using ICD-9 codes, the latter being a uniform and detailed numerical system of identifying specific diagnoses. This cross-linkage allows stratification of patients from these many hospitals by injury severity. The ISS stratified records were then limited to only those cases coded for an urgent or emergent admission to the hospital; records with ICD-9 codes representing cases of poisoning, medical errors, adverse drug reactions or late effects of injury were excluded. The resulting records were weighted using the proprietary KID 2000 weighting factor to produce a national estimate, yielding 176 348 evaluable cases. Cases were grouped by age (0–10 and 11–20 years), injury severity (ISS ≤ 15 vs. ISS > 15), and site of care (NACHRI-National Association of Children’s Hospitals and Related Institutions defined hospital type). This latter categorization allows comparison of children’s and adult hospitals. Measured outcomes included mortality, length of stay ( 15), most receive their care in adult oriented institutions.

Table 3.4. Types of institutions where injured children/ adolescents received care in the United States in the year 2000, stratified by injury severity Age 0–10 years

Children’s hospital Children’s unit in an adult hospital Adult hospital

Age 11–20 years

ISS ≤ 15

ISS >15

ISS ≤ 15

ISS > 15

20% 23%

28% 38%

5% 20%

5% 29%

56%

34%

75%

66%

Table 3.5 summarizes survival data which show a significantly higher probability of death for seriously injured children less than 11 years of age when treated in an adult general hospital, as compared to those treated in a children’s hospital.16 There was no apparent difference in hospital length of stay or charges. Taken together, the preceding data highlight a major problem with regard to the design of trauma systems in the United States. However, this is not unique. All around the world, a large number of injured children are cared for in institutions with an adult focus, either of necessity or because of historical referral practices. This is in spite of the fact that data from several different large databases make it clear that risk-adjusted survival for seriously injured young patients is higher in pediatric specific trauma centers and that functional outcomes are improved. A similar point can be documented with regard to splenic injuries. Splenic preservation was first proposed in pediatric patients in 1968.17 Over the ensuing years, it became clear that most children, and indeed many adults, with blunt splenic injury could be safely managed nonoperatively, yielding a very high probability of splenic salvage. Guidelines have been published detailing this non-operative approach.18,19 Generally, the overall splenic

Outcomes analysis and systems of children’s surgical care

Table 3.5. Outcomes of patients aged 0–10 years with an ISS > 15: comparison of children’s and adult hospitals Age 0–10 years, ISS >15 (n = 46 002)

Children’s hospital Adult general hospital ∗ Chi square, †t-test

Mean ISS

Percent mortality

Percent LOS ≥ 8d

Percent charges ≥$ 15K

18.9 ± 9.1 19.4 ± 9.3 P = 0.08†

5.3 7.6 P = 0.002∗

45 45 P = 0.84∗

36 35 P = 0.26∗

LOS, length of stay.

salvage rate in children with blunt abdominal trauma should be well over 90%. In the light of these published data, what actually happens in the United States at present? Given the perspective detailed above, that most injured children are cared for in adult trauma systems, we and others examined the incidence of splenic surgery following childhood trauma. Rothstein et al.,20 used the KID 2000 database (referenced above) to explore the question of whether the type of institution in which treatment was provided influenced the type of care with specific regard to the probability of splenectomy in injured children. They found that the crude rate of splenectomy in the USA was 3% (11/338) at dedicated (freestanding) children’s hospitals, 9% (45/525) in children’s units within adult general hospitals, and 15% (197/1327) in adult general hospitals. Adjusted for severity of injury, children treated at an adult hospital had a 2.8 times higher probability of splenectomy than if treated in a children’s hospital, and this likelihood was 2.6 times higher in a children’s unit within a general hospital (P = 0.003 and 0.013, respectively). Using similar analytic techniques, Densmore et al.21 found that the probability of operative management for pediatric splenic injury was 17% in a general hospital, 2% in a children’s hospital, and 4% in a children’s unit in a general hospital. This variance in practice could not be explained by injury severity. Stylianos et al.22 examined the issue of the management of pediatric splenic injury in administrative data sets obtained from California, Florida, New Jersey and New York, and Mooney et al.23 did so for New England. Both reports show substantial variation in practice with regard to the likelihood of operative management for splenic injury, indicating that published “best practice” benchmarks are not routinely achieved in institutions treating injured children. Once again, small scale studies indicate that these findings are not unique to the USA.24 In aggregate, these data make clear a clinical care issue that is not satisfactorily resolved at present by policy makers either in government or in medicine. The data summarized here indicate that injured children have a significantly

better outcome when cared for in designated pediatric trauma centers. This is demonstrably true with respect to overall survival, functional outcomes, and the likelihood of splenic preservation in the specific case of pediatric blunt splenic injury. One would hope that these observations would become the basis for changes in trauma system design and practice. Dr. Arnold Epstein writing on this general subject offered several principles for such change in a thoughtful essay entitled “Volume and outcome – it is time to move ahead”.25 1. The least intrusive action is education . . . Public dissemination of performance reports can lead to improvements in quality . . . Efforts to decrease the proportion of procedures performed in low volume hospitals seem appropriate. 2. Use financial incentives or even regulatory means to promote the use of high-volume centers for certain types of care. Initial restrictions should be confined to metropolitan areas and focused on surgical procedures for which the differences are greatest, with (volume) thresholds set very conservatively. 3. The notion of regulating the broad-scale regionalization of medical care prompts worries about the details and unintended consequences (who and how will this be done?). Demonstration and evaluation should precede broad-scale policy changes. 4. Better measures of quality, such as risk-adjusted mortality, should be used when available. 5. Finally, increasing the proportion of care provided in high-risk centers is only one step toward improving quality of care. Efforts at improvement by individual hospitals will be critical.

Biliary atresia Biliary atresia (BA) is a rare congenital obliterative cholangiopathy of unknown etiology (see Chapter 35). Reliable incidence figures are available from France (1 in 19 500 live

33

34

Keith T. Oldham, Mark D. Stringer, Pierre D. E. Mouriquand

births), the UK and Eire (1 in 16 700 live births), Georgia (United States) and Sweden (1 in 14 000 live births).26–29 Untreated infants succumb to liver failure within a year or two. In the late 1950s, Morio Kasai, a Japanese surgeon, reported the presence of patent microscopic biliary channels at the porta hepatis in young infants with BA. Exposure of these channels by radical excision of atretic extrahepatic biliary remnants could result in effective drainage of bile, especially if the operation was performed before 8 weeks of age. The Kasai portoenterostomy operation is now accepted as the standard operation for BA. The success of the operation as determined by clearance of jaundice (plasma bilirubin < 20 µmol/l) is related to several factors including age at presentation, type of BA, and the presence of associated anomalies. Children who clear their jaundice and remain anicteric for the first 3 years of life have about an 80% chance of reaching adulthood with their native liver, i.e., without a liver transplant. Infants who fail to clear their jaundice after portoenterostomy and those who develop complicated or end-stage chronic liver disease despite an initially successful Kasai procedure require liver transplantation. Biliary atresia is the commonest indication for liver transplantation in children. Most of these cases require their transplant in the first few years of life. Techniques such as split-liver grafting and living-related liver transplantation have minimized the risk of these small children dying on the waiting list (see Chapter 65). The combination of Kasai portoenterostomy and liver transplantation has transformed a disease that was almost invariably fatal in the 1960s into one with a current overall 5-year survival of about 90%. Furthermore, long-term studies have shown a relatively good quality of life in BA survivors after portoenterostomy alone30 and after liver transplantation.31 Initial clearance of jaundice is thus an important step in avoiding the need for liver transplantation and its attendant risks. In the UK, a series of three national studies have demonstrated that the outcome of infants with BA is markedly affected by the experience of the center.27,32,33 This organizational dimension to the delivery of surgical care is an important additional factor in determining the long-term outcome of BA and prompted a major change in the management of affected infants in the UK. The first of these UK surveys was conducted between 1980 and 1982 (Table 3.6).32 There was a significant difference in jaundice-free survival between the highest volume center treating more than five cases per year and the 15 low volume centers treating fewer cases. A similar survey of all infants with BA in UK and Eire conducted between 1993 and 1995 again demonstrated that outcome was related to center volume and experience.27 In this second study,

survival without liver transplantation and overall survival were both significantly greater in the two centers managing more than five cases per year than in the 13 centers treating fewer cases. After taking into account age at surgery, sex, gestational age, and presence of the biliary atresia splenic malformation syndrome, stepwise multivariate regression showed that center size was the only significant independent factor predictive of overall survival. There was much debate within the profession and the media about the dissemination and interpretation of these findings.34 However, in 1999 the Department of Health decided that the management of BA in England and Wales would be centralized and limited to three supra-regional pediatric liver units in the south (London), midlands (Birmingham) and north (Leeds) of England. Scotland has an independent body regulating these matters. Within pediatric surgery in the UK, there was understandable opposition to this decision.35 Nevertheless, the decision was implemented, and, most importantly, the subsequent outcome was critically audited. Thus, the third national survey of infants with BA was published (Table 3.6).33 This showed that, within the three centers, 57% of infants cleared their jaundice after Kasai portoenterostomy. Significantly, there was no difference in outcome between the three centers – all three had uniformly good results equivalent to those previously reported from centers treating five or more cases annually. Estimated actuarial 4-year survival was 89%. Another measure of success after Kasai portoenterostomy is survival with the native liver; this was estimated to be 51% at 4 years in this recent UK survey in comparison to an overall 5-year actuarial survival without liver transplantation of around 44% in the 1993–1995 survey. It is likely that small improvements soon after Kasai portoenterostomy will progressively translate into larger more significant benefits with the passage of time. Thus, in a recent follow-up of the 1993–1995 cohort of BA patients reassessed at a median age of 8 years, actuarial native liver survival without liver transplantation was statistically significantly more likely in centers treating more than five cases annually (56% vs. 27%, P = 0.004).36 The advantages of centralizing the management of BA into specialized hepatobiliary units, each dealing with the whole spectrum of liver disease including liver transplantation, are numerous. They include seamless care from rapid evaluation of neonatal obstructive jaundice through to prompt Kasai portoenterostomy and long-term management of the sequelae of chronic liver disease (metabolic, nutritional, infective, etc.); concentration of medical and surgical expertise into a multidisciplinary team with facilities for supporting patients and their families; parental education and advice in a center familiar with the whole

Outcomes analysis and systems of children’s surgical care

35

Table 3.6. Summary of three consecutive national surveys of biliary atresia in the United Kingdom Number of infants undergoing a portoenterostomy-type procedure

Median age at operation (days)

Duration of follow-up (months)

Author

Survey period (duration)

Number of infants with BA

McClement et al. 198532

1980–1982 (UK, 3 y)

114

107

63

Mean 27

McKiernan et al. 200027

1993–1995 (UK and Eire, 2 y)

93

91

54

Davenport et al. 200433

1999–2002 (England and Wales, 3.5 y)

148

142

spectrum of the disorder; optimal timing of liver transplantation (the majority of children with BA will eventually need a new liver); and potential opportunities for better scientific and clinical research. However, there are disadvantages such as the travelling for families and its social and economic burden, the potential deskilling of referring pediatric specialists, and the loss of training opportunities in regional units. Some of these problems are surmountable by good communication, shared care (including outreach clinics) and by arranging for trainees to visit specialist units. As with the trauma debate, the issues surrounding the management of BA are not unique. For example, in France the situation has been comparable to that in the UK before 1999. Thus, between 1986 and 1996, 472 infants with BA were recorded. The Kasai procedure was performed in 32 centers (nine of which offered pediatric liver transplantation) and overall 5-year survival and native liver survival were 70% and 32%, respectively.26 Univariate analysis showed significantly better 5-year survival with or without liver transplantation in the single center which performed 20 or more portoenterostomies per year compared with centers treating up to five cases per year. Thus, 5-year survival with the native liver was 39% in the center treating 20 or more new infants with BA each year, 31% in centers treating three to five cases annually and 24% in centers treating no more than two new cases each year.26 Pediatric surgeons in the UK were understandably concerned about the centralisation of BA but the carefully audited results of this change in practice have helped to recognize the advantages of this approach (category IIa evidence). Such a model of care for BA is unlikely to be appropriate for all healthcare systems because of variations in population density, geography, socioeconomic factors,

54

Center/volume effect and jaundice-free outcome

Median 42

> 5 cases/yr (n = 1) 2–5 cases/yr (n = 15) ≤ 1 case/yr > 5 cases/yr (n = 2)

: 43% : 29% : 11% : 62%

Median 25

2–5 cases/yr (n = 5) ≤1 case/yr (n = 8) > 5 cases/yr (n = 3)

: 58% : 17% : 57%

healthcare resources, etc. The perceived benefits in the management of BA in the UK are not simply related to surgical volume although, as with many other conditions with a major surgical component, this is undoubtedly an important factor. The situation is much more complex, involving the whole package of long-term multidisciplinary care for these complex patients. It is not known within pediatric surgery how many conditions will benefit from this type of approach. Generally, these are rare and complex disorders where there is evidence that concentrating medical and surgical expertise into designated units is beneficial to the outcome of the patient.

Pediatric urology Urogenital surgery in children represents between 50% and 75% of the activity of most European units dealing with general pediatric surgery and urology. Fewer and fewer adult surgeons operate on children, especially infants, mainly because regulations for anesthetizing these patients have become more restrictive. This explains the decreasing number of children operated on in district general hospitals (DGH) and small institutions, and the increasing flow of so-called “minor cases” to large university hospitals. Clinical practice in France has also been significantly affected by the new financing system for the National Health Service which was adopted in January 2004. In short, these new arrangements strongly encourage hospitals to operate on as many cases as possible to ensure an increased income. Hence, minor cases which used to be referred to private hospitals or DGHs are now retained in major institutions to increase income. This system is rather perverse since the number of doctors and nurses has not changed

36

Keith T. Oldham, Mark D. Stringer, Pierre D. E. Mouriquand

appreciably but the number of health administrators has increased dramatically. The most obvious failure of this system is that health authorities wrongly equate the number of patients seen and the quality of care. With the huge increase of hypospadias cases in Western countries, operations for this condition have become one of the commonest procedures in pediatric urology. Contrary to an old statement, there is no minor hypospadias and this type of surgery should be performed by very experienced surgeons. Testicular surgery is also extremely common and should only be performed after adequate, supervised training in order to avoid relatively high rates of morbidity. Surgery for ambiguous genitalia needs to be concentrated in a few centers with relevant multidisciplinary expertise (including pediatric endocrinologists, geneticists, biologists, psychiatrists, etc.). A country of 60 million inhabitants should require no more than four or five expert centers in this field. A large part of pediatric urology is made up of patients with prenatally detected congenital impairment of urine flow. Most of these cases (80–85%) will not require any surgical intervention but they do need close urological followup involving imaging and isotope departments. Here again, close collaboration between specialties (radiology, obstetrics, nephrology) is required. Another growing area of pediatric urology is represented by all aspects of bladder dysfunction including the challenging problems of neurogenic incontinence and the far more frequent non-neurogenic bladder dysfunction. These highly specialized fields have fostered the development of non-surgical pediatric urology where urodynamics and consultations by clinical nurse specialists are essential. Finally, specific, rare, major congenital malformations such as bladder exstrophy and cloacal malformations ought to be concentrated in specific centers. The UK now has two centers dealing with bladder exstrophy (London and Manchester). France is not so well organized but most of these malformations are currently referred to Paris or Lyon. Here again, it is multidisciplinary teamwork and the clinical nurse specialist that are essential. After this brief review of the organizational aspects of pediatric urology which may affect the long-term outcomes of our patients, the question of training in pediatric urology must be considered. In most Western countries there are two main avenues – via urology and pediatric surgery. Efforts are gradually being made to develop a specific career pathway to pediatric urology. At present, there is no consensus about this route but European institutions are working toward a designated pathway with accredited centers and, subsequently, accredited specialists. It will take time and much effort. The Joint Committee for Pediatric Urol-

ogy, which brings together urologists, pediatric surgeons and pediatric urologists, is a first step in coordinating the actions of different countries. This will not be an easy task since training schemes and standards of practice vary widely between countries. How many pediatric urologists do we need? The answer is extremely variable from one country to another and it will be the task of these institutions to regulate the number of specialists to avoid saturation. There is an increasingly apparent trend in many countries that pediatric urology is becoming relatively independent from adult urology and from pediatric surgery. This is a controversial issue because both pediatric urology and pediatric surgery are small disciplines and, for many, they need to stay together to have a stronger political voice. Each cannot survive without the other. Most countries have a national pediatric surgical association that includes pediatric surgeons, pediatric urologists, pediatric orthopedic surgeons, and other pediatric subspecialists. However, the two dominant forums for pediatric urology are the urology section of the American Academy of Pediatrics and the European Society for Pediatric Urologists, attended largely by pediatric urologists. There are parallels in pediatric orthopedic surgery. So, how long will pediatric urology and pediatric surgery remain so closely linked? Probably for quite a long time, not just for political reasons but also because children need a special environment to receive the best care. There is the common denominator between the pediatric subspecialties and it is important to preserve it.

Conclusions To summarize, this brief review highlights the importance of systems of care in affecting outcomes in pediatric surgery and urology. Organizational aspects at different levels influence outcome. Thus, in the case of pediatric trauma, injured children have a significantly better outcome when cared for in designated pediatric trauma centers rather than within adult care facilities. In pediatric urology, there is the need for broad subspecialist care by trained pediatric urologists. Finally, for certain rare, complex conditions such as biliary atresia, centralization within pediatric surgery may have advantages. We conclude that systems of care that recognize the needs of the child; that consolidate particular types of patients into regional centers; that provide and then require specialized training; and that offer comprehensive multidisciplinary access to children’s specific providers in all surgical, medical and allied health areas, yield the best outcomes for our patients.37 While data are

Outcomes analysis and systems of children’s surgical care

still limited at present, we are increasingly able to organize pediatric health care on the basis of demonstrable quality, that is clinically relevant outcomes. Our challenge now is to continue to generate relevant data and to have the wisdom and the will to use these outcomes to provide a rational system of care that is optimal for our patients and their families.

15.

16.

17.

REFERENCES 1. Luft, H. S., Bunker, J. P., & Enthoven, A. C. Should operations be regionalized? The empirical relation between surgical volume and mortality. N. Engl. J. Med. 1979.1; 301:1364–1369. 2. Hughes, R. G., Hunt, S. S., & Luft, H. S. Effects of surgeon volume and hospital volume on quality of care in hospitals. Med Care 1987; 29:1094–1107. 3. Birkmeyer, J. D. High-risk surgery – follow the crowd. J. Am. Med. Assoc. 2000; 283(9): 1191–1193. 4. Birkmeyer, J. D., Siewers, A. E., Finlayson, E. V. et al. Hospital volume and surgical mortality in the United States. N. Engl. J. Med. 2002; 346(15) :1128–1137. 5. Birkmeyer, J. D., Stukel, T. A., Siewers, A. E., Goodney, P. P., Wennberg, D. E., & Lucas, F. L. Surgeon volume and operative mortality in the United States. N. Engl. J. Med. 2005; 349(22): 2117–2127. 6. Halm, E. A., Lee, C., Chassin, M. R. Is volume related to outcome in health care? A systemic review and methodologic critique of the literature. Ann. Intern. Med. 2002; 137:511–520. 7. Center for Disease Control and Prevention, Epidemiology Program Office, Division of Public Health Surveillance; available online at: http://www.epo.cdc.gov/wonder. 8. Center for Disease Control and Prevention. Web-based injury statistics query and reporting system (WISQARS) [Online]. (2003) National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (producer). Available from URL: www.cdc.gov/ncipc/wisqars. 9. Knudson, M. M., Shagoury, C., & Lewis, F. R. Can adult trauma surgeons care for injured children? J. Trauma. 1992; 32(6):729– 739. 10. Fortune, J. B., Sanchez, J., Graca, L. et al. A pediatric trauma center without a pediatric surgeon: a four-year outcome analysis. J. Trauma. 1992; 33(1):130–139. 11. Bensard, D. D., McIntyre, R. C., Moore, E. E., & Moore, F. A. A critical analysis of acutely injured children managed in an adult level I trauma center. J. Pediatr. Surg. 1994; 29(1): 11–18. 12. Hall, J. R., Reyes, H. M., Meller, J. L., Loeff, D. S., & Dembek, R. The outcome for children with blunt trauma is best at a pediatric trauma center. J. Pediatr. Surg. 1996; 31(1):72–77. 13. Nakayama, D. K., Copes, W. S., Sacco, W. Differences in trauma care among pediatric and nonpediatric trauma centers. J. Pediatr. Surg. 1992; 27(4):427–431. 14. Potoka, D. A., Schall, L. C., & Ford, H. R. Improved functional outcome for severely injured children treated at pedi-

18.

19.

20.

21.

22. 23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

atric trauma centers. J. Trauma. – Injury Infect. Crit. Care. 2001; 51(5):824–832; discussion 832–4. Potoka, D. A., Schall, L. C., Gardner, M. J. et al. Impact of pediatric trauma centers on morality in a statewide system. J. Trauma 2000; 49:237–245. Densmore, J. S., Oldham, K. T., & Guice, K. S. Relationship of pediatric trauma mortality rates to type of institution providing care. J. Pediatr. Surg. 2006; 41(1):92–98. Upadhyaya, P. & Simpson, J. S. Splenic trauma in children. Surg, Gynecol. Obstet. 1968; 126:781–790. Stylianos, S. and the APSA Trauma Committee. Evidence-based guidelines for resource utilization in children with isolated spleen or liver injury. J. Pediatr. Surg. 2000; 35:164–169. Stylianos, S. and the APSA Liver/Spleen Trauma Study Group. Compliance with evidence-based guidelines in children with isolated spleen or liver injury: a prospective study. J. Pediatr. Surg. 2002; 37:453–456. Rothstein, D. H., Forbes, P. W., & Mooney, D. P. Variation in the management of pediatric splenic injuries in the United States (unpublished data). Densmore, J. S., Oldham, K. T., & Guice, K. S. Splenic surgery rates following pediatric trauma in the United States (unpublished data). Stylianos, S. Variation in practice for splenic injury (unpublished data). Mooney, D. P. & Forbes, P. W. Variation in the management of pediatric splenic injuries in New England. J. Trauma. 2004; 56: 328–333. Godbole, P. & Stringer, M. D. Splenectomy after paediatric trauma: could more spleens be saved? Ann. R. Coll. Surg. Engl. 2002; 84:106–108. Epstein, A. M. Volume and outcome – it is time to move ahead. N. Eng. J. Med. 2002; 346(15):1161–1164. Chardot, C., Carton, M., Spire-Bendelac, N. et al. Prognosis of biliary atresia in the era of liver transplantation: French national study from 1986 to 1996. Hepatology 1999; 30:606–611. McKiernan, P. J., Baker, A. J., & Kelly, D. A. The frequency and outcome of biliary atresia in the UK and Ireland. Lancet 2000; 355: 25–29. Yoon, P. W., Bresee, J. S., Olney, R. S., James, L. M., & Khoury, M. J. Epidemiology of biliary atresia: a population-based study. Pediatrics 1997; 99:376–382. Fischler, B., Haglund, B., & Hjern, A. A population-based study on the incidence and possible pre- and perinatal etiologic risk factors of biliary atresia. J. Pediatr. 2002; 141:217–222. Howard, E. R., MacClean, G., Nio, M., Donaldson, N., Singer, J., & Ohi, R. Biliary atresia: survival patterns after portoenterostomy and comparison of a Japanese with a UK cohort of longterm survivors. J. Pediatr. Surg. 2001; 36:892–897. Bucuvalas, J. C., Britto, M., Krug, S. et al. Health-related quality of life in pediatric liver transplant recipients: a single-center study. Liver Transpl. 2003; 9:62–71. McClement, J. W., Howard, E. R., & Mowat, A. P. Results of surgical treatment for extrahepatic biliary atresia in United Kingdom 1980–2. Br. Med. J. 1985; 290:345–347.

37

38

Keith T. Oldham, Mark D. Stringer, Pierre D. E. Mouriquand

33. Davenport, M., De Ville de Goyet, J., Stringer, M. D. et al. Seamless management of biliary atresia in England and Wales (1999– 2002). Lancet 2004; 363:1354–1357. 34. Davison, S., Miller, V., Thomas, A., Bowen, J., & Bruce, J. The profession, not the media, should assess where Kasai portoenterostomy should be performed. Br. Med. J. 1999; 318: 1013.

35. Lloyd, D., Jones, M., & Dalzell, M. Surgery for biliary atresia. Lancet 2000; 355:1099–1100. 36. McKiernan, P., Baker, A., Mieli-Vergani, G., & Kelly, D. The BPSU study of biliary atresia – outcome after 8 years. J. Pediatr. Gastroenterol. Nutr. 2003; 36:529 (abstract) 26. 37. Arul, G. S. & Spicer, R. D. Where should paediatric surgery be performed? Arch. Dis. Child. 1998; 79:65–72.

4 Perinatal mortality and morbidity: outcome of neonatal intensive care Janet M. Rennie Department of Neonatal Medicine, University College London Hospitals, UK

Introduction: historical aspects Perinatal mortality Perhaps no other medical subspecialty has achieved such a dramatic improvement in survival as that documented in neonatal medicine over the last 40 years. Since the 1960s the survival rate for infants born weighing less than 1500 g (very low birthweight, VLBW) has increased from 45% to over 80%. For the small group born weighing less than 1 kg (extremely low birthweight, ELBW) the increase in survival has been from 20% to almost 70%. These changes have occurred against a background of improving perinatal, infant and childhood mortality in the United Kingdom and elsewhere, although it remains true that a VLBW infant is 100 times more likely to be stillborn or die during the first month of life than an infant born weighing 3000 g or more (Table 4.1). The UK definition of a stillbirth was changed to include all fetuses delivered dead after 24 complete weeks of pregnancy in October 1992. This caused a step up of about 1 per 1000 in the UK perinatal mortality rate, which at 8.0 per 1000 total births remains similar to that in other European countries and the USA (Fig. 4.1). Whilst prematurity remains the leading cause of perinatal and neonatal death, significant contributions continue to be made from perinatal asphyxia, sepsis and congenital malformations. Group B streptococcal infection and chorioamnionitis, where the organism is rarely isolated, are important causes of fetal and neonatal deaths.

Congenital malformations As the number of deaths from infection has declined dramatically in the last 50 years, the proportion of deaths due

to congenital anomalies has increased. In 1998 28% of UK infant deaths were coded as due to congenital anomaly, compared to 13% in 1948. Exactly the same effect has been seen in the USA. This has occurred in spite of the fact that termination of pregnancy, when a malformation which is known to carry a substantial risk of handicap is diagnosed antenatally, has been a legal option in England, Scotland and Wales for some years, although this is still not the case in Eire or many other countries in Europe. English law was amended in 1990 to allow such terminations to be carried out at any stage of pregnancy, and 1813 such terminations were notified in 1999; 449 for CNS malformations and 596 for chromosomal anomalies (data from the Office for National Statistics, www.statistics.gov.uk). Liveborn anencephalic babies are now rarely seen in England, Wales or Scotland unless they are the product of a twin pregnancy, and only 12 were reported in 2001. Periconceptional folate supplementation has the potential to prevent the conception of infants with neural tube defects and is used increasingly by women who are planning to become pregnant. Dietary supplementation is particularly important for epileptic women on treatment. An unlooked-for benefit of dietary supplements has been a reduction in infants with cleft palate, and a protective effect against acute lymphoblastic leukemia in childhood has also been reported. The reported rate of congenital malformations in the UK in 2001 was 114/10 000 births, which is an increase from 81/10 000 reported in 1994 (Table 4.2). The current monitoring system provides the most comprehensive data yet, and has been in place since 1964 when it was introduced after the thalidomide epidemic in order to recognize any similar hazard. It is a voluntary system,

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


39

40

J. M. Rennie

Table 4.1. Perinatal mortality rates in the UK 2002

Birthweight

Stillbirth rate per 1000 total births

Neonatal mortality rate per 1000 live births

Postneonatal mortality rate per 1000 live births

Infant mortality rate per 1000 live births

30 mL/day) Increased crying/fussiness Discomfort with spitting up Frequent back arching

21.02

9.92–44.53

7.26

2.68–19.70

2.77 2.5 2.37

1.39–5.55 1.15–5.46 1.44–3.91

∗ Odds ratio obtained from multiple logistic regression; P < 0.001 From Nelson SP, Chen EH, Syniar GM, Christoffel KK. Prevalence of symptoms of gastroesophageal reflux during infancy. A pediatric practice-based survey. Pediatric Practice Research Group. Arch Pediatr Adolesc Med 1997; 151(6):569–572.

80

mean of 9.7 years of age. They found that 41% of children between 3 and 4 months of age were regurgitating or vomiting about half of their feedings daily. By 13–14 months of age, 10% if adult survivors included Not increased by postoperative radiation therapy In those undergoing operative intervention

Figures refer to episodes of obstruction requiring operative intervention during variable follow-up periods.

in four patients.40 Seventy-four percent of obstructions developed within 6 months of the primary operation and 87% within one year. Type of operation The incidence of adhesive small bowel obstruction complicating various intra-abdominal operations is detailed in Table 20.2. Several procedures merit further comment. Open appendectomy Adhesive small bowel obstruction is uncommon after appendectomy. Janik et al. (1981) reported an incidence of less than 1% during a 12-year period.32 However, because appendectomy is such a frequent operation, numerically it is an important cause of adhesive obstruction.39 Perforated appendicitis is associated with a doubling of the risk of adhesion-related small bowel obstruction.32,41 It has been suggested that adhesive small bowel obstruction after appendectomy in adults is usually due to a band adhesion and rarely settles with conservative management.48 Open transabdominal Nissen fundoplication Wilkins and Spitz identified a 10.3% incidence of adhesive small bowel obstruction occurring after a mean interval of 10 months in 156 children undergoing Nissen fundoplication.42 Additional procedures performed concomitantly with the fundoplication substantially increased

the risk of developing obstruction; the risk was 21% for patients who had a Ladd’s procedure and 12% for those who had an incidental appendectomy. The incidence in those patients who underwent no additional procedure other than the fundoplication was 1.8%. Inability to vomit in most of the children led to delay in diagnosing intestinal obstruction and this proved fatal in two cases. Other series have reported adhesive small bowel obstruction rates of 2%–10% after Nissen fundoplication.43 In a personal series of 80 infants and children undergoing Nissen fundoplication in Leeds (two-thirds of whom had a Stamm gastrostomy) there was not a single case of adhesive bowel obstruction during a median follow-up of more than 2 years. Restricting the operative field to the left upper quadrant of the abdomen and deliberately avoiding handling of the small bowel, which is retained within the infracolic compartment, appear to be important factors in preventing this complication. Nephrectomy for Wilms’ tumor In a report from the Third National Wilms’ Tumor Study, 104 (5.4%) of 1910 children developed adhesive small bowel obstruction after transperitoneal nephrectomy for Wilms’ tumor.38 Important risk factors were higher local tumor stage, extrarenal vascular involvement, and en bloc resection of other organs at the time of nephrectomy. The incidence of adhesive bowel obstruction was not increased in children who received postoperative radiation therapy. In neonates, meconium ileus, necrotizing enterocolitis and intestinal malrotation are more likely to be complicated by adhesive obstruction than other pathologies.40 Morbidity and mortality Postoperative adhesive small bowel obstruction is associated with a mortality of between 2 and 6%32,33,42,49 and a much greater morbidity from intestinal resection, intraabdominal sepsis, and wound complications. In addition, there are potential but controversial long-term risks from pelvic adhesions such as infertility and chronic abdominal/pelvic pain (see chapter on Appendix).50 The morbidity associated with surgical treatment has encouraged some authors to challenge the traditional surgical approach to adhesive bowel obstruction in children and advocate a more conservative policy in selected cases.39,49 However, in the absence of reliable clinical signs of intestinal strangulation, this approach must be adopted with caution. For those that require surgery, laparoscopic adhesiolysis is possible in selected cases.51 Prevention Attempts to prevent the development of adhesions have focused on (i) minimizing peritoneal injury during surgery,

Abdominal surgery: general aspects

(ii) separating adhesiogenic surfaces and, (iii) promoting fibrinolysis. Minimizing peritoneal trauma This requires meticulous surgical technique with gentle handling and careful hemostasis. The use of powderfree surgical gloves, wet gauze swabs and fine absorbable sutures may help to reduce the risk of postoperative adhesions. Avoiding tissue desiccation and covering raw surfaces with omentum may also be beneficial. Although omental adhesions are the most common, they are at low risk of producing intestinal obstruction.28 Laparoscopic techniques result in fewer adhesive complications than open procedures but the risk of adhesions is not eliminated. Several large studies of children undergoing laparoscopic appendectomy have been reported but there are no long-term data on the risk of subsequent adhesion obstruction. In one French multicenter study of 1379 pediatric laparoscopic appendectomies followed up for only 15 days, two cases required adhesiolysis and gut resection.52 Moore et al. (1995) reviewed 41 pediatric patients who had undergone second-look procedures after laparoscopic urologic interventions.53 Adhesions were noted in 4 (10%) patients either at the operative site or at trocar sites; the risk of adhesions was related to the extent of dissection. These authors considered that the incidence of adhesions was lower than that which would be expected with open surgery. A reduction in foreign body contamination of the peritoneum may be one factor contributing to the lower incidence of adhesions after laparoscopic surgery.54 Separating adhesiogenic surfaces Two approaches have been used to separate adhesiogenic surfaces after abdominal surgery: (a) the use of liquid or mechanical barriers and (b) the administration of prokinetic agents. Peritoneal lavage with normal saline or macromolecular dextran solution has no major role in preventing adhesions.55 Hyaluronan, a polysaccharide found throughout the body, facilitates cell migration and increases mesothelial cell proliferation.56 A hyaluronic acid–carboxymethylcellulose membrane (Seprafilm) has been shown to reduce intra-abdominal adhesion formation in adults undergoing colectomy without increasing the risk of intra-abdominal sepsis.57 A recent randomized controlled study of 122 children in Japan showed that the incidence and severity of intra-abdominal adhesions under the abdominal incision site was significantly reduced by the use of Seprafilm® but the frequency of clinical adhesive small bowel obstruction was not affected.57a Seprafilm® did not cause an increase in surgical complications. Wrapping a bowel anastomosis or staple line with this bioresorbable membrane should be avoided since this may increase the

risk of leakage.58 This material is approved in the United States and Europe for use in adults undergoing laparotomy for benign, non-infectious conditions but it is not currently licensed for use in children. In Europe, intraperitoneal 4% icodextrin solution (Adept® ) is also licensed for use in adults to reduce adhesions after abdominal surgery. At present, it is uncertain whether these agents will result in fewer episodes of adhesive bowel obstruction. Recently, fibrin-sealant plication of the small bowel has been proposed as a method of preventing the formation of obstructing adhesions in children but evidence for this technique is from retrospective, uncontrolled data only.59 In experimental animals, early introduction of enteral feeding and the use of Cisapride, a prokinetic agent, have been shown to reduce the duration of postoperative ileus and adhesion formation but the clinical validity of these findings is uncertain.60 In comparison with intravenous opiates, epidural analgesia is associated with earlier recovery of gut function after major abdominal surgery and this may be beneficial in reducing postoperative adhesion formation.61

Promoting fibrinolysis The rationale for the primary prevention of adhesions with fibrinolytic agents is based on the recognition that damaged peritoneum is unable to generate plasmin from plasminogen, and consequently fibrinolytic activity is impaired. Studies of recombinant tissue plasminogen activator have shown promise in experimental animals but there are no reports as yet of randomized controlled clinical trials.28 Attempts to minimize the inflammatory response using corticosteroids, thereby enhancing intrinsic fibrinolytic activity, have not been successful. There are a few reports on the secondary prevention of adhesive intestinal obstruction in children using intraluminal tube-stent plication of the small bowel.62 Short followup periods and uncontrolled data limit conclusions on the efficacy of this technique but it may have a place as an adjunct to adhesiolysis in patients requiring repeated operations for the relief of obstruction due to extensive, dense adhesions.

Intestinal resections The physiology and long-term complications of short bowel syndrome are discussed elsewhere (see Chapter 28). However, extensive gut resection may have significant longterm consequences in children without overt intestinal failure. The length of resected ileum and the presence of an ileocecal valve or colon are important factors affecting outcome.63 There is less potential for adaptation in the

261

262

M. D. Stringer

jejunum than in the ileum. Potential long-term sequelae of intestinal resection in infants include the following. Vitamin deficiencies Ileal resection leads to the loss of some or all of the capacity for receptor-mediated absorption of intrinsic factor–cobalamin complex. Vitamin B12 deficiency is usually associated with a low serum concentration and an abnormal Schilling test result. Valman and Roberts found impaired vitamin B12 absorption in seven of ten infants who had more than 45 cm of ileum resected but they documented normal vitamin B12 absorption in two children in whom only 15 cm of terminal ileum was preserved.64 In their patients with impaired absorption, the serum level of vitamin B12 did not fall below the normal range for several years, puberty being a particularly vulnerable time. Parashar et al. (1990) found only one child with vitamin B12 deficiency in a series of 27 children who had undergone ileocolic resection for various etiologies followed up for a mean of 6 years.65 Impaired absorption of vitamin B12 may recover after several years, probably because of intestinal adaptation.66 Vitamin B12 status should be checked periodically in children who have undergone ileal resection. Disturbance of the normal enterohepatic circulation of bile salts after ileal resection predisposes to fat-soluble vitamin deficiencies. This may be exacerbated by cholestyramine used to control bile salt diarrhea. Vitamin D deficiency and rickets can occur in children who achieve full enteral feeding with a short gut67 but does not appear to be a particular hazard after extensive ileal resection alone.68 Vitamin E malabsorption may manifest many years after apparently successful surgery for jejunal atresia.69 Deficiency of vitamin E may lead to mitochondrial degeneration and the deposition of brown lipofuscin pigment within intestinal smooth muscle which exacerbates intestinal dysmotility. A combination of intestinal tapering and vitamin E supplementation appears to be an effective treatment. Gallstones Cholelithiasis is a well-recognized late complication of ileal resection, disease or bypass in adults and, to a lesser extent, in children. Pathogenesis is multifactorial and may be related to perioperative gallbladder stasis and starvation, the disturbance of the normal enterohepatic circulation of bile salts, and increased absorption of bilirubin from the gut. Hyperoxaluria and renal stones Fat malabsorption secondary to ileal resection encourages the formation of insoluble calcium soaps in the lumen of the gut. One consequence of this is increased absorption of dietary oxalate which, under normal circumstances,

Fig. 20.2. Resected ileocolic anastomosis from a boy with bleeding from an ileocolic anastomotic ulcer 5 years after neonatal surgery. Reproduced by kind permission of Dr John WL Puntis, Leeds General Infirmary, UK.

would form insoluble calcium oxalate in the bowel lumen; bile salts in the colon may also increase oxalate absorption. Hyperoxaluria and renal oxalate stones may develop in patients with a retained functioning colon after extensive small bowel resection. A quarter of adults with less than 200 cm jejunum anastomosed to colon develop symptomatic renal stones.70 The risk in children is uncertain but in one small series approximately half the patients had evidence of hyperoxaluria.71 Bohane et al. (1979) found hyperoxaluria in four out of ten children who had undergone ileal resection but could not correlate this with the length of resected ileum or the interval since surgery.72 Perianastomotic ulceration Perianastomotic ulceration is a rare long-term complication of ileocolic or jejunocolic anastomosis in infancy (Fig. 20.2).75 All reported cases have presented between 3 and 13 years after their original surgery with severe iron deficiency anemia from overt or occult lower gastrointestinal bleeding, with or without abdominal pain and diarrhea (Table 20.3). Colonoscopy is the most useful investigation. The etiology of this complication is unknown but the pathology of the ulceration suggests a process of chronic inflammation and repair. Treatment is problematic: antibiotics, H2 -blockers and sulphasalazine are not curative and resection and reanastomosis may be complicated by recurrence.78 Other potential late sequelae Anastomotic strictures and other mechanical problems are rarely encountered as late complications of intestinal

M

Arnbjornsson77

Paterson76

M M M F

Hamilton75

9m

9w 2d 3y 2d

Infancy

M F M

M>F

2d 6w 4m

M M

Parashar73

Couper74

1m 4m

Sex

Author

Age at operation

Gastroschisis

Colonic atresia NEC Gastroschisis Gastroschisis Ileal atresia Gastroschisis NEC Gastroschisis Volvulus NEC

Volvulus Intussusception

Primary condition

Ileo-colic

Ileo-colic Jejuno-colic Ileo-colic Ileo-colic Jejuno-colic Ileo-colic Ileo-colic Ileo-colic Ileo-colic Ileo-colic

Jejuno-colic Ileo-colic

Anastomosis

–

Dexon Dexon – – – – Silk Silk CCG & silk CCG & silk

Silk –

Suture material

Table 20.3. Reported cases of peri-anastomotic ileo/jejuno-colic ulceration78

8

12 11 3 8 10 5 7 5 10 8

4 13

Age (y) at presentation

Multiple perianastomotic

Single, large, anastomotic Multiple, small, ileal & colic – Single, anastomotic Two small bowel Single, large, ileal Single, large, ileal Single, large, ileal Single, large, ileal Single, anastomotic

Multiple, small, ileal & colic Multiple, small, ileal & colic

Type & site of ulcer Resection Sulphasalazine & iron Resection – – Resection Resection – Resection Resection Resection Lactose-free diet Cholestyramine & iron Resection Ranitidine

Treatment

Recurrence 4 m Asymptomatic 2 y

Well 18 m – – Recurrence 4 m Well 7 m – Recurrence 3 m Recurrence 8 y Well 7 m Well 3 y

Well 18 m “Symptoms relieved”

Outcome

264

M. D. Stringer

resection but should be excluded in a child with chronic diarrhea, malabsorption or failure to thrive. Infants undergoing extensive ileal resection survive with few developmental scars,79 although some authors have suggested a more guarded prognosis.80 These children may be shorter and lighter than their siblings.80 Assessment of these patients is complicated by differences in maternal and intrauterine environments, prematurity, variability of neonatal nutrition, and complications other than those related to the gut. Until recently, only extensive small bowel resection in infants was considered to be associated with significant long-term sequelae. Davies et al. (1999) studied the longterm nutritional and metabolic effects of limited ileal resection (13 months following splenectomy, suggesting that splenectomy may provide durable remission in selected patients with refractory cytopenias or symptoms related to splenomegaly, and that some patients may experience a prolonged disease stabilization after splenectomy.56 It should be noted, however, that correction of pancytopenias is somewhat dependent on the reserve of the bone marrow, which may have been heavily treated with chemotherapy or radiation therapy. This is difficult to

predict prior to splenectomy, however, and the only means of assessing adequate reserve is a favorable response following splenectomy.

Gaucher’s disease Gaucher’s disease is a familial disorder in which there is a deficiency of the enzyme -glucosidase, leading to an abnormal storage or retention of glycolipid cerebrosides within reticuloendothelial cells, thereby resulting in significant splenomegaly. Splenectomy had historically been the procedure of choice for these patients, but the concern for postsplenectomy sepsis led many to advocate partial splenectomy for this condition.57–59 However, partial splenectomy has been less successful for this condition compared to more common congenital anemias, and some have hypothesized that the splenic remnant tends to enlarge with Gaucher’s disease due to tissue regeneration and/or continued deposition of the glucocerebrosides in the reticuloendothelial system.57,59 Some suggest that by leaving a splenic remnant for continued deposition of lipid, the bone marrow is spared from deposition and therefore protected, offering an additional benefit to partial splenectomy in these patients.5 With alglucerase enzyme therapy now widely instituted, rates of splenomegaly have decreased ameliorating much of the need for splenectomy in this disease.60,61

Overwhelming postsplenectomy sepsis Overwhelming postsplenectomy infection (OPSI) is a fulminate, life-threatening complication occurring after anatomic splenectomy or functional asplenia. King and Shumaker first reported the association between splenectomy and fulminate, lethal sepsis in their 1952 report of five splenectomized infants who dramatically succumbed to infection.3 Since that report, awareness of the association between splenectomy and risk of infection has led to improved prophylactic measures, better education of patients and families, and aggressive treatment. Surgical removal of the spleen clearly results in reduced clearance of extracellular and intracellular antigens, diminished response to new polysaccharide antigens, and impaired phagocytosis of unopsonized and opsonized bacteria.62 The most common organism is Streptococcus pneumonia, accounting for 50%–90% of cases and 60% of the mortality associated with OPSI, but other encapsulated organisms such as Haemophilus influenzae and Neisseria meningitidis are also significant

Splenectomy

pathogens.1,62–67 Less common infectious agents causing OPSI include Escherichia coli, Pseudomonas aeruginosa, capnocytophaga canimorsus (formerly called DF-2), group B streptococci, Enterococcus spp., Bacteroides spp., Bartonella, and others.67 Despite the widespread concern for OPSI, the true overall incidence of serious infections following splenectomy is somewhat difficult to ascertain, although the risk of fulminate OPSI has been reported to be between 0.1% and 8.5%.1 A large meta-analysis of postsplenectomy sepsis series that included 19 680 cases reported a serious infection rate of 3.2% and an overall mortality of 1.4%.67 The incidence of infection among children and adults was found to be similar, although the incidence of deaths was significantly higher among children (1.7%) than adults (1.3%) (P < 0.001). Other studies have supported these numbers, reporting overall risks of OPSI of 0.13% to 8.1% in children and 0.28% to 1.9% in adults.1,63–66 The majority of OPSI presents within the initial 2 years following splenectomy, with 50%–70% of cases occurring during that time period.1 However, splenectomized patients are at lifelong risk, as postsplenectomy sepsis has been reported more than 40 years after surgery.1 Overall lifetime risk of OPSI has been estimated to be 5%,62 with a mean time interval between splenectomy and infection of 22.6 months.67 The incidence of OPSI is dependent on the underlying disease process, and several studies have stratified the risk of OPSI by the disease state for which splenectomy was performed.1,67 Infection risk is highest among patients with thalassemia major (8.2%) and sickle cell anemia (7.3%) who exhibit alterations in complement activation, abnormal levels of immunoglobulin, and impaired phagocytic function placing them at increased risk for severe infection and death following splenectomy.1,67 The lowest infection rates are observed among patients with ITP (2.1%) and trauma (2.3%).67 The clinical manifestations of OPSI begin with a prodrome of fever and mild, non-specific symptoms including chills, malaise, myalgias, headache, nausea, vomiting, diarrhea, and abdominal pain, rapidly evolving into overwhelming septic shock. Although the onset may be insidious and non-specific, rapid deterioration remains the hallmark of OPSI with hypotension, respiratory distress, disseminated intravascular coagulation, coma, and death occurring within hours of presentation. A high index of suspicion with any febrile illness leading to rapid diagnosis and aggressive treatment is critical in preventing decompensation and death in splenectomized patients. Diagnostic work-up should not delay the initiation of empiric antibi-

otic therapy. Initial diagnostic work-up should include a peripheral blood smear evaluation for the presence of bacteria. Visualization of organisms on Gram stain of the buffy coat suggests a quantitative bacteremia of >106 organisms/ml. Blood cultures should always be collected and are generally positive within 24 hours, aiding in the identification of the pathogen and guiding subsequent antibiotic therapy. Additional studies should be obtained as appropriate, in efforts to identify all possible sources of infection. The critical point in management of these patients is early recognition followed by aggressive intervention. All asplenic patients with fever of unidentifiable source should be treated as medical emergencies. Initial empiric therapy should always include a broad-spectrum antibiotic with activity against S. pneumonia, H. influenzae, and N. meningitidis. Intravenous penicillin has been the cornerstone of antibiotic therapy due to its excellent activity against pneumococci and meningococci, but local resistance patterns must be taken into account, and inclusion of an antibiotic active against penicillin-resistant pneumococci and betalactamase-resistant organisms, such as ceftriaxone, should be considered.1,68 In areas with concern for highly resistant pneumococci, vancomycin and rifampicin should be considered as empiric therapy.1,68

Prevention of OPSI In general, risk reduction for OPSI in asplenic patients is multifaceted, including immunization, the use of prophylactic antibiotic regimens, education of the patient and family, and prompt recognition of early signs and symptoms of illness with institution of aggressive treatment. Although general clinical guidelines for the use of these risk reduction strategies exist, complete agreement on all aspects of risk reduction has not been reached. No prophylactic regimen will completely eliminate the risk of sepsis in asplenic patients, but with the institution of adequate vaccinations, the use of prophylactic antibiotics when appropriate, and the education of patients and families the morbidity and mortality of severe infection may be significantly reduced. Immunoprophylaxis The major approach to the prevention of severe infections in the asplenic host is vaccination against the most frequent causative agents. Highly immunogenic capsular vaccines are available for the three major organisms responsible for the morbidity and mortality of OPSI: S. pneumoniae (23 most common serotypes accounting for 85% of infections),

441

442

J. H. Aldrint and H. E. Rice

H. influenzae (type B), and N. meningiditis (groups A, C, Y, and W135). In children less than two years of age, a pneumococcal 7-valent conjugate vaccine should be used.68 All pathologic strains are not included in the vaccines, however, so infection may occur despite immunization. Whenever possible, it is recommended for vaccinations to be given two weeks prior to elective splenectomy, in order to optimize antibody response.69,70 In cases of emergent splenectomy, patients should be vaccinated prior to discharge from the hospital to ensure that they receive the vaccine, as many of these patients become lost to follow-up.36 Reimmunization with pneumococcal and meningococcal vaccines is recommended at 5–10 year intervals, and after 2 years, respectively.68 The need for reimmunization with the H. influenzae vaccine is not clear. The influenza vaccine is recommended annually for immunocompromised patients and may be valuable to asplenic patients by reducing the risk of secondary bacterial infection.68 Vaccination does not always result in protective levels of antibodies. Konradsen et al. evaluated pneumococcal and Haemophilus influenzae type B antibody levels following vaccination in splenectomized patients, and reported that only 52% of patients who had been properly vaccinated had protective levels of pneumococcal antibodies, and that only 60% of patients had Hib antibody levels high enough to confer long-term protection.71 Pneumococcal vaccination without antibiotic prophylaxis does not provide full protection against late sepsis, as up to 20–40% of the septicemias may be caused by bacteria other than pneumococci, emphasizing the need for additional prophylactic measures in splenectomized patients.72

Chemoprophylaxis Because vaccinations do not completely protect against infection, many authorities have recommended antibiotic prophylaxis for asplenic patients, especially for the first two years following splenectomy.1,68–70 Although the value of prophylactic antibiotics has been demonstrated in a clinical trial setting only for sickle cell disease, most investigators advocate their use in all children who have undergone splenectomy regardless of the indication until 5 years of age, and many propose continuation of antibiotic chemoprophylaxis until adulthood.68,70–73 Current guidelines for chemoprophylaxis include 125 mg twice-daily oral penicillin VK until age 3, and 250 mg twice daily thereafter.66 Problems with the use of orally administered penicillin in children have included poor compliance, the risk of selecting antibiotic-resistant strains, and providing the patient with a false sense of security should a febrile illness develop.70

Education Patient education in the prevention and early management of serious infection has a great potential to decrease the morbidity and mortality in OPSI. Studies have documented that from 11% to 50% of postsplenectomy patients remain unaware of the increased risk for serious infection, and the necessity for preventative strategies.74–79 Deodhar et al. found that in their series of asplenic patients, less than half had received pneumococcal immunization, were currently on antibiotic prophylaxis, or had received education and instructions regarding their increased risk for serious infection.69 Patients and families must be clearly informed about the risk of rapidly progressing, life-threatening infections, that the risk of infection is higher in the first 2 years after splenectomy but is still present life long, and about the need to inform doctors of splenectomizied status.70 It must be emphasized that despite the use of antibiotic prophylaxis and vaccination against S. pneumoniae, H. influenzae type B, and N. meningiditis, splenectomy still places the patient at life long risk for serious infection. Despite all of these measures, the true benefits of these risk reduction strategies remain poorly defined. The vaccines for S. pneumoniae and N. meningiditis are not effective against all bacterial strains, antimicrobial resistance develops, and the use of prophylactic antibiotics may generate a false sense of security in patients, families, and physicians. Therefore, careful and thorough education of patients and families, a rigorous vaccination policy, knowledge of regional antibiotic resistance patterns from which to tailor prophylactic coverage, and early recognition and aggressive initiation of therapy with the onset of any febrile illness are each vitally important in reducing the incidence and mortality of overwhelming postsplenectomy sepsis in asplenic patients.

Conclusions Splenectomy remains a critical procedure for several hematologic and non-hematologic diseases, although an emphasis is now placed on splenic preservation whenever possible. The current surgical repertoire includes total and partial splenectomy, open and laparoscopic techniques, splenorraphy, and splenic embolization. When making decisions regarding which approach to use, it is imperative that patients and families are aware of the risks and benefits of each procedure, as well as the long-term consequences of splenectomy. The major long-term complication following splenectomy is overwhelming sepsis, but through awareness, education, and proper use of immunoprophylaxis

Splenectomy

and chemoprophylaxis, the risk of this devastating outcome may be minimized.

REFERENCES 1. Lynch, A. M. & Kapila, R. Overwhelming postsplenectomy infection. Infect. Dis. Clin. N. Am. 1996; 10:693–707. 2. Morris, D. H. & Bullock, F. D. The importance of the spleen in resistance to infection. Ann. Surg. 1919; 50:513. 3. King, H. & Shumacker, H. B. Splenic studies. I. Susceptibility to infection after splenectomy performed in infancy. Ann. Surg. 1952; 136:239. 4. Sadler, T. W. Langman’s Medical Embryology, 7th edn. Baltimore: Williams and Wilkins, 1995: 245–250. 5. Fraker, D. L. Spleen. In Greenfield, L. J. ed. Surgery Scientific Principles and Practice, ed. 3rd edn. Philadelphia, PA: Lippincott, Williams and Wilkins, 2001; 1236–1257. 6. Cullis, J. O. & Mufti, G. J. Splenectomy. In Stringer, M. D. Oldham, K. T. Mouriquand, P. D. E. Howard, E. R. eds. Pediatric Surgery and Urology: Long Term Outcomes, London: W. B. Saunders, 1998; 394–401. 7. Schwartz, S. I. The Spleen. In Cameron, J. L. Current Surgical Therapy, 1998; 545–556. St. Louis: Mosby, Inc. 8. Carroll, A. & Thomas, P. Decision-making in surgery: Splenectomy. Br. J. Hosp. Med. 1995; 54:147. 9. Uranus, S., Pfeifer, J., Schauer, C. et al. Laparoscopic partial splenic resection. Surg. Laparosc. Endosc. 1995; 5:133. 10. Farhi, D. C. & Ashfaq, R. Splenic pathology after traumatic injury. Am. J. Clin. Pathol. 1996; 105:474. 11. Petroianu, A. Subtotal splenectomy for treatment of patients with myelofibrosis and myeloid metaplasia. Int. Surg. 1996; 81:177. 12. Beanes, S., Emil, S., Kosi, M. et al. A comparison of laparoscopic versus open splenectomy in children. Am. Surg. 1995; 61:908– 910. 13. Berman, R. S., Yahanda, A. M., Mansfield, P. F. et al. Laparoscopic splenectomy in patients with hematologic malignancies. Am. J. Surg. 1999; 178:530–536. 14. Cogliandolo, A., Berland-Dai, B., Pidoto, R. R. et al. Results of laparoscopic and open splenectomy for nontraumatic diseases. Surg. Laparosc. Endosc. Percutan. Tech. 2001; 11:256– 261. 15. Winslow, E. R. & Brunt, L. M. Perioperative outcomes of laparoscopic versus open splenectomy: a meta-analysis with an emphasis on complications. Surgery 2003; 134:647–653. 16. Yoshida, K., Yamazaki, Y., Mizuno, R. et al. Laparoscopic splenectomy in children. Preliminary results and comparison with the open technique. Surg. Endosc. 1995; 9:1279–1282. 17. Curran, T. J., Foley, M. I., Seanstrom, L. L., & Campbell, T. J. Laparoscopy improves outcomes for pediatric splenectomy. J. Pediatr. Surg. 1998; 33:1498–1500. 18. Watson, D. I., Coventry, B. J., Chin, B. J. et al. Laparoscopic versus open splenectomy for immune thrombocytopenic purpura. Surgery 1997; 121:18–22.

19. Lozano-Salazar, R. R., Herrera, M. F., Vargas-Vorackova, F., & Lopez-Karpovitch, X. Laparoscopic versus open splenectomy for immune thrombocytopenic purpura. Am. J. Surg. 1998; 176:366–369. 20. Knauer, E. M., Ailawadi, G., Yahanda, A. et al. 101 laparoscopic splenectomies for the treatment of benign and malignant hematologic disorders. Am. J. Surg. 2003; 186:500–504. 21. Tchernia, G., Gauthier, F., Mielot, F. et al. Initial assessment of the beneficial effect of partial splenectomy in hereditary spherocytosis. Blood 1993; 81:2014–2020. 22. Tchernia, G., Bader-Meunier, B., Berterottiere, P. et al. Effectiveness of partial splenectomy in hereditary spherocytosis. Curr. Opin. Hematol. 1997; 4:136–141. 23. Bader-Meunier, B., Gauthier, F., Archambaud, F. et al. Longterm evaluation of the beneficial effect of subtotal splenectomy for the management of hereditary spherocytosis. Blood 2001; 97:399–403. 24. Guzzetta, P. C., Ruley, E. J., Merrick, H. F. W. et al. Elective subtotal splenectomy – indications and results in 33 patients. Ann. Surg. 1990; 211:34–42. 25. Svarch, E., Vilorio, P., Nordet, I. et al. Partial splenectomy in children with sickle cell disease and repeated episodes of splenic sequestration. Hemoglobin 1996; 20:393–400. 26. Rice, H. E., Oldham, K. T., Hillery, C. A. et al. Clinical and hematologic benefits of partial splenectomy for congenital hemolytic anemias in children. Ann. Surg. 2003; 237:281–288. 27. Brandt, C. T., Rothbarth, L. J., Kumpe, D., Karrer, F. M., & Lilly, J. R. Splenic embolization in children: long-term efficacy. J. Pediatr. Surg. 1989; 24:642–644. 28. Petersons, A., Volrats, O., & Bernsteins, A. The first experience with non-operative treatment of hypersplenism in children with portal hypertension. Eur. J. Pediatr. Surg. 2002; 12:299– 303. 29. Tajiri, T. Long-term hematological and biochemical effects of partial splenic embolization in hepatic cirrhosis. Hepatogastroenterology 2002; 49:1445–1448. 30. Nio, M., Hayashi, Y., Sano, N., Ishii, T., Sasaki, H., & Ohi R. Longterm efficacy of partial splenic embolization in children. J. Pediatr. Surg. 2003; 38:1760–1762. 31. Shanmuganathan, K., Mirvis, S. E., Boyd-Kranis, R., Takada, T., & Scales, T. M. Nonsurgical management of blunt splenic injury: use of CT criteria to select patients for splenic arteriography and potential endovascular therapy. Radiology 2000; 217:75–82. 32. Sclafani, S. J., Weisberg, A., & Scalea, T. M. Blunt splenic injuries: nonsurgical treatment with CT, arteriography, and transcatheter arterial embolization of the splenic artery. Radiology 1991; 181:189. 33. Sclafani, S. J., Shaftan, G. W., Scalea, T. M. et al. Nonoperative salvage of computed tomography – diagnosed splenic injuries: utilization of angiography for triage and embolization for hemostasis. J. Trauma 1995; 39:818–827. 34. Delaunay, J. Genetic disorders of the red cell membrane. Crit. Rev. Oncol. Hematol. 1995; 19:79–110. 35. Sackey, K. Hemolytic anemia: part I. Pediatr. Rev. 1999; 20:152– 158.

443

444

J. H. Aldrint and H. E. Rice

36. Rice, H. E. Pediatric spleen surgery. In Oldham, K. T., Colombani, P. M., Foglia, R. P., Skinner, M. A., eds. Surgery of Infants and Children, Philadelphia: Lippincott Williams & Wilkins, 2005; 1511–1522. 37. Dover, G. & Valle, D. Therapy for -thalassemia – a paradigm for the treatment of genetic disorders. N. Engl. J. Med. 1994; 331:609–610. 38. Gaziev J. & Lucarelli G. Stem cell transplantation for hemoglobinopathies. Curr. Opin. Pediatr. 2003; 15:42–31. 39. Forget B. G. Thalassemia syndromes. In Hematology: Basic Principles and Practice, ed. R. Hoffman, 3rd edn. pp.488–499. Oxford: Churchill Livingstone, 2000. 40. Hoe, T. S., Lammi, A., & Webster, B. Homozygous betathalassemia: a review of patients who had splenectomy at the Royal Alexandra Hospital for Children, Sydney. Singapore Med. J. 1994; 35:59–61. 41. Idowu, O. & Jordan-Hayes, A. Partial splenectomy in children under 4 years of age with hemoglobinopathy. J. Pediatr. Surg. 1998; 33:1251–1253. 42. deMontalembert, M., Girot, R., Revillon, Y. et al. Partial splenectomy in homozygous beta thalassemia. Arch. Dis. Child. 1990; 65:304–307. 43. Politis, C., Spigos, D. G., Georgiopoulou, P. et al. Partial splenic embolization for hypersplenism of thalassemia major: fiveyear follow-up. Br. Med. J. 1987; 294:665–667. 44. Pinca, A., Di Palma, A., Soriani, S. et al. Effectiveness of partial splenic embolization as treatment of hypersplenism in thalassemia major: a 7-year follow-up. Eur. J. Haematol. 1992; 49:49–52. 45. Al-Salem, A. H., Naserullah, Z., Qaisaruddin, S. et al. Splenic complications of the sickling syndromes and the role of splenctomy. J. Pediatr. Hematol. Oncol. 1999; 21:401–406. 46. Aquino, V. M., Norvell, J. M., Buchanan, G. R. Acute splenic complications in children with sickle cell-hemoglobin C disease. J. Pediatr. 1997; 130:961–965. 47. Owusu-Ofori, S. & Riddington, C. Splenectomy versus conservative management for acute sequestration crises in people with sickling cell disease. Cochrane Database Syst. Rev. 2002; 4:CD003425. 48. Kar, B. C. Splenectomy in sickle cell disease. J. Assoc. Phys. India 1999; 47:890–893. 49. Svarch, E., Nordet, I., Valdes, J. et al. Partial splenectomy in children with sickle cell disease. Haematologica 2003; 88:222– 223. 50. DiPaola, J. A. & Buchanan, G. R. Immune thrombocytopenic purpura. Pediatr. Clin. North. Am. 2002; 49:911–928. 51. Mandeiros, D. & Buchanan, G. R. Idiopathic thrombocytopenic purpura: beyond consensus. Curr. Opin. Pediatr. 2000; 12:4–9. 52. Tura, S., Fiacchini, M., Zinzani, P. L., Brusamolino, E., & Gobbi, P. G. Splenectomy and the increasing risk of secondary acute leukemia in Hodgkin’s disease. J. Clin. Oncol. 1993; 11:925–930. 53. Dietrich, P. Y., Henry-Amar, M., Cosset, J. M., Bodis, S., Bosq, J., & Hayat, M. Secondary primary cancers in patients continuously disease-free from Hodgkin’s disease: a protective role for the spleen? Blood 1994; 84:1209–1215.

54. Brodsky, J., Abcar, A., & Styler, M. Splenectomy for nonHodgkin’s lymphoma. Am. J. Clin. Oncol. 1996; 19:558–561. 55. Morel, P., Dupriez, B., Gosselin, B. et al. Role of early splenectomy in malignant lymphomas with prominent splenic involvement (primary lymphomas of the spleen). A study of 59 cases. Cancer 1993; 71:207–215. 56. Yoong, Y., Kurtin, P. J., Allmer, C., Geyer, S. et al. Efficacy of splenectomy for patients with mantle cell non-Hodgkin’s lymphoma. Leuk Lymphoma 2001; 42:1235–1241. 57. Rubin, M., Yampolski, I., Lambrozo, R. et al. Partial splenectomy in Gaucher’s disease. J. Pediatr. Surg. 1986; 21:125–128. 58. Fleshnew, P. R., Aufses, A. H., Grabowski, G. A. et al. A 27-year experience with splenectomy for Gaucher’s disease. Am. J. Surg. 1991; 161:69–75. 59. Zer, M. & Freud, E. Subtotal splenectomy in Gaucher’s disease: towards a definition of critical splenic mass. Br. J. Surg. 1993; 80:399. 60. Freud, E., Cohen, I. J., Mor, C. et al. Splenic “regeneration” after partial splenectomy for Gaucher disease: histological features. Blood Cells, Molec. Dis. 1998; 24:309–316. 61. Dweck, A., Abrahamov, A., Hadas-Halpern, I. et al. Type I Gaucher disease in children with and without enzyme therapy. Pediatr. Hematol. Oncol. 2002; 19:389–397. 62. Weinreb, N. J., Charrow, J., Andersson, H. C. et al. Effectiveness of enzyme replacement therapy in 1028 patients with type I Gaucher disease after 2 to 5 years of treatment: a report from the Gaucher registry. Am. J. Med. 2002; 113:112–119. 63. Davidson, R. N. & Wall, R. A. Prevention and management of infections in patients without a spleen. Clin. Microbiol. Infect. 2001; 7:657–660. 64. Horan, M. & Colebatch, J. H. Relation between splenectomy and subsequent infection: a clinical study. Arch. Dis. Child. 1962; 37:398. 65. Ellis, E. F. & Smith, R. T. The role of the spleen in immunity with special references to the post-splenectomy problems in infants. Pediatrics 1966; 37:111. 66. Ellison, E. C. & Fabri, P. J. Complications of splenectomy. Surg. Clin. North. Am. 1983; 63:1313. 67. Holdsworth, R. J., Irving, A. D., & Cuschieri, A. Postsplenectomy sepsis and its mortality rate: actual versus perceived risks. Br. J. Surg. 1991; 78:1031. 68. Bisharat, N., Omari, H., Lavi, I., & Raz, R. Risk of infection and death among post-splenectomy patients. J. Infect. 2001; 43:183–186. 69. Sumaraju, V., Smith, L. G., & Smith, S. M. Infectious complications in asplenic hosts. Infect. Dis. North Am. 2001; 15(2) 551– 565. 70. Deodhar, H. A., Marshall, R. J., & Barnes, J. N. Increased risk of sepsis after splenectomy. Br. Med. J. 1993; 307:1408. 71. Castagnola, E. & Fioredda, F. Prevention of life-threatening infections due to encapsulated bacteria in children with hyposplenia or asplenia: a brief review of current recommendations for practical purposes. Eur. J. Haematol. 2003; 71:319–326. 72. Konradsen, H. B., Rasmussen, C., Ejstrud, P., & Hansen, J. B. Antibody levels against Streptococcus pneumoniae and

Splenectomy

Haemophilus influenzae type B in a population of splenectomized individuals with varying vaccination status. Epidemiol. Infect. 1997; 119:167–174. 73. Eber, S. W., Langendorfer, C. M., Ditzig, M. et al. Frequency of very late fatal sepsis after splenectomy for hereditary spherocytosis: impact of insufficient antibody response to pneumococcal infection. Ann. Hematol. 1999; 78:524–528. 74. Gaston, M. H., Verter, J. I., Woods, G. et al. Prophylaxis with oral penicillin in children with sickle cell anemia: a randomized trial. N. Eng. J. Med. 1986; 314:1593–1599. 75. White, K. S., Covington, D., Churchill, P. et al. Patient awareness of health precautions after splenectomy. Am. J. Infect. Contr. 1991; 19:36–41.

76. Kinnersley, P., Wilkinson, C. E., & Srinivasan, J. Pneumococcal vaccination after splenectomy: Survey of hospital and primary care records. Br. Med. J. 1993; 307:1398–1399. 77. Rasmussen, C., Ejstrud, P., Hansen, J. B., & Konradsen, H. B. Asplenic patients’ knowledge of prophylactic measures against severe infections (brief report). Clin. Infect. Dis. 1997; 25 (3):738. 78. Wright, J. G., Hambelton, I. R., Thomas, P. W. et al. Postsplenectomy course in homozygous sickle cell disease. J. Pediatr. 1991; 134:304–309. 79. Brigden, M. L. & Pattullo, A. L. Prevention and management of overwhelming postsplenectomy infection: an update. Crit. Care Med. 1999; 27:836–842.

445

35 Biliary atresia Edward R. Howard King’s College Hospital, London, UK

Introduction Biliary atresia presents in the neonatal period, occurs with a frequency of between 1 in 8000 and 1 in 16 000 live births and accounts for more than 50% of pediatric liver transplantation.1 The cause of the disease remains obscure but current evidence suggests that there is more than one etiologic factor. Both the intra- and extrahepatic bile ducts are affected and affected infants present with jaundice and pale stools within the first few weeks of life. The intrahepatic pathology, which has been likened to sclerosing cholangitis,2 is accompanied by an inflammatory sclerosing lesion of the extrahepatic bile ducts, which results in obstruction of the lumen and, in some cases, complete disappearance of segments of the biliary tract. Death from cirrhotic liver failure occurs within 2 years in untreated cases and biliary atresia represents the most frequent reason for liver transplantation in childhood.

Historical issues Thomson3 published the first major review in 1892. He collected 49 cases from the literature and added a further case of his own. He recorded that “the children themselves are either jaundiced at birth, or they become so within the first week or so of life; otherwise they are healthy and well nourished.” He concluded that, whatever the etiology of the condition, it was characterized by a progressive destructive inflammatory lesion of the biliary tract. A majority of the affected infants he described died from complications of spontaneous hemorrhage during the first few months of life and the increased risk of bleeding was ascribed to

an “impoverishment of the blood”. He suggested that the inflammatory component of biliary atresia was secondary to a congenital abnormality of the ducts, which interferes with bile flow and that the inflammation is progressive and can result in a variable pattern of duct occlusion. He also suggested that the obliteration of the ducts becomes complete at an early, but variable period during intrauterine life. Holmes, in 1916,4 first suggested that some cases of biliary atresia might be surgically correctable. He reviewed more than 100 case reports and found evidence of residual segments of bile ducts in 16%. Diagnostic confusion is evident in some of the early case reports. Ladd,5 for example, reported surgical correction in eight out of 20 cases in 1928 but it appears that two were examples of choledochal cysts and two were patients with neonatal hepatitis and hypoplastic bile ducts rather than biliary atresia. Historically, cases of biliary atresia were described as “correctable” or “non-correctable” depending on the presence or absence of residual patent segments of proximal bile ducts. Successful surgical treatment was uncommon; Hays and Snyder reported in 19636 that less than 5% of children with biliary atresia survived beyond early childhood. A variety of ineffective surgical procedures were tried, including partial hepatectomy, the insertion of a variety of tubes into the hepatic parenchyma and the drainage of lymph into the gastrointestinal tract via an anastomosis between the thoracic duct and the esophagus. More than 85% of biliary atresia is of the “noncorrectable” type in which there is occlusion of the hepatic ducts at the porta hepatis. In 1959, radical resection of the extrahepatic biliary tree was proposed after histologic studies had revealed that microscopic ductular structures

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


446

Biliary atresia

Table 35.1. Chromosomal abnormalities associated with biliary atresia r r r r

Trisomy 18 Trisomy 18 Trisomy 17–18 Trisomy 21

– – – –

Ikeda et al.10 Buffa et al.11 Alpert et al.12 Danks13

were present in tissue resected from the porta hepatis. The ductules, which measured up to 300 µm in diameter, were shown to communicate with intrahepatic ducts. Successful drainage of bile was observed after resection of all residual biliary tissue in the porta hepatis and anastomosis of the resected area to a Roux loop of bowel allowed satisfactory transit of bile into the gut (the portoenterostomy operation).7,8 The operation was shown to be most effective when performed before eight weeks of age. An increasing number of portoenterostomy patients have since survived into adult life. The chance of long-term survival was further increased by the development of liver transplantation. Failure to achieve satisfactory early bile flow after portoenterostomy, or progression to cirrhotic liver failure later in life, are now recognized indications for transplantation.

Studies of etiology The etiology of biliary atresia remains unknown, but current evidence suggests that there may be a genetic defect in morphogenesis in approximately 20% of cases. Hypotheses for the remaining 80% have included viral, autoimmune and immune-mediated pathogeneses as well as damage to the biliary tract by toxic monohydroxy bile acids.9

Embryologic and genetic factors There are a few reports of an association with chromosomal abnormalities which have included trisomy 17, 18, and 21 (Table 35.1). Twin studies, however, have not suggested any significant genetic transmission with reports of only two pairs of offspring concordant for biliary atresia,14,15 compared with 21 discordant pairs.16 Reports of a familial incidence of biliary atresia are also unusual and have included two affected siblings in each of ten families, and three affected siblings in each of four families.15 A study of human leukocyte antigens (HLA) in 55 children was reported to show a significantly higher frequency of HLA-B12, A9-B5 and A28-B35, and of their disequilib-

rium values compared with controls.17 It was concluded that immunogenetic factors may have a role in determining susceptibility to the disease. However, the later studies of Jurado et al.18 and Donaldson et al.19 failed to identify any relationship between biliary atresia and HLA status. Desmet and Callea20 observed that the majority of intrahepatic bile ducts affected by epithelial damage and destruction in biliary atresia have a mature tubular shape. However, 20%–25% of cases show features reminiscent of “ductal plate malformation” which is represented by partial or complete persistence of the original embryonic form of the intrahepatic ducts. Human embryologic studies have shown a striking similarity between the appearances of normal developing bile ducts during the first trimester of pregnancy and the abnormal ductules observed within the residual tissue resected from the porta hepatis during portoenterostomy.21 These studies demonstrate that at 6 weeks’ gestation the biliary epithelium of the ductal plate arises from hepatocyte precursor cells at the hepatic–mesenchymal margin adjacent to the portal vein. At 11 weeks’ gestation the common hepatic duct, which arises from the hepatic diverticulum of the foregut, is in communication with the lumen of the ductal plate. Intrahepatic ducts develop towards the periphery of the liver at around 12 weeks and bile first appears at this time. Comparison of the embryologic features at 12 weeks with sections of tissue taken from the porta hepatis of infants with biliary atresia reveals similar histologic features. The histologic appearances in biliary atresia might result from a disruption in the process of remodelling of the ductal plate during this phase of development together with an inflammatory reaction secondary to bile leakage from abnormal ducts. Evidence for an early gestational origin of biliary atresia has been provided by antenatal ultrasound diagnosis on at least five occasions.22 Hinds et al.23 reported that nine of 194 cases of atresia had been detected on routine antenatal scanning. At surgery six had type 3 disease, two had type 2, and one had type 1. All nine cases underwent successful portoenterostomy. Furthermore, estimations of gammaglutamyl transpeptidase (GGT) activity in amniotic fluid samples taken at different points in gestation from 10 000 pregnant women showed low levels of the enzyme between 18 and 20 weeks in three cases. The infants of these three women were found to have biliary atresia.24 Further support for the relationship of low amniotic GGT levels and biliary atresia was reported by Ben-Ami et al.25 GGT is synthesized in the liver, reaches the gut in bile, and passes into the amniotic fluid during defecation. The observation of low levels of the enzyme in amniotic fluid is consistent with an onset of biliary atresia early in gestation.

447

448

E. R. Howard

Infective factors Although there were early reports of a possible relationship between biliary atresia and infection with cytomegalovirus (CMV),26 rubella27 and Epstein–Barr virus (EBV),28 the evidence remains confusing. This confusion is illustrated by studies of human papillomavirus (HPV). Drut et al.29 reported evidence of HPV infection in liver biopsies from 16 out of 18 cases of BA compared with no evidence of infection in 30 control cases. In contrast, Domiati-Saad et al.30 showed no association with HPV infection in 19 cases of neonatal hepatitis or BA. These children were also negative for EBV, although three cases were positive for CMV. Jevon and Dimmick31 also examined bile duct tissue from 12 cases of BA for evidence of CMV infection but all were negative. Most experimental work on infection has related to the hepatotropic RNA viruses, reovirus and rotavirus. Weanling mice exposed to reo 3 virus develop chronic liver disease and obstructive jaundice32 and histologic examination of the porta hepatis shows changes with similarities to biliary atresia.33 Morecki et al.34 reported reovirus antigens in 68% of a group of infants with biliary atresia compared with 8% of age-matched controls, but other studies failed to confirm the results.35,36 There is also conflicting evidence for the role of reovirus infection from studies using reverse transcriptase polymerase chain reaction (RT-PCR) techniques. Tyler et al.37 detected reovirus RNA in hepatic and/or biliary tissues from 55% of 23 infants with biliary atresia compared with 12%–21% of control cases but these findings were at variance with a previous study.38 The case for a viral etiology in biliary atresia remains unproven.

Table 35.2. Biliary atresia and associated abnormalities in 251 patients (Carmi et al.39 ) 30 (59%) 15 (29%) 6 (12%) Total

Isolated abnormalities (cardiac, renal, etc.) Multiple abnormalities (polysplenia, preduodenal portal vein, situs inversus, etc.) Situs inversus alone 51 (20%)

investigation of 85 cholestatic infants in Sweden, showed that the mothers of biliary atresia patients had a higher mean age and were more commonly treated for gestational diabetes than mothers of patients with other causes of jaundice. Davenport et al.41 suggested that the BASM syndrome might be a subgroup with a separate etiology from the more common cases of biliary atresia without extrahepatic abnormalities and that the insult to the bile duct might occur early in gestation at a time coinciding with the development of the spleen and gut rotation. Organ asymmetry in the embryo starts to occur before 30 days’ gestation and would be disturbed by suppression of left- or right-sided development. Deletion of the inversin (INV) gene in transgenic mice43 results in situs inversus and atretic bile ducts. It was suggested that this might imply a role for the inversin gene in normal human bile duct development and that such a gene mutation might support a genetic basis of the BASM syndrome. However, a study of 65 patients with laterality defects did not reveal any mutation of the human inversin gene and the authors concluded that “the absence of mutation in a series of 7 cases with lateralization defects and biliary anomalies (BASM), demonstrates that INV is not frequently involved in such a phenotype in humans.”44

Associated abnormalities Presentation and immediate management A review of ten major series of biliary atresia showed that an average of 21% had additional abnormalities outside of the hepatobiliary system.16 The most frequent abnormalities affected the cardiovascular system, the gut (situs inversus) and the spleen (Table 35.2).39 A variety of splenic malformations have been described which include the polysplenia syndrome,40 double spleen, and asplenia.41 The polysplenia syndrome includes multiple spleens, situs inversus, preduodenal portal vein and malformations of the inferior vena cava. Davenport et al.41 reported a 7.5% incidence of the syndrome in 308 cases of atresia and proposed the term “biliary atresia splenic malformation” (BASM). It was noted that four infants with BASM were born to mothers with diabetes mellitus. Fischler et al.,42 in an

Physiologic jaundice (unconjugated hyperbilirubinemia) occurs in up to 90% of healthy neonates and is defined as a temporary inefficient hepatic excretion of bilirubin. Conjugated hyperbilirubinemia (hepatitis syndrome) is always pathological and may be caused by hepatocellular dysfunction secondary to infective, metabolic, drug, and endocrine disorders. The clinical presentation of structural defects such as biliary atresia is similar to the presentation of hepatocellular disease in the young infant and because of the urgent need for surgical correction of atresia the hepatocellular causes of conjugated hyperbilirubinemia such as alpha-1-antitrypsin deficiency, cystic fibrosis and endocrine disorders must be rapidly excluded.45

Biliary atresia

As well as conjugated hyperbilirubinemia and nonpigmented, white stools in biliary atresia patients there may be some hepatomegaly and, in a minority of cases, a range of associated abnormalities that can include cardiac defects, situs inversus and renal problems. Splenomegaly becomes more obvious after the first eight to 12 weeks of life and poor weight gain and growth retardation also occur with increasing age. It is advised that all infants who remain jaundiced after 14 days of age should have urinalysis for bilirubin and measurement of total and direct bilirubin in serum. Physiologic jaundice has usually disappeared by this time except in a small number of breast fed infants. If conjugated bilirubin is present and the stools are non-pigmented, investigations for biliary atresia should proceed immediately. It should be emphasized that most infants with biliary atresia are entirely well during the first 4 to 8 weeks of life apart from moderate jaundice. The diagnosis will only be suspected by the demonstration of conjugated hyperbilirubinemia and this must be investigated in an infant with yellow, rather than colourless, urine and pale stools.46

Investigations Early diagnosis is imperative for good results after portoenterostomy (see below) and infective, metabolic, endocrine, and drug-related causes of conjugated hyperbilirubinemia are excluded as quickly as possible so that surgery may be performed within the first 8 to 10 weeks of life. In early infancy the clinical features and biochemical tests of liver function are very similar in hepatocellular disease, hypoplasia of the bile ducts, obstruction of the extrahepatic ducts caused by choledochal cysts, and biliary atresia. Biochemical tests of liver function, such as the intracellular enzymes aspartate aminotransferase (AST), and alanine aminotransferase (ALT), give a measure of the severity of the liver disease but are of little diagnostic help. All four conditions present with jaundice, pale stools and dark urine and may be complicated by bleeding from malabsorption of vitamin K. Ultrasound examination is useful for assessing the structure of the gall bladder and the presence or absence of a choledochal cyst, inspissated bile, or gall stones. Intrahepatic bile duct dilatation is not a feature of biliary atresia and this limits the value of sonography, although an abnormally small or absent gall bladder may be suggestive of the diagnosis. Technetium labelled compounds such as methylbromoiminodiacetic acid (BrIDA), which are taken up by hepatocytes even in the presence of jaundice, are excreted into the biliary tract in high concen-

trations. No excretion of the isotope within 24 hours suggests a differential diagnosis of severe cholestasis or biliary atresia. The sensitivity of the investigation is increased by the prior administration of a 3-day course of phenobarbitone (5 mg/kg per day). Percutaneous biopsy of the liver is mandatory and although equivocal histologic findings are present in approximately 13% of cases,47 this failure rate may be improved by combining the biopsy with a laparoscopy and laparoscopic-guided cholangiography.48 Alternatively, endoscopic retrograde cholangiography (ERC) may be indicated in an infant who has shown no excretion of isotope from the liver but in whom the histology is doubtful. In an early series of nine cases of neonatal jaundice investigated in this way, normal bile ducts were demonstrated in four and atresia suggested in four.49 However, one of the latter cases proved to have patent bile ducts at surgery. One examination was a technical failure. Accurate results were therefore achieved in seven out of nine cases (78%). In a larger series, the diagnostic success rate with ERC was 86%.50 Iinuma et al.51 reported successful visualization of the biliary tract in 43 of 50 ERCPs without complication. Biliary atresia was diagnosed after successful cannulation of the ampulla and opacification of the pancreatic duct or a small residual segment of bile duct. Diagnoses of neonatal hepatitis, biliary hypoplasia or choledochal cyst were achieved after complete imaging of the biliary tract in 14 infants. Biliary atresia was later confirmed at surgery in six of the seven examinations classified as technical failures. A further advantage of ERCP is the ability to observe and aspirate the contents of the duodenum as the presence of bile excludes a diagnosis of atresia. A firm diagnosis of biliary atresia therefore requires a series of tests, all of which must be completed in a short period of time. A preoperative diagnosis is very important to the surgeon because the lumen of the biliary tract in an infant with either hepatocellular dysfunction or biliary hypoplasia is very small and failure to adequately visualize the bile ducts with intraoperative cholangiography could result in unnecessary surgery.52

Pathology and influence on long-term outcome The liver architecture in biliary atresia is preserved during the first few weeks of life but occlusion of the extrahepatic ducts leads to widening of the portal tracts with edema and increased amounts of fibrous tissue. Proliferation of bile ductules is accompanied by bile stasis within both canaliculi and hepatocytes. In contrast, typical findings in neonatal hepatitis include hepatocellular necrosis with

449

450

E. R. Howard

an inflammatory cell infiltrate in the hepatic parenchyma and multinucleate giant cell formation. However, these features are not pathognomonic and may be observed in some cases of biliary atresia, reducing the accuracy of liver biopsy to between 83% and 87%.47,53 Alpha-1-antitrypsin deficiency is diagnosed in 10%–20% of infants who present with the neonatal hepatitis syndrome and percutaneous liver biopsy in this condition may show features similar to those of biliary atresia. The difficulty with histologic diagnosis of alpha-1-antitrypsin deficiency may be compounded by the absence of PAS-positive inclusions, which are characteristic of this condition in older patients, and urgent phenotyping is therefore essential as part of the investigation of persistent neonatal hyperbilirubinemia. Difficulty in the interpretation of liver biopsies may also occur in infants with biliary hypoplasia. Inflammatory features may be present within the parenchyma and portal tracts but, in contrast to biliary atresia, bile ductules are absent or very sparse. However, the not infrequent syndromic association with vertebral anomalies, cardiovascular defects such as pulmonary stenosis, and a characteristic facial appearance aids the diagnosis of hypoplasia.54 The appearances of the extrahepatic bile ducts in biliary atresia vary greatly from case to case and may lead the inexperienced surgeon to an incorrect diagnosis. Abnormal segmental dilatations, for example, may superficially resemble a choledochal cyst. Proximal dilatation of the common hepatic duct may be secondary to an atretic occlusion of the distal common bile duct and although satisfactory bile drainage may be achieved after hepaticojejunostomy, the intrahepatic changes of biliary atresia may still affect the long-term prognosis. Non-communicating segmental dilatations containing clear mucus can also occur at any level of the extrahepatic biliary tract and may be misdiagnosed as a choledochal cyst and hence treated with an ineffectual anastomosis of bile duct to bowel. The historical classification of “correctable” and “noncorrectable” atresia was replaced by a classification devised by the Japanese Society of Pediatric Surgeons more than 25 years ago.55 The lesions were classified into three principal types: type 1: a bile-containing dilatation of a residual segment of proximal bile duct communicating with intrahepatic ducts type 2: residual lumen in undilated right and left hepatic ducts communicating with intrahepatic ducts but with atresia of the common hepatic duct type 3: complete atresia of the extrahepatic biliary tree including the right and left hepatic ducts

An analysis of 643 cases showed a 10% incidence of type 1 and an 88% incidence of type 3 atresias. Type 2 lesions were rare and made up only 2% of the series.56 Histologic assessment of the residual biliary tissue found at the porta hepatis may show persisting segments of the right and/or left hepatic ducts, multiple ductules, inflammatory cell infiltrates and fibrosis but even the larger duct remnants show at least partial loss of epithelium. The histologic appearances of the tissue at the porta hepatis have been classified into three main types.57 In type 1 the bile ducts are replaced completely by fibrous tissue with an inflammatory cell infiltrate and there is a complete absence of ducts and ductules. Type 2 tissue contains small ductules, approximately 50 µm in diameter, which are lined by cuboidal epithelium. In type 3 cases, bile ducts lined by columnar epithelium and bile-containing macrophages are present in more than two-thirds of resected specimens. Three-dimensional reconstruction studies have shown that, during the first few weeks of life, intrahepatic bile ducts are in communication with the residual tissue in the porta hepatis. The terminal portions of these bile ducts are the small ductules observed in tissue resected during the operation of portoenterostomy. The number of intrahepatic ducts decreases progressively with age but this process may be arrested if satisfactory bile drainage can be established within the first few weeks of life.58 It has been suggested that satisfactory bile flow may be anticipated after portoenterostomy when histological examination reveals residual ducts at the porta hepatis with diameters greater than 150 µm.59,60 However, good bile flow can be achieved in cases with much smaller ducts and histologic analysis is not always a good guide to prognosis.61 No statistically significant difference in bile ductule histology in tissue from the porta hepatis could be identified by Tagge et al.62 who found, in a series of 34 cases, that 43% with residual ducts less than 150 µm in diameter became anicteric compared with 36% of those with ducts over 150 µm. The response to surgery cannot be predicted from the histology of the hepatic parenchyma. Analyses of hepatic fibrosis, inflammatory change, and giant cell formation have not shown any clear correlation with postoperative bile flow.63,64 It is disappointing that hepatic fibrosis may even increase after achieving good biliary drainage; this progression has been demonstrated in 70% of long-term jaundice-free patients.65 The original Japanese classification of bile duct morphology in biliary atresia included many subtypes which recorded the patency or occlusion of the distal common bile duct, the anatomic features of the gall bladder and the appearance of residual tissue at the

Biliary atresia

Fig. 35.1. Figures 1–7 illustrate the operation of portoenterostomy. The operation commences with mobilization of the liver to expose the porta hepatis. Reproduced with permission, from Howard ER. Rob and Smith’s Operative Surgery 5th edn – Paediatric Surgery; Chapman & Hall;1995:551–561).

Fig. 35.3. The common bile duct is divided and the hepatic artery and portal vein are exposed.

Surgical techniques Portoenterostomy

Fig. 35.2. The mobilized gall bladder is used as a guide to the fibrous remnants of the extrahepatic bile ducts.

porta hepatis. Tan et al.66 analyzed the pattern and extent of the obliteration of the extrahepatic bile ducts in 205 cases and classified them into seven types but they found no correlation with long-term prognosis. It is concluded that these subtypes have little prognostic significance. The macroscopic appearances of the liver, porta hepatis, and portal venous system were recorded and correlated prospectively with subsequent bile flow.67 After a mean follow-up period of 32 months, 20 infants had a successful outcome compared with ten failures. Statistical analysis revealed that the only significant positive correlation with a successful outcome was the size of the tissue mass at the porta hepatis.

The operation was first devised by Kasai68 who observed that excision of all residual bile duct tissue from the porta hepatis could result in satisfactory drainage of bile from residual bile ducts. The essential steps in the operation are detailed in Figs. 35.1–35.7. It is important that the porta hepatis is fully visualized during the operation and in many cases this is only achieved by full mobilization of the liver and rotation of its inferior surface. The dissection extends behind the bifurcation of the portal vein and small tributaries of the vein to the caudate lobe are ligated to avoid hemorrhage during resection of the residual bile duct tissue. The biliary tissue at the porta hepatis is excised, together with the gallbladder and common bile duct, in a plane which is parallel with the liver capsule. The capsule itself should be damaged as little as possible as this might cause fibrous occlusion of the ductules at the porta hepatis. A long Roux loop of jejunum is used for the anastomosis at the porta hepatis; cutaneous enterostomies and antireflux valves are no longer in general use as they have not proved effective in the prevention of recurrent cholangitis (see below). Attempts to replace

451

452

E. R. Howard

Fig. 35.6. A 40 cm Roux-en-Y loop of jejunum is prepared and passed in a retrocolic manner to the hilum of the liver.

Fig. 35.4. The bifurcation of the portal vein is mobilized with division of tributaries to the caudate lobe.

Fig. 35.5. The fibrous tissue in the porta hepatis is divided flush with the capsule of the liver. All tissue within the main divisions of the vessels is removed.

the Roux loop with a conduit fashioned from the appendix (porto-appendiceal-duodenostomy), have also proved less effective than the standard operation, and a comparison of the results of the two procedures showed effective bile drainage in 31% of cases treated with an appendiceal conduit and 82% in those with a standard Roux loop of jejunum.69 Accurate dissection of the porta hepatis

Fig. 35.7. An anastomosis is fashioned between the cut edge of the transected tissue in the porta hepatis and the side of the Roux loop.

remains the most important factor in achieving successful bile drainage. In common with many other abdominal operations, portoenterostomy has now been performed laparoscopically in children.70 Four trochar ports were used for access and the Roux-en-Y jejunal anastomosis was performed through the umbilical port. The authors claimed that the technique avoids liver mobilization and the consequent risk of extensive adhesion formation.

Biliary atresia

Hepaticojejunostomy In cases of type 1 atresia the remnant of proximal bile duct may be large enough to allow the construction of a conventional hepaticojejunostomy. However, this is possible in only a small minority of cases.

Portocholecystostomy The gallbladder has been used for biliary drainage after cholangiographic confirmation of the patency of the cystic and distal common bile ducts. Complications are more frequent than after the standard portoenterostomy operation and include bile leaks, gallbladder obstruction, and kinking of the common bile duct.71,72 However, there does appear to be a reduction in postoperative cholangitis compared with portoenterostomy. A survey of 670 children from the North American Biliary Atresia Registry reported a 35% incidence of cholangitis after portocholecystostomy compared with 55% after other types of reconstruction (P = 0.02). The data were collected from more than 100 institutions.73

Postoperative management Fat-soluble vitamins are essential in preventing metabolic and coagulation abnormalities after surgery. Phenobarbitone and cholestyramine are often used to try and maximize bile flow although small prospective randomized controlled trials have not shown any definite advantage.74,75 Steroids have been suggested to prevent fibrosis at the porta hepatis but there are, as yet, no published controlled trials to prove their efficacy. Reports of the use of steroids in patients who have suffered from either cholangitis or diminishing bile flow have been encouraging. Kasai et al.76 recommended prednisolone 20 mg/day with a gradual reduction over 2 to 3 weeks and Altman and Anderson77 used 2 mg/kg per day with a reduction over 12 days. Karrer and Lilly,78 gave 10 mg/kg as an initial dose to 16 patients and discontinued the treatment after 4–7 days. The results were compared with 16 infants who were treated with antibiotics alone. In comparison with the controls, the patients treated with steroids seemed to show significant increases in their daily bile output; reductions in serum alkaline phosphatase and bilirubin levels were also noted. However, interpretation of these results is not straightforward as seven of the children had undergone revisional surgery before starting steroid therapy and the long-term results of the treatment were not given. Dillon et al.79 reported freedom from jaundice in 76% of 25 portoenterostomies who had been given long-term

postoperative oral steroids. Unfortunately, these cases were not part of a randomized trial and similar results have been reported in steroid-free series (see below). A further study comparing 14 infants treated with portoenterostomy and steroids (prednisone 2 mg/kg per day) with 14 who were not given steroids showed a 79% success rate in the former group compared with 21% in the latter.80 As in the previous report, this was not a randomized trial and treatment depended on surgeon preference. In summary, a beneficial effect of corticosteroid therapy after portoenterostomy remains unproven.

Long-term complications following surgery Cholangitis Early cholangitis Most surgeons give some form of antibiotic prophylaxis during the perioperative period to prevent attacks of bacterial cholangitis. The antibiotics are continued for a variable length of time but good evidence of any long-term benefit is not clear. Chaudary and Turner,81 recommended that trimethoprim-sulphamethoxazole (TMP/SMZ) be used for all cases after portoenterostomy and based this advice on the successful treatment of four children, who had suffered repeated attacks of cholangitis. Bu et al.82 compared two groups of infants treated with either TMP/SMZ or neomycin prophylaxis and compared both with a group of prophylaxis-free patients over a period of 14 months. The two treatment regimens were equally effective in reducing the rates of cholangitis compared with the control patients (P = 0.042 and 0.011). However, other authors have found no significant benefit from prophylaxis. For example, in one series of 41 infants, nine developed cholangitis, five of whom had received antibiotics and four no prophylaxis.83 A large multicentre prospective trial will be required to resolve this problem. Cholangitis, which is believed to be an ascending infection from the bowel affecting partially drained segments of liver, continues to be a major postoperative problem. (It is very rarely a problem after the surgical correction of choledochal cysts in which there is free drainage from all intrahepatic bile ducts.) Episodes of infection occur after portoenterostomy in 30%–60% of cases84 and can cause at least a transient deterioration in liver function. A wide range of enteric organisms may be identified and include E. coli, Proteus, and Klebsiella spp., Pseudomonas aeruginosa, Acinobacter baumanni and Salmonella typhi.85 Treatment of acute attacks is empirical and includes the prompt use of wide spectrum antibiotics such as cephalosporins and gentamicin.

453

454

E. R. Howard

Houwen et al.86 reported that the incidence of cholangitis was the most important determinant of successful surgical outcome. A 5-year survival rate of 54% was recorded in patients who suffered cholangitis compared with 91% for patients who remained free of infection. Prevention of enteric infection by separating the biliary and gastrointestinal tracts has resulted in the development of many types of external biliary diversion.87 However, the cholangitis rate has been little affected by these procedures and when a group of 12 patients with biliary diversion was compared with a group of 19 who did not have diversion the incidence of cholangitis was almost identical (33% vs. 32%).84 A difference in cholangitis rates for different operations does not necessarily influence outcome. For example, Tagge et al.62 performed a standard Kasai operation in 32% of 34 cases and a complete diversion of bile in 68%. Although there was a higher incidence of cholangitis in the biliary diversion group, there was no difference in the oneyear survival. Ohi88 illustrated and evaluated the results of various modified procedures performed throughout Japan between 1981 and 1985. He found that 43% had suffered attacks of cholangitis and that the first attack commonly occurred from one to three months after portoenterostomy. Only 10% developed infection for the first time more than 6 months after surgery. Importantly, 45% of 153 patients who had the original unmodified type of portoenterostomy had at least one attack of cholangitis compared with 41% of those who had modified procedures which included Rouxen-Y loops with tube enterostomies, complete diversion of bile via cutaneous stomas, interposition biliary conduits, and gastric tube reconstructions. One small prospective study investigated the value of an antireflux valve, constructed by intussuscepting a segment of bowel in the Roux-en-Y loop of jejunum; the valve made no significant difference to the incidence of subsequent attacks of cholangitis.89 Most of these techniques are now of historical interest only as they appear to be of little benefit and they increase the difficulties of transplantation, should this be necessary at a later date. External stomas may also be complicated by excessive electrolyte losses and by bleeding from varices that readily form on the stomal margins. Patency of the gallbladder and distal common bile duct was observed in 28% of boys and 18% of girls in 904 cases collected by the Biliary Atresia Registry of N. America.73 This variant of biliary atresia allows the use of the gallbladder as a conduit for biliary drainage (see above). Infants undergoing portocholecystostomy had a 35% incidence of cholangitis compared with 55% for other types of reconstruction. Ohi88 also reported a relatively low incidence of cholangitis (20%) in his review of this procedure. Unfortu-

nately, portocholecystostomy may be complicated by poor drainage of bile and by kinking of the cystic duct and Ohi did not recommend it as the procedure of choice in spite of the low incidence of subsequent infection. Late cholangitis Bacterial cholangitis decreases with age and is less frequent after 2 years.89 Gottrand et al.90 described four children who presented with infection between seven and 13 years of age. The cases represented 5% of 76 long-term survivors who had become anicteric after surgery. The infections were confirmed from either the histologic appearances of a liver biopsy or by culture of the liver tissue. All four presented with fever, jaundice and abdominal pain and all responded to antibiotic therapy and remained free of jaundice between 4 months and 4 years later. Occasionally, cholangitis may be precipitated by mechanical obstruction within the Roux loop, caused either by stenosis of an anastomosis or by an intra-abdominal adhesion. Surgical correction of the obstruction led to resolution of cholangitis in three of my own long-term survivors. Late presenting cholangitis should be investigated with a liver biopsy to confirm the diagnosis, and with a liver nuclide excretion scan to confirm the presence or absence of mechanical obstruction in the Roux loop. It is recommended that delayed excretion on the nuclide scan should be investigated by percutaneous cholangiography to give a more accurate visualization of the site of obstruction.

Portal hypertension Portal hypertension has been demonstrated in 68% of infants with biliary atresia between two and four months of age.91 Stringer et al.92 found varices in 67% of 61 children during endoscopies performed 2.5 years or more after portoenterostomy and variceal hemorrhage had occurred in 17 (28%). Portal hypertension was more common in children with jaundice than in those who were anicteric (86% vs. 62%) and also occurred more frequently in those who had suffered recurrent cholangitis. Furthermore, the risk of death or transplantation among 134 patients was approximately 50% within 6 years of the initial episode of bleeding from esophageal varices.93 Ohi et al.94 documented effective endoscopic obliteration of varices with 5% ethanolamine oleate in all children who completed their course of sclerotherapy; only one of the treated children died from uncontrolled hemorrhage. Complications, however, included three cases of symptomatic esophageal stricture, which responded satisfactorily to simple dilatation. There were no serious bleeding episodes from gastric varices in this report.

Biliary atresia

Concern with the possible complications of injection sclerotherapy stimulated the development of the technique of variceal ligation, described by Hall et al.95 in six children, two of whom were cases of biliary atresia. Each varix is ligated with an elastic band mounted on a modified gastroscope. The varix is occluded and necrosed by the band and sloughs in 5 to 10 days. Randomized trials in adults have now shown that banding is associated with fewer complications than sclerotherapy and with improved variceal obliteration. A preliminary study by Celinska-Cedro et al.96 has confirmed the efficacy of variceal ligation in 37 children, seven of whom were cases of biliary atresia. Eradication of the varices was achieved in 28 patients after two sessions of ligation with no significant post-treatment complications. Portosystemic shunting has also been effective in the management of portal hypertension in biliary atresia patients. Valayer97 reported 14 cases treated in this manner before the advent of sclerotherapy and liver transplantation. Shunting remains a useful, occasional, option in patients with bleeding ectopic varices in regions of the gut other than the esophagus, for example, the stomach or at the sites of surgical anastomoses. The formation of an intrahepatic portosystemic shunt by creating a passage between the hepatic and portal veins via a transjugular approach was made possible by the introduction of expandable metal stents.98 The procedure, known as TIPPS (transjugular intrahepatic portosystemic stent-shunt), is associated with complications of shunt thrombosis, encephalopathy, damage to the portal vein and shunt occlusion related to intimal proliferation within the stent which may occur in up to 50% of patients at 12 months.99 The technique has been used in children after portoenterostomy as young as 2 years.100 Schweizer et al.101 reported results in seven such cases. However, liver transplantation is now generally regarded as the most effective way of treating severe recurrent variceal bleeding in the esophagus or other parts of the gastrointestinal tract secondary to progressive liver disease from biliary atresia. Splenomegaly, associated with portal hypertension, is observed in the majority of long-term survivors after portoenterostomy and may be complicated by abdominal pain and distension, thrombocytopenia and leucopenia. Low platelet counts are found in many long-term survivors; the prevalence after 2 and 10 years has been recorded as 43% and 25%, respectively.102 The platelet count may be low enough to cause severe spontaneous bruising, particularly in the lower limbs. Stellin et al.103 treated a 6-year-old girl by splenic embolization using a percutaneous femoral artery catheter to introduce Gelfoam particles into the splenic artery. The procedure was complicated by abdominal pain,

ileus and fever but platelet and white blood cell counts returned rapidly to normal. Chiba et al.102 confirmed the effectiveness of the procedure in 19 postportoenterostomy children aged between 3 and 13 years of age in whom platelet counts ranged from 26 000 to 110 000/mm3 . Between 45% and 90% of the splenic tissue was embolized with no serious morbidity. Nine of the children were restudied 4 years later by scintigraphy and measurements of platelet counts. In some patients considerable regeneration of splenic tissue had occurred. In seven, the regeneration ranged from 23% to 76% with a mean of 41%, whilst in two cases there was no recovery of splenic volume. The platelet count increased more than fourfold in seven patients but two children showed signs of hypersplenism that required further embolization. A recent study of 32 children with biliary atresia again confirmed the longterm effectiveness of the technique. The average volume of spleen embolized was 70% and hematologic improvements were maintained in 17 out of 24 survivors who did not undergo liver transplantation.104 In summary, portal hypertension is a problem in a significant proportion of the survivors of the portoenterostomy operation. Endoscopic sclerotherapy and banding are excellent methods for the control of esophageal varices although portosystemic shunting may be required very occasionally for ectopic varices which have occurred at anastomotic sites within the gut (provided that the patient’s liver function is well preserved). Hypersplenism may be controlled with embolization rather than with splenectomy or portosystemic shunting. Complications of portal hypertension occurring in patients with biliary cirrhosis and poor liver function are best treated by liver transplantation.

Intrahepatic cystic change Large intrahepatic cysts, easily detectable with ultrasonography, may develop in both jaundiced and anicteric long-term survivors. Tsuchida et al.105 reported 29 welldocumented cases of whom 12 were males. The cysts were located near the hepatic hilum in 20 cases and at the periphery of the liver in eight. Cystic change was classified into three types: type A: non-communicating type B: communicating with the Roux-en-Y loop at the porta hepatis type C: multiple cystic dilatations of irregular bile ducts Clinical symptoms caused by the cysts included cholangitis and jaundice; in 66% of patients these occurred within 4 years of the portoenterostomy operation. Treatment included antibiotics and percutaneous drainage. Cystenterostomy was possible in six cases. However, five of the

455

456

E. R. Howard

six cases of type C disease developed symptoms between 10 and 28 years after surgery and the prognosis of this group was compromised by repeated infection. The development of this type of cystic change is probably an indication for liver transplantation. Metabolic problems The extent to which malabsorption improves after portoenterostomy is related to the adequacy of bile drainage. Maximal bile flow may not be achieved for one year after operation106 and even then bile salts seem to be excreted preferentially to cholesterol and phospholipid.107 The latter may take many months to normalize in successful cases. Lipid absorption can be improved by giving milk which contains fat in the form of medium-chain trigylcerides. The consequences of malabsorption are further reduced by regular dietary supplementation with fat-soluble vitamins, particularly D, E and K, and a multivitamin preparation of thiamine, riboflavine, pyridoxine, ascorbic acid, and folic acid.108 Bile is important for the intestinal absorption of calcium and magnesium because it is necessary for the absorption of vitamin D109 but cirrhosis may also interfere with the hydroxylation of vitamin D in the liver. In one study, rickets was found in 23 out of 39 patients (59%) with surgically uncorrected atresia and in four out of 15 (27%) who had undergone surgery. Osteoporosis was also found to develop a little later in both groups. In another report rickets was detected in 32% of long-term survivors older than 3 years.73 Florid rickets can be corrected rapidly by giving large doses of vitamin D and ensuring that there is adequate calcium and phosphate intake and absorption. Chronic liver disease may also lead to vitamin E deficiency in young children and this has been associated with a progressive neurologic disorder characterised by loss of tendon reflexes, a reduction in proprioception, abnormal eye movements and intellectual deterioration.110 Long-term biliary atresia survivors with severe liver fibrosis or cirrhosis may also suffer effects from poor absorption of soluble vitamins, protein, iron, calcium, zinc, and copper.

Hepatopulmonary syndrome and pulmonary hypertension In common with other forms of chronic liver disease, hypoxia with cyanosis on standing and exertion, dyspnea and finger clubbing may result from diffuse intrapulmonary shunting and intrapulmonary vascular dilatation in patients with biliary atresia. Valayer111 has pointed out that this complication, known as hepatopulmonary syn-

drome, is seen most frequently in the group of patients with an associated polysplenia syndrome and that the condition can be reversed by liver transplantation. Pulmonary arterial hypertension, which is recognized as a complication of cirrhosis in adults, may also occur in long-term survivors with biliary atresia, as well as in patients with prehepatic venous block, and congenital hepatic fibrosis.112 Possible etiologic factors include vasoactive substances such as endothelin or prostaglandin F2, which are either not metabolized by the liver, or are secreted by endothelial cells. Moscoso et al.113 reported the sudden death of a 10-year-old boy with pulmonary artery hypertension after a successful portoenterostomy. He had been asymptomatic and had been able to take part in normal sporting activities, such as swimming. At autopsy, the pulmonary arteries showed widespread thickening of the muscular media of the pulmonary arteries as well as proliferation of muscle and fibrous tissue within the subendothelial layer. Fresh fibrin thrombi were seen in some of the small arterioles and right ventricular hypertrophy was severe. The authors suggested that, in view of the asymptomatic nature of the pulmonary hypertension in this case, children with chronic liver disease such as biliary atresia should be followed up with assessments of pulmonary and cardiovascular function, and that liver transplantation might be the only option for those showing increasing vascular resistance in the pulmonary arteries.

Malignant change in the liver Malignant change in the cirrhotic liver of patients with biliary atresia has been reported on at least 15 occasions, eight at autopsy and seven during hepatic transplantation,111,114 with an age range of 3 to 16 years. The tumors have included cholangiocarcinoma (2 cases), hepatoma (12 cases) and a single case of hepatoblastoma, and all were associated with advanced biliary cirrhosis. On the basis of these cases, Tatekawa et al.114 advised routine screening of long-term biliary atresia patients with alpha-fetoprotein levels, CT scans and MR imaging. Six of the seven transplant cases were reported to show no evidence of recurrent tumor. It is likely that more cases will occur as the number of long-term survivors increases.

Long-term outcomes A baseline for survival of patients with untreated biliary atresia was defined by Hays and Snyder6 in a review of 41 cases seen at the Los Angeles Children’s Hospital between

Biliary atresia

Table 35.3. The incidence of esophageal varices and bleeding in long-term survivors after portoenterostomy Survival numbers 62

Tagge et al. Stringer et al.92 Miga et al.93 Valayer111 Laurent et al.119 Karrer et al.120 Howard & Davenport121 Totals

34 61 134 80 40 35 51 435

Varices ? 41 (67%) ? 24 (60%) ? 21 (41%)

Bleeding 6 (18%) 17 (28%) 39 (29%) 19 (24%) 15 (37%) 20 (55%) 8 (16%) 124 (28.5%)

1947 and 1961, before the introduction of the portoenterostomy procedure. Successful surgical correction had been possible in only one case and the average length of survival for the rest of the infants was only 19 months. Prolonged survival followed the introduction of the portoenterostomy procedure and an increasing number of long-term outcome studies are now appearing. Unfortunately, as Davenport has pointed out,115 an accurate comparative analysis of all long-term results is difficult as authors have tended to present their own results in an idiosyncratic manner. More uniformity in reporting is required but it is possible to deduce from the literature that long-term survival depends on a rapid postoperative fall of serum bilirubin into the normal range. It is also clear, however, that the attainment of a normal bilirubin level does not guarantee prolonged survival. At least 35% of cases suffer postoperative cholangitis but the effect of these infections on long-term survival remains debatable. Three reports86,87,116 appeared to demonstrate decreased survival after episodes of cholangitis but three other studies failed to confirm this relationship.62,73,117 Data from King’s College Hospital, London, revealed a history of cholangitis in 41% of the ten-year survivors and 36.6% of the 5-year survivors.117 The type of biliary drainage technique used in the portoenterostomy operation does not seem to influence the incidence of postoperative bacterial cholangitis. In a review of 1370 cases from the Japanese Biliary Atresia Registry, Nio et al.118 reported no significant differences in the incidence of cholangitis between the various techniques of biliary drainage recommended in the past and stated that since 1999, 96% of Japanese patients have undergone a standard portoenterostomy without modification.118 The prevalence of portal hypertension is similar in all of the published long-term results. Endoscopic evidence

457

Table 35.4. Collected 5-year survival figures after portoenterostomy for biliary atresia

Tagge et al.62 Houwen et al.86 Ohi et al.87 Howard et al.121 Kobayashi et al.122 Gauthier et al.123 Totals

Patient numbers

5-year survival

No jaundice

34 71 214 184 132 69 704

19 (37%) 17 (24%) 52 (24%) 71 (39%) 32 (24%) 31 (45%) 222 (32%)

7 (20%) 11 (15%) 47 (22%) 68 (37%) 17 (13%) 23 (33%) 173 (25%)

Table 35.5. Collected 10-year survival figures after portoenterostomy for biliary atresia

Caccia et al.116 Valayer111 Karrer et al.120 Howard et al.121 Nio et al.118 Totals a

Patient numbers

10-year survival

No jaundice

46 271 98 39 108 562

13 (28%) 80 (30%) 23 (23%) 17 (43%) 57 (53%) 190 (34%)

8 (17%) 38 (14%) 19 (19%) 9 (23%) ? 74 (16%)a

Denominator derived from the four series with complete data.

of esophageal varices is found in 40% to 60% of cases but esophageal bleeding is a problem in only about 28% (Table 35.3). Five-year survival figures after surgery for biliary atresia are shown in Table 35.4. Most of these cases were treated by portoenterostomy but sometimes it is not clear from the text whether or not an occasional case of hepaticojejunostomy is included. The figures are reasonably consistent in the various reports with a range of 24% to 45% and an average survival of 32%. Freedom from jaundice at 5 years is approximately 25%. The 10-year survival figures reveal that, although the survival rate of 34% is similar to the 5-year figure of 32%, there is a reduction in the proportion of cases who are free of jaundice. At 10 years approximately 16% are anicteric compared with 25% at five years. This figure could be useful when predicting the overall proportion of biliary atresia patients for whom transplantation might eventually be necessary (Table 35.5). Transplantation after failed portoenterostomy significantly increases the figure for long-term survival. The overall survival in Japan is now 75.3% at 5 years and 66.7% at 10 years.118 An illustration of the difficulty in comparing published results is provided by the report of Carcellar et al.124 who described the results of portoenterostomy in

458

E. R. Howard

Table 35.6. Prognostic studies concerning portoenterostomy Lawrence et al.61 Lopez Gutierrez et al.125 Lopez Gutierrez et al.126 Kang et al.127 Trivedi et al.128 Davenport & Howard67

Bile drainage not related to histology of porta hepatis Bile flow not related to histology of liver or porta hepatis Prognosis related to portal and hepatic arterial flow No relation between hepatic histology at portoenterostomy and development of varices Progressive liver damage associated with increasing levels of serum hyaluronic acid Macroscopic features at time of surgery related to effective postoperative bile drainage

Table 35.7. Relationship between the age at which portoenterostomy is performed and the restoration of bile flow Mieli-Vergani et al.64 (< 8 weeks) 12/14 (86%) (< 11 weeks) 40/59 (68%) Houwen et al.86 (< 10 weeks) 241/728 Nio et al.118 (33%) Gauthier et al.123 (< 8 weeks)31/50 (62%) ( 8 weeks) 13/36 (36%) (> 11 weeks) 6/12 (50%) (>10 weeks) 157/453 (34%) (> 8 weeks) 52/143 (36%) (>10 weeks) 4/16 (25%) 232/660 = 36%

77 patients at a mean follow-up of 8 years and 8 months. Seventeen (26%) were alive with their native liver and 25 after a transplant and therefore 42 (77%) of the original cohort were alive with a combination of portoenterostomy and transplantation. Unfortunately, it is not possible to compare these Canadian results directly with the previously quoted Japanese series as the method of reporting is so different. Prognostic factors for long-term survival Histologic studies of the liver and of atretic remnants of the biliary tract have not revealed any consistent prognostic features for either postoperative bile drainage or long-term survival after portoenterostomy. Some of these attempts to define prognostic criteria are listed in Table 35.6. Davenport and Howard67 analyzed the macroscopic features of the biliary tract and liver at the time of surgery; each feature was assigned a score. Outcome after surgery was assessed prospectively. Although the size of the portal remnant at the time of surgery was the most accurate prognostic feature, the discriminative power of the scoring system was not high. However, further analysis did show it to be a more accurate indicator of outcome than age of the infant at the time of surgery.

The prognosis following portoenterostomy has also been related to the speed with which the serum bilirubin falls into the normal range. Eighty one patients were separated into two groups 6 months after portoenterostomy, depending on whether or not their serum bilirubin had fallen into the normal range. The positive predictive value of a successful outcome using this categorization was 96% at 2 years and 95% at 5 years. The negative predictive value of failure was 76% and 74%, respectively. The authors also noted that bridging liver fibrosis and episodes of cholangitis were interdependent risk factors for failure after portoenterostomy.129 It is generally believed that early operation, under 8 weeks of age, provides the best chance of effective postoperative bile drainage in biliary atresia; some reports have also demonstrated a relationship between early surgery and long-term survival.73,87,120 Although several major series have shown this relationship it has been less convincing in others. Tagge et al.62 for example, showed that three out of 12 cases who underwent operation before 8 weeks of age were alive at 5 years compared with seven out of nine operated after 12 weeks. Similarly, Volpert et al.130 reported 92 cases, nine of whom underwent portoenterostomy before 30 days of life. Transplantation was required in seven (78%) of these “early” cases at a mean age of 11 months compared with a transplantation rate of 53% at a mean age of 32 months in the children who were treated after 30 days. The possibility of successful surgery in older age groups was also emphasized by Schoen et al.131 who achieved good bile drainage in five of six children over 10 weeks of age, although the significance of these figures is diminished by the small number of older cases available for analysis. In a large study of all French cases treated between 1986 and 1996, Chardot et al.132 compared the results of portoenterostomy in 380 infants under 90 days of age with 60 older infants and found no significant difference at 10 years. The 5-year survival of the two groups without transplantation was 35% and 25%, respectively, and the 10-year survival 25% and 22%. Valayer111 emphasized that, of 38 anicteric patients surviving for 10 years, 24 had undergone surgery later than 8 weeks of age. Davenport et al.117 also failed to find a statistically significant difference between age at the time of surgery and survival at 5 years. In summary, the previous belief in the benefit of early surgery before eight or ten weeks of age appears to have lost some of its significance as larger numbers of cases have become available for analysis (Table 35.7). Further data on the relationship of age at portoenterostomy and long-term survival are clearly needed to solve the apparent discrepancies between some of the published studies.

Biliary atresia

Table 35.8. Survival rates after liver transplantation for biliary atresia

Beath et al.134 Martinez et al.135 Valayer et al.136 Goss et al.137 Totals

Number

Survivors

39 32 72 190 333

28 (72%) 22 (70%) 62 (86%) 154 (80%) 266 (80%)

Liver transplantation It is now clear that the majority of patients will, sooner or later, require hepatic transplantation after portoenterostomy. A proportion of cases will completely fail to drain bile and will require urgent transplantation in the first year of life. Others, in whom satisfactory bile drainage is established, remain at risk from portal hypertension and progressive liver disease and may become candidates for liver replacement after several years of reasonable health. The outcome after liver transplantation for biliary atresia in four series of patients is shown in Table 35.8. Survival after transplantation is approximately 80%. Future qualityof-life studies in long-term survivors will be increasingly important. Goss et al.137 reported results in 190 cases; 1-, 2- and 5-year actuarial survival rates were 83%, 80%, and 78%, respectively. More recently, Diem et al.138 analyzed 328 transplants and reported survival figures at 1, 5 and 10 years of 87%, 83% and 81%, respectively. In this series the child’s age at the time of transplantation was a factor in long-term survival: the long-term survival of those treated before 6 years of age was approximately 85% compared with 100% for older age groups. The improved survival in those over 6 years further supports the policy of recommending portoenterostomy as the primary treatment of choice in young infants with biliary atresia. Quality-of-life and economic considerations A review of 80 patients who survived for more than 10 years without transplantation revealed complications affecting quality of life in 23 (29%) (Table 35.9).139 Forty-four were in school and 34 at work. Four women had given birth to normal children. The comments of Nio et al.140 on the status of 22 patients who survived for more than 20 years were very encouraging. One had died at 28 years of age but 16 had led near normal lives and three of the women had married. One of the latter had borne a normal child. Three patients had required a portosystemic shunt for variceal bleeding and one patient had intrahepatic calculi. Physical growth was judged to be normal in all but one case. Girls had had a normal menarche and secondary sexual

Table 35.9. Complications in 23 of 80 patients who survived ten years after portoenterostomy without transplantation (Ohi139 ) Increased fatigue Pruritus Abdominal pain Hepatopulmonary syndrome Jaundice Recurrent pyrexia

23 (29%) 10 (12.5%) 4 (5%) 3 (4%) 3 (4%) 2 (2.5%)

characteristics had developed normally in the boys. The overall quality of life was judged to be satisfactory in 16 of the patients but three required occasional hospital admission. The authors stated that progressive liver disease had not been observed in the 21 surviving patients. All of the patients were in employment. Other measures of the quality of life were supplied by Kobayashi et al.122 who reported satisfactory weights in 28 out of 32 survivors beyond five years and satisfactory heights in 17. Ohi et al.87 reported that 48 out of 52 cases surviving more than 5 years (the oldest of whom was 27 years) were leading normal lives and both growth and development were normal in 49. In Valayer’s report,111 school performance assessed in 26 ten-year survivors was judged to be normal in eight, 1 year below normal in 11, and 2 to 3 years below normal in seven. Twelve of the teenagers showed good sporting ability and professional activity for those in their twenties was said to be “normal.” Karrer et al.120 discussed the quality of life in 30 patients who had survived at least ten years after portoenterostomy, 25 of whom were within the normal expected range for height. Age-appropriate school or employment had been achieved in 75% of cases. One patient was pregnant at the time of the report. A unique study of the quality of life (QoL) of long-term survivors compared 30 Japanese over 14 years of age with 25 UK patients in the same age range.141 Blood and liver function tests did not show any significant differences between the two groups. Overall, there were no significant differences in the QoL measures which included functional status, well-being and an overall evaluation of health. Furthermore, there was a very satisfactory comparison with normative population data in both countries. The study confirmed that excellent long-term survival is possible after portoenterostomy in at least a proportion of patients. Successful pregnancy in patients treated for biliary atresia has now been reported by at least five authors.87,111,120,142,143 Questionnaires sent to 134 institutions affiliated to the Japanese Biliary Atresia Society

459

460

E. R. Howard

included questions on menstrual irregularities, numbers of pregnancies, problems related to pregnancy and delivery, and the health of the baby. Sixteen patients had given birth on 23 occasions (nine delivered once and seven delivered twice). However, only three (19%) had had a problem-free pregnancy. Deterioration of liver function occurred after delivery in six (37.5%), and one of these women required urgent transplantation. Cholangitis required treatment in four (25.0%). Other problems included two abortions, one of which was related to shock from bleeding esophageal varices. A further case of variceal bleeding was treated by endoscopic variceal ligation.143 In common with other reports, none of the babies of biliary atresia mothers had any liver problem or other congenital abnormality. This large collected series indicates that pregnancy in biliary atresia patients carries a high degree of risk and requires close surveillance. Barkin and Lilly144 examined the stresses on parents of children with biliary atresia. Paradoxically, the stresses appeared to be greater in families whose children had undergone successful surgery. Problems included “marital discord” (4), divorce (2), abandonment of the child (1), and unemployment and financial pressures (3). All of the families interviewed had some difficulty coming to terms with the chronic illness. Unfortunately, no comparison was made with families caring for children with other forms of chronic disease. The authors mentioned the existence of a support organization in North America for the families of children with biliary atresia. The Children’s Liver Disease Foundation charity in the UK performs a similar role in providing clinical information and support for the families of affected children. The satisfactory outcome after portoenterostomy in a proportion of children and the low rehospitalization rate suggest that this procedure should continue as first-line treatment of biliary atresia. Shortage of donor organs, complications of immunosuppression, chronic rejection and repeat transplantation remain limiting factors for primary transplantation particularly in children under six years of age. The overall mortality rate of liver transplantation in children over 3 years of age is approximately 20%. Portoenterostomy continues to be the primary treatment for biliary atresia, reserving transplantation for the failures and complications of the procedure.

Summary r Before the introduction of the portoenterostomy oper-

ation, the mortality rate for biliary atresia was virtually 100%.

r The introduction and refinement of portoenterostomy

r

r

r

r

resulted in long-term survival with a good quality of life in approximately 25% of the children. Complications in long-term survivors of portoenterostomy include cholangitis, portal hypertension, bleeding esophageal varices, intrahepatic cystic dilatation and cholelithiasis, hypoxia from pulmonary artery pathology, and malignant change in the liver. Liver transplantation is a successful mode of treatment for infants who fail to achieve satisfactory bile drainage after portoenterostomy and for complications in longterm survivors, including progressive liver disease and cirrhosis. The overall survival for biliary atresia patients treated in combined programmes of portoenterostomy and transplantation is greater than 80%. The quality of life in the majority of long-term survivors is very satisfactory and as yet there are no reports of transmission of bilary atresia to the offspring of patients.

REFERENCES 1. Balistreri, W. F., Grand, R., Hoofnagle, J. H. et al. Biliary atresia: current concepts and research directions. Hepatology 1996; 23:1682–1692. 2. Hays, D. M. & Kimura, K. et al. Biliary atresia: new concepts of management. Curr. Prob. Surg. 1981; 18:546. 3. Thomson, J. Congenital Obliteration of the Bile Ducts. Edinburgh:Oliver and Boyd:1892. 4. Holmes, J. B. Congenital obliteration of the bile duct: diagnosis and suggestions for treatment. Am. J. Dis. Child. 1916; 11:405– 43l. 5. Ladd, W. E. Congenital atresia and stenosis of the bile duct. J. Am. Med. Assoc. 1928; 91:1082–1084. 6. Hays, D. M., Snyder, W. H. Life-span in untreated biliary atresia. Surgery 1963; 64:373–375. 7. Kasai, M. & Suzuki, S. A new operation for ‘non-correctable’ biliary atresia: hepatic portoenterostomy. Shuzutsu 1959; 3:733–739. 8. Kasai, M. Treatment of biliary atresia with special reference to hepatic portoenterostomy and its modifications. Prog. Pediatr. Surg. 1974; 6:5–52. 9. Jenner, R. E. & Howard, E. R. Unsaturated monohydroxy bile acids as a cause of idiopathic obstructive cholangiopathy. Lancet 1975; 2:1073–1074. 10. Ikeda, S., Sera, Y., Yoshida, M. et al. extrahepatic atresia associated with trisomy 18. Pediatr. Surg. Int. 1999; 15:137–138. 11. Buffa, V., Tancredi, F., Pierro, M. et al. Case of trisomy of chromosome 18 associated with hypermethioninemia and biliary atresia. Pediatria (Napoli) 1972; 80:159–169. 12. Alpert, L. I., Strauss, L., & Hirschorn, K. Neonatal hepatitis and biliary atresia associated with trisomy 17–18 syndrome. N. Engl. J. Med. 1969; 280:16–20.

Biliary atresia

13. Danks, D. M. Prolonged neonatal obstructive jaundice. A survey of modern concepts. Clin. Pediatr. 1965; 4:499– 510. 14. Khaimin, V. M. Congenital atresia of the bile ducts in twin boys (Russian). Vopr. Okhr. Materin. Det. 1969; 14:184. 15. Smith, B. M., Laberge, J. M., Schreiber, R. et al. Familial biliary atresia in three siblings including twins. J. Pediatr. Surg. 1991; 26:1331–1333. 16. Silveira, T. R., Salzano, F. M., Howard, E. R. et al. Congenital structural abnormalities in biliary atresia: evidence for etiopathogenic heterogeneity and therapeutic implications. Acta Paediatr. Scand. 1991; 80:1192–1199. 17. Silveira, T. R., Salzano, F. M., Donaldson, P. T. et al. Association between HLA and extrahepatic biliary atresia. J. Pediatr. Gastroenterol. Nutr. 1993; 16:114–117. 18. Jurado, A., Jara, P., Camarena, C. et al. Is extrahepatic biliary atresia an HLA-associated disease? J. Pediatr. Gastroenterol. Nutr. 1997; 25:557–558. 19. Donaldson, P. T., Clare, M., Constantini, P. K. et al. HLA and cytokine gene polymorphisms in biliary atresia. Liver 2002; 22:213–219. 20. Desmet, V. & Callea, F. Ductal plate malformation (DPM) in extrahepatic bile duct atresia (EHBDA). In Ohi, R., ed. Biliary Atresia. ICOM Associates Inc. 1991; 27–31. 21. Tan, C. E. L., Driver, M., Howard, E. R. et al. Extrahepatic biliary atresia: a first-trimester event? Clues from light microscopy and immunohistochemistry. J. Pediatr. Surg. 1994; 29:808– 814. 22. Tsuchida, Y., Kawarasaki, H., Iwanaka, T. et al. Antenatal diagnosis of biliary atresia (type 1 cyst) at 19 weeks gestation: differential diagnosis and etiologic implications. J. Pediatr. Surg. 1995; 30:697–699. 23. Hinds, R., Davenport, M., Mieli-Vergani, G. et al. Antenatal presentation of biliary atresia. J. Pediatr. 2004; 144:43–46. 24. MacGillivray, T. E. & Scott Adzick, N. Biliary atresia begins before birth. Pediatr. Surg. Int. 1994; 9:116–117. 25. Ben-Ami, M., Perlitz, Y., Shalev, S. et al. Prenatal diagnosis of extrahepatic biliary duct atresia. Prenat. Diagn. 2002; 22:583– 585. 26. Tarr, P. I., Haas, J. E., & Christie, D. L. Biliary atresia, cytomegalovirus, and age at referral. Pediatrics 1996; 97:828– 831. 27. Strauss, L. & Bernstein, J. Neonatal hepatitis in congenital rubella; a histopathological study. Arch. Pathol. 1968; 86:317– 327. 28. Weaver, L. T., Nelson, R., & Bell, T. M. The association of extrahepatic bile duct atresia and neonatal Epstein–Barr virus infection. Acta. Pediatr. Scand. 1984; 73:155–157. 29. Drut, R., Drut, R. M., Gomez, M. A. et al. Presence of human papillomavirus in extrahepatic biliary atresia. J. Pediatr. Gastroenterol. Nutr. 1998; 27:530–535. 30. Domiati-Saad, R., Dawson, D. B., Margraf, L. R. et al. Cytomegalovirus and human herpesvirus 6, but not human papillomavirus, are present in neonatal giant cell hepatitis and extrahepatic biliary atresia. Pediatr. Dev. Pathol. 2000; 3:367–373.

31. Jevon, G. P. & Dimmick, J. E. Biliary atresia and cytomegalovirus infection: a DNA study. Pediatr. Dev. Pathol. 1999; 2:11–14. 32. Phillips, P. A., Keast, D., Papadimitriou, J. M. et al. Chronic obstructive jaundice induced by reovirus type 3 in weanling mice. Pathology. 1969; 1:193–203. 33. Morecki, R., Glaser, J. H., & Horwitz, M. S. Etiology of biliary atresia: the role of reo 3 virus. In Daum, F., ed. Extrahepatic Biliary Atresia. New York: Marcel Dekker, 1983; 1–9. 34. Morecki, R., Glaser, J. H., Cho, S. et al. Biliary atresia and reovirus type 3 infection. N. Eng. J. Med. 1982; 307:481–484. 35. Dussaix, E., Hadchouel, M., Tardieu, M. et al. Biliary atresia and reo virus type 3 infection. N. Engl. J. Med. 1984; 311;658– 66l. 36. Brown, W. R., Sokol, R. J., Levin, M. R. et al. Lack of correlation between infection with reovirus 3 and extrahepatic biliary atresia or neonatal hepatitis. J. Pediatr. 1988; 113:670– 676. 37. Tyler, K. L., Sokol, R. J., Oberhaus, S. M. et al. Detection of reovirus RNA in hepatobiliary tissues from patients with extrahepatic biliary atresia and choledochal cysts. Hepatology 1998; 27:1475–1482. 38. Steele, M. I., Marshall, C. M., Lloyd, R. D. et al. Reovirus 3 not detected by reverse transcriptase-mediated polymerase chain reaction analysis of preserved tissue from infants with cholestatic liver disease. Hepatology 1995; 21:697–702. 39. Carmi, R., Magee, C. A., Neill, C. A. et al. Extrahepatic biliary atresia and associated anomalies: etiologic heterogeneity suggested by distinctive patterns of associations. Am. J. Med. Genet. 1993; 45:683–693. 40. Chandra, R. S. Biliary atresia and other structural anomalies in the congenital polysplenia syndrome. J. Pediatr. 1974; 85:649– 655. 41. Davenport, M., Savage, M., Mowat, A. P. et al. Biliary atresia splenic malformation syndrome: an etiological and prognostic subgroup. Surgery 1993; 113:662–668. 42. Fischler, B., Papadogiannakis, N., & Nemeth, A. Aetiological factors in neonatal cholestasis. Acta. Paediatr. 2001; 90:88–92. 43. Mazziotti, M. V., Willis, L. K., Heuckroth, R. O. et al. Anomalous development of the hepatobiliary system in the INV mouse. Hepatology 1999; 30:372–378. 44. Schon, P., Tsuchiya, K., Lenoir, D. et al. Identification, genomic organization, chromosomal mapping and mutation analysis of the human INV gene, the ortholog of a murine gene implicated in left–right axis development and biliary atresia. Hum. Genet. 2002; 110:157–165. 45. Mieli-Vergani, G. & Mowat, A. P. Paediatric liver disease: medical aspects. In Millward-Sadler, G. H., Wright, R. Arthur, M. J. P. eds. Wright’s Liver and Biliary Disease, 3rd edn. London: WB Saunders, 1992; 1189–207. 46. Hussein, M., Howard, E. R., Mieli-Vergani, G. et al. Jaundice at 14 days of age: exclude biliary atresia. Arch. Dis. Child. 1991; 66:1177–1179. 47. Davenport, M., Betalli, P., D’Antiga, L. et al. The spectrum of surgical jaundice in infancy. J. Pediatr. Surg. 2003; 38:1471– 1479.

461

462

E. R. Howard

48. Hay, S. A., Soliman, H. E., Sherif, H. M. et al. Neonatal jaundice: the role of laparoscopy. J. Pediatr. Surg. 2000; 35:1706–1709. 49. Wilkinson, M. L., Mieli-Vergani, G., Ball, C. et al. Endoscopic retrograde cholangiopancreatography in infantile cholestasis. Arch. Dis. Child. 1991; 66:121–123. 50. Takahashi, H., Kuriyama, Y., Maiae, M. et al. ERCP in jaundiced infants. In Ohi, R., ed. Biliary Atresia. Tokyo: Professional Postgraduate Services, 1987; 110–113. 51. Iinuma, Y., Narisawa, R., Iwafuchi, M. et al. The role of endoscopic retrograde cholangiopancreatography in infants with cholestasis. J. Pediatr. Surg. 2000; 35:545–549. 52. Mason, G. R., Northway, W., & Cohn, R. B. Difficulties in the operative diagnosis of congenital atresia of the biliary ductal system. Am. J. Surg. 1966; 112:183–187. 53. Manolaki, A. G., Larcher, V. F., Mowat, A. P. et al. The prelaparotomy diagnosis of extrahepatic biliary atresia. Arch. Dis. Child. 1983; 58:59l–594. 54. Alagille, D., Odievre, M., Gautier, M. et al. Hepatic ductular hypoplasia associated with characteristic facies, vertebral malformation, retarded physical, mental and sexual development and cardiac murmur. J. Pediatr. 1975; 86:63–7l. 55. Hays, D. M. & Kimura, K. Biliary Atresia: The Japanese Experience. Cambridge, MA: Harvard University Press, 1980; 22. 56. Ohi, R., Chiba, T., Ohkochi, N. et al. The present status of surgical treatment for biliary atresia: report of the questionnaire for the main institutions in Japan. In Ohi, R., ed. Biliary Atresia. Tokyo: Professional Postgraduate Services, 1987; 125–30. 57. Gautier, M. & Eliot, N. Extrahepatic biliary atresia: morphological study of 98 biliary remnants. Arch. Path. Lab. Med. 1981; 105:397–402. 58. Kasai, M., Ohi, R., & Chiba, T. Intrahepatic bile ducts in biliary atresia. In Kasai, M., & Shiraki, K. eds. Cholestasis in Infancy. Baltimore: University Park Press, 1980; 181–188. 59. Altman, R. P., Chandra, R., & Lilly, J. R. Ongoing cirrhosis after successful porticoenterostomy in infants with biliary atresia. J. Pediatr. Surg. 1975; 10:685–689. 60. Ohi, R., Shikes, R. H., Stellin, G. P. et al. In biliary atresia duct histology correlates with bile flow. J. Pediatr. Surg. 1984; 19:467–470. 61. Lawrence, D., Howard, E. R., Tzanatos, C. et al. Hepatic portoenterostomy for biliary atresia. Arch. Dis. Child., 1981; 56:460–463. 62. Tagge, D. U., Tagge, E. P., Drongowski, R. A. et al. A long-term experience with biliary atresia. Reassessment of prognostic factors. Ann. Surg. 1991; 214:590–598. 63. Altman, R. P. The portoenterostomy procedure for biliary atresia. Ann. Surg. 1978; 188:357–36l. 64. Mieli-Vergani, G., Howard, E. R., Portmann, B. et al. Late referral for biliary atresia – missed opportunities for effective surgery. Lancet 1989; 1:42l–423. 65. Gautier, M., Valayer, J., Odievre, M. et al. Histological liver evaluation 5 years after surgery for extrahepatic biliary atresia: a study of 20 cases. J. Pediatr. Surg. 1984; 19:263–268. 66. Tan, C. E. L., Davenport, M., Driver, M., & Howard, E. R. Does the morphology of the extrahepatic biliary remnants in biliary

67.

68. 69.

70.

71. 72. 73. 74.

75. 76.

77.

78. 79.

80.

81.

82.

83.

84.

85.

atresia influence survival? A review of 205 cases. J. Pediatr. Surg. 1994; 29:1459–1464. Davenport, M. & Howard, E. R. Macroscopic appearance at portoenterostomy – a prognostic variable in biliary atresia. J. Pediatr. Surg. 1996; 31:1387–1390. Kasai, M., Kimura, S., Asakura, Y. et al. Surgical treatment of biliary atresia. J. Pediatr. Surg. 1968; 3:665–675. Tsao, K., Rosenthal, P., Dhawan, K. et al. Comparison of drainage techniques for biliary atresia. J. Pediatr. Surg. 2003; 38:1005–1007. Esteves, E., Clemente Neto, E., Ottaiano Neto, M. et al. Laparoscopic Kasai portoenterostomy for biliary atresia. Pediatr. Surg. Int. 2002; 18:737–740. Lilly, J. R. Hepatic portocholecystostomy for biliary atresia. J. Pediatr. Surg. 1979; 14:30l–304. Freitas, L., Gauthier, F., & Valayer, J. Second operation for repair of biliary atresia. J. Pediatr. Surg. 1987; 22:857–60. Karrer, F. M., Lilly, J. R., Stewart, B. A. et al. Biliary atresia registry, 1976 to 1989. J. Pediatr. Surg. 1990; 25:1076–1081. Vajro, P., Couterier, M., Lemmonier, F. et al. Effects of postoperative cholestyramine and phenobarbital administration on bile flow restoration in infants with extrahepatic biliary atresia. J. Pediatr. Surg. 1986; 21:262–265. Nittono, H., Tokita, A., Hayashi, M. et al. Ursodeoxycholic acid in biliary atresia. Lancet 1988; 1:528. Kasai, M., Suzuki, H., Ohashi, E. et al. Technique and results of operative management of biliary atresia. World J. Surg. 1978; 2: 571–580. Altman, R. P. & Anderson, K. D. Surgical management of intractable cholangitis following successful Kasai procedure. J. Pediatr. Surg. 1982; 17:894–900. Karrer, F. M., Lilly, J. R. Corticosteroid therapy in biliary atresia. J. Pediatr. Surg. 1985; 20:683–695. Dillon, P. W., Owings, E., Cilley, R. et al. Immunosuppression as adjuvant therapy for biliary atresia. J. Pediatr. Surg. 2001; 36:80–85. Meyers, R. L., Book, L. S., O’Gorman, M. A. et al. Highdose steroids, ursodeoxycholic acid, and chronic intravenous antibiotics improve bile flow after Kasai procedure in infants with biliary atresia. J. Pediatr. Surg. 2003; 38:406–411. Chaudhary, S. & Turner, R. B. Trimethoprimsulfamethoxazole for cholangitis following hepatic portoenterostomy for biliary atresia. J. Pediatr. 1981; 99:656–658. Bu, L. N., Chen, H. L., Chang, C. J. et al. Prophylactic oral antibiotics in prevention of recurrent cholangitis after the Kasai portoenterostomy. J. Pediatr. Surg. 2003; 38:590–593. Lally, K. P., Kenegaye, J., Matsumura, M. et al. Perioperative factors affecting the outcome following repair of biliary atresia. Pediatrics 1989; 83:723–726. Burnweit, C. A. & Coln, D. Influence of diversion on the development of cholangitis after hepatoportoenterostomy for biliary atresia. J. Pediatr. Surg. 1986; 21:1143–1146. Wu, E. T., Chen, H. L., Ni, Y. H. et al. Bacterial cholangitis in patients with biliary atresia: impact on short-term outcome. Pediatr. Surg. Int. 2001; 17:390–395.

Biliary atresia

86. Houwen, R. H. J., Zwierstra, R. P., Severijnen, R. S. et al. Prognosis of extrahepatic biliary atresia. Arch. Dis. Child. 1989; 64:214–218. 87. Ohi, R., Hanamatsu, M., Mochizuki, I. et al. Progress in the treatment of biliary atresia. World J. Surg. 1985; 9:285–293. 88. Ohi, R. Biliary atresia: long-term results of hepatic portoenterostomy. In Howard, E. R. ed. Surgery of Liver Disease in Children, Oxford:Butterworth-Heinemann, 1991; 60–71. 89. Ecoffey, C., Rothman, E., Bernard, O. et al. Bacterial cholangitis after surgery for biliary atresia. J. Pediatr. 1987; 111:824–829. 90. Gottrand, F., Bernard, O., Hadchouel, M. et al. Late cholangitis after successful surgical repair of biliary atresia. Am. J. Dis. Child. 1991; 145:213–215. 91. Kasai, M., Okamoto, A., Ohi, R. et al. Changes of portal vein pressure and intrahepatic blood vessels after surgery for biliary atresia. J. Pediatr. Surg. 1981; 16:152–159. 92. Stringer, M., Howard, E. R., & Mowat, A. P. Endoscopic sclerotherapy in the management of esophageal varices in 6l children with biliary atresia. J. Pediatr. Surg. 1989; 24:438–442. 93. Miga, D., Sokol, R. J., Narkewicz, M. R. et al. Survival after first esophageal variceal hemorrhage in patients with biliary atresia. Hepatology 2002; 26:498–499. 94. Ohi, R., Mochizuki, I., Komatsu, K. et al. (l986) Portal hypertension after successful hepatic portoenterostomy in biliary atresia. J. Pediatr. Surg. 1986; 21:271–274. 95. Hall, R. J., Lilly, J. R., & Stiegmann, G. V. Endoscopic esophageal varix ligation: technique and preliminary results in children. J. Pediatr. Surg. 1988; 23:1222–1223. 96. Celinska-Cedro, D. Teisseyre, M., Woynarowski, M. et al. Endoscopic ligation of esophageal varices for prophylaxis of first bleeding in children and adolescents with portal hypertension: preliminary results of a prospective study. J. Pediatr. Surg. 2003; 38:1008–1011. 97. Valayer, J. Portosystemic shunt surgery. In Howard, E. R. ed. Surgery of Liver Disease in Children. Oxford:ButterworthHeinemann. 1991; 171–80. 98. Rossle, M., Richter, G. M., Noldge, G. et al. Performance of an intrahepatic portocaval shunt (PCS) using a catheter technique: a case report. Hepatology 1988; 8:1348. 99. Benner, K., Sahagun, G., Meyer, R. A. et al. Shunt patency after transjugular intrahepatic portosystemic shunt. Gastroenterology 1993; 104:A876. 100. Kimura, B. T., Hasegawa, T., Oue, T. et al. Transjugular intrahepatic portosystemic shunt performed in a 2-year-old infant for uncontrollable intestinal bleeding. J. Pediatr. Surg. 2000; 35:1597–1599. 101. Schweizer, P., Brambs, H. J., Schweizer, M. et al. TIPS: a new therapy for esophageal variceal bleeding caused by EHBA. Eur. J. Pediatr. Surg. 1995; 5:211–215. 102. Chiba, T., Ohi, R., Yaoita, M. et al. Partial splenic embolization for hypersplenism in pediatric patients with special reference to its long-term efficacy. In Ryoji, O. ed. Biliary Atresia, Tokyo:ICOM Associates Inc., 1991; 154–158. 103. Stellin, G., Kumpe, D. A., & Lilly, J. R. Splenic embolization in a child with hypersplenism. J. Pediatr. Surg. 1982; 17:892–893.

104. Nio, M., Hayashi, Y., Sano, N. et al. Long-term efficacy of partial splenic embolization in children. J. Pediatr. Surg. 2003; 38:1760–1762. 105. Tsuchida, Y., Honna, T., & Kawarasaki, H. Cystic dilatation of the intrahepatic biliary system in biliary atresia after hepatic portoenterostomy. J. Pediatr. Surg. 1994; 29:630–634. 106. Howard, E. R. & Mowat, A. P. Hepatobiliary disorders in infancy: hepatitis; extrahepatic biliary atresia; intrahepatic biliary hypoplasia. In Thomas, H. C. & McSween, R. N. M. eds. Recent Advances in Hepatology. London: Churchill Livingstone, 1984; 153. 107. Lilly, J. R. & Javitt, N. B. Biliary lipid excretion after hepatic portoenterostomy. Ann. Surg. 1976; 184:369–375. 108. Greene, H. L. Nutritional aspects in the management of biliary atresia. In Daum, F., ed. Extrahepatic Biliary Atresia. New York: Marcel Dekker, 1983; 133–143. 109. Kobayashi, A., Kawai, S., Utsunomiya, T. et al. Bone disease in infants and children with hepatobiliary disease. Arch. Dis. Child. 1974; 49:641–646. 110. Nelson, J. S., Rosenblum, J. L., Keating, J. P. et al. Neuropathological complications of childhood cholestatic liver disease. In Daum, F. ed. Extrahepatic Biliary Atresia. New York: Marcel Dekker, 1983; 153–157. 111. Valayer, J. Conventional treatment of biliary atresia. J. Pediatr. Surg. 1996; 31:1546–1551. 112. Soh, H., Hasegawa, T., Sasaki, T. et al. Pulmonary hypertension associated with postoperative biliary atresia: report of two cases. J. Pediatr. Surg. 1999; 34:1779–1781. 113. Moscoso, G., Mieli-Vergani, G., Mowat, A. P. et al. Sudden death caused by unsuspected pulmonary arterial hypertension, 10 years after surgery for extrahepatic biliary atresia. J. Pediatr. Gastroenterol. Nutr. 1991; 12:388–393. 114. Tatekawa, Y., Asonuma, K., Uemoto et al. Liver transplantation for biliary atresia associated with malignant tumors. J. Pediatr. Surg. 2001; 36:436–439. 115. Davenport, M. Biliary atresia (letter). J. Pediatr. Surg. 2001; 36:1318. 116. Caccia, G., Dessanti, A., Alberti, D. et al. More than 10 years survival after surgery for biliary atresia. In Ohi, R., ed. Biliary Atresia. Tokyo:ICOM Associates Inc, 1991; 246–249. 117. Davenport, M., Kereker, N., Mieli-Vergani, G. et al. (1997) Biliary atresia: the King’s College Hospital experience (1974–95) J. Pediatr. Surg. 1997; 32:1–8. 118. Nio, M., Ohi, R., Miyano, T. et al. Five and 10-year survival rates after surgery for biliary atresia: a report from the Japanese Biliary Atresia Registry. J. Pediatr. Surg. 2003; 38:997–1000. 119. Laurent, J., Gauthier, F., Bernard, O. et al. Long-term outcome after surgery for biliary atresia. Study of 40 patients surviving for more than 10 years. Gastroenterology 1990; 99:1793– 1797. 120. Karrer, F. M., Price, M. R., Bensard, D. D. et al. Long-term results with the Kasai operation for biliary atresia. Arch. Surg. 1996; 131:493–496. 121. Howard, E. R. & Davenport, M. The treatment of biliary atresia in Europe – 1969–1995. Tohuko. J. Exp. Med. 1997; 181:75–83.

463

464

E. R. Howard

122. Kobayashi, A., Itabashi, F., & Ohbe, Y. Long-term prognosis in biliary atresia after hepatic portoenterostomy: analysis of 35 patients who survived beyond 5 years of age. J. Pediatr. 1984; 105: 243–246. 123. Gauthier, F., Laurent, J., Bernard, O. et al. Improvement of results after Kasai operation: the need for early diagnosis and surgery. In Ohi, R., ed. Biliary Atresia. Tokyo:ICOM Associates Inc. 1991; 91–95. 124. Carcellar, A., Blanchard, H., Alvarez, F. et al. Past and future of biliary atresia. J. Pediatr. Surg. 2000; 35:717–720. 125. Lopez, Gutierrez. J. C., Vazquez, J., Ros, Z. et al. Histopathology of biliary atresia: correlation with biliary flow. Cir. Pediatr. 1991; 4:16–18. 126. Lopez, Gutierrez J. C., Vazquez, J., Prieto, C. et al. Portal venous flow as a prognostic factor in biliary atresia. A preliminary study. Cir. Pediatr. 1992; 5:17–9. 127. Kang, N., Davenport, M., Driver, M. et al. Hepatic histology and the development of esophageal varices in biliary atresia. J. Pediatr. Surg. 1993; 28:63–66. 128. Trivedi, P., Cheeseman, P., & Mowat, A. P. Serum hyaluronic acid in healthy infants and children and its value as a marker of progressive hepatobiliary disease starting in infancy. Clin. Chim. Acta. 1993; 215:29–39. 129. Wildhaber, B. E., Coran, A. G., Drongowski, R. A. et al. The Kasai portoenterostomy for biliary atresia: a review of a 27-year experience with 81 patients. J. Pediatr. Surg. 2003; 38:1480– 1485. 130. Volpert, D., White, F., Finegold, M. J. et al. Outcome of early hepatic portoenterostomy for biliary atresia. J. Pediatr. Gastroenterol. Nutr. 2001; 32:265–269. 131. Schoen, B. T., Lee, H., Sullivan, K. et al. The Kasai portoenterostomy: when is it too late? J. Pediatr. Surg. 2001; 36:97–99. 132. Chardot, C., Carton, M., Spire-Bendelac, N. et al. Is the Kasai operation still indicated in children older than 3 months diagnosed with biliary atresia? J. Pediatr. 2001; 138:224–228. 133. Caccia, G., Dessanti, A., & Alberti, D. An 8-year experience on the treatment of extrahepatic biliary atresia: results in 72

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

cases. In Kasai, M., ed. Biliary Atresia and its Related Disorders. Amsterdam: Excerpta Medica, 1983:181–184. Beath, S., Pearmain, G., Kelly, D. et al. Liver transplantation in babies and children with extrahepatic biliary atresia: J. Pediatr. Surg. 1993; 28:1044–1047. Martinez Ibanez, V., Iglesias, J., Lloret, J. et al. 7 years experience with hepatic transplantation in children. Cir. Pediatr. 1993; 6:7–10. Valayer, J., Gauthier, F., Yandza et al. Biliary atresia: results of long-term conservative treatment and of liver transplantation. Transpl. Proc. 1993; 25:3290–3292. Goss, J. A., Shackleton, C. R., Swenson, K. et al. Orthotopic liver transplantation for congenital biliary atresia. An 11 year single-centre experience. Ann. Surg. 1996; 224:276–284. Diem, H. V., Evrard, V., Vinh, H. T. et al. Pediatric liver transplantation for biliary atresia: results of primary grafts in 328 recipients. Transplantation 2003; 75:1692–1697. Ohi, R. Biliary atresia: long-term outcomes. In Howard, E. R., Stringer, M. D., & Colombani, P. M., eds. Surgery of the Liver, Bile ducts and Pancreas in Children, 2nd edn. London:Arnold, 2002; 133–147. Nio, M., Ohi, R., Hayashi, Y. et al. Current status of 21 living patients surviving more than 20 years after surgery for biliary atresia. J. Pediatr. Surg. 1996; 31:381–384. Howard, E. R., MacLean, G., Nio, M. et al. Survival patterns in biliary atresia and comparison of quality of life of long-term survivors in Japan and England. J. Pediatr. Surg. 2001; 36:892– 897. Hadzic, N., Davenport, M., Tizzard, S. et al. Long-term survival following Kasai portoenterostomy: is chronic liver disease inevitable? J. Pediatr. Gastroenterol. Nutr. 2003; 37:430– 433. Shimaoka, S., Ohi, R., Saeki, M. et al. Problems during and after pregnancy of former biliary atresia patients treated successfully by the Kasai procedure. J. Pediatr. Surg. 2001; 36:349– 351. Barkin, R. M. & Lilly, J. R. Biliary atresia and the Kasai operation: continuing care. J. Pediatr., 1980; 96:1015–1019.

36 Choledochal cyst Takeshi Miyano Department of Pediatric General and Urogenital Surgery, Juntendo University School of Medicine, Bunkyo-ku, Tokyo, Japan

Introduction Choledochal cyst is a congenital dilatation of the common bile duct and has the potential to become malignant. It has a hereditary predisposition, which may explain the higher incidence seen in Asia, particularly in Japan, and its familial occurrence in siblings and twins.1,2 It was first reported by Douglas in 18523 and may or may not be associated with congenital dilatation of the intrahepatic bile ducts (Fig. 36.1). Choledochal cyst is almost always associated with an abnormal junction between the pancreatic and common bile duct, i.e., pancreaticobiliary malunion. This allows pancreatic secretions to reflux into the biliary tree, and bile to flow into the pancreatic duct, causing various pathologic changes in the biliary tract, pancreas, and liver.4,5 Because concurrent abnormalities of the pancreatic duct and intrahepatic bile ducts are almost always present, the importance of cholangiography in the planning of surgical management cannot be overemphasized. If these anomalies go unnoticed, aberrant pancreatic anatomy might be damaged, causing serious postoperative morbidity. Internal drainage of the choledochal cyst, a procedure popular in the past, has been abandoned for more than a decade now because of a prohibitive incidence of postoperative complications such as recurrent cholangitis, cholelithiasis, and biliary duct cancer.6 Primary cyst excision with biliary reconstruction to avoid two-way reflux of bile and pancreatic secretions is now the standard surgical treatment of choice.7 Surgery for choledochal cyst is generally successful and a satisfactory surgical outcome with low morbidity is expected in the short term. However, on mid- to long-term

follow-up, there are increasing reports of post-cyst excision complications, including recurrent cholangitis, intrahepatic bile duct stone formation, relapsing pancreatitis, stone formation in the intrapancreatic residual terminal choledochus, and malignancy.6–22 Long-term follow-up is mandatory and must be thorough. Recently, laparoscopic cyst excision and biliary reconstruction techniques have been introduced.23–26 These advanced laparoscopic techniques are not widely used as yet, and there are no results available for mid- to long-term follow-up.

Classification There are various types of choledochal cyst. Alonso-Lej et al.,27 Todani et al.,28 and Komi et al.29 each described classifications of choledochal cysts based on anatomy, cholangiography of the hepatic ducts, and the pancreaticobiliary junction. However, we prefer to classify choledochal cysts into groups according to the presence or absence of pancreaticobiliary malunion (Fig. 36.2). Cystic or fusiform types are the most common while other types such as diverticulum of the common bile duct, choledochocele and Caroli’s disease are extremely rare in children in the author’s experience. Todani et al.30 also reported only cystic or fusiform types in his series of 103 cases.

Protocol for surgical management Before surgery, the precise anatomy of the entire hepatobiliary-pancreatic tract, including the intrahepatic and

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


465

466

T. Miyano

Fig. 36.1. Intraoperative cholangiograms. (a) Cystic choledochal dilatation (b) Fusiform choledochal dilatation (c) Forme fruste choledochal cyst. P: Pancreatic duct. Arrow: Junction between the pancreatic duct and the common bile duct. Arrowheads: long common channel.

(vi) hepaticojejunostomy (end-to-end anastomosis) (vii) roux-en-Y biliary reconstruction

Cholangiography A

D

B

E

C

F

Fig. 36.2. Classification of choledochal cysts. With pancreaticobiliary malunion: (a) Cystic type. (b) Fusiform type (c) Forme fruste. Without pancreaticobiliary malunion: (d) Cystic diverticulum of the common bile duct. (e) Choledochocele (diverticulum of the distal common bile duct). (f ) Intrahepatic bile duct dilatation alone (Caroli’s disease). (Reproduced from Miyano T, Yamataka A. Choledochal cysts. Curr. Opin. Pediatr. 1997; 9:284.)

extrahepatic bile ducts, and the intrapancreatic duct must be obtained. Major steps in the surgical management of choledochal cyst are: (i) cholangiography (ii) cyst excision (iii) intraoperative endoscopy (iv) dissection and excision of the distal common bile duct (v) adequate excision of the common hepatic duct at the correct level

Recent developments in imaging technology such as the introduction of magnetic resonance cholangiopancreatography (MRCP) and improvements in endoscopic retrograde cholangiopancreatography (ERCP) now allow the anatomy of the hepato-biliary-pancreatic ductal system to be visualized preoperatively in most choledochal cyst cases.31 Before cyst excision, detailed information about intrahepatic bile duct anomalies such as ductal stenosis, dilatation, and the presence of debris/stones (Fig. 36.3), as well as intrapancreatic bile duct anomalies such as the type of pancreaticobiliary malunion, presence of debris/protein plugs in the common channel (Fig. 36.4(a)), and dilatation of the pancreatic duct must be obtained. MRCP is generally reliable in children over 3 years old (Fig. 36.4(a)), but if the patient is an infant or younger child, pancreaticobiliary malunion may not be demonstrated clearly. If pancreatitis is present, ERCP is generally contraindicated, so intraoperative cholangiography is then the only option for obtaining information about the entire anatomy of the hepato-biliary-pancreatic duct system.

Cyst excision There are usually more adhesions between a cystic type choledochal cyst and surrounding vital structures such as the portal vein and hepatic artery compared with a fusiform

Choledochal cyst

A

C

A

X

b

O a

c B

B Fig. 36.3. (a) Preoperative MRCP showing extensive debris within dilated intrahepatic bile ducts (large arrows) and non-dilated intrahepatic bile ducts (small arrows). C: choledochal cyst. (b) Intrahepatic bile duct debris seen through the pediatric cystoscope. (Reproduced from Shimotakahara A. et al. Pediatr. Surg. Int. 2004; 20:68.)

type choledochal cyst, especially if the patient is an older child. In adolescents and adults with a cystic type choledochal cyst, adhesions are often very dense, and great care is required during cyst excision. Prior to dissection of the cyst, we always open the anterior wall of the choledochal cyst transversely (Fig. 36.5(a)). Because anatomical variants of the common hepatic duct are sometimes associated with a cystic type choledochal cyst, this incision should be made below the middle of the cyst. By opening the cyst anteriorly, the posterior wall of the cyst is visible directly from the inside (Fig. 36.5(b), (d)), and dissection of the choledochal cyst from surrounding

Fig. 36.4. (a) MRCP showing a fusiform choledochal cyst, long common channel, protein plugs (arrowheads), and pancreatic duct (P). V: Papilla of Vater. Arrow indicates junction between the common bile duct and the pancreatic duct. (Reproduced from Miyano, T. & Yamataka, A. Choledochal cysts. Curr. Opin. Pediatr. 1997; 9:285.) (b) a. O is the appropriate level of excision. X is inadequate. b. Incomplete excision leads to residual cyst formation and protein plugs. c. Intraoperative endoscopy allows safe and complete cyst excision and irrigation of the common channel. (From Miyano, T. (2006). Choledochal cyst. In Pediatric Surgery (Springer Surgery Atlas Series), ed. P. Puri and M. Hollwarth, Berlin: Springer-Verlag.)

tissues is easier than dissecting the cyst in toto. If the cyst is extremely inflamed and adhesions are very dense, a mucosectomy (Fig. 36.5(c), (d)) rather than full-thickness dissection should be performed (Fig. 36.5(b), (d)) to minimize the risk of damaging surrounding structures such as the portal vein and hepatic artery. Compared with the cystic type, adhesions between the extrahepatic duct and surrounding structures are less dense in fusiform choledochal cysts, especially in young children. As a result, the common hepatic duct can be safely dissected free from the portal vein and hepatic artery, after

467

468

T. Miyano

A

B1

B2

C1

C2

Intraoperative endoscopy is also useful for identifying debris in dilated intrahepatic bile ducts, determining the ideal level of resection for the common hepatic duct, and for assessing the severity of any intrahepatic bile duct stenosis. A high incidence (26%) of intrahepatic bile duct debris not usually detected by preoperative radiologic investigations has been reported from the author’s institution.32 Without endoscopy, there is a tendency to overlook debris in the intrahepatic bile ducts noted on preoperative imaging. Furthermore, debris can be present in non-dilated intrahepatic bile ducts, although it is far more common when these ducts are dilated. These facts indicate that intraoperative endoscopic examination of intrahepatic bile ducts is mandatory even if preoperative radiologic investigations have shown no debris or abnormality.

Dissection and excision of the distal common bile duct

D Fig. 36.5. (a) Opening the anterior wall of the choledochal cyst transversely. (b1) The posterior wall of the cyst is visible from the inside and this facilitates dissection of the choledochal cyst. (b2) Full-thickness dissection of the cyst. (C1, C2) If the cyst is extremely inflamed and adhesions dense, mucosectomy should be performed. (d) Arrows indicate the appropriate layer for full-thickness dissection and mucosectomy. (From Miyano, T. (2006). Choledochal cyst. In Pediatric Surgery (Springer Surgery Atlas Series), ed. P. Puri and M. Hollwarth, Berlin: Springer-Verlag.)

the anterior wall of the common bile duct is incised as described above.

Intraoperative endoscopy Before the introduction of intraoperative endoscopy, it was difficult to safely excise the pancreatic portion of a fusiform type choledochal cyst at an adequate level for fear of injuring the pancreatic duct (Fig. 36.4(b)). We began performing intraoperative endoscopy routinely using a neonatal or pediatric cystoscope in 1986.17 Intraoperative endoscopy allows safe and complete excision of the intrapancreatic portion of a fusiform type choledochal cyst without damaging the pancreatic duct; it also enables the common channel to be irrigated and cleared of any debris or protein plugs. We believe that intraoperative endoscopy reduces the risk of postoperative complications such as recurrent pancreatitis, stone formation, and carcinoma.17

A fusiform type choledochal cyst is usually associated with complicated pancreaticobiliary malunion as well as debris/protein plugs in the common channel (Fig. 36.4). Pancreatic duct anomalies are also often present.27,33 Thus, dissection of the distal common bile duct in a fusiform type choledochal cyst should be performed with great care so as to avoid leaving behind any intrapancreatic portion of the common bile duct and to prevent injury to the pancreatic duct. Intraoperative endoscopy of the distal common bile duct should be performed in all cases, but especially in patients with a fusiform type choledochal cyst. Fusiform type choledochal cyst is associated with a specific complication shown in Fig. 36.4. If the distal common bile duct is resected along line X (Fig. 36.4(b)a), a cyst will gradually reform around the distal duct left within the pancreas, leading to recurrent pancreatitis and/or stone formation or malignancy within the cyst (Fig. 36.4(b)b).33 In contrast, if the distal duct is resected along line O just above the pancreaticobiliary ductal junction, this complication is unlikely to develop (Fig. 36.4(b)a). With a cystic type choledochal cyst (Fig. 36.5), the distal common bile duct is often narrow; occasionally, it is so narrow that it cannot be identified and the cyst appears blind-ended. In such cases, mucosectomy or full-thickness cyst excision need only be performed to the level of the pancreaticobiliary junction.

Adequate excision of the common hepatic duct at the correct level The ideal length of the common hepatic duct required for anastomosis is approximately 10 mm, since a longer common hepatic duct may become kinked leading to bile stasis in the intrahepatic bile ducts. However, the lumen of

Choledochal cyst

the common hepatic duct should be inspected before it is shortened, because ductal anomalies such as luminal stenosis, an aberrant duct opening, or a septum may be present. These anatomical variations have been described in the literature and encountered by the author16,30 and can affect the success of hepaticoenterostomy after cyst excision. Intraoperative endoscopy can also be used to inspect the common hepatic duct and minimize difficulties arising from duct anomalies.17 A cystic type choledochal cyst is often associated with stenosis and dilatation of intrahepatic bile ducts. The incidence of postoperative complications such as recurrent cholangitis, stone formation, and anastomotic stricture is increased in patients with intrahepatic bile duct dilatation.8–10 Surgical correction is required for severe congenital intrahepatic duct stenosis, especially if the proximal intrahepatic bile ducts are very dilated. However, the incidence of late complications secondary to stenosis or dilatation of intrahepatic ducts is generally low, especially in young children.8 Therefore, excessive surgical intervention may be unnecessary except in cases where there is massive dilatation of peripheral intrahepatic bile ducts with severe downstream stenosis. If intrahepatic bile duct dilatation persists after definitive surgery, careful follow-up is mandatory. When ductal stenosis is present at the hepatic hilum or at the first branch of the left or right hepatic ducts, surgical dilatation or duct plasty can be performed.16,34 However, if the stenosis is located more peripherally, treatment is difficult. Large diffusely dilated intrahepatic bile ducts in both lobes cannot be treated but if cholangitis or stone formation affects a localized area of dilated intrahepatic ducts, liver resection may be indicated at a later stage.

Hepaticojejunostomy (end-to-end anastomosis) After cyst excision and Roux-en-Y hepaticojejunostomy, a good short-term outcome is expected. However, with time, complications related to Roux-en-Y hepaticojejunostomy in childhood develop often enough to be of concern; they are mostly related to elongation of the blind pouch at the hepaticojejunostomy or at the Roux-en-Y jejunal anastomosis as a result of the child’s growth.35 Based on experience and complications reported in the literature,6–13 the author’s institution developed the following recommendations to prevent complications related to the Roux-en-Y hepaticojejunostomy.35 An end-to-end anastomosis at the Roux-en-Y hepaticojejunostomy is recommended if the ratio between the diameters of the common hepatic duct and the proximal Rouxen-Y jejunum at the proposed site of anastomosis is less

A b/a ≤ 2.5

a b

b/a ≤ 2.5

B

O

C

×

D

Fig. 36.6. Hepaticojejunostomy. (a) End-to-end anastomosis. (b) End-to-side anastomosis as close as possible to the closed end of the Roux loop. (c) If an end-to-side anastomosis is performed far from the closed end of the Roux loop, elongation of the blind pouch will occur leading to complications. (d) An adequate Roux-en-Y hepatico-jejunostomy. Arrowheads indicate the approximated native jejunum and distal Roux limb. Arrows indicate smooth passage without reflux of small bowel contents. (a)–(c) from Miyano T. (2006). Choledochal cyst. In Pediatric Surgery (Springer Surgery Atlas Series), ed. P. Puri and M. Hollwarth, Berlin: Springer-Verlag. (d) from Yamataka A. et al. Recommendations for preventing complications related to Roux-en-Y hepatico-jejunostomy performed during excision of choledochal cyst in children. J. Pediatr. Surg. 2003; 38:1831.)

than or equal to 1 (common hepatic duct): 2.5 (jejunum) (Fig. 36.6(a)) – this prevents elongation of the blind pouch. If an end-to-side anastomosis is unavoidable, the common hepatic duct should be anastomosed as close as possible to the closed end of the Roux loop (Fig. 36.6(b)); if this anastomosis is constructed some distance away from the blind end of the Roux loop (Fig. 36.6(c)), elongation of the blind pouch will occur later in life as the child grows, causing bile stasis in the pouch and intrahepatic bile ducts (especially if they are dilated) and subsequent cholangitis and

469

470

T. Miyano

stone formation. An end-to-end hepaticojejunostomy or a modified end-to-side anastomosis as mentioned above prevents these complications.

Debris and stones

Endoscope

Roux-en-Y biliary reconstruction Tortuous

Some surgeons predetermine the length of the Roux-en-Y jejunal limb (e.g., 30 cm, 40 cm, 50 cm, or 60 cm) without considering the size of the child, which results in an unnecessarily long Roux-en-Y jejunal limb, especially in infants and younger children. Redundancy of the Roux-enY limb is likely to occur later in life as the patient grows and may cause bile stasis in the Roux loop itself as well as the intrahepatic bile ducts, leading to cholangitis and/or stone formation. Thus, the length of the Roux-en-Y limb should be individualized so that the Roux-en-Y jejunojejunostomy fits naturally into the splenic flexure after it is returned to the peritoneal cavity (Fig. 36.6(d)). In this way, redundancy of the Roux-en-Y limb can be prevented. When a jejunojejunostomy and Roux-en-Y limb are used, both the native jejunum and the Roux-en-Y jejunal limb proximal to the jejunojejunostomy should be approximated for up to 8 cm to ensure that both the bile in the Roux-en-Y limb and the contents of the native jejunum flow smoothly down into the distal jejunum (Fig. 36.6(d)). If this approximation is not performed, the jejunojejunostomy tends to be T-shaped, and there may be reflux of jejunal contents into the Roux-en-Y limb, leading to dilatation and biliary stasis in the Roux loop, a situation the author recently encountered in an 18-year-old girl who had had cyst excision and Roux-en-Y hepaticojejunostomy at another hospital 17 years earlier.36 Biliary stasis in the Rouxen-Y limb in this patient resulted in stone formation in the intrahepatic bile ducts (Fig. 36.7). She required endoscopic removal of intrahepatic bile duct stones and revision of the Roux-en-Y jejunojejunostomy.

Short-term outcome The postoperative course of primary cyst excision with biliary reconstruction is generally uneventful in children, and the short-term outcome is excellent with few exceptions reported in the literature.6,37–40 Perioperative death due to primary cyst excision is extremely rare in children. The author’s institution and affiliated hospitals have performed primary cyst excision with no operative mortality in 176 children with choledochal cyst aged 15 years or less. Early post excision complications seen occasionally include minor leakage at the hepaticojejunostomy, elevation in pancreatic or hepatic enzymes (asymptomatic in the author’s series), and adhesive bowel obstruction.

Dilated

T–shaped jejunojejunostomy

Fig. 36.7. In this case, a flexible endoscope was inserted into the intrahepatic bile ducts and massive debris and multiple biliary stones were removed using an electro hydraulic lithotripsy device inserted through the endoscope. The patient had a tortuous, dilated Roux-en-Y loop and T-shaped jejunojejunostomy, and there appeared to be reflux (arrow) of jejunal contents up the Roux loop. The jejunojejunostomy and the dilated distal part of the Roux loop (between the dotted lines) were revised. (Reproduced from Shima, H. et al. Intracorporeal electrohydraulic lithotripsy for intrahepatic bile duct stone formation after choledochal cyst excision: a case report. Pediatr. Surg. Int. 2004; 20:71.)

All can usually be managed conservatively and rarely require emergency surgery except for severe adhesive bowel obstruction.6,8 The incidence of minor leakage at the hepaticoenterostomy anastomosis has been reported to range from 4 to 14%,6,41 although this complication was not encountered in the author’s series. Leakage is probably a result of severe inflammation in the wall of the common hepatic duct at the anastomosis rather than technical failure, and is therefore more likely in adolescents and adults. Appropriate placement of a Penrose drain is extremely important in keeping the hepaticoenterostomy anastomosis drained postoperatively, thereby helping to prevent abscess formation due to a small bile leak. After cyst excision, bleeding from the operative site, gastrointestinal bleeding due to peptic ulceration (possibly secondary to Roux-en-Y hepaticojejunostomy), acute pancreatitis, and pancreatic fistula have all been reported30 but the author has not personally experienced these postoperative complications.

Long-term outcome Based on the author’s experience, the risk of post-cyst excision complications is reduced in patients who undergo cyst

Choledochal cyst

Table 36.1. Type of cyst excision and hepaticoenterostomy and postoperative complication rate for children vs. adults8 Type of excision

Number of children Number of adults

Primary cyst excision 176 (13) [7.4%] Cyst excision after 5 (0) [0%] cystoenterostomy Cyst excision after other 19 (5) [26.3%] biliary surgerya

22 (6) [27.3%] 9 (5) [55.6%]

Total

40 (17) [42.5%]

200 (18) [9.0%]

9 (6) [66.7%]

Note: Numbers in parentheses indicate number of patients who had post-cyst excision complications, percent in brackets indicate post-cyst excision complication rate. a Percutaneous transhepatic biliary drainage, percutaneous cyst drainage, T-tube drainage, cholecystectomy, choledochotomy or sphincterotomy.

excision as young children (at 5 years of age or less) compared to older children or adults.8 Many published series of choledochal cysts6–15,30 include a large number of adults and other patients who had secondary cyst excision after previous cystoenterostomy (no longer recommended). As a result, these reports may not reflect the true incidence of complications after primary cyst excision in children. This chapter provides a more accurate description of long-term complications after primary cyst excision in children with a choledochal cyst. Thus, data from children are presented separately from adult data and from patients who underwent secondary cyst excision after previous cystoenterostomy or other drainage procedures.

“Cystic”- or “fusiform”-type choledochal cysts Children vs. adults The author has reviewed 200 children and 40 adults who had a choledochal cyst excised between 1964 and 1995 and who have been followed up for more than 8 years. Patients were classified as children if they underwent surgery when less than 16 years of age, and as adults if they had a cyst excised at 16 years of age or older. Choledochal cysts were either cystic or fusiform. Complications and their surgical management were of particular interest. Children Of the 200 children, 145 were aged 5 years or less at the time of cyst excision; the overall mean age at cyst excision was 4.2 years. One hundred and seventy-six had primary cyst excision, 5 had cyst excision after previous cystoenterostomy, and 19 had cyst excision after other types of biliary surgery (Table 36.1). Of the 200 hepaticoenterostomies, 159

were end-to-side anastomoses (147 Roux-en-Y hepaticojejunostomies, 11 standard hepaticoduodenostomies, and 1 jejunal interposition hepaticoduodenostomy), and 41 were end-to-end Roux-en-Y hepaticojejunostomies. The age at onset of symptoms (abdominal pain and/or jaundice and/or abdominal mass) was 5 years or less in 175 children, and between 6 and 15 years in the remaining 25 children (overall mean 3.0 years). Intraoperative endoscopy was performed in 70 children treated between 1986 and 1995. The mean duration of follow-up was 17.9 years. Adults The mean age at cyst excision was 35 years. Of the 40 adults, 22 had primary cyst excision, 9 had cyst excision after previous cystoenterostomy, and 9 had cyst excision after other types of biliary surgery (Table 36.1). Of the 40 hepaticoenterostomies, 38 were end-to-side Roux-en-Y hepaticojejunostomies, and 2 were end-to-end hepaticojejunostomies. Eleven adults were symptomatic as children (less than 16 years of age) and the remaining 29 became symptomatic at a later age (mean 26 years). The mean duration of follow-up was 17.7 years. The time interval between the onset of initial symptoms and cyst excision was significantly less in children than in adults (P < 0.0001: ANOVA). The number of children and adults with post-cyst excision complications was 18/200 (9.0%) and 17/40 (42.5%), respectively (Table 36.1). The complication rate in children was significantly lower than in adults (P < 0.0001: 2 ). Complications after cyst excision in children Twenty-five complications after cyst excision were seen in 18 children (Table 36.2). Fifteen of the 18 children required surgical intervention (Table 36.3). Intrahepatic bile duct stone formation occurred in three children who had primary cyst excision; two had stones at the porta hepatis, and one had stones at the porta hepatis and in the left lobe of the liver. All three had intrahepatic bile duct dilatation prior to cyst excision that persisted postoperatively. Two of the three children with stones had a stricture at the end-toside hepaticojejunostomy (aged 7 and 12 years at cyst excision). The remaining child had no anastomotic stricture but there was an abnormally long blind pouch of jejunum at the end-to-side hepaticojejunostomy (Fig. 36.8), bile stasis in dilated intrahepatic bile ducts exacerbated by this long blind pouch was considered to be the cause of stone formation. The two children with strictures presented with ascending cholangitis and the other patient with recurrent epigastric pain. Recurrent ascending cholangitis without intrahepatic bile duct stone formation was seen in the child who was treated at another hospital by cyst excision and a jejunal interposition hepaticoduodenostomy. Revision

471

472

T. Miyano

Table 36.2. Complications after cyst excision and hepaticoenterostomy: children vs. adults8

Complication Ascending cholangitis Intrahepatic bile duct stones Intrapancreatic terminal choledochal calculi Pancreatic duct calculus Stones in the blind pouch of the end-to-side Roux loop Bowel obstruction Cholangiocarcinoma Liver dysfunction Pancreatitis Total

Incidence in Incidence in 200 children 40 adults 3 3 3a

9 5 1

1 1a

1 0

9 0 0 5

3 2 1 5

25 (18)

A

27 (17)

Note: The numbers in parentheses indicate the number of patients who had post-cyst excision complications (18 children and 17 adults had 25 and 27 complications, respectively). a One patient with intrapancreatic terminal choledochal calculi also had a stone in the blind pouch of an end-to-side hepaticojejunostomy.

dilated HD stones Hepaticojejunostomy

Table 36.3. Surgical management of post-cyst excision complications in children vs. adults8

Surgical management Revision of hepaticoenterostomy Percutaneous transhepatic cholangioscopic lithotomy Left hepatic lobectomy Excision of residual intrapancreatic terminal choledochus Endoscopic sphincterotomy Pancreaticojejunostomy (Puestow procedure) Exploratory laparotomy for malignancy Laparotomy for bowel obstruction Total

Number of children

Number of adults

4a 1

2 2

0 2b

1 0

1 1

2 0

0

2

6

1

15

10

a

One child also required left lateral segmentectomy of the liver. One child also required excision of the blind pouch of the end-toside Roux-en-Y hepaticojejunostomy which contained stones. b

of the hepaticoenterostomy was performed in three cases and percutaneous transhepatic cholangioscopic lithotomy in one. Stone formation in the residual intrapancreatic

B

elongated blind pouch

Fig. 36.8. (a) Percutaneous transhepatic cholangiogram. Arrowheads show an elongated blind pouch at the end-to-side hepaticojejunostomy. The arrow shows stones in the dilated intrahepatic bile ducts. (b) Diagram of the findings.

terminal choledochus developed postoperatively in three children after primary cyst excision (Fig. 36.9). Pancreatic duct stones developed in one patient 11 years after secondary cyst excision. Four patients presented with recurrent pancreatitis; two had excision of a remnant intrapancreatic terminal choledochus, one had stones removed by endoscopic sphincterotomy, and the remaining patient had a pancreaticojejunostomy. One additional child had pancreatitis without stone formation after primary cyst excision, and was treated medically. Nine children developed adhesive bowel obstruction. One required revision of the Roux-en-Y hepaticojejunostomy; in this case, cyst excision with hepaticojejunostomy had been performed at another hospital and dense adhesions had formed between the Roux-en-Y limb, the

Choledochal cyst

Table 36.4. Stone formation after cyst excision and hepaticoenterostomy: children vs. adults8

Number of patients with stones

145 children aged < or = 5 yra

55 children >5 yra

40 adults >15 yra

0 (0%)b

7c (12.7%)

7 (17.5%)

Note: Percentages in parentheses indicate the incidence of postcyst excision stone formation. a Age at cyst excision. b P < 0.0001 for children aged < or = 5 yr vs. children aged >5 yr or adults. c One child formed two stones (one in the residual intrapancreatic terminal choledochus and the other in the blind pouch of the endto-side Roux loop).

Fig. 36.9. ERCP showing stones (arrow) in the residual intrapancreatic terminal choledochus after excision of fusiform type choledochal cyst. (Reproduced from Yamataka, A. et al. Complications after cyst excision with hepaticoenterostomy for choledochal cyst and their surgical management in children versus adults. J. Pediatr. Surg. 1997; 32:1099.)

duodenum, and an abnormally elongated blind pouch of the end-to-side hepaticojejunostomy. After adhesiolysis and excision of the redundant pouch, the end-to-side anastomosis was converted to an end-to-end anastomosis. In the remaining eight patients, six required adhesiolysis and two were treated conservatively. Stone formation after cyst excision occurred in seven (3.5%) children, all of whom underwent cyst excision at the age of 6 years or more (Table 36.4). One patient had stone formation both in the residual intrapancreatic terminal choledochus and in the blind pouch of the end-to-side hepaticojejunostomy. Neither stone formation nor anastomotic stricture was seen in the 145 children who underwent cyst excision at 5 years of age or less. The incidence of stone formation after cyst excision in children aged 5 years or less was significantly lower than in older children (12.7%) and adults (17.5%) (P < 0.0001: 2 ). It is worth emphasizing that there was neither stone formation, anastomotic stricture, nor cholangitis after cyst excision in the 70 children who had intraoperative endoscopy, and in the 41 children who underwent endto-end hepaticojejunostomy.

Complications after cyst excision in adults Twenty-five complications were seen after cyst excision in 17 (42.5%) adults (Table 36.2). Ten of these 17 adults required surgical intervention (Table 36.3). Intrahepatic bile duct stone formation at the porta hepatis occurred in four cases, all of whom had secondary cyst excision after a previous cystoenterostomy. Two of these patients had an anastomotic stricture of an end-to-side hepaticojejunostomy. Intrahepatic bile duct stones developed in the left lobe of the liver in one patient 10 years after secondary cyst excision (Fig 36.10). Four of these five patients also had intrahepatic bile duct dilatation before cyst excision that persisted postoperatively. All five presented with ascending cholangitis. Revision of the hepaticoenterostomy was performed in two, percutaneous transhepatic cholangioscopic lithotomy in two, and left hepatic lobectomy in one. Stone formation in the residual intrapancreatic terminal choledochus developed in one adult and a pancreatic duct stone in another. The former had undergone secondary cyst excision after cystoenterostomy, and the latter had had primary cyst excision. Both presented with recurrent pancreatitis. Both patients underwent endoscopic sphincterotomy and stone removal. Two adults died of cholangiocarcinoma at 54 years and 29 years of age, respectively. The former had secondary cyst excision at the age of 48 years after a previous cholecystectomy at the age of 32 years, and developed carcinoma in a dilated intrahepatic bile duct in the left lobe. The other patient underwent primary cyst excision at the age of 25 years, and carcinoma developed within a large remnant intrapancreatic terminal choledochus. Both patients presented with ascending cholangitis and both had an exploratory laparotomy.

473

474

T. Miyano

Dilated intrahepatic bile ducts Intrahepatic bile duct dilatation typically improves or resolves after cyst excision, particularly in young children.39 However, it tends to persist in adolescents and adults. This is because histologic damage to intrahepatic bile ducts appears to be reversible in young children but not in adolescents and adults. Dilatation of peripheral intrahepatic bile ducts in adults is associated with late complications such as stone formation and recurrent cholangitis, and is thus occasionally managed by liver resection, intrahepatic cystoenterostomy, or balloon dilatation of duct stenoses at the time of cyst excision.18,28,41,42 In contrast, the incidence of late complications is much less in children with intrahepatic bile duct dilatation than in adults, and so similar surgical interventions may be unnecessary, although these children must be followed-up very carefully.

Fig. 36.10. Preoperative cholangiogram and pathologic sections of resected specimen from an adult who required a left lobectomy for hepatolithiasis. (a) Cholangiogram suggests stones (arrowheads) in the left hepatic lobe and a patent hepaticojejunostmy (arrows). (b) Sections of resected left hepatic lobe showing stones in dilated intrahepatic bile ducts (arrows). (Reproduced from Yamataka A. et al. Complications after cyst excision with hepaticoenterostomy for choledochal cyst and their surgical management in children versus adults. J. Pediat. Surg. 1997; 32:1100.)

Two adults with intrahepatic bile duct dilatation had ascending cholangitis without stone formation; both had had secondary cyst excision. Three adults had recurrent pancreatitis without stone formation after secondary cyst excision and one further adult had postoperative liver dysfunction. These six subjects were treated medically. Three adults developed adhesive bowel obstruction, one of whom required laparotomy and adhesiolysis.

Stone formation in intrahepatic bile ducts In the author’s experience, the risk factors for intrahepatic bile duct stone formation are intrahepatic bile duct dilatation, anastomotic stricture, and residual debris in the intrahepatic bile ducts.32

Anastomotic stricture In the author’s series, an anastomotic stricture developed in four patients who had had cyst excision at the ages of 7, 12, 16, and 19 years, respectively. There appears to be an increased incidence of stricture in older children undergoing cyst excision; no anastomotic strictures developed in the 145 children who had cyst excision at 5 years of age or less. Inflammation of the cyst wall tends to be mild in children under 10 years of age and more severe in older patients, indicating that histologic damage to the common hepatic duct at the site of the hepaticoenterostomy is more marked in older children and adults.39 Bile stasis in dilated intrahepatic bile ducts probably further exacerbates inflammatory changes at the anastomosis. A previous cystoenterostomy is also likely to have caused irreversible histologic damage to the common hepatic duct and intrapancreatic ducts. Our four patients who had an anastomotic stricture either had intrahepatic bile duct dilatation or a previous cystoenterostomy. Anastomotic leakage secondary to technical problems can cause the development of granulation tissue that may result in an anastomotic stricture. The fate of a hepaticoenterostomy in young children (5 years of age or less at cyst excision) is probably related more to surgical skill than histologic damage to the common hepatic duct. Mortality or serious morbidity secondary to an anastomotic stricture and intrahepatic bile duct stone formation have been widely reported in young children.9–12,43 Common hepatic ducts are generally larger in older children and adults and the hepaticoenterostomy is usually easier than in young children. However, cyst excision is more difficult in adolescents and adults because of the presence of severe pericyst inflammation, and histologic damage of the common

Choledochal cyst

hepatic duct is more marked. Consequently, an anastomotic stricture occurring in older children and adults is more likely to be caused by inflammatory changes rather than surgical inexperience. Residual debris in the intrahepatic bile ducts Stones may occasionally develop in dilated intrahepatic bile ducts even when there is no anastomotic stricture or bile duct stenosis.9,11–13 Choledochal cysts in adults are frequently associated with stones or debris in the biliary tract before surgery.14,22,44 Debris in the intrahepatic bile ducts is also commonly seen in children at the time of cyst excision.32 This can sometimes be impacted. Intraoperative endoscopy is beneficial in such cases to ensure that the ducts are cleared of debris. In the author’s series, there was no instance of intrahepatic bile duct stone formation in the 70 children who had intraoperative endoscopy. In contrast, in the subjects who developed intrahepatic bile duct stones in our series and in previous reports, intraoperative endoscopy had not been performed.9–11 Thus, residual debris in the intrahepatic bile ducts at the time of cyst excision is likely to be a risk factor for stone formation after cyst excision.32 Stone formation in the intrapancreatic residual choledochus Stone formation may occur as a result of stasis of pancreatic secretions in the intrapancreatic terminal choledochus (from incomplete excision of the terminal choledochus at the time of cyst excision) or from residual protein plugs/debris within the common channel. Many surgeons divide the distal part of the choledochal cyst just above the pancreatic head during cyst excision, even if the cyst is fusiform. However, by doing this, a large intrapancreatic terminal choledochus, which is the distal end of the choledochal cyst, often ends up being left behind, particularly when the choledochal cyst is fusiform. This problem can also occur in cystic type choledochal cysts in adults, because of the difficulty in excising the distal part of the cyst in the presence of severe pericystic inflammation around the pancreas. To prevent stone formation in the residual intrapancreatic choledochus, complete excision of the distal part of the choledochal cyst should be performed under intraoperative endoscopic control.33,45 Malignancy Although excision of the cyst removes a potential site of malignant change, it does not prevent the development of carcinoma38 (see also Chapter 20). The relative risk for carcinoma after excision is still higher than that in the

general population, although the risk is decreased by approximately 50% by cyst excision.46 The interval between primary cyst excision and the development of cancer ranges from 1.8 to 25 years.47 This indicates that the epithelium of the remnant bile duct wall may have already progressed to a precancerous stage at the time of surgery. There are case reports of cholangiocarcinoma arising from the intrahepatic bile ducts, within the intrapancreatic residual terminal choledochus, and at the hepaticojejunostomy anastomosis.19–21 The author has had no cases of cholangiocarcinoma in children but two adults have died from cholangiocarcinoma, one after primary cyst excision at the age of 25 years. There is one report of a cholangiocarcinoma developing in the intrapancreatic residual terminal choledochus after primary cyst excision at the age of 14 years.19 Chronic bile stasis in the intrahepatic bile ducts may be responsible for the development of carcinoma in intrahepatic bile ducts.6 Carcinoma in the intrapancreatic residual choledochus could be due to continuous reflux of pancreatic secretions into this segment. If so, complete excision of the intrapancreatic terminal choledochus should reduce this risk.

“Forme fruste” choledochal cyst Okada et al.48 were the first to report a case of a common pancreaticobiliary channel with minimal dilatation of the common bile duct in 1981 and called it “common channel syndrome.” It was later renamed “forme fruste choledochal cyst” by Lilly et al.49 in 1985 who described pancreaticobiliary malunion with little or no dilatation of the common bile duct (Fig. 36.1(c)). The optimal treatment of this condition in adults is controversial50,51 but, in children, the common bile duct should always be excised because of the long-term risk of bile duct malignancy from pancreaticobiliary reflux associated with the common channel.52 There are only a few reports on forme fruste choledochal cyst in the English-language literature.53–55 We have experience of 17 (6%) cases from a total of 281 choledochal cysts treated by cyst excision between 1965 and 2002 at the author’s institute or affiliated hospitals. The maximum diameter of the normal common bile duct in children has been reported to vary from 3 to 6 mm.56,57 Minimal dilatation of the common bile duct is therefore defined as a maximum diameter of less than 10 mm. Pancreaticobiliary malunion was identified in all 17 of our patients using cholangiography. The mean diameter of the common bile duct was 7.8 mm (range 4 to 10 mm). Age at diagnosis ranged from 1 to 7 years (mean 2.5 years). Presenting

475

476

T. Miyano

symptoms were those of pancreatitis, i.e. abdominal pain associated with an elevated serum amylase, with or without jaundice. Mean age at common bile duct excision was 2.9 years. Fourteen patients had Roux-en-Y hepaticojejunostomy (end-to-end anastomosis in 5; end-to-side in 9), and 3 had hepaticoduodenostomy, using a single layer of 5–0 absorbable sutures. Interrupted sutures were used for both the anterior and posterior walls of the anastomosis in 8 cases, continuous sutures for both walls in 6 cases, and a combination of a continuous suture for the posterior wall and interrupted sutures for the anterior wall in 3 cases. Histopathology of the excised common bile duct showed mucosal ulceration and/or slough in 35%, fibrosis in 53%, and inflammatory cell infiltration in 41%. The common bile duct in forme fruste choledochal cyst is therefore fragile and inflamed.48,58 Consequently, hepaticoenterostomy can be difficult and tenuous, with a greater potential for anastomotic complications.49 The common channel contained protein plugs/debris in nine cases (54%), identified either by MRCP preoperatively or by endoscopic examination of the common channel intraoperatively. Intrahepatic bile ducts were dilated in 10 cases (59%) on preoperative MRCP and half contained debris. Intraoperative endoscopy was successfully used to remove all debris. Intrahepatic bile duct dilatation resolved in nine of ten cases postoperatively. All patients are currently well after a mean postoperative follow-up period of 10.8 years (range 3 to 19 years), although one has occasional attacks of mild epigastric pain. No postoperative morbidity (e.g., ascending cholangitis, anastomotic stricture, or pancreatitis) was observed. The majority of forme fruste choledochal cysts in adults have been found as an incidental finding during investigation of an abnormality of the gall bladder such as a tumor, polyp, or thickening of the wall.59,60 Simple cholecystectomy without biliary reconstruction is often performed for adult cases.55 In contrast, the majority of children with forme fruste choledochal cyst present with pancreatitis. A forme fruste choledochal cyst must be considered in a child with recurrent pancreatitis of unknown cause. MRCP or ERCP should confirm or exclude the presence of pancreaticobiliary malunion, which is a pathognomonic feature.

Recent research Hepaticojejunostomy vs. hepaticoduodenostomy Primary cyst excision with biliary reconstruction is generally accepted as the treatment of choice for choledochal

cyst. The type of biliary reconstruction is based on the personal preference of the surgeon – some use Roux-en-Y hepaticojejunostomy, others hepaticoduodenostomy, and a few use other methods such as jejunal graft interposition hepaticoduodenostomy.15,16,35,61 A case of hilar bile duct carcinoma developing 19 years after primary cyst excision and hepaticoduodenostomy at the age of 13 months has recently been reported.62 The author’s institution reviewed 86 children who had primary cyst excision with either Roux-en-Y hepaticojejunostomy (n = 74: end-to-end anastomosis in 58; end-to-side anastomosis in 16) or hepaticoduodenostomy (n = 12) between 1986 and 2002, with special emphasis on postoperative complications related to the type of biliary reconstruction.63 Forty-six hepaticojejunostomy cases were associated with intrahepatic bile duct dilatation and were excluded in order to focus on the results of biliary reconstruction alone. Hepaticoduodenostomy was not used for biliary reconstruction if intrahepatic bile duct dilatation was present. Thus, 28 children who had hepaticojejunostomy were compared with 12 children who had hepaticoduodenostomy. Mean duration of follow-up was 8.7 years after hepaticojejunostomy and 7.9 years after hepaticoduodenostomy. Differences between the hepaticojejunostomy and hepaticoduodenostomy groups with respect to type of choledochal cyst, age at cyst excision and length of follow-up were not statistically significant. However, the incidence of postoperative complications such as endoscopy-proven bilious gastritis due to duodenogastric reflux of bile (4/12 [33%] of the hepaticoduodenostomy group) and adhesive bowel obstruction (2/28 (7%) of the hepaticojejunostomy group) were significantly different. It would appear that hepaticoduodenostomy is not an ideal biliary reconstructive technique after choledochal cyst excision because of the high incidence of duodenogastric bile reflux. Currently, Roux-enY hepaticojejunostomy is the technique of choice in such patients. Hepaticoenterostomy at the hepatic hilum is used by some surgeons because they consider that a wider anastomosis created by incising the lateral wall of both transected hepatic ducts at the hepatic hilum may prevent anastomotic stricture and may also allow better drainage of bile.9,64 Although hepaticoenterostomy at the hepatic hilum is appealing technically, it is a more complicated procedure and is indicated in selected cases only such as adolescents with severe inflammation of the common hepatic duct requiring secondary cyst excision, and children with intrahepatic bile duct dilatation and relative stenosis at the porta hepatis.

Choledochal cyst

Prenatally diagnosed choledochal cyst There are increasing reports on the management of the asymptomatic choledochal cyst detected prenatally. Some pediatric surgeons recommended primary cyst excision soon after birth in such cases or soon after diagnosis in neonates.65,66 The author’s experience is that cyst excision need not be performed hastily in these infants; rather they should be thoroughly assessed and surgery should be planned and performed by experienced, well-trained pediatric surgeons.66–68 If obstructive jaundice is severe, external biliary drainage is recommended either by percutaneous transhepatic cholangiocatheter or direct percutaneous cyst drainage. Delayed primary cyst excision may then be carried out 1 or 2 months later.

Conclusions With improvements in diagnostic imaging and prenatal ultrasound screening more cases of choledochal cyst are being detected. The author recommends choledochal cyst excision and hepaticojejunostomy with an end-to-end anastomosis. Intraoperative endoscopy is easy and highly effective in preventing many long-term complications and should be regarded as routine. The key to successful management of a choledochal cyst is early diagnosis and a clean hepatobiliary–pancreatic system. Surgical expertise, a good understanding of the anatomy of the hepatobiliary– pancreatic system, and careful long-term follow-up are also important factors in achieving good long-term outcomes.

REFERENCES 1. Miyano, T. & Yamataka, A. Choledochal cysts. Curr. Opin. Pediatr. 1997; 9:283–288. 2. Lane, G. J. Different types of congenital biliary dilatation in dizygotic twins: a case report. Pediatr. Surg. Int. 1999; 15:403– 404. 3. Douglas, A. H. Case of dilatation of the common bile duct. Monthly. J. Med. Sci. 1852; 14:97–99. 4. Babbitt, D. P. Congenital choledochal cyst: new etiologic concepts on anomalous relationships of the common bile and pancreatic bulb. Ann. Radiol. 1969; 12:231–241. 5. Miyano, T., Suruga, K., & Suda, K. Abnormal choledochopancreatico ductal junction related to the etiology of infantile obstructive jaundice diseases. J. Pediatr. Surg. 1979; 14:16–25. 6. Todani, T., Watanabe, Y., Toki, A. et al. Reoperation for congenital choledochal cyst. Ann. Surg. 1987; 207:142–147.

7. Metcalfe, M. S., Wemyss-Holden, S. A., & Maddern, G. J. Management dilemmas with choledochal cysts. Arch. Surg. 2003; 138:333–333. 8. Yamataka, A., Ohshiro, K., Okada, Y. et al. Complications after cyst excision with hepaticoenterostomy for choledochal cyst and their surgical management in children versus adults. J. Pediatr. Surg. 1997; 32:1097–1102. 9. Todani, T., Watanabe, Y., Urushihara, N. et al. Biliary complications after excisional procedure for choledochal cyst. J. Pediatr. Surg. 1995; 30:478–481. 10. Ohi, R., Yaoita, S., Kamiyama, T. et al. Surgical treatment of congenital dilatation of the bile duct with special reference to late complications after total excisional operation. J. Pediatr. Surg. 1990; 25:613–617. 11. Arima, T., Suita, S., Kubota, M. et al. Endoscopic treatment of intrahepatic gallstones after surgery for choledochal cyst. Pediatr. Surg. Int. 1995; 10:218–220. 12. Schier, F., Clausen, M., Kouki, M. et al. Late results in the management of choledochal cysts. Eur. J. Pediatr. Surg. 1994; 4:141– 144. 13. Chijiwa, K. & Tanaka, M. Late complications after excisional operation in patients with choledochal cyst. J. Am. Coll. Surg. 1994; 179:139–144. 14. Nagorney, D. M., MacIrath, D. C., & Adson, M. A. Choledochal cysts in adults: clinical management. Surgery 1984; 96:656–663. 15. Todani, T., Watanabe, Y., Mizuguchi, T. et al. Hepaticoduodenostomy at the hepatic hilum after excision of choledochal cyst. Am. J. Surg. 1981; 142:584–587. 16. Miyano, T., Yamataka, A., Kato, Y. et al. Hepaticoenterostomy after excision of choledochal cyst in children: A 30year experience with 180 cases. J. Pediatr. Surg. 1996; 31:1417– 1412. 17. Miyano, T., Yamataka, A., Kato, Y. et al. Choledochal cysts: special emphasis on the usefulness of intraoperative endoscopy. J. Pediatr. Surg. 1995; 30:482–484. 18. Ando, H., Ito, T., Kaneko, K. et al. Intrahepatic bile duct stenosis causing intrahepatic calculi formation following excision of a choledochal cyst. J. Am. Coll. Surg. 1996; 183:56–60. 19. Yoshikawa, K., Yoshida, K., Shirai, Y. et al. A case of carcinoma arising in the intrapancreatic terminal choledochus 12 years after primary excision of a giant choledochal cyst. Am. J. Gastroenterol. 1986; 81:378–384. 20. Yamamoto, J., Shimamura, Y., Ohtani, I. et al. Bile duct carcinoma arising from the anastomotic site of hepaticojejunostomy after the excision of congenital biliary dilatation: a case report. Surgery 1996; 119:476–479. 21. Chaudhuri, P. K., Chaudhuri, B., Schuler, J. J. et al. Carcinoma associated with congenital cystic dilatation of bile ducts. Arch. Surg. 1982; 117:1349–1351. 22. Deziel, D. J., Rossi, R. L., Munson, J. L. et al. Management of bile duct cysts in adults. Arch. Surg. 1986; 121:410–415. 23. Shimura, H., Tanaka, M., Shimizu, S. et al. Laparoscopic treatment of congenital choledochal cyst. Surg. Endosc. 1998; 12:1268–1271.

477

478

T. Miyano

24. Mori, T., Abe, N., Sugiyama, M. et al. Laparoscopic hepatobiliary and pancreatic surgery: an overview. J. Hepatobiliary. Pancreat. Surg. 2002; 9:710–722. 25. Tanaka, M., Shimizu, S., Mizumoto, K. et al. Laparoscopically assisted resection of choledochal cyst and Roux-en-Y reconstruction. Surg. Endosc. 2001; 15:545–552. 26. Watanabe, Y., Sato, M., Tokui, K. et al. Laparoscope-assisted minimally invasive treatment for choledochal cyst. J. Laparoendosc. Adv. Surg. Tech. A. 1999; 9:415–418. 27. Alonso-Lej, F., Rever, W. B., & Pessagno, D. J. Congenital choledochal cyst, with a report of 2, and an analysis of 94 cases. Int. Abst. Surg. 1959; 108:1–30. 28. Todani, T., Narusue, M., Watanabe, Y. et al. Management of congenital choledochal cyst with intrahepatic involvement. Ann. Surg. 1977; 187:272–280. 29. Komi, N., Takehara, H., Kunitomo, K. et al. Does the type of anomalous arrangement of the pancreaticobiliary ducts influence the surgery and prognosis of choledochal cyst? J. Pediatr. Surg. 1992; 27:728–731. 30. Todani, T. Choledochal cysts. In Stringer, M. D., Oldham, P., Mouriquand, D. E., Howard, E. R. eds. Pediatric Surgery and Urology: Long Term Outcomes. London: WB Saunders; 1998; 417–429. 31. Yamataka, A., Kuwatsuru, R., Shima, H. et al. Initial experience with non-breath-hold magnetic resonance cholangiopancreatography: a new noninvasive technique for the diagnosis of the choledochal cyst in children. J. Pediatr. Surg. 1997; 32:1560–1562. 32. Shimotakahara, A., Yamataka, A., Kobayashi, H. et al. Massive debris in the intrahepatic bile ducts in choledochal cyst: possible cause of postoperative stone formation. Pediatr. Surg. Int. 2004; 20:67–69. 33. Yamataka, A., Segawa, O., Kobayashi, H. et al. Intraoperative pancreatoscopy for pancreatic duct stone debris distal to the common channel in choledochal cyst. J. Pediatr. Surg. 2000; 35:1–4. 34. Ando, H., Kaneko, K., Ito, F. et al. Operative treatment of congenital stenoses of the intrahepatic bile ducts in patients with choledochal cysts. Am. J. Surg. 1997; 173:491–494. 35. Yamataka, A., Kobayashi, H., Shimotakahara, A. et al. Recommendations for preventing complications related to Rouxen-Y hepatico-jejunostomy performed during excision of choledochal cyst in children. J. Pediatr. Surg. 2003; 38:1830– 1832. 36. Shima, H., Yamataka, A., Yanai, T. et al. Intracorporeal electrohydraulic lithotripsy for intrahepatic bile duct stone formation after choledochal cyst excision: a case report. Pediatr. Surg. Int. 2004; 20:70–72. 37. Lilly, J. R. Total excision of choledochal cyst. Surg. Gyn. Obst. 1978; 146:254–256. 38. Joseph, V. T. Surgical techniques and long-term results in the treatment of choledochal cyst. J. Pediatr. Surg. 1990; 25:782– 787. 39. O’Neill, J. A., Jr. Choledochal cyst. Curr. Prob. Surg. 1992; 29:361– 410.

40. Stringer, M. D., Dhawan, A., Davenport, M. et al. Choledochal cysts: lessons from a 20 year experience. Arch. Dis. Child. 1995; 73:528–531. 41. Warren, K. W., Kune, G. A., Hardy, K. J. et al. Biliary duct cysts. Surg. Clin. North. Am. 1968; 88:567–577. 42. Eagle, J. & Salmon, P. A. Multiple choledochal cysts. Arch. Surg. 1964; 88:345–349. 43. Csentino, C. M., Luck, S. R., & Raffensperger, J. G. et al. Choledochal duct cyst. Resection with physiological reconstruction. Surgery 1992; 112:740–748. 44. Voyles, C. R., Smadja, C., Shands, W. C. et al. Carcinoma in choledochal cysts. Arch. Surg. 1983; 118:986–988. 45. Long, L., Yamataka, A., Segawa, O. et al. Coexistence of pancreas divisum and septate common channel in a child with choledochal cyst. J. Pediatr. Gastroenterol. Nutr. 2001; 32:602– 604. 46. Kobayashi, S., Asano, T., Yamasaki, M. et al. Risk of bile duct carcinogenesis after excision of extrahepatic bile ducts in pancreaticobiliary maljunction. Surgery 1999; 126:939–944. 47. Ng, W. T. In “Letters to the editors”. Surgery 2000; 128: 492– 494. 48. Okada, A., Nagaoka, M., Kamata, S. et al. “Common channel syndrome”: anomalous junction of the pancreatico-biliary ductal system. Z. Kinderchir. 1981; 32:144–151. 49. Lilly, J. R., Stellin, G. P., Karrer, F. M. Forme fruste choledochal cyst. J. Pediatr. Surg. 1985; 20:449–451. 50. Aoki, T., Tsuchida, A., Kasuya, K. et al. Is preventive resection of the extrahepatic bile duct necessary in cases of pancreaticobiliary maljunction without dilatation of the bile duct? Jpn J. Clin. Oncol. 2001; 31:107–111. 51. Murakami, T., Kodama, T., Takesue, Y. et al. Anomalous arrangement of the pancreaticobiliary ductal system without the dilatation of the biliary tract. Surg. Today 1992; 22:276– 279. 52. Morishita, H., Kamei, K., Funabiki, T. et al. A case of pancreaticobiliary maljunction without bile duct dilatation associated with gallbladder cancer and minute bile duct cancer. Tando 1992; 6:176–183. 53. Miyano, T., Ando, K., Yamataka, A. et al. Pancreaticobiliary maljunction associated with nondilatation or minimal dilatation of the common bile duct in children: Diagnosis and treatment. Eur. J. Pediatr. Surg. 1996; 6:334–337. 54. Ando, H., Ito, T., Nagaya, M. et al. Pancreaticobiliary maljunction without choledochal cysts in infants and children: Clinical features and surgical therapy. J. Pediatr. Surg. 1995; 30:1658– 1662. 55. Tanaka, K., Nishimura, A., Yamada, K. et al. Cancer of the gallbladder associated with anomalous junction of the pancreaticobiliary duct system without bile duct dilatation. Br. J. Surg. 1993; 80:622–624. 56. Witcombe, J. B. & Cremin, B. J. The width of the common bile duct in childhood. Pediatr. Radiol. 1978; 7:147–149. 57. Schulman, M. H., Ambrosino, M. M., Freeman, P. C. et al. Common bile duct in children: sonographic dimensions. Radiology 1995; 195:193–195.

Choledochal cyst

58. Pushparani, P., Redkar, R. G., & Howard, E. R. Progressive biliary pathology associated with common pancreato-biliary channel. J. Pediatr. Surg. 2000; 35:649–651. 59. Sugiyama, M. & Atomi, Y. Anomalous pancreaticobiliary junction without congenital choledochal cyst. Br. J. Surg. 1998; 85:911–916. 60. Mori, K, Akimoto, R., Kanno, M. et al. Anomalous union of the pancreaticobiliary ductal system without dilatation of the common bile duct or tumor: case reports and literature review. Hepato-Gastroenterology. 1999; 46:142–148. 61. Rao, K. L. N., Mitra, S. K., Kochher, R. et al. Jejunal interposition hepaticoduodenostomy for choledochal cyst. Am. J. Gastroenterol. 1987; 82:1042–1045. 62. Todani, T., Watanabe, Y., Toki, A., et al. Hilar duct carcinoma developed after cyst excision followed by hepaticoduodenostomy. In Koyanagi, Y., Aoki, T., (eds) Pancreaticobiliary Maljunction, Tokyo: Igaku tosho shuppan; 2002; 17–21.

63. Shimotakahara, A., Yamataka, A., Yanai, T. et al. Roux-en-Y hepaticojejunostomy or hepaticoduodenostomy for biliary reconstruction during the surgical treatment of choledochal cyst; which is better? Pediatr. Surg. Int. 2005; 21:5–7. 64. Lilly, J. R. Surgery of coexisting biliary malformations in choledochal cyst. J. Pediatr. Surg. 1979; 14:643–647. 65. Burnweit, C. A., Birken, G. A., & Heiss, K. The management of choledochal cysts in the new born. Pediatr. Surg. Int. 1996; 11:130–133. 66. Lugo-Vicente, H. L. Prenatally diagnosed choledochal cysts: observation or early surgery. J. Pediatr. Surg. 1995; 30:1288– 1290. 67. Lane, G. J., Yamataka, A., Kohn, S. et al. Choledochal cyst in the newborn. Asian J. Surg. 1999; 22:310–312. 68. Miyano, T. Congenital biliary dilatation. In Puri P., ed. Newborn Surgery, Oxford, UK: Butterworth-Heinemann; 1996; 433– 439.

479

37 Biliary stone disease Faisal G. Qureshi,1 Evan P. Nadler,2 and Henri R. Ford3 1

Division of Pediatric Surgery, Children’s Hospital of Los Angeles, CA, USA New York University School of Medicine, NY, USA 3 Division of Pediatric Surgery, Children’s Hospital of Pittsburgh, University of Pittsburgh School of Medicine, PA, USA 2

Introduction Over the past decade, cholelithiasis and choledocholithiasis have been diagnosed with increasing frequency during infancy and childhood.1–3 The increased rate of diagnosis may be related to a true rise in the incidence of the disease, or, more likely, to an enhanced ability to detect gallstones.2–4 The prevalence of gallstones in the pediatric population has been reported to be between 0.13% and 0.22%.5,6 However, in children who undergo an abdominal sonogram for abdominal pain, the incidence of gallstones and sludge has been reported to be as high as 1.9%.7 The mean age for cholelithiasis in pediatric patients is between 7 and 10 years.2,7–9 Most authors report a slight preponderance of boys among pre-adolescents with cholelithiasis. However, this trend is completely reversed in the adolescent group.2,7 Although underlying hematologic diseases such as sickle cell anemia, hereditary spherocytosis and thalassemia have been implicated as major predisposing factors for childhood cholelithiasis, the majority of gallstones in children are believed to be idiopathic. Several series suggest that only 20% of gallstones are related to hematologic diseases.8,10,11 Other putative risk factors for childhood cholelithiasis include: total parenteral nutrition; ileal resection; ileal disorders; obesity; family history of gallstones; cystic fibrosis; biliary tract anomalies and medications (birth control pills, cyclosporin, ceftriaxone).7,8,12–17 Gallstones can be classified as pigment, cholesterol or mixed-type stones. Pigment stones are usually detected during infancy and early childhood, and typically are associated with hemolytic disorders. In contrast, cholesterol and mixed-type stones are more

commonly seen in adolescents.18,19 Because pigmented stones are more radio-opaque than cholesterol stones, plain abdominal radiographs are useful in diagnosing gallstones in up to 50% of children.20,21 Children with symptomatic gallstones present most commonly with right upper quadrant pain (75%–85%), followed by nausea or vomiting in 60%. Jaundice is less frequently seen and epigastric tenderness is found in only one-third of the patients. Gallstones can be asymptomatic in up to 17% of children.7,8 Medical therapy is ineffective in children with symptomatic cholelithiasis; cholecystectomy is the treatment of choice.11,22 Laparoscopic cholecystectomy (LC) has become standard of care in children.23 In this chapter, we will discuss the different aspects of management and the long-term outcomes of biliary stone disease in children.

Management of gallstones in asymptomatic patients The relatively frequent use of abdominal ultrasound in children with abdominal disorders in recent years has resulted in the diagnosis of gallstones in a large number of “asymptomatic” children. Despite the fact that they comprise up to 17% of children with gallstones, very little information exists regarding the appropriate management of children with asymptomatic cholelithiasis.7,8 In addition, because these data were collected on patients undergoing workup for abdominal symptoms, the true incidence of asymptomatic gallstones in children is unknown. In adults, previously asymptomatic patients develop biliary symptoms at a rate of 2% to 4% per year, with relatively

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


480

Biliary stone disease

little associated morbidity or mortality.24 The most recent reports studying the natural history of asymptomatic gallstones in children are summarized in Table 37.1. In one study by Bruch et al., 3 of 41 patients underwent cholecystectomy while another 4 developed symptoms but had not undergone cholecystectomy. However, mean followup was relatively short (21 months). Although the number of patients in each of the various series is relatively small, based upon the low rate of development of symptoms and the absence of complications, clinical follow-up of patients with asymptomatic gallstones without associated comorbidities is advisable. A similar recommendation can be made for patients with gallbladder sludge as this condition usually resolves spontaneously.7

Complications of biliary stone disease in children Patients with biliary stone disease may develop a number of complications (Table 37.2) which include: (i) Biliary colic or chronic abdominal pain (ii) Cholecystitis (iii) Common bile duct obstruction (iv) Pancreatitis (v) Acute cholangitis Based on several reports, chronic and intermittent abdominal pain are the most common complications of biliary stone disease in children.7–9 These symptoms usually consist of postprandial right upper quadrant or epigastric abdominal pain, nausea, vomiting, and jaundice. Acute cholecystitis complicates biliary stone disease in 3%–14% of the patients; the diagnosis relies on the clinical findings of acute right upper quadrant abdominal pain, fever, and leukocytosis (Murphy’s triad). Patients with cholangitis present with similar findings and jaundice; however, cholangitis complicating biliary stone disease is relatively rare, affecting only 2% of patients.9 The incidence of choledocholithiasis in pediatric patients has been reported to be between 0% and 5%.25,26 However, more recent reports suggest that the incidence may be as high as 18% among patients with gallstones.8,27 Common bile duct (CBD) stones are found more commonly in patients with hematological disorders and may be related to the smaller size of pigment stones which allow them to pass into the CBD more easily.28 Gallstone pancreatitis, resulting from the passage of gallstones into the CBD, has been seen in up to 7% of patients with biliary stone disease. The management of CBD stones, biliary stone disease in patients with hematological disorders, and the management and outcomes of patients with gallstone pancreatitis will be discussed in subsequent sections of this chapter.

481

Table 37.1. Follow-up of children with asymptomatic gallstones

Author

Follow-up Number that Asymptomatic Mean/range developed patients (months) symptoms Complications

Bruch, 200096 41 Wesdorp, 20007 16 10 Kumar, 20008

21 (3–41) 7 (33%) 54 (2–216) 1 (2%) 114 (12–228) 0 (0%)

0 0 0

Table 37.2. Complications of biliary stone disease in children. NA = not available Author, year and number of patients

Cholecystitis

CBD stones

Pancreatitis

Holcomb, 1999, n = 10023 Waldhausen, 1999, n = 1212 Miltenburg, 2000, n = 1329 Kumar, 2000, n = 1028 Wesdorp, 2000, n = 827 Bruch, 2000, n = 7496

7% 4% 14% 3% 4% 4%

2% 7% NA 18% 0% 3%

5% 7% 4% 1% 2% 3%

Safety, efficacy and cost effectiveness of laparoscopic cholecystectomy (LC) in children With an increasing number of children diagnosed with gallstones, the optimal management must be discussed. Laparoscopic cholecystectomy has now supplanted open cholecystectomy (OC) as the gold standard for management of symptomatic gallstones in adults.29 Similarly, LC has gained widespread acceptance in the management of symptomatic gallstones in children. Several authors have examined the safety, efficacy and cost effectiveness of this procedure in the pediatric population. Holcomb et al. first reported the safety and efficiency of LC for the treatment of acute cholecystitis in five children.30 These patients had no evidence of complications during the follow-up period of 16 months (range 2–24 months). In this early report, the authors also compared elective LC versus open cholecystectomy (OC) in children. Patients undergoing elective LC had shorter hospital stay, reduced analgesic requirement and decreased total hospital charges. Since that report, several authors have compared LC with OC in children (Table 37.3). As can be seen, postoperative analgesic requirements, length of hospital stay and overall hospital costs are improved in the LC group. Operative time and operative costs are higher in the LC group, but these are offset by the shorter hospital stay. LC has now become the standard of care in managing gallstone disease in children.

482

F. G. Qureshi, E. P. Nadler and H. R. Ford

Table 37.3. Laparoscopic vs. open cholecystectomy. Comparison of postoperative analgesic requirements, length of hospital stay, hospital costs and complication rates between LC and OC. Only data corresponding to the LC group are presented. NA = not available Operative Hospital Hospital Analgesics time stay costs Complications Kim, 199597 Al-Salem, 199728 Jawad, 199856 Luks, 199998

Decreased NA Decreased NA

Equal Greater Equal Greater

Shorter Shorter Shorter Shorter

Less NA Less Less

Equal Equal Equal NA

The complication rate of LC in children is 0%– 15.5%.2,23,31 Although this range seems rather wide, it probably reflects the more stringent classification of complications in these reports. For instance, Esposito et al. reported 17/110 patients with complications, which included 11 gallbladder perforations, 1 “fall of stones into the abdominal cavity” and 5 trocar site infections.31 The inclusion of gallbladder perforation and loss of stones into the abdominal cavity as complications of LC may have skewed the data. Most authors report a complication rate of less than 5%, with trocar site infections being most common.2,28 Patients undergoing urgent LC for acute cholecystitis and especially those patients with sickle cell disease or other significant comorbid conditions have a higher rate of complications.9,32 Patients with sickle cell disease are prone to respiratory compromise and re-hospitalization with abdominal pain, whilst those with cardiac comorbidity have a higher incidence of multisystem organ failure.9,33 Very little information regarding the true rate of bile duct injury following LC in children is known. In two series of over 100 patients each, there were no reports of bile duct injury or the development of a biloma.23,33

Diagnosis and treatment of common bile duct stones and gallstone pancreatitis Although most authors report the incidence of CBD stones to be around 5% among children with biliary stone disease, some suggest that CBD stones may be present in up to 18% of this population.8,25,26 In patients with hematologic disorders, the incidence approaches 26%, perhaps due to the smaller size of pigment stones that allows them to pass easily into the CBD.34,35 Clinical, laboratory and radiographic tools have been used to predict or diagnose CBD stones in adults. How-

ever, their utility in diagnosing CBD stones in children is not known. Two retrospective reviews of 182 patients revealed that clinical, laboratory and ultrasound findings were inaccurate in predicting CBD stones in up to 43% of children.7,27,36 Because these studies were retrospective in nature, a prospective study with stringent criteria will be required to determine whether non-invasive testing and clinical presentation can accurately predict CBD stones.

Endoscopic retrograde cholangiopancreatography (ERCP) and intraoperative cholangiography (IOC) Endoscopic retrograde cholangiopancreatography has been used both as a diagnostic and therapeutic tool in children with choledocholithiasis from 1 month to 18 years of age with a success rate of up to 95%.37–40 Reported complication rates are similar to those seen in adults, with postERCP pancreatitis occurring in up to 8% of children. The incidence of pancreatitis is even higher in children undergoing therapeutic ERCP. Hemorrhage and perforation are seen in 0.3% to 2% of children undergoing ERCP.38 A conventional duodenoscope can be used for children above 10 years of age, while a pediatric duodenoscope is used in children under 10 years of age. More recently a small caliber videoduodenoscope for ERCP in very young children and infants has been developed and successfully used for diagnosis and extraction of stones.41 In patients with biliary stone disease, ERCP has been noted to be safe and efficacious before, during or after laparoscopic cholecystectomy with CBD clearance achieved in 95% of patients.7,16,27,42 CBD stones were extracted by endoscopic sphincterotomy and the use of a balloon or basket. Newman et al. have suggested that preoperative ERCP may be most efficacious if preoperative work-up suggests or demonstrates CBD stones.36 Concomitant ERCP with LC can increase operative time by 86% and thus may impact on operative costs.23 The disadvantages of ERCP include complication rates of up to 8% and the need for general anesthesia in young children.43 Thus, intraoperative cholangiography (IOC) may be a more appropriate initial step in patients with biliary stone disease and a suspected CBD stone. Holcomb et al. performed IOC in 57 patients undergoing LC.23 These authors reported success in 49 patients, with an overall increase in operative time of 29%. Kumar et al. attempted 88 IOCs and reported a 100% success rate without any complication related to the IOC.8 To determine the role of IOC in children with biliary stone disease, Waldhausen et al. performed 63 IOCs in 100 children undergoing LC; 55 had successful studies.27 CBD stones were found in 18 patients. IOC increased operative time by 35% and did not result in any

Biliary stone disease

complications. Preoperative symptoms and/or laboratory findings were inaccurate in predicting CBD stones in 8 patients. Based on these findings, Waldhausen et al. concluded that routine IOC in all children undergoing LC was warranted since preoperative work-up was inaccurate. Furthermore, they argued that IOC could help avoid unnecessary ERCP and the obligatory second anesthetic. Callery et al. also recommend routine IOC in all children undergoing LC to demonstrate choledocholithiasis and define biliary tract anatomy.44 The ideal management of suspected choledocholithiasis in children in the era of laparoscopic cholecystectomy is unknown. Options include preoperative or postoperative ERCP with sphincterotomy and removal of stones; intraoperative ERCP during LC, or IOC during LC with laparoscopic or open common bile duct exploration.16 Newman et al. have suggested that preoperative ERCP should be performed on patients with suspected CBD stones followed by LC.36 Despite their recommendation, however, 42% of children had negative ERCP and therefore underwent an unnecessary procedure. The high rate of negative ERCP in patients with suspected CBD stones has been documented by other authors.27,33 Based on these findings, Tagge et al. and Shah et al. have recommended that children undergoing LC should also undergo IOC with concomitant laparoscopic common bile duct exploration either via the cystic duct or via the CBD if the IOC reveals CBD stones.33,104 Neither group reported any biliary tract injuries or longterm complications. Preoperative ERCP has been reserved for those children with fulminant pancreatitis, cholangitis, or those with CBD obstruction. Intraoperative ERCP is reserved for those patients whose CBD stones either cannot be managed via laparoscopy, or the surgical endoscopist is unavailable. Postoperative ERCP is reserved for those patients in whom CBD stones are retained or recur after LC (Fig. 37.1). Some authors have suggested that expectant management of small, asymptomatic CBD stones is acceptable, but this approach has not gained widespread acceptance.45

Magnetic resonance cholangiopancreatography (MRCP) Little is known about the utility of MRCP in predicting CBD stones in children. Topal et al. have shown that, in adults with a predicted probability of CBD stones of greater than 5%, MRCP can confirm the presence or absence of stones, with an observed sensitivity of 95%, a specificity of 100%, a positive predictive value of 100% and a negative predictive value of 98%.46 The model used by these authors to predict common duct stones included ultrasonography, which revealed CBD stones or bile duct dilatation, age greater than

Symptomatic Gallstones

Laparoscopic Cholecystectomy and IOC

+IOC

Cholangitis, Fulminant Pancreatitis or CBD obstruction

−IOC

Pre-op ERCP

Laparoscopic CBD Exploration

Successful

Unsuccessful

Intra- or Post-op ERCP

Fig. 37.1. Management of symptomatic gallstones and common bile duct stones in children. IOC = intraoperative cholangiogram, CBD = common bile duct, ERCP = endoscopic retrograde cholangiopancreatography.

60 years, fever, serum alkaline phosphatase level above 670 units/l and serum amylase level above 95 units/l. Although MRCP has been shown to be accurate in predicting pancreatobiliary disease in children, its role in predicting CBD stones in this population is not known at present.47

Gallstone pancreatitis Gallstone pancreatitis occurs in 1%–5% of children with cholelithiasis and is an absolute indication for cholecystectomy (Table 37.2). Affected children should undergo laparoscopic cholecystectomy with IOC following initial supportive care and resolution of the acute bout of pancreatitis. Holcomb et al. reported five patients with gallstone pancreatitis who underwent LC and IOC. None of these patients had documented CBD stones.23 Rescorla and colleagues suggest that children with gallstone pancreatitis should be treated initially with supportive care followed by delayed LC and IOC once laboratory values have normalized. This approach reduced the rate of positive IOC from 63%–75% to 2%–5%, and therefore obviated the need for CBD exploration/ERCP in a significant number of patients in their hands.11 Pancreatitis has also been associated with gallbladder sludge. Choi et al. reported their experience with biliary disease-related pancreatitis in eight children; six patients had sludge and two had gallstones.48 All patients with biliary sludge underwent ERCP with sphincterotomy and endoscopic naso-biliary drainage. Three of the six patients with gallbladder sludge subsequently developed recurrent pancreatitis within 3 years, while two patients had a followup of less than a year and one patient did not develop any

483

484

F. G. Qureshi, E. P. Nadler and H. R. Ford

Table 37.4. Biliary dyskinesia in children. Outcome in children undergoing cholecystectomy for biliary dyskinesia

Author, year

n

Ejection fraction

Lugo-Vicente, 199799 Dumont, 1999100 Gollin, 1999101 Al-Homaidhi, 2002102 Wood, 2004103

12 42 29 10 5

seven times per day) the syndrome is denoted overactive bladder or urge incontinence. There are three distinct clinical patterns of bladder dysfunction with possible evolvement from one pattern into another. The first is the overactive bladder or urge syndrome and urge incontinence with normal micturation; the second is dysfunctional voiding with staccato or fractionated micturition. The final step is detrusor underactivity with large residual urine volumes in a bladder with weak detrusor contractions. This latter condition is often complicated by recurrent UTIs and/or vesicoureteral reflux (VUR). If there is a direct relationship between these three “nonneurogenic bladder–sphincter dysfunctions” it is not clear, but it is likely that some of the patients with dysfunctional voiding may end up with detrusor underactivity which may continue into adulthood.4,17,18

Holding manoevers that will strengthen the pelvic floor and sphincter

??

? Dysfunctional voiding

??

? ??

VUR

Constipation

Table 50.1. Treatment of bladder dysfunction in childhood

Detrusor underactivity (“Lazy bladder”)

Fig. 50.1. Possible flowchart of “bladder dysfunction” in childhood.

Epidemiology Nocturnal enuresis ranks high among common childhood disorders, affecting approximately 10% of 7-yearolds.10,19 However, the prevalence data reported for “non-neuropathic bladder-sphincter dysfunction” are often difficult to interpret, since several different definitions and classifications have been used. The term has been used for all sorts of bladder malfunction between the extremes of simple urgency symptoms and severe cases of bladder dysfunction that may affect the upper urinary tract. Several surveys of incontinence in different age groups of children have recently been published.7–9,12,20 In a study on 3556 7-year-old Swedish schoolchildren, Hellstr¨om et al. found that 26% had at least one symptom suggesting incomplete bladder control, including nocturnal enuresis. Daytime wetting was seen in 6% of the girls and 2.0% of the boys, nocturnal enuresis without daytime incontinence in 2.3% of the girls and 7.0% of the boys, and combined day- and night-time wetting in 2.3% of the girls and 2.0% of the boys.8 Similar prevalence figures in young schoolchildren have been found by other investigators in different countries.7–9,12 Bakker et al. examined micturition problems and voiding habits in a slightly older population of 4332 schoolchildren aged 10–14 years and found daytime wetting with or without enuresis in 8% and only 1% with isolated nocturnal enuresis.7 Although there are many studies describing the prevalence of urinary incontinence in women and men in various age groups,21–24 the prevalence of bladder dysfunction and/or incontinence in adolescents or young men and women has not been studied in detail.

Non-neuropathic bladder–sphincter dysfunction

Enuresis Older notions about psychological causes behind this psychologically distressing condition have now been abandoned in favor of endocrinological, urological and sleeprelated mechanisms. The emerging consensus is that the enuretic child wets his or her sheets because of a combination of, on the one hand (i) nocturnal polyuria and/or (ii) nocturnal detrusor hyperactivity, and on the other (iii) high arousal thresholds. Proof of the first mechanism came with the pioneer studies by a Danish group who showed that a subgroup of enuretic children produce large amounts of dilute urine at night due to a relative lack of the antidiuretic hormone vasopressin. 25 Detrusor hyperactivity is implicated indirectly by the large overlap between nocturnal enuresis and urge incontinence8 and has been demonstrated directly in cystometric studies.26 Finally, the formerly controversial “disorder of arousal” theory, i.e., that enuretic children sleep more “deeply” than their night-dry peers, has been corroborated elegantly using graded arousal stimuli during well-defined sleep stages.27 Consequently, enuresis is a heterogenous disorder and different children need different treatments. Two therapies have emerged as safe and reasonably successful firstline alternatives: the vasopressin analogue desmopressin28 and the enuresis alarm.29 The former, taken at bedtime, reduces nocturnal polyuria by antidiuretic action, while the latter is an apparatus that addresses the underlying disorder of arousal by waking the child at the moment of micturition. Anticholinergic, detrusor-relaxant drugs such as oxybutynin or tolterodin are probably useful as a second-line treatment in children suffering from enuresis due to detrusor hyperactivity, but this has not, as yet, been substantiated by randomized, controlled studies.30

Urge-incontinence and overactive bladder As mentioned above, this disturbance is characterized by a recurrent, strong need to void, regardless of bladder filling. It is often associated with an increased voiding frequency (>7 times/day).2,17,31 The urine flow is however normal and there is no residual urine. Simple urge is a “symptom” that most schoolchildren can recognize and Nev´eus et al. found that 35% of 7–10-year-old children recognized urge symptoms.10 It is thus more or less a variant of normality and only the severe forms with urgency/frequency, often combined with wetting, may be treated with urotherapy

and/or anticholinergics. These children void with normal sphincter relaxation and they have a normal, but sometimes high pressure bladder emptying. The term “overactive bladder” has today been accepted as the descriptive term for urge and urge incontinence.14,32 Urge symptoms seem to peak at age 6–9 years and diminish towards puberty with an assumed spontaneous cure rate of at least 14% per year.7,17,33 In childhood it is common that urge symptoms may be provoked by physical activity, but this is not stress incontinence but a sign of an overactive bladder. Stress incontinence occurs at exertion and is caused by an underactive or damaged sphincter. The condition is very rare in neurologically healthy children but common in children with neurogenic bladder. Mixed incontinence is due to the combination of overactive detrusor and sphincter deficiency. This is also rare in children not suffering from neurogenic bladder. A special group of children with symptoms of urgency and holding maneuvers are those with “voiding postponement.”34 The child postpones the voiding until overwhelmed by urgency, then makes a rush to the toilet but often too late so that there is some urge incontinence on the way. A recent study, comparing a cohort of children with “classical” urge syndrome to another group considered to have voiding postponement, has reported a significantly higher frequency of clinically relevant behavioral symptoms in the postponers than in children with urge syndrome. The authors conclude that their findings “support the entity of voiding postponement as an acquired or behavioural syndrome.”

Dysfunctional voiding Dysfunctional voiding (detrusor–sphincter dyscoordination in earlier terminology) is defined as intermittent voiding due to contractions of the sphincter and the pelvic floor muscles, in neurologically normal individuals, often girls 6– 12 years of age. The cause of the dysfunctional voiding and also the urge syndrome is believed to lie in the persistence of an immature bladder control.35 In response to the uninhibited contractions during toilet training most children may “learn” to contract the sphincter to avoid wetting,4,36,37 and many of these girls evolve a form of urge-syndrome as they learn to suppress the urge and voiding by sphincter contractions and different holding maneuvers like squatting (Vincent’s sign).37,38 These patients void in small portions with a staccato or fractionated (interrupted) uroflow curve, leaving residual

645

646

¨ ´ G. Lackgren and T. Neveus

urine in the bladder. This disorder, as well as voiding postponement, is often combined with constipation and encopresis.39,40

Detrusor underactivity (“lazy bladder syndrome”) The term lazy bladder has been used to characterize particularly girls with infrequent voiding and poor detrusor function (detrusor underactivity). It affects somewhat older girls (>10 years of age). The girls have learned to suppress the urge to void and they empty their bladders 1– 3 times a day.6,41 The detrusor becomes decompensated and the bladder capacity is increased and it empties poorly. This syndrome most often presents as chronic or recurrent UTI, with or without incontinence and often vesicoureteral reflux. A characteristic finding is that the child has to strain and often use the abdominal muscles to start and maintain micturition. Postvoid residual is always present but may vary considerably in the same patient. Detrusor underactivity may, in some children, be the end point of a pathogenetic cascade, starting with urge syndrome due to idiopathic overactive bladder (probably genetically determined), continuing with dysfunctional voiding with sphincter overactivity (in an attempt to keep dry), and ending up with increasing residual urine volumes distending the bladder with subsequent loss of detrusor contractile force.42

Complications The most important complications to dysfunctional voiding and detrusor underactivity (“lazy bladder”) are incontinence, recurrent urinary tract infection and, in many cases, vesicoureteral reflux. Urinary tract infections and other bladder-related problems are common not only among children with daytime incontinence, but also bedwetters or former bedwetters.43 Regular and complete voidings are the most efficient ways to prevent urinary infection that may induce a temporary detrusor overactivity and more importantly girls with asymptomatic bacteriuria (ABU) have symptoms of an overactive bladder, such as urgency and incontinence, in a high percentage.40,44 Bladder dysfunction may, in some cases, be the cause of vesicoureteral reflux (VUR) in girls, and the VUR may aggravate and perpetuate the dysfunctional bladder.36,45–47 It is generally considered that conservative treatment of reflux in a patient with nonneuropathic bladder–sphincter dysfunction will be unsuccessful as long as the dysfunction persists. Therefore, reha-

bilitation of the bladder is important in the treatment of VUR in these patients. The presence of reflux in those patients(mainly girls) may also be the cause of bladder dysfunction and not the result. The anomaly of the ureteric entrance may in some children affect the trigone and the detrusor and thus a non-invasive early treatment of VUR may, in many cases, enhance the cure of bladder dysfunction.47,48 It may also be noted that some children with secondary incontinence and urge syndrome or dysfunctional voiding may be victims of child abuse. This may be difficult to prove but should be kept in mind. A significant proportion of adult women with complex urinary symptoms have reported sexual abuse as a child.17

Treatment Urotherapy has been defined as “non-surgical, nonpharmacological treatment for lower urinary tract function of neurogenic and non-neurogenic bladders.41,49–51 It is a cognitive-behavioral training, teaching children to recognize and employ their cerebral centers to achieve command over their lower urinary tract. It includes information about normal bladder function and explanations on why the child with bladder dysfunction deviates from the normal. Instruction about what to do about it, and support and encouragement to go through with the training program are essential ingredients in urotherapy. Urotherapy also includes pelvic floor training, biofeedback (in particular during uroflow), behavioral modification (advice for regular habits regarding meals, drinks, voiding, defecation, and sleep), electrical neurostimulation, and clean intermittent catheterization (instruction and follow-up of CIC in patients with neurogenic bladder and non-neurogenic underactive detrusor).31,52 Urotherapy cures about 75% of children with overactive bladder and dysfunctional voiding.49 However, the non-responders to bladder training and/or anticholinergics should be referred to specialized centers for further urodynamic investigations and eventually biofeedback training.35,42 Defecation and voiding are closely linked and the dysfunction of one cannot be properly evaluated without consideration of the other. Thus many of the children with bladder dysfunction also have constipation and therefore an active treatment of the constipation should be included in the urotherapy. In a recent study53 daytime wetting was strongly associated with fecal incontinence, whereas fecal incontinence and nocturnal enuresis lacked a significant association.

Non-neuropathic bladder–sphincter dysfunction

Table 50.2. Treatment I. Urge and urge incontinence Girls and boys 5–10 years. r “Information” about the good prognosis and the maturation of the bladder function and the high “spontaneous cure rate” (15%–20% per year) r Simple bladder training with “timed micturitions” r In severe cases controlled by urotherapist r Add Oxybutynin or Tolteridin in children with frequency and/or incontinence Most of the children with the urge syndrome have a good prognosis. Early recognition and information is essential.

Table 50.3. Treatment Dysfunctional voiding Mainly schoolgirls >6 years of age r Information – learn how to void r Urotherapy r Careful follow-up with regular control-visits and control of flow and residual urine r Treatment of constipation r Antibiotic prophylaxis 75% are cured by urotherapy. Early recognition and treatment essential for the outcome.

Table 50.4. Treatment III. Detrusor underactivity (“lazy bladder”) A severe condition mainly affecting schoolgirls >9 years of age r Urotherapy r Careful follow-up with regular control-visits and control of flow and residual urine r Treatment of constipation r Sometimes biofeedback-training r Antibiotic prophylaxis r In girls with the more severe emptying difficulties r Electrostimulation or r CIC In all children with signs and symptoms of detrusor underactivity a long-term follow-up into adulthood is mandatory.

Pharmacological treatment, i.e., Desmopressin, can be used in nocturnal enuresis. There are numerous controlled studies on the efficacy and safety of Desmopressin in children showing clinical good results with 50%–70% full response.3,19,54 Some children have nocturnal enuresis in combination with an overactive bladder where Desmopressin can be used in combination with anti-cholin ergic or anti-muscarinic drugs. Regarding the pharmacological treatment of overactive bladder, oxybutynin and tolteridin have been used and studied and have good tolerability

and efficacy particularly in children with urge incontinence combined with frequency.6,32,55 In almost all patients the pharmacological treatment should be combined with urotherapy.

Outcome measures for enuresis and “non-neuropathic bladder–sphincter dysfunction” Assessment of treatment outcome should be based on pretreatment baseline registration of the frequency of incontinent episodes, of bedwetting and/or day wetting. The outcome of different treatment regimens should be given as percentage reduction of the wetting episodes compared to baseline, and grouped as full response, response, partial response, or non-response to therapy. It should be clear that no therapy is fully successful until the child can stay dry after discontinuing treatment.3,7 The definition of immediate result of treatment: r non-response: 90% reduction. r full response: 100% or less than one incontinent episode per month. Post-treatment evaluation of surgical or medical therapy of incontinence should be conducted no less often than 1, 6 and 12 months after treatment. The definitions of the long-term result of treatment: r relapse: more than one accident per month r continued success: no relapse in 6 months after treatment r complete success: no relapse in two years after treatment Evaluation should be done at yearly intervals thereafter and continued as long as possible, preferably for at least 5 years. Questionnaires and frequency–volume charts, used in the diagnosis of incontinence, are the most important tools in reporting on outcome and results of therapeutic interventions.

Long-term follow-up studies The prevalence of urinary incontinence in adulthood has been extensively studied in different epidemiological reports but most of them concern adult women and only a few studies of young women from late adolescence have been included.16,56,57 It has been shown that the proportions of types of incontinence differ by age. A survey of older women suggests that mixed incontinence predominates.21 A survey of young and middleaged women suggests that pure stress incontinence predominates in that age group.23,28,58 In a study by Hannested et al.22 the entire age range was included and it demonstrates a fairly regular increase in prevalence of

647

648

¨ ´ G. Lackgren and T. Neveus

mixed incontinence across the age range, and a regular decrease in prevalence of stress incontinence. Several different risk factors for future urinary incontinence have been studied,58–60 but in only a few the bladder dysfunction and/or nocturnal enuresis in childhood have been included as a possible risk factor. Thus, these factors deserve more attention and children should be followed over many years into adulthood. Such a controlled study design is necessary because the effect of childhood incontinence may only become clear years later when the patient is older. The prevalence of nocturnal enuresis in adolescence and adulthood have been well studied, but the prevalence of daytime wetting or combined daytime and nocturnal wetting in childhood has been poorly investigated and no true longitudinal cohort studies have been published.

Enuresis: long-term outcome without treatment The prevalence of nocturnal enuresis at the age of 5, 7 and 10 years is approximately 25, 10 and 5%, respectively.10,61–63 A small percentage of teenagers still wet their beds.64 There are no major ethnic or geographic differences in this regard. The often quoted study by Forsythe and Redmond indicates a spontaneous resolution rate of 15% per year from the age of 5 to 19.65 The problem here is that this is almost the only estimation we have, since nowadays, fortunately, enuretic children are not left untreated. The short observational study by Monda and Husman, however, supports the findings by Forsythe.66 Although enuresis tends to improve with age, there are still as many as 0.5% of the adult population who experience more or less frequent nocturnal accidents.67 Accordingly, a wait-and-see attitude towards a 5-year-old child with enuresis is warranted, but not with a teenager. The chance that the latter will “grow out of” his or her problem within a year or two is not high.

Long-term outcome with treatment The efficacy and safety of desmopressin has been evaluated in a large number of trials.28,54,68 It is clear that the frequency of nocturnal accidents is reduced in a majority of patients, but the proportion of children becoming completely dry is less than 50%.54 Desmopressin may be regarded as symptomatic treatment in children with enuresis due to polyuria, and relapse is the rule after cessation of treatment: the research community is divided as regards its possible curative effects. Fortunately, desmopressin has been shown to be remarkably safe, even for long-term treatment.19,69,70 As long as the

drug is not combined with excessive fluid intake, no shortor long-term safety concerns have as yet emerged. The enuresis alarm is successful in between 50% and 80% of children.71–73 Motivation, information, and followup are the major determinants for treatment success. The majority of children successfully treated with the alarm can be considered cured, but approximately 20% relapse after treatment and may need additional sessions.74 The risks of treatment are nil, but the therapy puts great stress on the family and may be remembered many years afterwards.75 There are no physical risks or adverse somatic consequences entailed with bedwetting per se, although underlying detrusor hyperactivity may certainly cause morbidity via urinary tract infections.

Long-term outcome of bladder dysfunction in childhood Our knowledge of the future outcome of children with bladder dysfunction is based on a few studies that have followed a population into adolescence, but no particular group of patients have been followed into adulthood. Longitudinal study designs are needed to estimate incidence of voiding dysfunction in adolescence and adulthood and to describe the course of the condition and its different forms. The urge syndrome is calculated to decrease by 14% per year7,8,17 and thus most of the children with urge symptoms are cured spontaneously (particularly boys). Furthermore, an active “bladder training” in school-age children may add to the low incidence of urinary problems in late adolescence.7,18,32 Swithinbank completed a prospective longitudinal study in 1176 British schoolchildren using a questionnaire at 11–12 years and again at 15–16 years. The authors noted a significant steady decrease in the prevalence of urinary symptoms with age. Enuresis decreased from 5% in the 11–12-year-olds to 1% in 15–16-year-olds and daytime symptoms and/or wetting decreased from 12% to 3% in 15–16-year-olds.76 Hellstr¨om et al. performed a follow-up study of 3556 Swedish schoolchildren33 10 years later (when they had reached the age of 17 years). A new questionnaire was sent to 1096 randomly selected adolescents from the total group and, in addition, all the children who had reported symptoms (imperative urge, day wetting, emptying difficulties, or bedwetting) were contacted. Of the 151 girls with symptoms at 7 years of age, only 16 still had bladder dysfunction (i.e., 20% cure rate/year), most of them with signs and symptoms of “dysfunctional voiding.” In the 208 boys with symptoms at 7 years of age (of whom 194 had nocturnal enuresis), only 8 still had symptoms of whom 4 still were bedwetting and the remainder had some urge and emptying difficulties (i.e., 25% cure rate/year).

Non-neuropathic bladder–sphincter dysfunction

From the randomized group, 21 teenagers had symptoms of whom 14 reported them as new. It was concluded that symptoms of bladder disturbances occurred at a low frequency in children aged 17 years and they were only mild or moderate. Similar prevalence was found in studies from other countries.7,9,12 This clearly shows that most of the minor “bladder problems” tend to disappear with time and few of the children with symptoms and signs of an overactive bladder will have problems continuing into adulthood. However, some of these children with urge incontinence may very well turn into “dysfunctional voiders” later in childhood and the fate of these children in adulthood is not well known. In a recent follow-up study on girls with severe dysfunctional problems and incontinence in adolescence Yang et al.76 found that77 (30%) of the girls (27 patients) still suffered from the same problems in adulthood. However, 30% were cured with “conservative” treatment and 30% were lost to follow-up, but still had unresolved problems. The patients in this study comprised a group of 27 children with severe bladder dysfunctional problems and may indicate that the natural course in this group of children may be worse than had previously been expected.42 There is a lack of the identification of prognostic risk factors and further prospective studies are needed to identify the children at risk. In a Danish study78 2613 women aged 30–59 years were randomly selected and responded to a questionnaire. They found that childhood bedwetting was associated with prevalent urge incontinence and also incontinence during sleep as a sign of a persisting overactive bladder. Kuh et al. found that childhood enuresis and/or childhood day-time incontinence is over-represented among middleaged woman who suffer from urge incontinence.62 Detrusor hyperactivity supposedly may persist even after the patient has become continent. Moore et al.79 studied a group of adults with idiopathic “detrusor instability” and found that a significant number had suffered from bedwetting as children (63% of men and 38% of women), probably those patients who suffered from detrusor-dependent enuresis and also had daytime urge symptoms. However, this is only speculation and the only way to show the future outcome of these children is to perform controlled prospective studies to find the natural course of bladder dysfunction in childhood. Thus, the long-term outcome of these conditions is not well known and one can only speculate about the future outcome of girls with childhood bladder dysfunction and deduce the outcome from recent retrospective, interview studies on incontinent adult men and women. Today, in the Western world, nocturnal enuresis and daytime wetting are beginning to be recognized as disorders that can be cured if they are properly diagnosed and treated. However, even

in developed countries, the problem is often neglected and parents do not seek help. Therefore, correct treatment may be delayed and in particular the more severe cases with bladder dysfunctions may persist in the long term.77,80 The psychological consequences of untreated or unsuccessfully treated nocturnal enuresis and/or bladder dysfunction should, however, not be underestimated. Enuretic children have been shown to have significantly lower selfesteem than unaffected children and children with combined daytime wetting and enuresis even more so – differences that disappear when the children became dry.81 We can only speculate about the long-term social effects of this, but the impact of chronic low self-esteem in a growing individual may be considerable. Thus it is very important to recognize enuresis and bladder dysfunction in children over the age of 5 years and provide a correct diagnosis leading to information and treatment for the child.

REFERENCES 1. Abidari, J. M. & Shortliffe, L. M. D. Urinary incontinence in girls. Urol. Clin. N. Am. 2002; 29:661–675. 2. Bloom, D. A., Faerber, G., & Bomalaski, M. D. Urinary incontinence in girls. Evaluation, treatment, and its place in the standard model of voiding dysfunctions in children. Urol. Clin. North. Am. 1995; 22(3):521–38. 3. Hjalmas, K., Arnold, T., Bower, W. et al. on behalf of the International Children’s Continence Society (ICCS). Nocturnal enuresis: an international evidence based management strategy. J. Urol. 2004; 171:2545–2561. 4. Rushton, G. Wetting and functional voiding disorders. Urol. Clin. North Am. 1995; 22:75–93. 5. Bloom, D. A., Sweley, W. M., Ritchey, M. L., & McGuire, E. J. Toilet habits and continence in children: an opportunity sampling in search of normal parameters. J. Urol. 1993; 149:1087–90. 6. Mattsson, S. Urinary incontinence and nocturia in healthy schoolchildren. Acta Pediatr. 1994; 83:950–4. 7. Bakker, E., van Sprundel, M., Van der Auwera, J. C., van Goold, J. D., & Wyndaele, J. J. Voiding habits and wetting in a population of 4,332 Belgian schoolchildren aged between 10 and 14 years. Scand. J. Urol. Nephrol. 2002; 36(5):354–362. 8. Hellstrom, A. L., Hanson, E., Hansson, S., Hjalmas, K., & Jodal, U. Micturition habits and incontinence in 7-year-old Swedish school entrants. Eur. J. Pediatr. 1990; 149:434–437. 9. Kajiwara, M., Inoue, K., Usui, A., Kurihara, M., & Usui, T. The micturition habits and prevalence of daytime urinary incontinence in Japanese primary school children. J. Urol. 2004; 171(1):403–407. 10. Nev´eus, T., Hetta, J., Cnattingius, S. et al. Depth of sleep and sleep habits among enuretic and incontinent children. Acta Pædiatr. 1999; 88:748–752. 11. Sureshkumar, P., Bower, W., Craig, J. C., & Knight, J. F. Treatment of daytime urinary incontinence in children: a system-

649

650

¨ ´ G. Lackgren and T. Neveus

12.

13. 14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

atic review of randomized controlled trials. J. Urol. 20003; 170(1):196–200, discussion 200. Sureshkumar, P., Craig, J. C., Roy, L. P., & Knight, J. F. Daytime urinary incontinence in primary school children: a populationbased survey. J. Pediatr. 2000; 137(6):814–818. Meadow, S. R. Day wetting. Pediatr. Nephrol. 1990; 4:178–184. Abrams, P., Cardozo, L., Fall, M. et al. The standardisation of terminology of lower urinary tract function: report from the Standardisation Sub-committee of the International Continence Society. Neurourol. Urodyn. 2002; 21:167–178. Hj¨alm˚as, K. Urinary incontinence in children: suggestions for definitions and terminology. Scand. J. Urol. Nephrol. Suppl., 1992; 141:1. Cheater, F. M. & Castleden, C. M. Epidemiology and classification of urinary incontinence. Clin. Obstet. Gynaecol. 2000; 14:183. Nijman, R. Classification and treatment of functional incontinence in children. Br. J. Urol. 2000; 85 Suppl. 3:43–46. Swithinbank, L. V., Brookes, S. T., Shepherd, A. M., & Abrams, P. H. The natural history of urinary symptoms during adolescence. Br. J. Urol. Suppl. 1998; 81:90. Knudsen, U. B., Rittig, S., Nørgaard, J. P., Lundemose, J. B., Pedersen, E. B., & Djurhuus, J. C. Long-term treatment of nocturnal enuresis with desmopressin. A follow-up study. Urol. Res. 1991; 19(4):237–240. Hansen, A., Hansen, B., & Dahm, T. L. Urinary tract infection, day wetting and other voiding symptoms in seven- to eightyear-old Danish children. Acta Paediatr. 1997; 86(12):1345– 1349. Hampel, C., Wienhold, D., Benken, N., Eggersmann, C., & Thurhoff, J. W. Definition of overactive bladder and epidemiology of urinary incontinence. Urology 1997; 50:4. Hannestad, Y. S., Rortveit, G., Sandvik, H., & Hunskaar, S. A community-based epidemiological survey of female urinary incontinence: the Norwegian EPINCONT study. J. Clin. Epidemiol. 2000; 53:1150. Stewart, W., Herzog, A. R., Wein, A. et al. Prevalence of overactive bladder in the US: results from the NOBLE program. 2nd International Consultation on Incontinence, abstract no 79, Paris 2001. Swithinbank, L. V., Donovan, J. L., Du Heaume, J. C. et al. Urinary symptoms and incontinence in women: relationships between occurrence, age, and perceived impact. Br. J. Gen. Pract. 1999; 49:897. Rittig, S., Knudsen, U. B., Nørgaard, J. P., Pedersen, E. B., & Djurhuus, J. C. Abnormal diurnal rhythm of plasma vasopressin and urinary output in patients with enuresis. Am. J. Physiol. 1989; 256:F664–F671. Yeung, C. K., Chiu, H. N., & Sit, F. K. Bladder dysfunction in children with refractory monosymptomatic primary nocturnal enuresis. J. Urol. 1999; 162(3 Pt 2):1049–1055. Wolfish, N. M., Pivik, R. T., & Busby, K. A. Elevated sleep arousal thresholds in enuretic boys: clinical implications. Acta Pædiatr. 1997; 86:381–384. Tuvemo, T. DDAVP in childhood nocturnal enuresis. Acta Pædiatr. Scand. 1978; 67:753–755.

29. Wille, S. Comparison of desmopressin and enuresis alarm for nocturnal enuresis. Arch. Dis. Child. 1986; 61:30–33. 30. Lovering, J. S., Tallett, S. E., & McKendry, B. I. Oxybutynin efficacy in the treatment of primary enuresis. Pediatrics 1988; 82; 104–106. 31. Hj¨alm˚as, K., Hoebeke, P. B., & dePaepe, H. Lower urinary tract dysfunction and urodynamics in children. Eur. Urol. 2000; 38:655. 32. Hjalmas, K., Hellstrom, A. L., Mogren, K., Lackgren, G., & Stenberg, A. The overactive bladder in children: a potential future indication for tolterodine. BJU Int. 2001; 87:569–574. 33. Hellstrom, A., Hanson, E., Hansson, S., Hjalmas, K., & Jodal, U. Micturition habits and incontinence at age 17 -reinvestigation of a cohort studied at age 7. Br. J. Urol. 1995; 76(2):231–234. 34. Lettgen, B., von Gontard, A., Olbing, H., Heiken-L¨owenau, C., Gaebel, E., & Schmitz, I. Urge incontinence and voiding postponement in children: somatic and psychosocial factors. Acta Pædiatr. 2002; 91:978–984. 35. Hoebeke, P., Van Laecke, E., Van Camp, C., Raes, A., & Van De Walle, J. One thousand video-urodynamic studies in children with non-neurogenic bladder sphincter dysfunction. BJU Int. 2001; 87:575–580. 36. Allen, T. D. Vesicoureteral reflux as a manifestation of dysfunctional voiding. In Hodson, J. & Kincaid-Smith, P. eds. Reflux Nephropathy. Chapter 18, New York: Masson Publishing Co, 1979:171–180. 37. Bauer, S. B. Special considerations of the overactive bladder in children. Urology 2002; 60(5 Suppl 1):43–48; discussion 49. 38. Chiozza, M. L. Dysfunctional voiding. Pediatr. Med. Chir. 2002; 24(2):137–140. 39. von Gontard, A. Enkopresis: Erscheinungsformen – Diagnostik – Therapie. Stuttgart: Kohlhammer Verlag, 2004. 40. Hansson, S. Urinary incontinence in children and associated problems. J. Urol. Nephrol. Suppl. 1992; 141:47–55. 41. Hoebeke, P., Renson, C., Raes, A., Vanlaecke, E., & Van de Walle, J. Pelvic floor spasms in children: an unknown condition responding well to pelvic floor therapy. BJU Int. 2003; 91 Suppl 1:19. 42. Hobeke, P. New insights in diagnosis and treatment of nonneuropathic bladder-sphincter dysfunction in children. Thesis. Gent. 1998. 43. Hellst¨om, A.-L., Hanson, E., Hansson, S., Hj¨alm˚as, K., & Jodal, U. Association between urinary symptoms at 7 years old and previous urinary tract infection. Arch. Dis. Child. 1991; 66:232– 234. 44. Hansson, S., Hjalmas, K., Jodal, U., & Sixt, R. Lower urinary tract dysfunction in girls with untreated asymptomatic or covert bacteriuria. J. Urol. 1990; 143:333–335. 45. Koff, S. A. Relationsship between dysfunctional voiding and reflux. J. Urol. 1992; 148:1703–1705. 46. Sill´en, U., Hj¨alm˚as, K., Aili, M., Bjure, J., Hansson, E., & Hansson, S. Pronunced detrusor hypercontractility in infants with gross bilateral reflux. J. Urol. 1992; 148:598. 47. Snodgrass, W. Relationship of voiding dysfunction to urinary tract infection and vesicoureteral reflux in children. Urology 1991; 38:341–344.

Non-neuropathic bladder–sphincter dysfunction

48. L¨ackgren, G. & Stenberg, A. Bladder dysfunction and VUR. Result of endoscopic treatment with Deflux. Abstract. ICCS Hongkong, 2002. 49. Hellstrom, A. L., Hjalmas, K., & Jodal, U. Rehabilitation of the dysfunctional bladder in children: method and 3-year followup. J. Urol. 1987; 138:847–849. 50. Hobeke, P., Van de Walle, J., Theunis, M., De Paepe, H., & Oosterlinck, Renson, C. Outpatient pelvic floor therapy in girls with daytime incontinence and dysfunctional voiding. Urology 1996; 48:923–927. 51. Vijverberg, M. A., Elzinga-Plomp, A., Messer, A. P., Van Gool, J., & de Jong, T. P. Bladder rehabilitation, the effect of a cognitive training programme on urge incontinence. Eur. Urol. 1997; 31:68–72. 52. Kuh, D., Cardozo, L., & Hardy, R. Urinary incontinence in middle aged women: childhood enuresis and other lifetime risk factors in a British prospective cohort. J. Epidemiol. Commun. Health 1999; 53(8):453–458. 53. Soderstrom, U., Hoelcke, M., Alenius, L., Soderling, A. C., Hjern, A. Urinary and faecal incontinence: a population-based study. Acta Paediatr. 2004; 93(3):386–389. 54. Hj¨alm˚as, K., Hanson, E., Hellstr¨om, A.-L., Kruse, S., & Sill´en, U. Long-term treatment with desmopressin in children with primary monosymptomatic nocturnal enuresis: an open multicentre study. Swedish Enuresis Trial (SWEET) Group. Br. J. Urol. 1998; 82(5):704–709. 55. Reinberg, Y., Crocker, J., Wolper, J., & Vandersten, D. Therapeutic efficacy of extended release oxybutynin chloride, and immediate release and long acting tolterodine tartrate in children with diurnal urinary incontinence. J. Urol. 2003; 169:317–319. 56. Hunskaar, S. One hundred and fifty men with urinary incontinence. I. Demography and medical history. Scand. J. Prim. Health Care 1992; 10:21. 57. Hunskaar, S., Burgio, K., Diokno, A., Herzog, A. R., Hjalmas, K., & Lapitan, M. C. Epidemiology and natural history of urinary incontinence inwomen. Urology 2003; 62:16–23. 58. Samuelsson, E., Victor, A., Tibblin, G. A population study of urinary incontinence and nocturia among women aged 20– 59 years. Prevalence, well-being and wish for treatment. Acta Obstet. Gynecol. Scand. 1997; 76:74. 59. Yarnell, J. W., Voyle, G. J., Richards, C. J., & Stephenson, T. P. The prevalence and severity of urinary incontinence in women. J. Epidemiol. Commun. Health 1981; 35:71. 60. Kalo, B. B. & Bella, H. Enuresis: prevalence and associated factors among primary school children in Saudi Arabia. Acta. Pædiatr. 1996; 85:1217–1222. 61. Laberge, L., Tremblay, R. E., Vitaro, F., & Montplaisir, J. Development of parasomnias from childhood to early adolescence. Pediatrics 2000; 106:67–74. 62. van der Wal, M. E., Pauw-Plomp, H., & Schulpen, T. W. Bedplassen bij Nederlandse, Surinaamse, Marokkanse en Turkse kinderen van 3–4, 5–6 enl 11–12 jaar. Ned. Tijdschr. Geneeskd. 1996; 140(48):2410–2414. 63. Marugan de Miguelsanz. J. M., Lapena Lopez de Armentia, S., Rodriguez Fernandez, L. M. et al. Analisis epidemiologico de la secuencia de control vesical y prelvalencia de enuresis noc-

64. 65.

66.

67.

68.

69.

70.

71.

72. 73.

74.

75.

76. 77. 78.

79.

80.

81.

turna en ninos de la provincia de leon. An. Esp. Pediatr. 1996; 44(6):561–567. Forsythe, W. I. & Redmond, A. Enuresis and spontaneous cure rate: study of 1129 enuretics. Arch. Dis. Child. 1974; 49:259–263. Monda, J. M. & Husmann, D. A. Primary nocturnal enuresis: a comparison among, observation., imipramine, desmopressin acetate and bed-wetting alarm systems. J. Urol. 1995; 154(2 Pt 2):745–748. Hirasing, R. A., van Leerdam, F. J., Bolk-Bennink, L., & Janknegt, R. A. Enuresis nocturna in adults. Scand. J. Urol. Nephrol. 1997; 31(6):533–536. Aladjem, M., Wohl, R., Boichis, H., Orda, S., Lotan, D., & Freedman, S. Desmopressin in nocturnal enuresis. Arch. Dis. Child. 1982; 57:137–140 Fjellestad-Paulsen, A., Laborde, K., Kindermans, C., & Czernichow, P. Water-balance hormones during long-term follow-up of oral DDAVP treatment in diabetes insipidus. Acta Pædiatr. 1993; 82(9):752–757. L¨ackgren, G., Nev´eus, T., Lilja, B., & Stenberg, A. Desmopressin in the treatment of severe nocturnal enuresis in adolescents – a 7-year follow-up study. Br. J. Urol. 1998; 81(Suppl 3):17–23. Berg, I. B., Forsythe, W. I., & McGuire, R. Response of bed wetting to the enuresis alarm: influence of psychiatric disturbance and maximum functional bladder capacity. Arch. Dis. Child. 1982; 57:394–396. Bollard, J. & Nettelbeck, T. A comparison of dry-bed training and standard urine-alarm conditioning treatment of childhood bedwetting. Beh. Res. Ther. 1981; 19(3):215–226. Devlin, J. B. & O’Cathain, C. Predicting treatment outcome in nocturnal enuresis. Arch. Dis. Child. 1990; 65(10):1158–1161. El-Anany, F. G., Maghraby, H. A., Shaker, S. E., & Abdel-Moneim, A. M. Primary nocturnal enuresis: a new approach to conditioning treatment. Urology 1999; 53(2):405–409. Bengtsson, B. S¨ok hj¨alp tidigt f¨or barn med enures. R˚ad fr˚an vuxna som haft sv˚ar nattv¨ata som barn. L¨akartidningen 1997; 94(4):245–246. Swithinbank, L. V., Carr, J. C., & Abrams, P. H. Longitudinal study of urinary symptoms in children. Scand. J. Urol. Nephrol. Suppl. 1994; 163:67. Yang, C. C. & Mayo, M. E. Morbidity of NNBSB syndrome. Urology 1997; 49:445–448. Foldspang, A. & Mommsen, S. Adult female urinary incontinence and childhood bedwetting. J. Urol. 1994; 152(1):85–88. Moore, K. H., Richmond, D. H., & Parys, B. T. Sex distribution of adult idiopathic detrusor instability in relation to childhood bed-wetting. Br. J. Urol. 1991; 68:479–482. Holst, K. & Wilson, P. D. The prevalence of female urinary incontinence and reasons for not seeking treatment. N. Z. Med. J. 1988; 101:756. H¨aggl¨of, B., Andr´en, O., Bergstr¨om, E., Marklund, L., & Wendelius, M. Self-esteem before and after treatment in children with nocturnal enuresis and urinary incontinence. Scand. J. Urol. Nephrol. 1997; 31(Suppl. 183):79–82. Schulman, S. L., Quinn, C. K., Plachter, N., & Kodman-Jones, C. Comprehensive management of dysfunctional voiding. Pediatrics 1999; 103(3):E31.

651

51 Undescended testes John M. Hutson Department of General Surgery, Royal Children’s Hospital, Parkville, Victoria, Australia

Since ancient times, it has been recognized that the testis needs to be fully descended in the scrotum for normal functioning. Indeed, the testis derives its name from the Latin word “witness,” following the custom in Roman times to hold the testicles when taking an oath.1 Hence, one of the primary concerns of surgeons has been the development of surgical procedures to place an undescended testis into the scrotum. In recent years there have been rapid changes in attitudes to long-term outcomes: not that long ago, success of surgery was measured by such crude criteria as cosmetic result or survival, which could be determined immediately. However, now the profession and the community require and expect a much higher standard, such that at present the yardstick for success in the management of undescended testes is normal fertility and a low risk of malignancy in adult life. Community attitudes, and knowledge about long-term outcomes, are changing rapidly, as evidenced by the fact that standard texts only 25 years ago did not contain any significant information about longterm malignancy risks.2

Etiology of the undescended testis Most undescended testes are probably caused by abnormal migration of the gubernaculum during the inguinoscrotal phase of testicular descent. Testicular descent normally begins at about 10 weeks of gestation, shortly after the onset of sexual differentiation. Two morphological steps in descent can be identified: initial relative movement of the testis (compared with the ovary) from the urogenital ridge to the inguinal region, known as transabdominal descent; and migration from the inguinal canal to the scrotum,

known as inguinoscrotal descent.3 During the first phase, which occurs at approximately 10–15 weeks of gestation, the gubernaculum or genito-inguinal ligament enlarges in the male, but not in the female. The enlarged gubernacular fold effectively anchors the testis near the inguinal region while the embryo is enlarging. This prevents the testis from moving away from the inguinal region as the ovary tends to do. In the second phase of descent, at between 28 and 35 weeks, the gubernaculum migrates from the inguinal canal across the pubic bone and into the scrotum. The end of the gubernaculum is gelatinous mesenchyme while the proximal gubernaculum is hollowed out by a peritoneal diverticulum, the processus vaginalis which contains the testis. The cremaster muscle develops in the outer rim of the gubernaculum outside the peritoneal diverticulum. The gubernaculum is supplied by the genitofemoral nerve, both before, during and after migration to the scrotum.4 Normal testicular descent is controlled by hormones. There is controversy about which hormones modulate the first and second stage.5 Androgens have some role in the first phase, by stimulating regression of the cranial suspensory ligament, which then allows the testis to descend. But it does not appear to have a role in stimulating enlargement of the gubernaculum. The second phase of descent is controlled by androgen, but its action on the gubernaculum appears to be indirect. The primary hormone controlling descent is now known to be insulin-like hormone 3 (INSL3) which is produced by Leydig cells (as well as androgen).6 Mullerian inhibiting substance/anti-Mullerian hormone appears to be a secondary factor.7 Recent evidence suggests that the direction of gubernacular migration may be controlled by androgen indirectly via the genitofemoral nerve. A neurotransmitter released from the nerve,

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


652

Undescended testes

Table 51.1. Etiology of cryptorchidism Unknown Mechanical failure of gubernacular migration Hormone deficiency syndrome (T, Insl3, MIS) Placental dysfunction (hCG deficiency) Abdominal wall defects Posterior urethral valve (prune belly syndrome) Spina bifida/cerebral palsy Chromosomal defects Multiple malformation syndromes

calcitonin gene-related peptide (CGRP), is implicated in normal testicular descent of rodents and in rodent models with undescended testes.8 One cause of abnormal inguinoscrotal migration of the gubernaculum may be a lack of CGRP released from the genitofemoral nerve. Alternatively, the nerve may be abnormally sited so that migration is stimulated in an ectopic direction (e.g., perineal testis).9 There are numerous causes for undescended testes (Table 51.1), as normal descent is multiphasic, with different morphology, hormones, and anatomical regulators. Rare hormonal deficiency syndromes of testosterone (e.g., partial androgen insensitivity) or Mullerian inhibiting substance (e.g., persistent Mullerian duct syndrome) do cause undescended testes, but in most children these are not present. Deficiency of INSL3 as a cause of cryptorchidism has not been identified.10 It is presumed that, in the common variety of undescended testis, there is a mechanical failure of gubernacular migration. This could be secondary to transient androgen deficiency in the third trimester, perhaps caused by placental insufficiency, CGRP deficiency within the genitofemoral nerve, or anatomical defects within the nerve itself. Also, the gubernaculum may have an intrinsic defect preventing it from responding to normal trophism by migration to the scrotum. Various mechanical causes, such as abdominal wall defects or massive prenatal bladder distension, can cause undescended testis. In addition, there are various chromosomal and multiple malformation syndromes, as well as neurological abnormalities, causing cryptorchidism. The multitude of putative causes of cryptorchidism may produce different long-term outcomes. Also, a significant percentage of undescended testes may be acquired during early childhood, and are not present from birth. This controversial view has profound implications for long-term prognosis and is discussed more later. From a clinical perspective, most of the identified specific causes of undescended teste are rare by comparison with the common

idiopathic variety. A simple mechanical failure of gubernacular migration is suspected, either because of an intrinsic mechanical defect or from secondary hormonal deficiency. It is this latter common form of undescended testis that prognostications should be centered around.

Classification of cryptorchidism Undescended testis can be classified by etiology or position. The first phase of “descent” requires passive anchoring of the testis as the embryo grows and, consequently, intraabdominal impalpable testes are uncommon. Most testes are palpable in the groin and might be called inguinal. The most common position for an undescended testis is just above and lateral to the external inguinal ring outside the abdominal musculature in the superficial inguinal pouch, described by Denis Browne.11 There is controversy about whether such a testis is truly ectopic or is merely displaced into a position of lower pressure. Truly ectopic testes in the femoral region, perineum, or prepenile regions are quite rare. Recent evidence suggests that some undescended testes present later in childhood, after being apparently normal in infancy. In the past, these testes were often called retractile.12 More recently, some have been classified as ascending testes, which are defined as testes which descend late into the scrotum, in the first 3 months postnatally, but then ascend out of the scrotum again later in childhood.13 Both ascending testes and severe retractile testes may represent acquired abnormality. In a recent study,14 dissection within the spermatic cord in these children has revealed the presence of a fibrous string deep to the cremaster muscle and fascia. Transection of this fibrous string has allowed adequate elongation of the vas deferens and gonadal vessels for scrotal placement of the testis. Histological examination of the fibrous string has revealed residual peritoneal cells consistent with incomplete obliteration of the processus vaginalis. It has been proposed that cryptorchidism presenting later in childhood is not only acquired, but may be common, and secondary to incomplete disappearance of the processus vaginalis. Although this diagnosis remains controversial, it has the potential to confound previous long-term outcome studies, which may include an amalgam of congenital and acquired cryptorchid testes.

Current treatments Current management of cryptorchidism is based on the premise that secondary testicular degeneration caused by

653

654

J. M. Hutson

Fig. 51.1. Schematic of germ cell maturation in the postnatal human testis, showing where UDT causes delay and/or disruption of development.

high temperature can be prevented by early intervention.15 The intra-abdominal core temperature is 3–4 degrees higher than the temperature of the scrotal testis.16 Recent evidence shows that early postnatal germ cell development is deranged in the undescended testis (Fig. 51.1).17,18 Previously, it was well known that after puberty, the intraabdominal testis suffered irreversible azoospermia, both in animals19 and in humans.15 Macroscopic atrophy in school-age children with cryptorchidism then became the criterion for intervention. Light microscopic changes in undescended testes can be identified in the third and fourth years.20,21 Tubular dysplasia on electron microscopy can be identified in the second year of life.22 Degeneration of testicular germ cells in the first 6–12 months23 along with physiological derangements in hormone production by the undescended testis in the first year,24–26 suggest that intervention must be very early if it is to prevent damage. Although the evidence for early degeneration, presumed secondary, is substantial, the timing for surgery remains controversial, because previous studies looking at outcome have not demonstrated an association with age. A more detailed account of these studies is given later. Despite the lack of evidence in humans that early surgery will prevent testicular degeneration, in animal models early surgery does prevent such degeneration.27–30 However, a difficulty in extrapolation from the experimental rat arises because testicular descent in rats does not occur until puberty, in contrast to prenatal descent in humans. Orchidopexies in rats therefore, need only be done at puberty to demonstrate an effect of “early” surgery. This could be interpreted as suggesting that surgery should be done soon after the time of normal scrotal occupancy by the testis, which is puberty in rats or the neonatal period in humans.

The author now recommends orchidopexy at 6 months of age. Orchidopexy in small infants can be challenging, but in specialized pediatric surgical centers, operation at 6–12 months of age is quite satisfactory and does not appear to increase the risk of complications.31 All newborn males should be examined carefully for the presence of testes within the scrotum. Those children in whom the testes are not located obviously at the bottom of the scrotum should be reviewed at 12 weeks of age. During that time, about half the cryptorchid testes will descend spontaneously. These children need no immediate intervention, but do require regular annual follow-up to ensure that they do not develop acquired ascending testes. In cases where the testes remain out of the scrotum at 12 weeks, referral for orchidopexy should occur soon after, with the exact timing of surgery dependent on the local circumstances and expertise of the surgeon. Testes that are in the scrotum at 3 months of age, but which do not have a completely dependent position, also should be followed carefully during childhood, to make sure they do not develop acquired maldescent. The author’s own choice of therapy is day surgical correction, with routine orchidopexy and mobilization of the inguinal testis, and surgical placement of the testis within the scrotum. Children presenting later in childhood with an apparent acquired undescended testis may be managed by a similar surgical approach or, alternatively, by using a scrotal approach as described by Bianchi and Squire.32 It is not the author’s own policy to use hormonal therapy with either human chorionic gonadotrophin (hCG) or luteinizing hormone releasing hormone (LHRH), but these therapies remain in wide use throughout Europe and the USA. Various protocols are described to which the reader is referred.33–36 Two simple protocols include intramuscular hCG administration at 100 IU twice a week for 3–4 weeks; alternatively, LHRH can be given as a spray at 100 g in each nostril six times a day for 3–4 weeks. The major dilemma in predicting long-term outcome for orchidopexy is that recent biological studies suggest that testicular function will be preserved only when surgery is done in infancy. As can be seen from Fig. 51.2, the recommended age for intervention has been falling progressively since the 1950s, and it has only been in the 1980s or 1990s that the recommended age lies within what would be regarded now as the correct range. If extrapolation from animal models and the current biological evidence acquired from human studies is correct, it means that we cannot predict optimal outcomes until children who have been operated on in the last 10–15 years have reached adult life. By contrast, adults operated on in a previous generation have had what is now regarded as inferior treatment,

Undescended testes

655

Table 51.2. Testicular volumes after orchidopexy

Series

Patient numbers

Age at operation (y)

Cortes et al.83

11

Puri and Sparnon38

159

3–7 10–12 10 (7–13.6)

Schreiber et al.39

54

–

Testicular volume (ml) No difference between groups 4.9 ± 3.5 abdominal (19) 9.8 ± 5.4 canalicular (36) 17.0 ± 4.9 SIP (119) 18.6 ± 4.0 normal 17.6 unilat. UDT 16.0 ± 4.9 bilat. UDT

SIP, superficial inguinal pouch. Numbers in parentheses in testicular volume column refer to testes. Fig. 51.2. Schematic graph of recommended age for orchidopexy (for congenital UDT) vs. age, showing a rapid fall in recent decades. Currently, orchidopexy is performed around 6 months of age in most pediatric centers.

and so the studies to be reported in the rest of this review need to be considered with some circumspection.

Early complications of surgery The risk of complications after orchidopexy should be less than 5%, in experienced hands.31 Atrophy of the testis after intraoperative damage to the testicular vessels is a serious but uncommon complication after orchidopexy. Infants who have a concomitant incarcerated or strangulated hernia are most at risk. Damage to the vas deferens with possible occlusion of the lumen may occur, but is not frequently reported because of difficulty in diagnosis. The use of diathermy or electrocautery has reduced the risk of hemorrhage from poor hemostasis. With orchidopexy now frequent in infancy, the most common complication is wound infection. Both inguinal and scrotal incisions are at risk of sepsis in infants, but usually they respond to simple treatment. Secondary ascent of the testis out of the scrotum may occur after infection, poor mobilization, or inadequate fixation within the scrotum. Postoperative lymphedema of the testis resolves spontaneously after a few months.

Long-term outcomes Testicular size and position Orchidopexy results in an intrascrotal testis of reasonable size in most boys. In a review of 135 testes in 121 boys, 85% of the testes were intrascrotal at long-term follow-up

after surgery at 6 years of age.37 Testes initially beyond the inguinal canal were larger than those within the canal or abdomen. The latter had a higher incidence of persisting abnormality following orchidopexy, particularly failure of the testes to reach the scrotum, inadequate growth, or atrophy of the testes. Complete infarction of the testis occurred in approximately 3% of patients with an impalpable undescended testis. In another 15%–20% of patients with a high intra-abdominal testis, significant atrophy occurred requiring subsequent excision. Puri and Sparnon38 found a significant correlation between the initial position of the testis and subsequent volume, in a group of 159 children operated on at approximately 9.8 years of age. Nineteen abdominal testes had an average testicular volume in adult life of 4.9 ml, compared with 9.8 ml in canalicular testes and 17 ml, in superficial inguinal pouch testes (Table 51.2). Most authors have found that, in adults, testicular volume after childhood orchidopexy is slightly lower than normal,39 although volume at operation has proven to be a poor predictor of germ cell count.40 Reduced testicular volume is more marked in those with a past history of bilateral orchidopexy.41,42 The highest incidence of testicular atrophy appears to occur after combined herniotomy and orchidopexy in infants presenting with a strangulated hernia. Exact figures are scarce, but the author’s own anecdotal experience is that this difficult dissection may compromise the testicular vessels much more than routine orchidopexy. However, the overall risk of subsequent testicular atrophy with surgery in infancy appears to be low, as long as the surgeon has been trained in the surgery of infancy.31 Preoperative and postoperative gonadal position was documented in 1209 cryptorchid testes in 961 boys.21 Only one-third of all “impalpable” testes were found in the abdomen, while a third were in the inguinal canal. Seven

656

J. M. Hutson

Table 51.3. Paternity after orchidopexy

Series Gilhooly et al.47 Cendron et al.48 Kumar et al.49 Lee44 Lee and Coughlin46

Successful paternity (%)

Patient numbers

Age at operation (y)

Unilateral

Bilateral

145 40 56 467 408

NK 7.0 7–18 NK NK

80 87 84 88 90

48 33 60 59 65 (P < 0.001)

Table 51.4. Semen analysis after orchidopexy

Series 60

Singer et al. Puri and O’Donnell84 Bremholm-Rasmussen et al.41 Okuyama et al.56 ””” Grasso et al.54 Cortes and Thorup55 Mandat et al.58 Lee44 Mayr et al.57 Cortes et al.53

Normal semen analysis (%)

Patient numbers

Age at operation (y)

Unilateral

Bilateral

25 142 45

6.2 7–13 No surgery

70 74 NK

40 30 33

167 43 91 90 135 NK 46 146

2–5 9–12 14–29 13 3–15 NK 2–12 ≤ 12

95 86 17 NK 53 57 46 83

24 20 NK 0 26 25 32 19

percent of impalpable testes were absent. Nearly all (96%) testes reached the scrotum, including 69% of initially intraabdominal testes and 94% of canalicular testes. The authors concluded that even high testes can have a good result, contrary to common belief. In a selected group of 77 ascending testes, 38 were following previous herniotomy and 39 occurred spontaneously.43 The final gonadal position was “excellent” in all but 6 instances. Fertility and subsequent paternity One way to assess long-term outcome for fertility is to document successful paternity (Table 51.3). Lee44 in an extensive review of the literature, noted that 15%–20% of normal married couples are infertile, and in a quarter to a third of these, the abnormality is identified in the male.45 Summarizing a large number of other studies, Lee found that 81.4% of married men with a past history of unilateral undescended testis had offspring. This compared with 50.8%, of married men with offspring who had a pre-

vious history of bilateral undescended testes. Lee compared these figures in the literature with the early results of his own follow-up study from Pittsburgh, where over 450 men are still being followed up for long-term outcome from undescended testes. Forty-four out of 66 men with bilateral undescended testes were married, and of those 19 (44%) already had children. Four hundred and seven men were being followed up with unilateral undescended testes, of whom 303 were married, and of these 66% had children. In a recent update of his figures, Lee46 found severely compromised paternity rates in adulthood in men born with bilateral cryptorchidism (65.3%) compared to those who were formerly unilaterally cryptorchid (89.7%) or control men (93.2%). Similar results were found 20 years ago by Gilhooly et al.46 He reviewed 45 men with a history of bilateral undescended testes, of whom 73% were attempting conception. Of these, 48% had been successful. This compared with 100 men with a unilateral undescended testis, of whom 90% were attempting conception and 80% had been successful. In a 1989 study, Cendron et al.48 followed up 40 men operated on between 1950 and 1960 when the operation was performed at a mean age of 7 years. Twenty-three out of 30 men with a past history of a unilateral undescended testis had attempted paternity, with an 87% success rate, compared with 3/9 men with bilateral undescended testes attempting paternity and a success rate of only 33%. In a different approach, Kumar et al.49 reviewed 56 men undergoing orchidopexy between 1950 and 1975; they recorded the age of the man at the birth of his first child, and found only very late treatment affected fertility adversely. However, only 3/113 orchidopexies were done at less than 6 years of age. Semen analysis Numerous studies have addressed the outcome for orchidopexy by semen analysis (Table 51.4). Because of their logistic difficulties, none of these studies has huge numbers of patients, with most having between 50 and 150 men enrolled. The age at initial orchidopexy varies from 1.5 to 29 years (!), but most series have an average age for operation of 5–10 years. Lee44,46 found that men with a past history of bilateral undescended testes were reported as having a wide range of subnormal sperm counts. Some papers reported a 44% incidence of abnormality, compared with others ranging right up to 100% abnormal. However, in all reports at least half the patients were azoospermic. On average, only 25% with a past history of bilateral undescended testes were reported as having normal sperm counts. This

Undescended testes

compares with those men with a past history of a unilateral undescended testis, where 20%–69% were reported as having subnormal counts, and 57% were found to have normal sperm counts. Not surprisingly, the results of the semen analysis appear worse than the success rates reported above for paternity. A number of factors are likely to account for this, including the fact that the time taken to achieve paternity has not been calculated, let alone the genetic proof of paternity. Also, it is hard to assess the importance of the deliberate redundancy of biological factors in the semen analysis. Lee reported that, on his review of the literature,44 there was no proven effect of decrease in age (at operation) on outcome, and no proven effect of location of the testis on outcome. Similarly, Gracia et al.50 found that, in 251 men with a history of undescended testis, the sperm quality was independent of the age of surgery or testicular location, but was influenced only by whether the cryptorchidism was unilateral or bilateral. Interpretation of such pessimistic conclusions about surgical benefit, however, should be guarded, since these reports described the outcome of operations done 30 years ago. By contrast, some studies, such as Taskinen et al.51 report significantly better results. They have reviewed 31 men operated on at ages between 10 months and 12 years. Normal semen analysis was found in 90% of those with a unilateral undescended testis, compared with 50% of those with bilateral undescended testes. In particular, those men undergoing surgery before 4 years of age were found to have no abnormalities on semen analysis. Hormone levels were recorded and serum FSH levels were found to be elevated in those with severe testicular damage. An improved fertility outcome with younger age at surgery may not be obvious with the present long-term follow-up results, but it is anticipated by recent pediatric studies. McAleer et al.52 reviewed 355 testicular biopsies in 226 children coming to orchidopexy at between 6 months and 16 years of age. They calculated the fertility index as the mean number of spermatogonia per cross-section of the tubule (S:T ratio) in 50 tubules, and compared these with age-matched controls. They found that those undergoing surgery at less than one year had a normal fertility index, while those coming to operation beyond one year had a decrease in fertility index compared with controls. The contralateral descended testes had a normal fertility index for operation age up to 6 years, but beyond that time they became abnormal. Huff et al.17 and Hadziselimovic and Herzog18 summarized their results of germ cell development in undescended testes and showed loss of the germ cells during the first year of life with defective or delayed transformation of neonatal gonocytes into type

A spermatogonia. Also, delayed or defective transformation of type A spermatogonia to primary spermatocytes was seen at 3–5 years of age. The contralateral testes showed some changes but much less severe than in the undescended testes. They concluded that these results were consistent with hypogonadotrophic hypogonadism, although equally they could indicate relatively normal testes initially undergoing secondary degeneration. Cortes et al.53 recently reviewed 1335 boys with cryptorchidism who underwent orchidopexy and biopsy for undescended testes below 12 years of age. One hundred and forty-six of these underwent detailed assessment in adult life. Lack of germ cells in the original biopsy was found from 18 months, and increased with age. Normal sperm counts were found in 19% (14/75) and 83% (54/65) of men who had undergone bilateral or unilateral orchidopexy, respectively. When surgery is delayed to adolescence or beyond54 azoospermia or oligospermia is likely. In 91 men with unilateral undescended testes having surgery at between 14 and 29 years, 83% were subsequently found to be oligoor azoospermic. Cortes and Thorup55 reviewed 90 boys with an average age of 13 years, undergoing bilateral orchidopexy with biopsy. Seventy-three of these were followed up at an average age of 25.6 years. The fertility index on initial biopsy (percentage of tubular cross-sections with spermatogonia) was correlated with the subsequent maximum sperm density and the volume of the testes, and inversely correlated with the serum FSH level. They found that those testes that were in the superficial inguinal pouch had significantly better outcomes than those that were intracanalicular or abdominal. Okuyama et al.56 compared 167 men operated on at 2–5 years of age, with 43 operated on at 9–12 years of age. Of those having surgery early in childhood, 95% had normal semen analysis with a unilateral undescended testis, and 24% had a normal sperm count with bilateral undescended testes. This compared with 86% with normal sperm counts in those having surgery in later childhood for a unilateral UDT, and 20% normal sperm counts with bilateral undescended testes. They concluded that operation did not reverse the poor fertility in bilateral undescended testes, and that the contralateral testis in those with a history of a unilateral undescended testis had poor function. Bremholm-Rasmussen et al.41 reviewed 45 men who had bilateral undescended testes that were never treated by surgery but descended spontaneously after 10 years of age. Clinically, they fell into two groups. In the first group, the testis was in the superficial inguinal pouch and it could be manipulated into the scrotal neck but immediately retracted. The second group had suprascrotal testes which could be manipulated to the middle of the scrotum, but

657

658

J. M. Hutson

Table 51.5. Relative risk of testicular malignancy in cohort studies

Series

Number of cancer patients

Number with previous UDT

UDT/cancer (%)

UDT in population (%)

Relative risk

Gilbert and Hamilton62 Campbell63

>7000 1422

840 165

11 11.6

0.23 0.23

48 50.4

UDT, undescended testis.

then retracted. Such testes would be regarded as retractile or ascending in many series. At 18 years of age they underwent semen analysis, and only 33% were found to have normal values. Forty-seven percent had azoospermia or oligospermia, and a large number had increased FSH levels. Perhaps these patients can be regarded as having acquired maldescent and being equivalent to the subgroup of adolescents whose undescended testes were known in the past to descend spontaneously at puberty. Their poor sperm counts in adult life suggest that, even later in childhood, 5–10 years of non-scrotal position does cause secondary degeneration of the testes. Mayr et al.57 reviewed 46 men undergoing operation at between 1.5 and 12 years of age. Eleven out of 24 with a unilateral undescended testis had normal sperm counts. By contrast, only 7/22 men with bilateral undescended testes had normal sperm counts; 13% of these had azoospermia. They found there was no correlation between semen analysis and either the age at surgery or the spermatogonia/tubule ratio. They found that there was a correlation between the initial gonadal position and both subsequent sperm counts and FSH levels, with retractile and high scrotal testes having significantly better outcomes than intraabdominal or inguinal testes. Mandat et al.58 reviewed 135 men who underwent surgery at between 2.5 and 15 years of age. They grouped them by age at operation: before 6 years of age; at 6–10 years of age; and over 10 years. The position of the gonad was recorded as either abdominal, canalicular or inguinal. The percentage of men with normal sperm counts decreased with increasing age at surgery for the 112 men with a unilateral undescended testis. In addition, the number of sperm also decreased with increasing age in this group. The number of men with bilateral undescended testes was too small to see any age effect. The effect of early operation was more clearly seen when the men were classified into those with gonads in a similar position, where it could be seen that early age was a significant advantage for testes in a similar position. Mandat and colleagues concluded that surgery would be optimal at 1–2 years of age in the hope of preventing this secondary degeneration.

Urry et al.59 looked at the effect of orchidopexy on the development of antisperm antibodies in adult life. Fifty-five men were reviewed by semen analysis for the presence of antisperm antibodies. In the context of an infertility clinic, it was known already that 66% of men with a past history of undescended testes had antibodies, compared with only 2.8% of controls. In their group of 55 men, antibodies were present in 52% after orchidopexy. Antibodies were present when orchidopexy was done at an average age of 11 years. By contrast, when orchidopexy was done at an average age of 8.6 years, antibodies were not present subsequently. It is difficult to interpret these results, but it might be an indicator that testicular degeneration which persists into adolescence allows the patient to be exposed to a testis with a deficient blood–testis barrier, which would then allow autoimmunization against sperm antigens to occur.60

Malignancy follow-up studies In a review of neoplasia in men with previous cryptorchidism, Whitaker61 reported that Percival Pott knew in 1777 that there was an increased risk of malignancy in the undescended testis. Since that time, however, the relative risk of malignancy related to different forms of cryptorchidism has remained uncertain (Table 51.5). In 1940, Gilbert and Hamilton62 reviewed records from the US Bureau of Census and found over 7000 cases of testicular malignancy. Review of their past histories revealed 840, or 11%, of the total, who had had previous undescended testis. They related this incidence to over nine million military recruits, of whom 22 665 had undescended testes, an incidence of 0.23%. They concluded therefore, that the relative risk of malignancy in a person with undescended testis was 48-fold. In a similar study, Campbell63,64 reviewed 1422 patients with testicular tumors, of whom 165 (11.6%) had a history of undescended testis. Using the same population statistics for undescended testis, this gave a relative risk of 50-fold. One difficulty with these studies was the potential bias in the number of patients in the population with undescended testis: the use of adult army recruits introduced

Undescended testes

659

Table 51.6. Relative risk of subsequent malignancy after cryptorchidism

Series

Number of men with previous UDT

Number who developed cancer (95% CI)

Number in population expected to develop cancer

Relative risk (95% CI)

Giwercman et al.65 ” ” Campbell63 Benson et al.66 ” ”

506 – 1413 224 –

6 (2.2–13.1) 22 2 –

1.3 – 0.07% 2% >35%

4%

? ? 0.2% dehiscence 0.02% glans injury 0.05% excess skin Removed 0.13% inappropriate circ. (hypospadias) ?

5882 1.07% 0.02% (retention) 0.42%

USA (1976) Retrospective over 10 y

Gee and Ansell23

31% 8% 1% (recurrent phimosis)

Canada (1966) Questionnaire and examination 6 mo-2y 100 4% (moderate) ? 8%

Patel26

Table 52.1. Complications of neonatal circumcision in five US studies

USA (1989) Armed forces retrospective study 100 157 0.083% (surgery in 0.03%) ? 0.07% (bacteremia in 0.008%) 0% 0% 0.02% UTI 0.007% dehiscence 0.001% circumcised hypospadias 0.001% urethral injury 0.01% pneumothorax 0.015% too little/ too much skin excised 0.052% (potentially from the data) 0.19%

Wiswell and Geschke27

1.7%

?

? ? 0.2% aspiration pneumonia 0.4% malignant hyperpyrexia

USA (1983) Armed forces retrospective study 476 (mean age 2.9y) 0.9% ? 0.2%

Wiswell et al.32

Circumcision

667

Table 52.2. Complications of circumcision in older children MacCarthy et al. 4 Country (and year) Type of study

UK (1952) Patient visit at age 4 y Number of patients 585 Age group 0–4 y Hemorrhage 1.5 Difficulty with micturition (%) 0.2 (retention) Infection (%) 1.4 Meatal ulcer ? Meatal stenosis (%) ? Inadequate circumcision (%) ?? Reoperation rate 1.2 Overall complication rate (%) 5.5

Leitch2

Fraser et al. 28

Stenram et al. 6

Griffiths et al. 30

Cuckow et al. 34

Australia (1970) Retrospective

UK (1981) Prospective randomized trial 100 4.7 y (mean) 20 43 7 ? ? 1 1 29

Sweden (1986) 5 y follow-up

UK (1985) Prospective

UK (1993) Prospective audit

117 0–10 y 6 ? 0.8 ? 11 1.7 1.7 19.6

140 4.3 y (mean) 1.4 readmitted 0.7 (retention) 4.3 ? 2.8 0.7 3.6 15

50 6.6 y (mean) 6 14 12 ? ? 6 (poor cosmesis) 6 40 (anesthetic comps. included)

200 2.3 y (mean) 7 ? 0.5 4 1.5 1 3.5 15.5

be anastomosed to the shaft skin. Hemostasis and closure are achieved in the same way.

Complications There are few prospective and long-term studies of the complications of circumcision. It is often performed by a junior surgeon,2 and in the USA is rarely performed by a surgeon at all.15 The assumed unimportance of the foreskin has led to an underestimate of the importance of the complications of circumcision. In the UK, between 1942 and 1947 there were an average of 16 deaths a year reported following circumcision in those under 5 years. Causes of death included anesthesia, hemorrhage, and systemic infection. There were two deaths attributable to circumcision in Australia between 1960 and 1966,2 and more recent reviews suggest that it is now extremely uncommon.3,15 Outside modern medical practice, there is still a significant mortality, as the experience with the results of ritual circumcision in one African tribe attests.31 The tables summarize the complications of circumcision from reported Western series. Table 52.1 relates to routine neonatal circumcision. Three of these studies contain impressively large numbers of patients with low overall complication rates of 0.06–1.9%.1,27 The performance of procedures in many different centers by unidentified practitioners, the use of retrospective data collection from hospital records, and lack of follow-up in these studies does call into question the validity of the oft-quoted complication rate of 0.06% from Speert’s series.1 There is one

prospective study of 100 patients that is limited to one center and a known group of operators, and this involved direct patient/parent contact.26 This showed a significantly higher complication rate of over 35%. Further comparison is provided in a second study by Wiswell, who identified an increased complication rate of 1.7% in boys circumcised later in life, compared with 0.19% in his series of neonatal circumcisions using the same method of inquiry.27,32 Many of these patients had developed a medical indication for the procedure, required general anesthetic, and were circumcised by a different technique, although he used these data to further justify the timing of the operation in the neonatal period.32 Comparison with clamp-type circumcisions has revealed an increase in infection around retained Plastibel devices and greater bleeding with the Gomco device that relies on skin crushing alone for hemostasis.33 One study comparing Plastibel to freehand circumcision did not support this increased infection rate and failed to show an appreciable difference in older boys.28 Table 52.2 shows a European and Australian experience of circumcisions performed for medical reasons in older boys over the same time period. While numerically smaller, these studies reflect the practice of individual surgeons or groups of surgeons, and their design has been either prospective or retrospective with some patient contact. There are significant differences in complication rates between them and the neonatal group, although the closer agreement with Patel’s neonatal series suggests that differences may be more related to the type of study than to the timing or technique of operation. Even within these series there is great variation, alluding to the subjectivity

668

Peter M. Cuckow

Fig. 52.1. Recurrent phimosis due to contraction of the circumcision scar, concealing the glans (see colour plate section).

of the assessment of complications, which have never been standardised. Thus, hemorrhage varies from 1.4% to 20%, requiring readmission and operation30,34 but retention requiring a catheter is reported in only 0.2–0.7%.4,30 Minor infection is common if looked for and results in visits to the family doctor and antibiotic prescription in up to 45% of cases,34 although it is rarely the cause of long-term problems. The literature contains reports of necrotizing fasciitis, Fournier’s gangrene, meningitis, and tetanus following circumcision.3,15 Such severe sepsis is rare but when encountered may result in death.35,36 As one might expect, the reoperation rate was highest (14.5%) in the prospective study with the longest follow-up (5 years) and was mostly for the correction of meatal stenosis.6

Long-term outcomes Poor long-term results from circumcision can be subdivided into: r skin complications: the removal of too much or too little skin, or unsatisfactory skin healing; r irreversible damage to the penis, glans and erectile mechanism;

r inappropriate circumcision of an abnormal penis (buried

penis or hypospadias), compromising reconstruction;

r trauma to the sensitive glans and urethral meatus follow-

ing their exposure.

Skin complications Leaving too much of the inner mucosal layer of foreskin followed by cicatrisation at the suture line is a cause of recurrent phimosis or burying of the penis and requires recircumcision28,37 (Fig. 52.1). The amount of foreskin to be retained after circumcision is set out precisely in Jewish law,38 but away from this there is considerable variation in opinion as to what constitutes an adequate or “normal” circumcision. Certainly, incomplete or inadequate circumcision is unsightly and often associated with the use of a circumcision clamp. Significant numbers of revision circumcisions are performed in centers where neonatal circumcision is practised, notably Atlanta where 46 were performed in 2 years.39 In spite of this, large neonatal series often omit this complication.1,27,33 Among Leitch’s 200 consecutive circumcisions, 19 were revisions following an inadequate initial procedure,2 and McCarthy et al. reported revising 1% of cases following freehand circumcision.4

Circumcision

Excessive loss of skin can occur with circumcision clamps and results in a denuded penile shaft.40 This can also be due to local infection and diathermy injury.3,15 In neonatal circumcision, healing can be achieved by secondary intention through granulation with satisfactory long-term appearance.29 In older patients, skin grafting gives excellent functional and cosmetic results. Burying the shaft in a scrotal tube is another option, although the cosmetic results using this hair-bearing skin are questionable.3,33,40 Glanular adhesions with the inner preputial skin due to reformation or failure to divide them properly at circumcision are well recognized15,53 and have a tendency to separate spontaneously with time. Adhesion of circumcision scar to a traumatized or ulcerated glans edge may result in a more permanent adhesion or skin-bridge, beneath which smegma accumulates and which can cause pain and deformity.15 This was seen in around 2% of patients in Ponsky’s study of 254 boys following neonatal circumcision53 and requires surgical release. Epidermal inclusion cysts are also reported,29 and penile lymphedema, the etiology of which is unclear, is seen extremely rarely.3,35 The assessment of cosmetic results is very subjective. A questionnaire sent to the parents of 50 circumcised boys found that 26% found its appearance to be normal, 53% acceptable and 21% slightly scarred with no complaints of a bad result. Surgeons appraising the same patients offered a revision circumcision in three for a poor cosmetic result.3 Certainly, the skin–mucosal junction may be scarred and unsightly, particularly where sutures produce a cross-hatching effect (see Fig. 52.2). There are no current reviews of this aspect of circumcision and undoubtedly many men put up with poor cosmetic results without complaint.

Major penile injury Amputation of part or all of the glans is rare but may arise when it is caught in the jaws of a circumcision clamp.39,40 The excised portion may be grafted back on the penis (with reported success),39,40 or the glans can be repaired.41 Pressure necrosis is also seen with the Plastibel if too small a bell is selected.33 Urethral damage is caused by overzealous hemostasis with diathermy on the ventral surface of the penis in the region of the frenular artery (Fig. 52.2). It can also be caught up in a Gomco or Plastibel clamp, leading to a subcoronal fistula, which requires secondary repair.54 Corporal fibrosos and total ablation of the penis are fortunately very rare occurrences and are often sequelae of the use of monopolar diathermy alone or in combination with a metal clamp such as the Gomco.42–44 In some patients,

Fig. 52.2. Urethrocutaneous fistula following circumcision, probably due to diathermy around the frenular artery. Note the prominent cross-hatching of the circumcision scar and poor cosmetic result (see colour plate section).

complex reconstruction is possible and reasonable results have been reported.43,44 Otherwise, gender reassignment, staged genital reconstruction and hormonal manipulation at puberty is advocated by Gearhart and Rock.42 Two of their four patients are sexually active following their reconstructions and the other two are awaiting the start of estrogen therapy at 11 years and their subsequent vaginoplasty at 15 years of age. All individuals and their families have had counseling and require help to come to terms with the consequences of this disastrous injury and its management.42 Although only reported in adults, failure of the erectile mechanism very rarely occurs following intracorporeal injection of local anesthetic agents, or may be a sequence of more serious penile injury.15

Inappropriate circumcision Circumcision should be avoided in hypospadias, where the foreskin is required for reconstruction. While perhaps more understandable in the intact prepuce megameatus variant of hypospadias,45 which may appear normal to the untrained eye, inappropriate circumcision has been reported in 36% of hypospadii from one series of neonatal circumcisions.33 This relates probably to its performance

669

670

Peter M. Cuckow

Fig. 52.3. Normal slit-like urethral meatus (see colour plate section).

by non-urologists, e.g., pediatricians, gynecologists, junior residents, and midwives. Buried penis, an anomaly comprising insufficient shaft skin, abundant inner preputial skin, phimosis retaining the penis within abdominal wall fat, and abnormal tethering of the corporal bodies, is a contraindication to circumcision.16 Removal of the usual amount of interpreputial skin in these cases leaves skin deficiency. Leaving enough inner preputial skin to cover the penile shaft simply results in recurrent burying of the penis if only a circumcision is performed. A more complex procedure is required to release the corporal bodies and obtain a satisfactory result.48,55

Meatal ulceration and stenosis The urethral meatus is arguably the most sensitive area of the penis, and close inspection reveals the normal eversion of urethral mucosa at this point (Fig. 52.3). Meatal ulceration has been documented in 20% of infants 2–3 weeks after circumcision,47 and Patel’s prospective study revealed this problem in 31%.26 It is thought to be due to chemical and physical irritation within the wet ammoniacal diaper,47,48 which is borne out by a lower incidence in older boys – 4% with circumcision performed at a mean

age of 2 years 4 months2 – and offered as a reason to postpone circumcision until after potty training.49 The natural history of meatal ulceration is either resolution or progression to chronic scarring and meatal stenosis, which is reported in 1.5%–11% of patients after circumcision in those series with sufficient follow-up. 2,6,26,30 This comprised almost 26% of Patel’s and 38% of Leitch’s patients with early meatal ulceration.2,26 In addition, Persad et al.48 have proposed that the relative ischemia of the meatus that follows division of the frenular artery is highly significant in the pathogenesis of meatal stenosis. Berry and Cross48 calibrated the urethral meatus in children and adults: 60% of circumcised men had a meatus of 20F or less, compared with only 25% of uncircumcised men. These differences were more significant with earlier circumcision. In circumcised infants of 12–18 months, only 30% had a meatal caliber greater than or equal to 14F, compared with 53% of those uncircumcised, whereas meatal size was the same in infants under 1 year of age. It appears from these data that meatal stenosis develops some time after circumcision, accounting for its absence as a complication in many series, and will be detectable in at least a subclinical form in most patients. Persad et al.48 reviewed the symptoms of 12 patients presenting between 4 months and 8 years after circumcision, finding pain at the initiation of micturition in them all, a narrow high-velocity stream in 8 cases, spillage of urine around the toilet bowl in 6 cases and urethral bleeding in one case. Clinical examination showed a pinhole meatus with a filmy membrane covering its lower margins (Fig. 52.4(a) and 52.4(b)). There was no history or clinical features of balanitis xerotica et obliterans. Meatotomy – opening the ventral aspect of the meatus proximally in the midline and suturing inner epithelium to outer skin with absorbable sutures – was effective in all cases, and there was no recurrence after a mean follow-up of 13 months. Rarely the bladder and upper tracts are secondarily affected by meatal stenosis, leading to bladder hypertrophy that may ultimately produce upper urinary tract dilatation and sepsis.49 Figure 52.5 demonstrates a case of meatal stenosis in a 7-year-old boy who presented several years after circumcision with a long history of difficulty in micturition and a poor urinary stream that progressed to chronic urinary retention. The cystourethrogram, performed via a suprapubic catheter, showed a grossly dilated urethra and trabeculated bladder resulting from chronic meatal stenosis. The upper tracts were not involved and he was treated successfully by meatotomy. With this picture and the high bladder pressures needed to produce effective micturition, it is not surprising that these patients complain of urethral pain.48

Circumcision

(a)

Fig. 52.5. A cystogram in a patient with chronic, severe meatal stenosis with secondary changes in the bladder.

(b) Fig. 52.4. (a) and (b). Reduction of the meatal caliber in meatal stenosis so only a fine lachrymal probe can pass (see colour plate section).

(Fig. 52.6). Complete removal of the foreskin can remove all the affected tissue14 and glanular lesions appear to regress spontaneously following circumcision. There were no recurrences or long-term complications in 10 children followed up to 5 years.14

Religious circumcision Balanitis xerotica et obliterans Otherwise known as lichen sclerosus et atrophicus, BXO is characterized by tight phimosis with a sclerotic whitish ring at the tip of the prepuce and characteristic lesions on the glans. It is occasionally associated with meatal stenosis, perhaps representing a cause of meatal stenosis following circumcision.30 Frank et al.50 have demonstrated that 10/12 boys with meatal and submeatal stenosis had previously been circumcised for BXO, although this is at variance with the clinical impression of Persad et al.48 for whose 12 patients BXO was not a known precursor of meatal stenosis. BXO is the truest indication for medical circumcision13 and is clinically recognized by its sclerotic preputial orifice

Although performed in the community by Mohels, admissions to hospital following Jewish circumcision of 8-dayold infants reveals a similar spectrum of complications, although the exact rate is unknown.29,41,51 If the circumcision dressing is too tight, urinary retention, infection, and septicemia may ensue.39,41 Its use probably explains the increased incidence of urinary tract infection in Jewish boys, which is highest in this group during the 12 days after circumcision.52 Severe penile injury is related to inexperienced Mohels,29 and recircumcision is required when too much skin is left.38 An extreme view from another culture is presented by Crowley and Kessner,31 who found a mortality of 9% among 45 consecutive youths presenting with severe complications of ritual circumcision amongst the Xhosa people of South Africa.

671

672

Peter M. Cuckow

were pleased with the operation. Success of preputial plasty depends on early mobilization of the foreskin, which is possible without discomfort in around 90% of boys within 2 weeks of surgery. This prevents contraction of the suture line and scarring that leads to secondary or true phimosis, and was the reason for circumcision in 4% of patients. 70% of parents judged the appearance of the penis to be normal after this procedure, and the long-term results also appear excellent, although they warrant more formal appraisal.34 Certainly no case of meatal stenosis has been reported following this procedure.48

Conclusions

Fig. 52.6. The sclerotic prepucial tip, which is characteristic of BXO (see colour plate section).

Preputial plasty If the complications of possessing a foreskin are related to its non-retractability and poor hygiene, then there are alternative surgical procedures. Preputial plasty has been used in Europe to produce a retractable yet intact foreskin. There are various techniques, the simplest consisting of a short dorsal incision at the narrowest point of the prepuce.34 The foreskin is retracted and the congenital adhesions freed. The narrowed distal prepuce produces a constricting band on the shaft of the penis, and this is divided by a dorsal longitudinal incision to the level of Buck’s fascia. The resulting defect allows widening of the prepuce at this point and is sutured transversely with absorbable sutures. The foreskin is then checked for mobility and the parents encouraged to commence regular retraction within days of the operation. Comparing 50 patients undergoing this preputial plasty with 50 undergoing circumcision, there was considerably less short-term morbidity. The minimal dissection required, and in particular avoidance of the ventral aspect of the penis and the frenular artery, results in greater postoperative comfort and elimination of bleeding as a serious complication. There were also fewer consultations with the patient’s general practitioner after surgery, and parents

Circumcision and its indications remain a hot discussion topic in both the medical and the lay literature. Arguments are laced with subjectivity, emotion, religious fervour and too little scientific study. There are significant short-term complications of the operation whichever timing, technique or setting is employed, and these are underestimated by many of the studies cited. Major injury to the penis, although fortunately rare, will have a profound effect on the boy’s life and, more frequently, loss of protection of the urinary meatus is a potent cause of long-term problems. Avoidance of these is simple – avoidance of circumcision. While most arguments for circumcision are found lacking under scrutiny, congenital urological anomalies diagnosed antenatally may provide the only indication for prophylactic neonatal circumcision to help protect against urinary tract infection. It is true to say that the long-term effects of this procedure are yet to be fully assessed over 3000 years since it was first performed.

Acknowledgment Marie-Klare Farrugia, Research Registrar in Urology, The Institute of Child Health.

REFERENCES 1. Speert, H. Circumcision of the newborn: appraisal of its current status. Obstet. Gynecol. 1953; 2:164–172. 2. Leitch, I. O. W. Circumcision: a continuing enigma. Austral. Pediatr. J. 1970; 6:59. 3. Williams, N. & Kapila, L. Complications of circumcision. Br. J. Surg. 1993; 80:1231–1236. 4. McCarthy, D., Douglas, J. W. B., & Mogford, C. Circumcision in a national sample of 4-year-old children Br. Med. J. 1952; ii:755– 756.

Circumcision

5. Rickwood, A. M. K. & Walker, J. Is phimosis overdiagnosed and are too many circumcisions performed in consequence? Ann. Roy. Coll. Surg. Engl. 1989; 71:275–277. 6. Stenram, A., Malmfors, G., & Okiman, L. Circumcision for phimosis: a follow-up study. Scand. J. Urol. Nephrol. 1986; 20:89–92. 7. O’Brien, T. R., Calle, E. E., & Poole, W. K. Incidence of neonatal circumcision in Atlanta, 1985–1986. South. Med. J. 1995; 88:411– 415. 8. Gairdner, D. The fate of the foreskin. Br. Med. J. 1949; 2:1433– 1437. 9. Øster, J. Further rate of the foreskin. Arch. Dis. Child. 1986; 43:200–203. 10. Rickwood, A. M. K., Hemalthala, V., Batcup, G., & Spitz, L. Phimosis in boys. Br. J. Urol. 1980; 52:147–150. 11. Harnes, J. R. The foreskin saga. J. Am. Med. Assoc. 1971; 217:1241–1242. 12. Escala, J. M. & Rickwood, A. M. K. Balanitis. Br. J. Urol. 1989; 63:196–197. 13. Anonymous. Medical indications for childhood circumcision. Drug Ther. Bull. 1993; 31:99–100. 14. Mueli, M., Briner, J., Hanniman, B., & Sacher, P. Lichen sclerosus et atrophicus causing phimosis in boys: a prospective study with 5-year follow-up after complete circumcision. J. Urol. 1994; 152:987–989. 15. Kaplan, G. W. Complications of circumcision. Urol. Clin. N. Amer. 1983; 10:543–549. 16. Alter, G. J., Horton, C. E., & Horton, C. E., Jr. Buried penis as a contraindication for circumcision. J. Am. Coll. Surg. 1994; 178:487–490. 17. Koo, H. P. & Duckett, J. W. Circumcision quo vadis? In: Williams, D. I., & Etker, S. (eds), Contemporary Issues in Paediatric Urology, in Memoriam Herbert B. Eckstein. Istanbul: Logos, 1996: 149–154. 18. Schoen, E. J. Report of the task force on circumcision. Pediatrics 1990; 84:388. 19. Winberg, J., Bollgren, I., Gothefors, L., Herthelius, M., & Tullus, K. The prepuce: a mistake of nature? Lancet 1989; i:598–599. 20. Ginsberg, C. M. & McCracken, G. H. Urinary tract infections in young infants. Pediatrics 1982; 69:409–412. 21. Wiswell, T. E. & Roscelli, J. D. Corroborative evidence for the decreased incidence of urinary tract infections in circumcised male infants. Pediatrics 1986; 78:96–99. 22. Craig, J. C., Knight, J. F., Sureshkumar, P., Mantz, E., & Roy, L. P. Effect of circumcision on incidence of urinary tract infection in preschool boys. J. Pediatr. 1996; 128:23–27. 23. Herzog, L. W. & Alvarez, S. R. The frequency of foreskin problems in uncircumcised children. Am. J. Dis. Child. 1986; 140:254– 256. 24. Fergusson, D. M., Lawton, J. M., & Shannon, F. T. Neonatal circumcision and penile problems: an 8 year longitudinal study. Pediatrics 1985; 75:901–903. 25. Wiswell, T. E., Smith, F. R., & Bass, J. W. Decreased incidence of urinary tract infections in circumcised male infants. Pediatrics 1985; 75:901–903.

26. Patel, H. The problem of routine circumcision. Can. Med. Assoc. J. 1966; 95:576–581. 27. Wiswell, T. E. & Geschke, D. W. Risks from circumcision during the first month of life compared with those of uncircumcised boys. Pediatrics 1989; 83:1011–1015. 28. Fraser, I. A., Allen, M. J., Bagshaw, P. F., & Johnstone, M. A. randomised trial of childhood circumcision with the Plastibel device compared to a conventional dissection technique. Br. J. Surg. 1981; 68:593–595. 29. Shulman, J., Ben-Hur, N., & Neuman, Z. Surgical complications of circumcision. Am. J. Dis. Child. 1964; 107:149–154. 30. Griffiths, D. M., Atwell, J. D., & Freeman, N. V. A prospective study of the indications and morbidity of circumcision in children. Eur. Urol. 1985; 11:184–187. 31. Crowley, I. P. & Kessner, K. M. Ritual circumcision (Umkhwetha) amongst the Xhosa of the Ciskei. Br. J. Urol. 1990; 66:318– 321. 32. Wiswell, T. E., Tencer, H. L., Welch, C. A., & Chamberlain, J. L. Circumcision in children beyond the neonatal period. Pediatrics 1993; 92:791–793. 33. Gee, W. F. & Ansell, J. S. Neonatal circumcision: a ten-year overview with comparison of the Gomco clamp and Plastibel devices. Pediatrics 1976; 58:824. 34. Cuckow, P. M. & Mouriquand, P. D. E. Saving the normal foreskin. Br. Med. J. 1993; 306:459–460. 35. Fredman, R. M. Neonatal circumcision: a general practitioner survey. Med. J. Austral. 1969; 1:117–120. 36. Cleary, T. G. & Kohl, S. Overwhelming infection with group B beta haemolytic streptococcus associated with circumcision. Pediatrics 1979; 64:301–303. 37. Redman, J. F., Scriber, L. J., & Bissada, N. K. Postcircumcision phimosis and its management. Clin. Pediatr. 1975; 14:407–409. 38. Sotolongo, J. R., Hoffman, S., & Gribetz, M. E. Penile denudation injuries after circumcision. J. Urol. 1985; 133:102–103. 39. Horrowitz, S. J. & Glassberg, K. I. Circumcision: successful glanular reconstruction and survival following traumatic amputation (abstract). American Academy of Pediatrics Annual Meeting, New Orleans, 1995. 40. Gluckman, G. R., Stoller, M. L., Jacobs, M. M., & Kogan, B. A. Newborn penile glans amputation during circumcision and successful reattachment. J. Urol. 1995; 153:778–779. 41. Menahem, S. Complications arising from ritual circumcision: pathogenesis and possible prevention. Israeli J. Med. Sci. 1981; 17:45–48. 42. Gearhart, J. P. & Rock, J. A. Total ablation of the penis after circumcision with electrocautery: a method of management and long term follow-up. J. Urol. 1989; 142:799–801. 43. Azmy, A., Boddy, S. A., & Ransley, P. G. Successful reconstruction following circumcision with diathermy. Br. J. Urol. 1985; 57:587– 588. 44. Stefan, H. Reconstruction of the penis after necrosis due to circumcision burn. Eur. J. Pediatr. Surg. 1995; 4:40–43. 45. Duckett, J. W. & Keating, M. A. Technical challenge of the megameatus intact prepuce hypospadias variant: the pyramid procedure. J. Urol. 1989; 141:1407.

673

674

Peter M. Cuckow

46. Joseph, V. T. A new approach to the surgical correction of buried penis. J. Pediatr. Surg. 1995; 30:727–729. 47. MacKenzie, A. R. Meatal ulceration following circumcision. Obstet. Gynecol. 1966; 28:221–223. 48. Persad, R., Sharma, S., McTavish, J., Imber, C., & Mouriquand, P. D. E. Clinical presentation and pathophysiology of meatal stenosis following circumcision. Br. J. Urol. 1995; 75:91–93. 49. Berry, C. D. & Cross, R. R. Urethral meatal calibre in circumcised and uncircumcised males. Am. Med. Assoc. J. Dis. Child. 1956; 92:152–156. 50. Frank, J. D., Pocock, R. D., & Stower, M. J. Urethral strictures in childhood. Br. J. Urol. 1998; 62:590–592.

51. Schlosberg, C. Thirty years of ritual circumcision. Clin. Pediatr. 1971; 10:205–209. 52. Cohen, H. A., Drucker, M. M., Vainer, S. et al. Postcircumcision urinary tract infection. Clin. Pediatr. 1992; 31:322–324. 53. Ponsky, L. E., Ross, J. H., Knippper, N., & Kay, R. Penile adhesions after neonatal circumcision. J. Urol. 2001; 165(3):915. 54. Baskin, L. S., Canning, D. A., Snyder, H. M. III & Duckett, J. W. Jr. Surgical repair of urethral circumcision injuries. J. Urol. 1997; 158(6): 2269–2271. 55. Smeulders, N., Wilcox, D. T., & Cuckow, P. M. Surgical correction of the buried penis – an anatomical approach. BJU Int. 2000; 86:523–526.

53 The single kidney Adrian S. Woolf Nephro-Urology Unit, UCL Institute of Child Health, London, UK

Introduction This review covers three clinical scenarios which feature the single kidney. The first category concerns patients who are born with only one kidney: they have congenital solitary functioning kidney, either caused by unilateral renal agenesis or regression of a malformed rudiment. After addressing the definition, incidence and diagnosis of this disorder, three aspects will be discussed: the putative developmental etiologies of this disorder; the occasional familial nature of the disorder suggesting a genetic basis to the disorder, and the long-term outcome of these individuals in terms of risk of subsequent disease in the kidney and the occurrence of hypertension (the “renal prognosis”). The second category concerns patients who, in childhood or adulthood, have had either a unilateral nephrectomy or a subtotal nephrectomy for intrinsic renal disease. The latter subgroup have had one kidney and a fraction of the contralateral organ removed and are said to have a “remnant kidney.” I will discuss the clinical evidence which addresses the renal prognosis of these individuals. The third category concerns the renal prognosis of the single kidney in otherwise healthy renal transplant donors.

Renal agenesis and the congenital single kidney Definition Renal agenesis implies the total absence of the kidney and can be considered as part of the spectrum of renal malformations which also include: (i) renal hypoplasia, a disorder

in which the kidney is small and contains fewer nephrons than normal; (ii) renal dysplasia in which the kidney contains undifferentiated tissue; (iii) the multicystic dysplastic kidney in which the dysplastic organ contains massive cysts and; (iv) renal aplasia which describes a tiny dysplastic organ.1,2 Unilateral renal agenesis is often associated with absence of the ipsilateral ureter and bladder trigone. On occasion, the ipsilateral adrenal gland or gonad may also be missing. With the advent of fetal renal ultrasound scanning, it has become clear that a solitary kidney may result from in utero or postnatal regression of a contralateral multicystic dysplastic kidney.3 In addition, small dysplastic, nonfunctional, kidneys (i.e., “renal aplasia”) can regress in a similar manner.4 This process is perhaps explained by the excess of programmed cell death, or apoptosis, which has been recorded in these organs.5

Clinical diagnosis Isolated, non-syndromic, renal agenesis is usually clinically silent. Therefore its incidence is probably underestimated. The complete absence of a kidney can only be definitively diagnosed by direct inspection at laparotomy or autopsy. While commonly used renal ultrasound scanning cannot exclude the presence of a tiny contralateral aplastic kidney, this would be the most common imaging technique. While intravenous pyelography and isotope renography would also fail to detect an aplastic contralateral organ, these techniques do have the advantage of detecting an ectopic contralateral organ (e.g., a pelvic kidney) which might be missed on ultrasound imaging.

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


675

676

Adrian S. Woolf

Incidence of solitary kidney and contralateral renal agenesis In a study of 9200 autopsies, unilateral renal agenesis was found in seven subjects.6 In contrast, a radiological study made this diagnosis in 0.3% of a control population of 682 adults.7 Bilateral renal agenesis presents with Potter’s sequence (oligohydramnios and lung hypoplasia) and is an order of magnitude less common than unilateral agenesis. Bilateral disease occurs in 0.1–0.3 per 1000 births.8–10 Interestingly, one group reported an excess of bilateral cases born in the spring, similar to neural tube defects.9

Urinary tract malformations associated with the congenital single kidney Malformations associated with the solitary renal tract are not uncommon. For example, Atiyeh and colleagues11 reviewed 16 pediatric patients with unilateral renal agenesis: of ten who had a micturating cystourethrogram, nine had anomalies of the single kidney with vesicoureteric reflux being the commonest abnormality. Other diagnoses were pelviureteric obstruction, megaureter, ureterovesical obstruction, ectopic ureter with partial obstruction, and hypoplasia.11 It is worth noting that some of these anomalies would not have been reliably detected on ultrasound scanning or even intravenous pyelography. Moreover, since vesicoureteric reflux can regress during childhood, it is difficult (or impossible) to be sure that a solitary kidney detected in later childhood or adulthood has indeed been connected to a normal urinary tract throughout life.

The developmental anatomy of renal agenesis The adult mammalian kidney is derived from two components:1,2 (i) the mesonephric duct epithelium and its outgrowth, the ureteric bud, which forms the branching collecting ducts and epithelia of the renal pelvis, ureter and trigone; (ii) the renal mesenchyme, a segment of intermediate mesoderm which differentiates to form the nephrons (glomeruli, proximal tubule and loop of Henle). In humans the metanephros appears at 5 weeks after conception and the first layer of glomeruli form by 9 weeks. Nephrogenesis continues in the cortex of the fetal kidney until the 34th week of gestation. In 1927 Boyden demonstrated that ablation of the mesonephric duct in an animal model caused failure of metanephric development, presumably because the ureteric bud did not grow to induce the renal mesenchyme.12 In 1932 the same investigator examined a 10 mm human embryo with renal agenesis: the ureteric bud had appar-

ently failed to branch from the mesonephric duct, hence the normal reciprocal inductive interaction with the renal mesenchyme could not occur.13 Others have postulated that an aberrant position of the ureteric bud can give rise to less severe malformations such as renal dysplasia.14 Of note, the prenatal obstruction of the ureter can generate renal dysplasia but not renal agenesis.15

The molecular etiologies of renal agenesis: clues from animal models Over 40 years ago Clifford Grobstein used organ culture experiments to demonstrate that the renal mesenchyme and ureteric bud fail to differentiate if cultured separately but if recombined they form a kidney in vitro.16 He concluded that reciprocal interactions, or induction events, are required for normal kidney development. We now know that cell proliferation, survival, and differentiation in normal nephrogenesis are controlled by the expression of genes which encode four classes of molecules: (i) transcription factors which regulate the expression of other genes; (ii) locally acting growth factors; (iii) survival factors; and (iv) cell–cell and cell–matrix adhesion molecules.1,2 Some of these molecules may act as the inductive signals postulated by Grobstein. Mutations of genes in mice result in renal agenesis because of the failure of induction of the renal mesenchyme, e.g., the WT1 gene17 or the failure of ureteric bud outgrowth from the mesonephric duct, e.g., RET gene18 and PAX2 gene.19 Animal studies also show that it may be necessary to ablate two related nephrogenesis genes to produce renal agenesis, e.g., the HOX20 and the retinoic acid receptor21 families. The experimental genetic ablation of the BCL2 survival factor22 provides a loose animal analogy to the involution of human aplastic/dysplastic kidneys.3,4 Here, the affected mice have fulminant apoptosis during nephrogenesis and are born with hypoplastic kidneys.22 Finally, teratogens such as drugs can cause various renal malformations in animals and in these cases the period of exposure to the insult determines the final effect on the developing excretory system. Glucose has been implicated as a teratogen that can cause renal agenesis in humans.23

Familial renal agenesis: a genetic basis for human disease Stephens24 found no familial cases in a review of over 200 individuals with congenital solitary kidney. However, in 1974 Cain et al. cited 12 previous reports and described a new kindred with two consecutive male infants: the first had bilateral renal agenesis while the second had

The single kidney

agenesis on one side and renal dysplasia in the contralateral system.25 Other reports of familial bilateral renal agenesis followed.9,10,26,27 McPherson and colleagues emphasised that dysplasia and absence of a kidney can be inherited within single kindreds in a probable autosomal dominant manner, and that both disorders could occur in the same patient.28 Roodhoft and coworkers7 found that siblings and parents of patients with unilateral or bilateral agenesis had an approximately tenfold higher incidence of congenital solitary kidney compared with controls. Murugasu et al.29 and Arfeen et al.30 described kindreds with congenital single kidney with a probable autosomal dominant inheritance.

Syndromes associated with renal agenesis Numerous human syndromes include unilateral or bilateral renal agenesis. These include those associated with chromosomal anomalies such as 4p- syndrome and partial trisomy 22.27 Other syndromes include: branchiooto-renal syndrome, cerebro-oculo-facio-skeletal syndrome, Di George syndrome (cardiac, face, thymus and parathyroid defects), Fraser syndrome (cryptophthalmos and syndactyly), Goldenhar syndrome (oculo-auriculovertebral dysplasia), Kallmann’s syndrome (olfactory bulb agenesis and infertility), Mayer–Rokitansky–Kuster– Hauser syndrome (aplasia of the female genital tract), renal agenesis with duplicated uterus and/or cervix and the thymic–renal–anal–lung syndrome.1,2 Several syndromes are of especial note because they are familial and the genetic bases are being unravelled. The branchio-oto-renal syndrome is an autosomal dominant condition with a genetic locus linked to chromosome 8. The affected individuals have renal, ear and neck malformations and some familes have a mutation of the EyesAbsent 1 (EYA1) gene which codes for a transcription factorlike molecule.31 Hepatocyte nuclear factor 1 (HNF1) is another transcription factor gene which is mutated in humans with congenital solitary functioning kidney; other family members can have renal dysplasia or hypoplasia, and some affected individuals develop diabetes requiring insulin therapy; the gene expressed in the developing kidney and pancreas.32,33 In the X-linked Kallmann’s syndrome the mutated gene is called KAL and it encodes a putative cell-signalling molecule that is expressed in the developing excretory and nervous systems.34 In Kallmann’s syndrome renal agenesis can be uni- or bilateral and there is apparently no left- or right-hand bias for the solitary kidney.35 Recently, Fraser syndrome was reported to be caused, at least in some families, by mutations of a gene called FRAS1, which codes for a protein which coats embryonic

kidney epithelia and mediates interactions with surrounding mesenchyme.36 No cases of human solitary kidney or bilateral agenesis have yet been associated with mutations of the mouse nephrogenesis genes such as PAX2, WT1 or BCL2 genes; however, families have been described with congenitally small kidneys and PAX2 mutations and there are other kindreds with renal agenesis and gut anomalies that resemble mice with RET mutations.

Hyperfiltration damage: theory and practice After nephrectomy, it is the normal response of the remaining kidney to hypertrophy (i.e., the size of its cells increases). The glomeruli increase in size and the glomerular filtration rate remarkably approach the level of excretory function for two organs.37 There is no convincing evidence that new nephron units can be generated in mammals in response to such an injury when it occurs after birth, although it has been considered possible, particularly in young animals, that some hyperplasia (increase of cell numbers) is involved in addition to hypertrophy. This response can be seen as compensatory and thus positive for the health of the patient or experimental animal: the total renal excretory capacity tends to be maintained. The stimulus for this hypertrophic response is unknown but it follows a rapid increase in renal blood flow in response to contralateral nephrectomy. Thus far no intrarenal or circulating “renotrophic” factor, which might rise in response to uninephrectomy, has been convincingly isolated. It is fascinating that “renal hypertrophy,” as assessed by the size of the single kidney, has also been documented to occur before birth.38,39 In these cases the increase in size cannot be linked to an increased functional need performed by the single kidney because the fetal circulation is effectively dialysed across the placental barrier. Recently, it was reported that fetal uninephrectomy in sheep, when performed two-thirds of the way through gestation, at a time when nephrogenesis is ongoing, leads to a compensatory 45% increase of nephron numbers in the solitary kidney.40 Hostetter and colleagues drew attention to the observation that hyperfiltration in remnant nephrons (i.e., increased single nephron glomerular filtration rate) was “a potentially adverse response to renal ablation.”41 Rats with subtotal renal ablation developed glomerulosclerosis and progressive renal damage. This group of investigators demonstrated that hyperfiltration was associated with an elevation of glomerular capillary hydrostatic pressure and, if this could be ameliorated by drugs like angiotensinconverting enzyme inhibitors which dilate the efferent glomerular arteriole, then the progressive injury could be

677

678

Adrian S. Woolf

minimized. O’Donnell and coworkers performed unilateral nephrectomy in both immature and adult rats; the younger animals developed focal glomerulosclerosis after a period of greater compensatory renal growth while the adult rats experienced a rise of glomerular capillary hydrostatic pressure but had no evidence of glomerulosclerosis.42 Similarly, other workers found that: (i) rats are more prone to develop proteinuria and reductions in glomerular filtration rate if uninephrectomy was performed in infancy vs. adulthood;43 (ii) nephron supply is a major determinant of long-term allograft outcome in rats.44 However, even healthy rats tend to develop some proteinuria with aging and therefore some investigators have questioned the relevance of these animal studies to clinical practice. On the other hand, Moritz et al.45 reported that fetal uninephrectomy in sheep led to postnatal hypertension associated with a decreased glomerular filtration rate. These animal studies suggest that reduced nephron number leads to progressive glomerular damage and have major implications for human diseases including the human solitary and remnant kidney.

Prognosis of the congenital solitary functioning kidney In a review of 586 renal biopsies, 29 individuals had a diagnosis of focal segmental glomerulosclerosis: of this group, 5 (17%) had unilateral renal agenesis.6 In 1984 Thorner and colleagues reported on two children with congenital solitary kidneys who had persistent proteinuria, renal impairment and glomerular sclerosis.46 One had a single pelvic kidney with no vesicoureteric reflux, while the other had a single hypertrophied kidney in the normal location and also did not have vesicoureteric reflux. Another report described four children with congenital single kidney who had proteinuria.47 These studies suggest that progressive glomerular damage can be associated with the congenital solitary kidney but do not provide an estimate of the risk of an individual with a congenital solitary kidney developing renal disease. Kiprov reviewed 9200 autopsies and found unilateral renal agenesis in seven subjects, two of whom developed end-stage renal disease due to focal segmental glomerulosclerosis.6 Of interest is a report by Arfeen and colleagues of a 22-year-old woman with a congenital solitary kidney who was found to have proteinuria during pregnancy. A renal biopsy showed that occasional glomeruli had segmental sclerosis and over the next 10 years her glomerular filtration rate fell from 88 ml/min to levels which required the initiation of dialysis. Two of her four children were also found to have a single kidney and proteinuria, and one affected offspring had a daughter with a single kid-

ney but no proteinuria.30 In addition, oligomeganephronia can be associated with progressive renal failure and proteinuria.48 In this disorder there is a severe reduction of nephron number at birth and these nephrons are considerably larger than normal. They are clearly subject to hyperfiltration and it is possible that the poor renal prognosis is determined by “hyperfiltration damage.” Argueso et al.49 reviewed the prognosis of 157 adult patients (mean age at diagnosis was 37 years) with unilateral “renal agenesis” and a normal contralateral kidney, i.e., there was compensatory hypertrophy, no vesicoureteric reflux, no parenchymal scarring and no hydronephrosis. In the whole study population, by the end of the study 27% had died, a survival rate similar to that of age- and sexmatched life tables; however, six had died of renal failure; of 37 tested in life, 19% had proteinuria (>150 mg/day); of 47 tested in life, hypertension developed in 47%; of 32 tested with creatinine clearance in life, 13% had decreased renal function. In a recent study,50 it was found that, compared with age-, height- and weight-matched controls, children with congenital solitary kidney had mildly elevated blood pressure, as assessed by 24 ambulatory monitoring, e.g., the average daytime systolic pressure was elevated by about 4 mmHg, and the diastolic pressure by about 2 mmHg. It could be argued that several of the above reports studied populations biased to have disease e.g., because they comprised referrals to a specialist clinic rather than to a primary care setting. There has not been a long-term follow-up (into late adulthood) of a large cohort of pediatric patients diagnosed, using a non-biased screening method, with congenital solitary kidney; certainly, if renal morbidity occurs, it tends to do so later in life, rather than in childhood. On the other hand, these reports do establish the fact that potentially important renal disease can occur in the congenital solitary kidney. It is possible to interpret these outcomes in several ways. First, it is possible that unilateral renal agenesis predisposes to the development of hyperfiltration and glomerular sclerosis in the normal congenital single kidney. Second, an alternative explanation would be that the single kidney was not, in fact, normal. As discussed above, the congenital solitary kidney is often attached to a structurally abnormal lower urinary tract and these abnormalities (e.g., vesicoureteric reflux) would not always be detected by ultrasound scanning or intravenous urography.11,51 Importantly, a renal biopsy is only performed when significant proteinuria is detected and hence the histology of the solitary kidney at birth is never known. Third, it is conceivable that both the renal agenesis and the glomerulosclerosis are the direct result of another, unknown, a primary event. For example, a yet-to-be defined mutation might perturb both kidney

The single kidney

development causing the agenesis and also predispose to glomerulosclerosis. In this context, heterozygous mutations of the WT1 gene cause glomerulosclerosis in the Denys–Drash syndrome in children whereas homozygous mutations cause renal agenesis in mice.

Nephrectomy for renal disease Renal prognosis after unilateral nephrectomy for renal disease In a retrospective study Zucchelli drew attention to the occurrence of focal glomerulosclerosis in seven male patients who, as young adults, had nephrectomies for unilateral renal disease such as obstruction or nephrolithiasis.52 None had systemic hypertension. Of these, four had a renal biopsy showing focal segmental glomerulosclerosis. Renal excretory function and the modest levels of proteinuria (441–1450 mg/day) were unchanged after a further 7 years of follow-up. Another study confirmed that renal excretory function, as assessed by creatinine clearance, was stable after mean follow-up of 23 years in 27 individuals who had undergone unilateral nephrectomy in childhood for hydronephrosis, Wilms’ tumor, renal dysplasia or kidney trauma.53 Another study reported a minimal decline in glomerular filtration rate with time during the follow-up of 36 patients with either unilateral renal agenesis or childhood nephrectomy for hydronephrosis followed for 7–47 years. Having said this, only one individual was found to have a glomerular filtration rate (slightly) lower than normal age-matched range.54 One problem with these studies is that it was sometimes unclear whether disease had been rigorously excluded from the remaining kidney at the time of the nephrectomy. In other words, did these patients really have unilateral and not bilateral disease? Argueso and colleagues reported on the prognosis of 30 children with solitary kidney after unilateral nephrectomy for obstruction, vesicoureteric reflux or Wilms’ tumor after a median follow-up of 25 years.55 These authors excluded from analysis those children with evidence of scars or vesicoureteric reflux into the opposite (unoperated) kidney. At follow-up 8 had proteinuria (160/95 mm Hg). Lent and Harth (1994) reported on 39 patients with unilateral nephrectomy after a follow-up of 2–23 years.56 The initial diagnoses were evenly distributed between nephrolithiasis, tumors or pyelonephritis. The investigators excluded patients with disease of the remaining kidney or who initially had proteinuria >150 mg/day.

Qualitative glomerular proteinuria was found in 32 using microelectrophoresis but only 4 had a rise in absolute values to over 150 mg/day. Another long-term study was recently published by Ohishi and coworkers regarding 21 patients who had undergone nephrectomy in adulthood for unilateral disease. After an average follow-up for 27 years the mean creatinine clearance sustained by the solitary kidney was 88 ml/ min/1.73 m2 , or 93% of that expected for a normal agematched individual! Protein excretion was not significantly raised (214 vs. 119 mg/day) and age at nephrectomy, length of time with a single kidney or sex had little influence on current renal status. Six developed de novo hypertension and tended to have a positive family history of raised blood pressure more often than those with normotension.57 Thus it appears to be highly unusual to develop progressive kidney damage of any significance after nephrectomy for unilateral disease in adulthood. On the other hand, there may be a small risk if a similar operation is performed in children. However, even in those cases the question of occult renal disease at the time of nephrectomy must remain open: vesicoureteric reflux might have regressed by the time a micturating cystogram was performed, we know that intravenous urography is less sensitive than the modern renal isotope scans (e.g., DMSA) at detecting renal scars, and tissue for renal histology of the contralateral kidney is clearly not reported or available at the time of nephrectomy for alleged unilateral renal disease.

Renal prognosis after subtotal nephrectomy: the human remnant kidney Based on experiments in rats, it is possible that humans with subtotal nephrectomy (e.g., single nephrectomy plus partial contralateral nephrectomy) would be susceptible to progressive renal damage.40 The following reports suggest that subtotal nephrectomy generally has a good renal prognosis in humans. Rutsky and coworkers reported on a 10-year follow-up of an individual with three-quarters renal ablation after vascular trauma. Creatinine clearance increased from 16 to 53 ml/min and remained stable but daily urine excretion rose from 90 to a modest level of 400 mg, the upper end of the normal range.58 Lhotta and colleagues reported on six patients who underwent enucleation for carcinoma in solitary kidneys: four had previously had a nephrectomy and two had an atrophic non-functioning kidney. None showed progressive deterioration of function or proteinuria over 10–23 months but two subjects had a rise in systemic blood pressure.59 Foster collected the records of patients with

679

680

Adrian S. Woolf

surgical resection resulting in a remnant kidney.60 All had resection for renal carcinoma and one for tuberculosis. Six patients were observed for at least 10 years, and 6 for 5–7 years: all had stable plasma creatinine levels. Moreover, the two longest surviving patients had pregnancies with no change in renal function. Only one patient had evidence of a progressive rise in plasma creatinine. In a retrospective analysis of 109 cases after nephrectomy for carcinoma only one individual had severe renal impairment after semitotal nephrectomy: in others the progressive impairment could be explained by concomitant renal disease.61 While these studies suggest a good renal prognosis in the human remnant kidney, another report suggests a somewhat different story. Novick evaluated 14 patients with a solitary kidney 5–17 years after partial nephrectomy for carcinoma.62 Twelve had stable renal function but two developed renal failure and nine had proteinuria, which in four was >900 mg/day. None had systemic hypertension. The extent of the proteinuria was proportional to length of follow-up and inversely correlated to amount of remaining renal tissue. The four patients with the most marked proteinuria had renal biopsies and focal segmental glomerulosclerosis was demonstrated in three and global sclerosis in the other patient. Thus, compared to a simple nephrectomy for disease or transplant donation (see below), there appears to be some evidence which indicates that there is a small but significant risk of developing proteinuria and progressive renal dysfunction after subtotal nephrectomy. At present, we do not know how to predict which patients will deteriorate and long-term follow-up regarding protein excretion and plasma creatinine is warranted in these patients.

Transplant kidney donation A relatively early study in this population by Hakim and colleagues assessed 52 donors who had been nephrectomized 10 years previously.63 They found no significant decline of renal function as assessed by levels of serum creatinine or creatinine clearance. There was, however, a higher incidence in mild proteinuria of less than 1 g/day and additionally there was an increase of systemic hypertension in male donors. It was suggested that these results might indicate that a modest degree of hyperfiltration damage had occurred. Talseth and coworkers examined 68 kidney donors who had their nephrectomies 9–15 years previously.64 Although 15% were hypertensive, it must be realized that the inci-

dence of high blood pressure in the general population is of this order of magnitude. Twenty-six individuals had a urinary albumin excretion over 10 mg/min or a total daily excretion over 185 mg. However, only four had proteinuria over 400 mg/day and of these three individuals had intercurrent disease. It was found that the compensatory rise in glomerular filtration rate was inversely proportional to age at nephrectomy. In contrast, other studies suggested an increased incidence of proteinuria and hypertension in donors 9–18 years after unilateral nephrectomy.65 In 1989 Fotino published a review of 25 studies of longterm follow-up of kidney transplant donors.66 Based on these studies it was concluded that the raised glomerular filtration rate of the remaining kidney was stable for decades. While some donors did develop mild nonprogressive proteinuria, there appeared no definitive evidence that they suffered an undue rise of systemic blood pressure. Since then, several other reports have been published. For example, Goldfarb et al67 reported outcomes 25 years after donor nephrectomy; they noted that, although blood pressure tended to rise, levels remained in the normal range and were similar to that expected for an age-matched population. There was a modest increase in protein excretion rates in male vs. female donors. Eberhard et al.68 monitored living related kidney donors an average of 11 years after donation; 7 of 29 developed microalbuminuria and, even though all donors were considered normotensive before donation, just over a quarter proved to be hypertensive (>130/80, as assessed by 24-hour blood pressure monitoring). Therefore, there appears to be little evidence that donation of a kidney has an adverse effect on excretory function of the remaining kidney over a period of up to three decades, at least when compared to the modest loss of excretory function which occurs in the normal aging kidney. On the other hand, there is mixed evidence that hypertension may be greater than expected, compared to appropriate controls. It should also be remembered that the populations described above are predominantly adult and it is conceivable that the reaction of the remaining kidney from an individual who donated in childhood might be different; however, it is currently rare to donate a kidney in early childhood.

Acknowledgment ASW acknowledges grant support from the Kidney Research Aid Fund.

The single kidney

REFERENCES 1. Risdon, R. A. & Woolf, A. S. Developmental defects and cystic diseases of the kidney. In Jennette, J. C., Olson, J. L., Schwartz, M. M., & Silva, F. G., eds. Heptinstall’s Pathology of the Kidney. 5th edn. Philadelphia-New York, USA: Lippincott-Raven, 1998:1149–1206. 2. Woolf, A. S., Welham, S. J. M., Hermann, M. M., & Winyard, P. J. D. Maldevelopment of the human kidney and lower urinary tract: an overview. In Vize, P. D., Woolf, A. S., & Bard, J. B. L., eds. The Kidney: From Normal Development to Congenital Disease. The Netherlands: Elsevier Science/Academic Press, 2003:377–393. 3. Dungan, J. S., Fernandez, M. T., Abbitt, P. L., Thiagarajah, S., Howards, S. S., & Hogge, W. A. Multicystic dysplastic kidney: natural history of prenatally detected cases Prenat. Diagn. 1990; 10:175–182. 4. Hiraoka, M., Tsukahara, H., Ohshima, Y., Kasuga, K., Ishihara, Y., & Mayumi, M. Renal aplasia is the predominant cause of congenital solitary kidney. Kidney Int. 2002; 61:1840–1844. 5. Winyard, P. J. D., Nauta, J., Lirenman, D. S. et al. Deregulation of cell survival in cystic and dysplastic renal development. Kidney Int. 1996; 49:135–146. 6. Kiprov, D. D., Calvin, R. B., & McLuskey, R. T. Focal and segmental glomeruloscerosis and proteinuria associated with unilateral renal agenesis. Lab. Invest. 1982; 46:275–281. 7. Roodhooft, A. M., Birnholz, J. C., Holmes, L. B. Familial nature of congenital absence and severe dysgenesis of both kidneys. N. Engl. J. Med. 1984; 310:1341–1345. 8. Potter, E. L. Bilateral renal agenesis. J. Pediatr. 1946; 29:68–76. 9. Carter, C. O., Evans, K., & Pescia, G. A familial study of renal agenesis. J. Med. Genet. 1979; 16:176–188. 10. Bankier, A., De Campo, M., Newell, R., Rogers, J. G., & Danks, D. M. A pedigree study of perinatally lethal renal disease. J. Med. Genet. 1985; 22:104–111. 11. Atiyeh, B., Husmann, D., & Baum, M. Contralateral renal abnormalities in patients with renal agenesis and noncystic renal dysplasia. Pediatrics 1993; 91:812–815. 12. Boyden, E. A. Experimental obstruction of the mesonephric ducts. Proc. Soc. Exp. Biol. Med. 1927; 24:572–576. 13. Boyden, E. A. Congenital absence of the kidney. An interpretation based on a 10 mm human embryo exhibiting unilateral renal agenesis. Anat. Rec. 1932; 52:325–349. 14. Mackie, C. G. & Stephens, F. D. Duplex kidneys: a correlation of renal dysplasia with position of the ureteric orifice. J. Urol. 1975; 114:274–280. 15. Woolf, A. S. Congenital obstructive nephropathy gets complicated. Kidney Int. 2003; 63:761–763. 16. Grobstein, C. Morphogenetic interaction between embryonic mouse tissues separated by a membrane filter. Nature 1953; 172:869–870. 17. Kreidberg, J. A., Sariola, H., Loring, J. M. et al. WT-1 is required for early kidney development. Cell 1993; 74:679–691. 18. Schuchardt, A., D’Agati, V., Larsson-Blomberg, L., Costantini, F., & Pachnis, V. Defects in kidney and enteric nervous system

19.

20.

21.

22.

23.

24. 25.

26. 27.

28. 29.

30.

31. 32.

33.

34.

35.

36.

of mice lacking the tyrosine kinase receptor Ret. Nature 1994; 367:380–383. Torres, M, Gomex-Pardo, E., Dressler, G. R., & Gruss, P. Pax-2 controls multiple steps of urogenital development. Development 1995; 121:4057–4065. Davis, A. P., Witte, D. P., Hsieh-Li, H. M., Potter, S. S., & Capecchi, M. R. Absence of radius and ulna in mice lacking hoxa-11 and hoxd-11. Nature 1995; 375:791–795. Mendelsohn, C., Lohnes, D., Decimo, D. et al. Function of the retinoic acid receptors (RAR) during development. Development 1994; 120:2749–2771. Veis, D. J., Sorenson, C. M., Shutter, J. R., & Korsmeyer, S. J. Bcl2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys and hypopigmented hair. Cell 1993; 75:229– 240. Novak, R. W. & Robinson, H. B. Coincident DiGeorge anomaly and renal agenesis and its relation to maternal diabetes. Am. J. Med. Genet. 1994; 50:311–312. Stevens, A. R. Pelvic single kidneys. J. Urol. 1937; 37:610–618. Cain, D. R., Griggs, D., Lackey, D. A., & Kagan, B. M. Familial renal agenesis and total dysplasia. Am. J. Dis. Child. 1974; 128:377– 380. Pashayan, H. M., Dowd, T., & Nigro, A. V. Bilateral absence of the kidneys and ureters. J. Med. Genet. 1977; 14:205–209. Schinzel, A., Homberger, C., & Sigrist, T. Bilateral renal agenesis in 2 male sibs born to consanguineous parents. J. Med. Genet. 1978; 15:314–316. McPherson, E., Carey, J., Kramer, A. et al. Dominantly inherited renal adysplasia. Am. J. Med. Genet. 1987; 26:863–872. Murugasu, B., Cole, B. R., Hawkins, E. P., Blanton, S. H., Conley, S. B., & Portman, R. J. Familial renal adysplasia. Am. J. Kidney Dis. 1991; 18:490–494. Arfeen, S., Rosborough, D., Luger, A. M., & Nolph, K. D. Familial unilateral renal agenesis and focal and segmental glomerulosclerosis. Am. J. Kidney Dis. 1993; 21:663–668. Rodriguez-Soriano, J. Branchio-oto-renal syndrome. J. Nephrol. 2003; 16:603–605. Bingham, C., Ellard, S. Cole, T. R. P. et al. Solitary functioning kidney and diverse genital tract malformations associated with hepatocyte nuclear factor-1 mutations. Kidney Int. 2002; 61:1243–1251. Kolatsi-Joannou, M., Bingham, C., Ellard, S. et al. Hepatocyte nuclear factor 1: a new kindred with renal cysts and diabetes, and gene expression in normal human development. J. Am. Soc. Nephrol. 2001; 12:2175–2180. Duke, V. M., Winyard, P. J. D., Thorogood, P., Soothill, P., Bouloux, P. M. G., & Woolf, A. S. KAL, a gene mutated in Kallmann’s syndrome, is expressed in the first timester of human development. Mol. Cell. Endocrin. 1995; 110:73–79. Kirk, J. M. W, Grant, D. B., Besser, G. M. et al.. Unilateral renal aplasia in X-linked Kallmann’s Syndrome. Clin. Genet. 1994; 46:260–262. McGregor, L., Makela, V., Darling, S., M. et al. Fraser syndrome and mouse blebbed phenotype caused by mutations in

681

682

Adrian S. Woolf

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

FRAS1/Fras1 encoding a putative extracellular matrix protein. Nat. Genet. 2003; 34:203–208. Fine, L. G., Kurtz, I., Woolf, A. S. et al. Pathophysiology and nephron adaptation in chronic renal failure. In Schrier, R. W. & Gottschalk, C. W., eds. Diseases of the Kidney, Boston: Little and Brown, 1992:2703–2742. Mandell, J., Peters, C. A., Estroff, J. A., Allred, E. N., & Benacerraf, B. R. Human fetal compensatory renal growth. J. Urol. 1993: 150:790–792. Glazebrook, K. N., McGrath, F. P., & Steele, B. T. Prenatal compensatory renal growth: documentation with US. Radiology 1993; 189:733–735. Douglas-Denton R., Moritz, K. M., Bertram, J. F., & Wintour, E. M. Compensatory renal growth after untilateral nephrectomy in the ovine fetus. J. Am. Soc. Nephrol. 2002; 13:406–410. Hostetter, T. H., Olson, J. L., Rennke, H. G., & Brenner, B. M. Hyperfiltration in remnant nephrons: a potentially adverse response to renal ablation. Am. J. Phys. 1981; 241:F85–F93. O’Donnell, M. P., Kasiske, B., Raij, L., & Keane, W. F. Age is a determinant of the glomerular morphologic and functional responses to chronic nephron loss. J. Lab. Clin. Med. 1985; 106:308–313. Celsi, G., Bohman, S.-O., & Aperia, A. Development of focal glomeruloscerosis after unilateral nephrectomy in infant rats. Pediatr. Nephrol. 1987; 1:290–296. Mackenzie, H. S., Tullius, S. G., Heemann, U. W. et al. Nephron supply is a major determinant of long-term allograft outcome in rats. J. Clin. Invest. 1994; 94:2148–2152. Moritz, K. M., Wintour, E. M., & Dodic, M. Fetal uninephrectomy leads to postnatal hyspertension and compromised renal function. Hypertension 2002; 39:1071–1076. Thorner, P. S., Arbus, G. S., Calermajer, D. S., & Baumal, R. Focal segmental glomerulosclerosis and progressive renal failure associated with a unilateral kidney. Pediatrics 1984; 73:806– 810. Gutierrez-Millet, V., Nieto, J., Praga, M. et al. Focal glomerulosclerosis and proteinuria in patients with solitary kidneys. Ann. Intern. Med. 1986; 146:705–709. Nomura, S. & Osawa, G. Focal glomerular sclerotic lesions in a patient with unilateral oligomeganephronia and agenesis of the contralateral kidney: a case report. Clin. Nephrol. 1990; 33:7–11. Argueso, L. R., Ritchey, M. L., Boyle, E. T. Jr, Milliner, D. S., Bergstralh, E. J., & Kramer, S. A. Prognosis of patients with unilateral renal agenesis. Pediatr. Nephrol. 1992; 6:412–416. Mei-Zahav, M., Koizets, Z., Cohen, I. et al. Ambulatory blood pressure monitoring in children with a solitary kidney – a comparison between unilateral renal agenesis and nephrectomy. Blood Pressure Monitoring 2001; 6:262–267. Cascio, S., Paran, S., & Puri, P. Associated urological anomalies in children with unilateral renal agenesis. J. Urol. 1999; 162:1081–1083.

52. Zucchelli, P., Cagnoli, L., Casanova, S., Donini, V., & Pasquali, S. Focal glomerulosclerosis in patients with unilateral nephrectomy. Kidney Int. 1983; 24:649–655. 53. Robitaille, P., Mongeau, J.-G., Lortie, L., & Sinnasammy, P. Long term follow-up of patients who underwent unilateral nephrectomy in childhood. Lancet 1985; 1:1297–1299. 54. Wikstad, I., Celsi, G., Larsson, L., Herin, P., & Aperia, A. Kidney function in adults born with unilateral agenesis or nephrectomised in childhood. Pediatr. Nephrol. 1988; 2:177–182. 55. Argueso, L. R., Ritchey, M. L., Boyle, E. T., Milliner, D. S., Bergstralh, E. J., & Kramer, S. A. Prognosis of children with solitary kidney after unilateral nephrectomy. J. Urol. 1992; 148:747– 751. 56. Lent, V. & Harth, J. Nephropathy in remnant kidneys: pathological proteinuria after unilateral nephrectomy. J. Urol. 1994; 152:312–316. 57. Ohishi, A., Suzuki, H., Nakamoto, H. et al. Status of patients who underwent uninephrectomy in adulthood more than 20 years ago. Am. J. Kidney Dis. 1995; 26:889–897. 58. Rutsky, E. A., Dubovsky, E. V., & Kirk, K. A. Long term follow-up of a human subject with a remnant kidney. Am. J. Kidney Dis. 1991; 18:509–513. 59. Lhotta, K., Eberle, H., Konig, P., & Dittrich, P. Renal function after tumour enucleation in a solitary kidney. Am. J. Kidney Dis. 1991; 17:266–270. 60. Foster, M. H., Sant, G. R., Donohoe, J. F., Harrington, J. T. Prolonged survival with a remnant kidney. Am. J. Kidney Dis. 1991; 17:261–265. 61. Grossman, H. B., Sommerfield, D., Konnak, J. W., & Bromberg, J. Long-term assessment of renal function following nephrectomy for stage I renal carcinoma. Br. J. Urol. 194; 74:279–282. 62. Novick, A. C., Gephardt, G., Guz, B., Steinmuller, D., & Tubbs, R. R. Long-term follow up after partial removal of a solitary kidney. N. Engl. J. Med. 1991; 325:1058–1062. 63. Hakim, R. M., Goldszer, R. C., & Brenner, B. M. Hypertension and proteinuria: long-term sequelae of uninephrectomy in humans. Kidney Int. 1984; 25:930–936. 64. Talseth, T., Fauchald, P., Skrede, S. et al. Long-term blood pressure and renal function in kidney donors. Kidney Int. 1986; 29:1072–1076. 65. Watnick, T. J., Jenkins, R. R., Rackoff, P., Baumgarten, A., & Bia, M. J. Microalbumin and hypertension in long-term renal donors. Transplant 1988; 45:59–65. 66. Fotino, S. The solitary kidney: a model of chronic hyperfiltration in humans. Am. J. Kidney Dis. 1989; 13:88–98. 67. Goldfarb, D. A., Matin, S. F., Braun, W. E. et al. Renal outcome 25 years after donor nephrectomy. J. Urol. 2001; 166:2043–2047. 68. Eberhard, O. K., Kliem, V., Offner, G. et al. Assessment of long-term risks for living related kidney donors by 24-h blood pressure monitoring and testing for microalbuminuria. Clin. Transpl. 1997; 11:415–419.

54 Multicystic kidney Gianantonio M. Manzoni1 and Anthony A. Caldamone2 1

Although multicystic kidney (MCK) is a common renal anomaly, the management of this entity remains controversial. Much of the controversy stems from a lack of longterm data on its natural history. Only recently have efforts been made to follow the MCK and accumulate long-term data on its natural history. The results of these efforts may have a significant impact on management. Schwartz is credited with the first description of an MCK in a 7-month-old child.1 In addition to describing the kidney as having been replaced by multiple cysts in a “bunch of grapes” arrangement, he also reported the ureter to be atretic. Spence’s classic article appeared in 1955, in which he further described the MCK, distinguishing it from other forms of renal cystic disease.2 This chapter reviews the presentation of MCK and looks at the available long-term data in an attempt to arrive at a management protocol.

Presentation and diagnosis The classic presentation of MCK was either a palpable mass in a newborn or infant or an incidental finding at autopsy. Pathak and Williams, in a series of 22 cases, reported that eight cases presented with an abdominal mass, seven had vomiting, and three had failure to thrive.3 With the advent of prenatal ultrasonography, many more asymptomatic lesions are being identified. A review of data accumulated by the National Multicystic Kidney Registry in the USA indicated that 72% of registered cases were discovered on prenatal ultrasound (J. Wacksman, personal communication). The diagnosis of MCK can be strongly suggested, if not made definitely, by prenatal ultrasound.4,5 The classic pic-

Division of Urology, Ospedale di Circol e Fondazione Macchi, Varese, Italy 2 Division of Pediatric Urology, Hasbro Children’s Hospital

ture of multiple fluid-filled cysts of varying size, randomly arranged with no communication, separated by echodense stroma, is often pathognomonic for MCK (Fig. 54.1). A hydronephrotic variant of MCK, however, may present the ultrasonographer with some concern (Fig. 54.2). In 1975, Felson and Cussen6 described a pattern in which cysts were located peripherally around a hydronephrotic renal pelvis, with evidence of communication between cysts and pelvis. They also noted some function between the cysts radiographically. The existence of this type of variant of the MCK lends credence to the embryological theory that MCK and obstructed hydronephrotic kidney are on a spectrum of disease. The duration, extent, and timing of the obstruction dictates the outcome (i.e., MCK or hydronephrosis). Unilateral multicystic kidney dysplasia (UMCK) is the second most common urinary tract abnormality diagnosed antenatally. While the isolated unilateral form UMCK has a good prognosis, a poor outcome must be expected only when it is associated with other complex abnormalities.7 Bilateral MCK has been reported in 7%–23% of all cases and usually results in an absence of renal function, incompatible with extrauterine life.8 In a recent Canadian review of 54 cases Aubertin4 reported that chromosomal abnormalities may be present in a low proportion (3%) and an amniocentesis is only recommended if associated anomalies are identified. Careful examination of the contralateral kidney is mandatory since up to 33% of genitourinary tract abnormalities may be present and interestingly, a positive family history of structural renal anomalies was found in 20% of the cases. Among non-renal anomalies (16%) congenital heart defects seem to be the most frequent (7%), therefore a careful assessment

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


683

684

G. M. Manzoni and A. A. Caldamone

Fig. 54.2. (a) Ultrasound scan of a hydronephrotic variant of a multicystic kidney. (b) DMSA nuclear scan. Fig. 54.1. (a) Ultrasound scan of a newborn male with a history of prenatal hydronephrosis. This shows the typical MCK appearance with cysts of various sizes, without communication and separated by dysplastic parenchyma. (b) DMSA nuclear scan, showing no uptake of tracer in the area of the right kidney.

of the fetal heart is always required. Complete prenatal involution of the MCK was documented in two cases. Ultrasound should be followed by a functional assessment in the renal unit. Virtually any radionuclide agent will determine whether there is functional renal tissue, but the most sensitive are dimercaptosuccinic acid (DMSA) or MAG3. Most often the renal scan will show a deficient area in the appropriate renal fossa on both early and delayed images. If one is dealing with the hydronephrotic variant, however, there may be a small percentage of function

seen on the delayed images. Wacksman9 has pointed out that accuracy in the preoperative diagnosis of MCK is close to 100%, with 97 of the initial 98 patients registered in the United States Multicystic Kidney Registry, who underwent nephrectomy, correctly diagnosed with ultrasound and renal scan. If further assessment is necessary, especially in cases where the ultrasound demonstrates a hydronephrotic pattern or there is function on renal nuclear scan, one can either place a percutaneous nephrostomy tube and later assess for function or surgically explore the kidney with a biopsy. Minevich et al.10 recently reported 4 children who had a presumptive diagnosis of MCK who on follow-up had enlargement of the renal unit which led to exploration. The enlarged renal units were

Multicystic kidney

The cause of MCK remains unknown but it may be related to abnormalities of one or more genes involved in the process of nephrogenesis.17 It has been suggested that at least some cases may be inherited in an autosomal dominant fashion18 . Furthermore non-syndromic MCK can occur as part of the spectrum of hereditary dysplasia with autosomal dominant inheritance.19

Natural history US data from the National Multicystic Kidney Registry Fig. 54.3. Ultrasound of a newborn female presenting with prenatal hydronephrosis, showing a duplicated right system with a multicystic upper pole system and hydronephrotic lower pole system.

due to Wilms’ tumor, mesoblastic nephroma, segmental MCK of a duplex system, and simple renal cysts. MCK may thus be segmental in nature and may be found in segments of duplicated systems (Fig. 54.3), as well as in systems associated with ureteroceles, or even with a contralateral MCK.11 It has been discovered recently that there is a high incidence of vesicoureteral reflux in association with MCK. Flack and Bellinger12 reported a 30% incidence of contralateral reflux in association with MCK. Similarly, Elder et al.13 found a 15% incidence of reflux in the contralateral renal unit, with a mean grade of 3/5 (Fig. 54.4). Considering that this represents reflux into a solitary functioning renal unit, the potential consequences of this reflux may, indeed, be significant. A voiding cystourethrogram (VCUG), therefore, has been recommended in the evaluation of the child with a MCK. The status of the contralateral kidney deserves further comment. Early reports indicated a high incidence of contralateral renal abnormalities, most commonly hydronephrosis. In 1992, Atyeh et al.14 found that 51% of patients with MCK had an associated abnormality, including reflux, pelviureteric junction obstruction, or a hypoplastic kidney. Data from the US MCK Registry indicated that the contralateral kidney was normal in 73% of cases. More recent reports indicate that multicystic kidney dysplasia, renal agenesis, and ureteropelvic junction obstruction are pathogenetically related,15 and that combinations of these urological malformations occur in the same individual or in different relatives in the same family.16

There are currently 660 patients registered in the National Multicystic Kidney Registry in the USA. Of the registered cases, 473 (72%) presented by antenatal ultrasound, 97(15%) had a palpable flank mass at presentation, and 25 (4%) presented after evaluation for a urinary tract infection (J. Wacksman, pers. commun.). Males were slightly more commonly afflicted than females (56% vs. 44%). Contralateral kidney abnormalities were present in 27% of cases, with vesicoureteral reflux seen in approximately 38% and hydronephrosis in the remainder. The majority of those with hydronephrosis involved the renal pelvis only, with one-third involving the ureter as well. The exact incidence of hypertension in association with MCK has not been determined. Of the 660 patients registered, five had hypertension. Of the registered cases, 234 have undergone nephrectomy, leaving 426 who are being followed. Of those who have their kidneys in place, 5 have hypertension, 15 have had at least a single urinary tract infection, 2 have had hematuria, and 1 had pain. There have been no cases in which a tumor has been identified as having developed. Most of the cases who have been followed are 1 to 5 years from presentation. With regard to the size of the MCK over time, most evaluations with ultrasound demonstrate a reduction (45%) in the size or no change in size (35%). Of the images, 5% have demonstrated enlargement, while in 14% the renal unit has not been detectable.

European data Clinical data on the natural history of MCK are available from published series in the British literature and from unpublished data of a cohort of 64 children referred to the Hospital for Sick Children in London between 1988 and 1992. The series will be referred to here as the London, Leeds,20 Liverpool,21 and Manchester22,23 series (Table 54.1).

685

686

G. M. Manzoni and A. A. Caldamone

(a)

(b)

(c)

In the London series, all cases had a prenatal diagnosis which was made at 18 to 20 weeks’ gestation in 33%, at 20 to 30 weeks in 44%, and at 30 to 40 weeks in 23% (overall, 77% before 30 weeks’ gestation). However, 36% of ultrasounds were reported as normal on the 18 to 20 weeks scan, which could indicate that this abnormality may become easier to identify ultrasonically later in the second trimester. This is supported by the fact that 11% were actually noted to be increasing in size in utero on sequential scans, while a further 11% disappeared, presumably owing to prenatal involution. It is relevant to note that in 33% the antenatal diagnosis was uncertain, oscillating from hydronephrosis to MCK and vice versa. All patients in the Liverpool and

Fig. 54.4. Newborn female presenting with prenatal hydronephrosis showing: (a) Ultrasound scan, hydronephrotictype MCK on the left, with no communication between cystic structures (and no functioning on DMSA scan). (b) Ultrasound scan of normal right kidney. (c) MCU, showing bilateral vesicoureteral reflx.

Manchester series had a prenatal diagnosis with similar findings. It appears that there is no difference in the prevalence for one sex or for site, and bilateral MCK was diagnosed in only three cases, always prenatally. It is a constant finding in all series that only one-third of the multicystic kidneys were palpable at birth. It is difficult to estimate how many of the prenatally diagnosed palpable kidneys would have been found on routine postnatal examination without the in utero diagnosis. It is realistic to expect that approximately three-quarters of unilateral MCKs remained undiagnosed in the period before prenatal ultrasound diagnosis.

Multicystic kidney

Table 54.1. European data on MCK evolution

Series

Leeds20 Liverpool21 Manchester22,23 London (unpublished data) Totals a

Total kidneys

22a 43a 62 64 191

CON

NX

13 37 13 51

7 5 49 13

114

74

Follow-up

1982–87 1979–89 1982–92 1988–96

Complete involution Pre

Post

0 0 0 6

2 10 4 16

↓

MCK size unchanged

↑

Hypertension

6 9 ? 17

3 14 8 20

− 0 ? 5

0 0 3 0

Leeds:2 bilateral; Liverpool:1 bilateral. CON, conservative management; NX, nephrectomy; pre, prenatal; post, postnatal

Postnatal outcome is demonstrated in Table 54.1. Overall there is data available on 114 patients followed with conservative treatment over a variable period from 5 to 10 years (mean 7). Only in the Liverpool series were 5 patients (13%) eventually lost to follow-up. Prenatal involution was clearly demonstrated in 6 cases in the London series, with 16 further cases involuting postnatally (11 in the first year and five subsequently). Complete involution may be expected in nearly 30%, while a substantial reduction or unchanged size is evident in different proportions (Fig. 54.5). According to the London data, the size of the MCK seems to have an important influence on the possibility of involution. It would seem reasonable to expect that under 5 cm in length all units either disappeared, became smaller, or remain unchanged. If the kidney is 6 cm or greater by the end of the first year, it would be unlikely to involute. Interestingly, over 80% of the MCKs did involute in the London study during the first year. Another important finding was that nearly 8% of the kidneys were enlarging, as observed by other authors.24,25 Associated urological abnormalities involved both upper and lower tracts (Table 54.2). As confirmed in other previous studies, there was a significant incidence of vesicoureteral reflux, but this was in most cases low grade and benign with a tendency to spontaneous resolution. Among the contralateral abnormalities, pelviureteric junction obstruction was the most common finding, followed by megaureter. Surgical treatment can be required, and in a few cases the renal function of the solitary kidney was already severely compromised (four cases). A concerning possibility is represented by the association of renal dysplasia contralateral to the MCK, with inevitable progression to renal failure (one case). Lower tract anomalies are mostly represented by bladder or paraureteric diverticula and ureteroceles, which most frequently are ipsilateral to the MCK. Only one patient in the Leeds series with multiple non-urological coexistent anomalies died of a cause unrelated to the MCK at 2 weeks after birth.

None of the children followed conservatively developed malignant change. Three patients in the Manchester series were found to be hypertensive on follow-up, and all became normotensive after nephrectomy.23 The only urological symptoms (urinary tract infections, renal failure) experienced were always related to a pathological condition elsewhere in the urinary tract rather than to the MCK.

Other reports In a report by Oliveira et al. of 19 patients followed after presenting prenatally with MCDK and confirmed by postnatal evaluation, 13 showed partial involution, 4 complete involution, and 2 an increase in size.26 A similar study by Kuwertz-Broeking et al. reported on the follow-up of 97 patients with MCDK, 82 of which presented prenatally.27 Of the 75 followed non-operatively, total involution occurred in 25%, reduction in size in 60%, and stable in size in 15%. No cases of malignancy were found. In a recent series of 29 children with prenatally detected MCDK, 52% demonstrated total involution at a median age of 6.5 years.28 Eckoldt et al. reported on the management of 93 infants prenatally diagnosed with MCDK. Of 37 patients followed non-operatively for an extended period of time, 28 underwent size reduction and 16 complete involution at a mean of 16.2 months.29 In a long-term study by Sukthankar and Watson, 52% of patients demonstrated a reduction in size or complete disappearance by 5 years.30

The role of nephrectomy Indications There is no uniformity of opinion on the role of nephrectomy in the management of MCK. There are a few situations where there is unanimous agreement to remove the kidney.

687

688

G. M. Manzoni and A. A. Caldamone

The presence of a large mass, responsible for compression Intestinal/respiratory compression Although percutaneous decompression has been proposed as a temporary relief,31 nephrectomy is still the preferred option.

(a)

To confirm the diagnosis The sonographic criteria for diagnosing a MCK are now well defined,32 and the addition of an isotope scan usually confirms the distinction between a non-functioning MCK and a poorly functioning hydronephrotic kidney. On rare occasions it may be critical to establish an unequivocal and early diagnosis of MCK, and if any doubt still remains surgical exploration is indicated. Increase in volume In the sonographic follow-up, an increase in volume of a small MCK may indicate a pathological evolution (malignant degeneration, abscess) or a misdiagnosis (cystic nephroma, multilocular cyst).10 Surgical exploration is mandatory to establish a correct diagnosis.

(b)

To relieve symptoms It seems difficult to argue against the removal of a symptomatic MCK. In practice, pain, hematuria and infection as complications of MCK have been reported only in adults, later in life.33 Should these symptoms be attributed to MCK, excluding a pathological condition elsewhere in the urinary tract, a nephrectomy is justified. However, such cases perhaps represent a minority in relation to the true incidence of the condition. On the other hand, the presence of hypertension is a clear and absolute indication for surgical treatment. Concomitant surgical treatment In the event of concomitant surgery for a urological (or non-urological) problem, removal of the MCK should be considered as part of the program, avoiding a potential subsequent unnecessary procedure.

(c) Fig. 54.5. MCK postnatal involution. (a) Prenatal longitudinal ultrasound scan of a male fetus with unilateral MCK. (b) Ultrasound scan of left MCK at 18 months. (c) Ultrasound scan of left MCK at 36 months.

Poor parental compliance The decision to maintain an observational approach is based on the assumption that long-term follow-up is required. Although rare or unlikely, the possibility of the development of hypertension or malignant change does exist and it would be incorrect and unethical to withhold this information from parents in the discussion of therapeutic options. In the event of questionable long-term follow-up, nephrectomy should be strongly considered.

Multicystic kidney

Table 54.2. Associated urological anomalies in the European series Upper tract

Lower tracts

Other information

Series

Ipsilateral

Contralateral

Leeds20

VURa 1

PUJ 2(S) VURa 2

1 death unrelated causes (duodenal atresia, heart defect)

Liverpool21

MCK in ectopic kidney 1

PUJ 1 Megaureter 1

1 aortic coarctation, 1 hip dislocation

Manchester22,23

Ureterocele 1 MCK upper pole duplex system 3 VURa 8

Ureterocele 1 Megaureter 1 4PUJ 7 (3S)

Paraureteric diverticula 1 Bladder diverticula 2

London (unpublished)

Ureterocele 5VURa 2

VURa 10 Renal cyst 1 PUJ 3 (S) Megaureter 2 VURa 10 (1S)

Paraureteric diverticula 2 Bladder diverticula 3

1 renal failure (renal dysplasia)

a

Leeds:1 bilateral; Manchester: 5 bilateral; London: 2 bilateral. MCK, multicystic kidney; PUJ, pelviureteric junction obstruction; S, surgery; VUR, vesicoureteral reflux.

Nephrectomy technique While the indications for nephrectomy in MCK continue to be debated, the approach for nephrectomy itself is an additional area of controversy. The traditional approach for nephrectomy of a multicystic kidney is extraperitoneal. This can be accomplished by an anterior muscle-splitting technique or a dorsal lumbotomy technique. Either approach affords a low morbidity and a low complication rate, with hospital stays of 48 hours or less in infants.13 Laparoscopic and robotic nephrectomy has been reported by several authors.34–36 This technique has been reported in infants as well as older children. Each of the reports comments on the superior postoperative course in patients approached laparoscopically, with regard to pain management, length of hospital stay, and cosmetic concerns. Most surgery has been carried out as a day case or with an overnight stay. Operative times are quite variable, from as short as under 1 hour to as long as 5 hours. One argument against laparoscopic nephrectomy in the pediatric population is the violation of the peritoneal cavity, thus creating potential intraperitoneal complications. Ono et al. 37 have advocated a retroperitoneal approach to nephrectomy by dilating the retroperitoneal space, which has been reported in children as well.38 While this approach avoids potential intraperitoneal complications, it is technically more challenging. Just because a procedure can be accomplished laparoscopically does not necessarily mean that it should be

done that way. There have been no controlled nor randomized studies comparing the laparoscopic approach to a traditional open approach for any of the pediatric laparoscopic procedures. Several factors should be considered. It is obvious that the generic benefits of laparoscopic surgery (i.e., smaller incision and faster recovery) carry somewhat less significance for the very young child, and infants recover from open surgery quite quickly. There is a steep learning curve to laparoscopic surgery, especially in children where the margins for error are smaller in the confined space. Instrumentation, in addition, presents some limitations, although technology is catching up in pediatric laparoscopy. Finally, except in the most able hands, laparoscopic surgical procedures in children require more anesthetic time especially in the younger pediatric patient. In conclusion, the present authors consider that laparoscopic nephrectomy for MCK is becoming the technique of choice in experienced hands. However, the technology should not change the indications.

The dilemma Surgeons who routinely advise nephrectomy often cite the potential risk of malignant change and the development of hypertension as the main reasons. The following discussions will therefore focus on an analysis of these two aspects.

689

690

G. M. Manzoni and A. A. Caldamone

Table 54.3. Malignancy related to MCK Series Wilms’ tumor Raffensberger and Aboulseiman46 Hartman et al.47 Oddone et al.48 Minevich et al.10 De Oliveira et al.49 Homsy et al.50 Beckwith41

Renal cell carcinoma Barret and Wineland52 Burgler and Hauri53 Birken et al.54 Shirai et al.55 Rackley et al.56 Transitional cell carcinoma Mingin et al.51 Mesothelioma Gutter and Hermanek57

Year Age

Sex Tumor

Country

1968 10 m

M

Stage I

USA

1986 4 yrs F 1994 17 m M 1996 11 m M 1997 18 m M 1997 3 m F 5m F 1997 23 m M 46 m F 6.5yrs M 18 m F

Stage II Stage I Stage III Stage II Stage I Stage II Stage II Stage I Stage IV Unknown stage

USA Italy USA Brazil Canada Canada USA USA USA France

1980 1983 1985 1986 1994

Renal cell Ca Renal cell Ca Renal cell Ca Renal cell Ca Renal cell Caa

USA Switzerland USA Japan USA

2000 63 yrs M

Grade I

USA

1957 68 yrs F

Mesothelioma

Switzerland

26 yrs 68 yrs 15 yrs 33 yrs 53 yrs

F M F F M

a

Involuted MCK. (Table adapted from Perez et al. J. Urol. 1998; 160:1216.)

blastoma, while in the adults five had a renal cell carcinoma, one a mesothelioma, and one transitional cell carcinoma.10,41,46–57 From these reports there is evidence of only one case of a renal cell carcinoma in an adult arising in a regressed MCK, with a concomitant orthotopic ureterocele and a blind ending ureter.56 To date, there have not been any documented cases of MCK in which renal tumors developed in 660 cases followed by the Multicystic Kidney Registry of the Section of Urology of the American Academy of Pediatrics.58 An analysis by Gordon et al.20 has calculated that an incidence of MCK of 1 in 4300 live births gives a figure of approximately 1000 new cases per annum for the UK and the USA, a theoretical total of nearly 30 000 for the period covered by the literature review. Although this is only a hypothetical figure, it may serve to put into context the actual risk of malignant change. It should be added that the estimated risk of dying from a tumor in a multicystic kidney has to be balanced against the morbidity of nephrectomy and ultimately against the risk of dying from an anesthetic or some other complication of the surgery. Whether or not laparoscopy and robotic technology should alter the paradigm is debatable. In a comprehensive review Husmann concludes that available literature dictates that the incidence of renal tumors in MCK is so sporadic that the surgical extirpation of dysplastic renal tissue should not be performed solely on the basis of its presence.59

Hypertension Potential for malignant change It has been demonstrated that the absence of renal tissue determined radiographically does not preclude the presence of residual renal dysplasia.22,39,40 Involution seems to occur only through reabsorption of cystic fluid, leaving cellular elements, and the potential hazards of neoplastic change remain lifelong. The incidence of nodular renal blastema in multicystic dysplasia has been reported with a range between 2% and 5%,41–43 but similar histological specimens have been identified in obstructed kidneys or in dysplastic upper poles of duplex kidneys.44,45 The clinical significance of this histological finding is unclear, and no correlation has yet been established between nodular renal blastema and the actual development of malignancy. The progression rate of nodular renal blastema into Wilms’ tumor has been estimated at approximately one in 100 cases,41 indicating that nearly 2000 nephrectomies for MCK would be necessary to prevent one Wilms’ tumor from occurring.43 A review of the literature covering the last 30-year period identified 18 reports of malignancy associated with MCK (Table 54.3). In all 11 pediatric cases there was a nephro-

Among the justifications for prophylactic nephrectomy, the risk of developing hypertension is probably the one cited most commonly. It is known that there is a link between MCK and hypertension, and there is growing interest in the problem of trying to estimate the magnitude of the risk. Table 54.4 summarizes the reported pediatric cases in the literature of the last 30 years. Gordon and associates undertook a computerized literature search covering a 20-year period and found 9 cases.20 Of these, 3 had responded to removal of the MCK while 6 had not. The present authors have been able to identify an additional 31 cases reported in the last 10 years, of which 16 definitely responded to nephrectomy.13,20,23,24,53,60–68 Webb et al.23 in their report of a case of hypertension in a 14-year-old girl already discharged from follow-up as a result of “disappearance” of the cystic kidney on ultrasound exam, raise the issue of balancing the costs of nephrectomy with a 24-hour hospital stay and one postoperative followup visit vs. multiple ultrasound examinations and office visits in the first 5 years of life and annually thereafter. Emmet and King68 postulate that the absence of response to nephrectomy does not mean that the hypertension

Multicystic kidney

was not caused initially by the MCK. A similar report by Snodgrass et al. attributes sustained hypertension after nephrectomy to either damage to the contralateral kidney or other factors such as cardiac hypertrophy.67 With sustained hypertension, contralateral arteriolar thickening of the “normal” kidney may produce continued hypertension persisting, therefore, even after the removal of the MCK. The published literature on hypertension in childhood, on the other hand, does not seem to identify any case related to MCK, as reported by Taylor et al.69 and Hendren et al.70 in their series of 42 and 22 children, respectively, with “surgical” hypertension. The role of MCK as a source of hypertension in adult life and the interpretation of the literature on hypertension in adults is much more difficult to evaluate. A few earlier studies have already been analyzed by Gordon et al.20 and their conclusion was that the available evidence does not support the view that MCK poses a significant threat of hypertension later in life or in adult life. We still do not know all the mechanisms by which renal disease induces hypertension. More recently attention has focused on the hypothesis of hyperfiltration damage in solitary kidneys. The mechanism of hypertension in multicystic dysplasia is still not understood. Multicystic dysplasia results in global kidney aberration affecting nephron development and the vascular tree. Renin and the renin/angiotensin system are central players in hypertension but there is little known about the role of renin in multicystic dysplastic kidneys. It is proposed that renin in the fetal kidney may have a role in vascular development and may modulate the kidney morphogenesis. Renin localization studies were performed by Konda71 on non-hypertensive and hypertensive multicystic dysplasia demonstrating increased renin expression. Lapis72 conversely showed a quantitatively decreased renin in multicystic dysplastic kidneys but a pronounced ectopic renin expression within arterioles and interstitial macrophages was present in hypertensive multicystic dysplastic kidneys (two kidneys) compared to non-hypertensive kidneys (12 kidneys). With increased sensitivity in the array technology and development of other competing methods, our understanding will hopefully improve in the near future. Further expression profiling analyses of series of multicystic dysplastic and normal kidneys will be helpful in identifying links between renin and other gene families involved in kidney development. The true incidence of hypertension in children with MCK may be unknown because of the difficulty in determining the blood pressure in a neonate and infant and mostly by the fact that it is not routinely done or reported in these children. As we continue to follow expectantly most

691

Table 54.4. Hypertension related to MCK Series

Year n

Age

Management Outcome

Javadpour et al.60 Burgler and Hauri53 Chen et al.61 Vinocur et al.24 Gordon et al.20 Susskind and King62 Angermeier et al.63 Wacksman (unpublished data) Elder et al.13 Andretta et al.64 Petterson and Klauber65 Webb et al.23 Rudnik-Schoneborn et al.66 Snodgrass67

1970 1983 1985 1988 1988 1989 1992 1993

1 1 1 1 9 2a 1 5

6yrs

Nx Nx Nx Cons. Nx Nx Nx Nx

Improved Improved Improved NR Impr. (3) Improved Improved ? NR

1995 1995 1996 1997 1998

3 1 6 yrs 1 3 3 m to 14 yrs 6b 1 m to 1 yr

Nx Nx Nx Nx Cons(4) Nx(2)

Improved Improved NR Improved Impr. (2)

Oliveira et al.

26

1m 9 yrs 1m 5m

2000 4

1 m to 4 yrs

Nx

Impr. (2)

2001 1

1m

Cons.

Improved

(a) One patient had multiple complications associated with prematurity including intraventricular hemorrhage and the other had ureteropelvic junction obstruction on the contralateral renal unit. (b) Two patients had cardiovascular anomalies and 1 patient developed hypertension after nephrectomy. NR = not recorded. (Table adapted from Oliveira et al. Pediatr. Nephrol. 2002; 17:954–958.)

Fig. 54.6. Basic algorithm of current management of MCK.

of these children, it would be important to include routine blood pressure measurements and possibly to document any form of vascular flow or subsequent changes in the affected kidney. Ideally, whenever possible, additional measurements of plasma renin activity (PRA) and circulating angiotensins (RAS) should be added for all the hyertensive patients. Only in a systematic study may we avoid underestimating the real entity of this potential problem.

692

G. M. Manzoni and A. A. Caldamone

Fig. 54.7. Algorithm for treatment.

The present evidence is that long-term morbidity of MCK is minimal but there is very little data on such long-term outcomes. Undoubtedly, only appropriate long-term studies will eventually confirm this clinical impression. If one chooses a non-operative approach to management of MCK, one must be assured of adequate follow-up by the family, since the data is not clear on how long follow-up should be undertaken. LaSalle et al. reported in a poll of selected insurance companies that only 15% would issue life insurance if a MCK was left in situ, while 70% would issue coverage after nephrectomy.73 While these policies are clearly not based on data, we need to factor these into the care of patients. From an economic standpoint, a cost analysis by Perez et al. found that nephrectomy was more cost-effective than ultrasound surveillance if surveillance was performed every 3 months.74 While the authors acknowledge that the risk of Wilms’ tumor in a MCK is low (estimated 0.1%), if one chooses to screen, very frequent ultrasounds should be performed.

Conclusions It is clear that our understanding of MCK is still incomplete, particularly in terms of the real incidence of its potential complications. The advent of prenatal ultrasound imaging has undoubtedly increased the number of cases with the diagnosis and has presented a potentially new spectrum of patients. The institution of a registry of patients with MCK and a change of attitude was needed to begin to understand better the natural history of this entity. The authors’ current management protocol is summarized by the algorithms in Figs. 54.6 and 54.7.

Neonatal and perinatal evaluation should confirm the diagnosis and carefully identify any contralateral abnormality. Prophylactic antibiotic treatment and MCU screening evaluation should be reserved for those patients whose ultrasound evaluation has clearly demonstrated an abnormal contralateral kidney or any ureteric dilation. The real clinical significance of the vast majority of associated VUR is questionable, being mostly of a low grade and prone to spontaneous resolution. Some clinicians may choose an MCU on all patients with MCK owing to the higher incidence of VUR identified in these patients, but the clinical significance of the low grade of reflux remains to be determined. Whenever a hydronephrotic solitary kidney is identified (i.e., pelviureteric or vesicoureteric junction obstruction), prompt surgical treatment is indicated to maximize preservation of renal function. In only a few selected situations does it seem reasonable to offer surgery as a definitive step in the management of an otherwise healthy infant with MCK: with a very large MCK(> 6 cm); when the retained mass appears to be growing; when the diagnosis is in question; when adequate follow-up cannot be assured; and if hypertension or symptoms develop. Only the future will eventually provide a better understanding of the biology of MCK, its spectrum of occurrence and its developmental origins, as well as its relationship to malignancy. Prophylactic routine nephrectomy is, therefore, unjustified, although it is still very important to recommend lifelong monitoring of blood pressure of all patients with a retained MCK until definitive information on the risk of hypertension becomes available.

Multicystic kidney

Acknowledgment We would like to acknowledge both Ms. H. K. Dhillon and Dr J. Wacksman for their cooperation in providing unpublished data on MCK patients.

REFERENCES 1. Schwartz, J. An unusual unilateral multicystic kidney in an infant. J. Urol. 1926; 35:259. 2. Spence, H. M. Congenital unilateral multicystic kidney. J. Urol. 1955; 74:693. 3. Pathak, I. G. & Williams, D. I. Multicystic and cystic dysplastic kidneys. Br. J. Urol. 1964; 36:318. 4. Aubertin, G., Cripps, S., Coleman, G. et al. Prenatal diagnosis of apparently isolated unilateral multicystic kidney: implications for counselling and management. Prenat. Diagn. 2002; 22:388. 5. Eckoldt, F., Woderich, R., Smith, R. D., & Heling, K. S. Antenatal diagnostic aspect of unilateral multicystic kidney dysplasiasensitivity, specificity, predictive values, differential diagnoses, associated malformations and consequences. Fetal Diagn. Ther. 2004; 19:163. 6. Felson, B. & Cussen, L. J. The hydronephrotic congenital multicystic disease of the kidney. Semin. Roentgenol. 1975; 10:113. 7. Feldenberg, L. R. & Siegel, N. J. Clinical course and outcome for children with multicystic dysplastic kidneys. Pediatr. Nephrol. 2000; 14:1098–1101. 8. Al-Khaldi, N., Watson, A. R., Zuccollo, J., Twining, P., & Rose, D. H. Outcome of antenatally detected cystic dysplastic kidney disease. Arch. Dis. Child. 1994; 70:520–522. 9. Wacksman, J. Dilemma of the multicystic kidney. Prob. Urol. 1990; 4:574–582. 10. Minevich, E., Wacksman, J., Phipps, L. et al. The importance of accurate diagnosis and early close follow-up in patients with suspected multicystic dysplastic kidney. J. Urol. 1997; 158:1301–1304. 11. Agrawal, L., Millard, M. L., Fairhurst, J., & Gilbert, R. D. Bilateral multicystic kidneys – an unusual case. Pediatr. Nephrol. 2002; 17:964. 12. Flack, C. E. & Bellinger, M. F. The multicystic dysplastic kidneys and contralateral vesicoureteral reflux: protection of the solitary kidney. J. Urol. 1993; 150:1873–1874. 13. Elder, J., Kladky, D., & Selzman, A. A. Outpatient nephrectomy for nonfunctioning kidneys. J. Urol. 1995; 154:712–715. 14. Atyeh, B., Husmann, D., & Baum, M. Contralateral renal abnormalities in multicystic dysplastic kidneys disease. J. Pediatr. 1992: 121:65–67. 15. Robson, W. L. M., Rogers, R. C., & Leung, A. K. C. Renal agenesis, multicystic dysplasia, and uretero-pelvic junction obstruction – a common pathogenesis? (letter) Am. J. Med. Genet. 1994; 53:302. 16. Roodhoft, A. M., Birnholz, J. C., & Homes, L. B. Familial natural of congenital absence and severe dysgenesis of both kidneys. N. Engl. J. Med. 1984; 310:1341–1345.

17. Groenen, P. M., Vanderlinden, G., Devriendt, K. et al. Rearrangement of the human CDC5L gene by a t(6;19) (p21; q13.1) in a multicystic renal dysplasia. Genomics 1998; 49:218. 18. Srivastava, T., Garola, R. E., & Hellerstein, S. Autosomal dominant inheritance of multicystic dysplastic kidney. Pediatr. Nephrol. 1999; 13:481. 19. Bernstein, J. & Gilbert-Barness, E. Developmental abnormalities of the kidney. In Stenberg, S. S. (ed.) Diagnostic Surgical Pathology, 3rd edn. Philadelphia: Lippincott, Williams & Wilkins. 2000: 1685–1700. 20. Gordon, A. C., Thomas, D. F. M., Arthur, R. J. et al. Multicystic dysplastic kidney: is nephrectomy still appropriate? J. Urol. 1998; 140:1231–1234. 21. Rickwood, A. M. K., Anderson, P. A. M., & Williams, M. P. L. Multicystic renal dysplasia detected by prenatal ultrasonography: natural history and results of conservative management. Br. J. Urol. 1992; 69:538–540. 22. Gough, D. C. S., Postlewaite, R. J., Lewis, M. A. et al. Multicystic renal dysplasia diagnosed in the antenatal period: a note of caution. Br. J. Urol. 1995; 76:244–248. 23. Webb, N. J. A., Lewis, M. A., Bruce, J. et al. Unilateral multicystic dysplastic kidney: the case for nephrectomy. Arch. Dis. Child. 1997; 76:31–34. 24. Vinocur, L., Slovis, T. L., Perlmutter, A. et al. Follow-up studies of multicystic dysplastic kidneys. Radiology 1998; 167:311– 315. 25. Strife, J. L., Souza, A. S., Kirks, D. R. et al. Multicystic dysplastic kidney in children:US follow-up. Radiology 1993; 186:785– 788. 26. Oliveira, E. A., Diniz, J. S., Vilasboas, A. S., Rabelo, E. A., Silva, J. M., & Filgueiras, M. T. Multicystic dysplastic kidney detected by fetal sonography: conservative management and follow-up. Pediatr. Surg. Int. 2001; 17(1) 54–57. 27. Kuwertz-Broeking, E., Brinkmann, O. A., Von Lengerke, H. J. et al. Unilateral multicystic dysplastic kidney: experience in children. Br. J. Urol. Int. 2004; 93(3):388–392. 28. Belk, R. A., Thomas, D. F., Mueller, R. F., Godbole, P., Markham, A. F., & Weston, M. J. A family study and the natural history of prenatally detected unilateral multicystic dysplastic kidney. J. Urol. 2002; 167(2 Pt 1):666–669. 29. Eckoldt, F., Woderich, R., Wolke, S., Heling, K. S., Stover, B., & Tennstedt, C. Follow-up of unilatral multicystic kidney dysplasia after prenatal diagnosis. J. Matern. Fetal Neonatal Med. 2003; 14(3):177–186. 30. Sukthankar, S. & Watson, A. R. Unilateral multicystic dysplastic kidney disease: defining the natural history. Angelia Paediatric Nephrourology Group. Acta. Paediatr. 2000; 89(7):811–813. 31. Holloway, W. R. & Weinstein, S. H. Percutaneous decompression: treatment for respiratory distress secondary to multicystic dysplastic kidney. J. Urol. 1990; 144:113. 32. Stuck, K. J., Koff, S. A., & Silver, T. M. Ultrasonic features of multicystic dysplastic kidney: expanded diagnostic criteria. Radiology 1982; 143:217. 33. Ambrose, S. S., Gould, R. A., Trulock, T. S. et al. Unilateral multicystic renal disease in adults. J. Urol. 1990; 128:366–369.

693

694

G. M. Manzoni and A. A. Caldamone

34. Koyle, M. A., Woo, H. H., & Kavoussi, L. R. Laparoscopic nephrectomy in the first year of life. J. Pediatr. Surg. 1993; 28:693. 35. Erlich, R. M., Gershman, A., Mee, S., & Fuchs, G. Laparoscopic nephrectomy in a child: expanding horizons for laparoscopy in pediatric urology. J. Endourol. 1992; 6:463. 36. Cilento, B. G., Jr. & Peters, C. Laparoscopic and robotic surgery in pediatric urology. In Pediatric Surgical Procedures, A. A. Caldamone (ed.), Atlas Urol. Clin. N. Am. 2004; 12:143. 37. Ono, Y., Katoh, N., Kinukana, T. et al. Laparoscopic nephrectomy via the retroperitoneal approach. J. Urol. 1996; 150:101– 1104. 38. Diamond, D. A., Price, H. M., McDougall, E. M., & Bloom, D. A. Retroperitoneal laparoscopic nephrectomy in children. J. Urol. 1995; 153:1966–1968. 39. Pedicelli, G., Jequier, S. Bowen, A. D. et al. Multicystic dysplastic kidneys: spontaneous regression demonstrated with US. Radiology 1986; 160:23. 40. Avni, E. F., Thoua, Y., Lalmand, B. et al. Multicystic dysplastic kidney: natural history from in utero diagnosis and post-natal follow-up. J. Urol. 1987; 138:1420–1424. 41. Beckwith, J. B. Wilms’ tumor and other renal tumors in childhood: an update. J. Urol. 1986; 136:320. 42. Johnson, H. W., Coleman, G. U., & Dimmick, J. E. Wilms’ tumorlet, nodular renal blastema and multicystic renal dysplasia: a spectrum. J. Urol. 1989; 142:484–485. 43. Noe, H. N., Marshall, J. H., & Edwards, O. P. Nodular renal blastema in the multicystic kidney. J. Urol. 1989; 142:486–488. 44. Gaddy, C. D., Gibbons, M. D., Gonzales, E. T., Jr. et al. Obstructive uropathy, renal dysplasia and nodular renal blastema: is there a relationship to Wilms’ tumor? J. Urol. 1985, 134:330. 45. Craver, R., Dimmick, J. E., Johnson, H. W. et al. Congenital obstructive uropathy and nodular renal blastema. J. Urol. 1986; 136:305. 46. Raffensperger, J. & Aboulseiman, A. Abdominal masses in children under one year of age. Surgery 1968; 63:514. 47. Hartman, G. E., Smolik, L. M., & Schochat, S. J. The dilemma of the multicystic dysplastic kidney. Am. J. Dis. Child. 1986; 140:925. 48. Oddone, M., Marino, C., Sergi, C. et al. Wilms’ tumor arising in a multicystic kidney. Pediatr. Radiat. 1994; 24:236. 49. deOliveira-Filho, A. G., Carvalho, M. H., Sbragia-Neto, L. et al. Wilms’ tumor in a prenatally diagnosed multicystic kidney. J. Urol. 1997; 158:1926–1927. 50. Homsy, Y. Personal communication. In Elder, J., Hladky, D., & Selzman, A. A. Outpatient nephrectomy for nonfunctioning kidneys. J. Urol. 1995; 154:712–715. 51. Mingin, G. C., Gilhooly, P., & Sadeghi-Nejad, H. Transitional cell carcinoma in a multicystic dysplastic kidney. J. Urol. 2000; 163:544. 52. Barrett, D. M. & Wineland, R. E. Renal cell carcinoma in multicystic dysplastic kidney. Urology 1980; 15:152. ¨ 53. Burgler, W. & Hauri, D. Vital complications in multicystic kidney degeneration (multicystic dysplasia). Urol. Int. 1983: 38:251– 256.

54. Birken, G., King, D., & Vane, D. Renal cell carcinoma arising in a multicystic dysplastic kidney. J. Pediatr. Surg. 1985; 20:619. 55. Shirai, M., Kitagawa, T., Nakata, H. et al. renal cell carcinoma originating from dysplastic kidney. Acta Path. Jap. 1986; 36:1263. 56. Rackley, R. R., Angermeier, K. W., Levin, H. et al. Renal cell carcinoma arising in a regressed multicystic dysplastic kidney. J. Urol. 1994; 152:1543–1545. 57. Gutter, W. & Hermanek, P. Maligner Tumor der nierengengend uter dem bilde der Knollenniere (Nierenblastemcystein). Urol. Int. 1957; 4:164. 58. Wacksman, J. Editorial comment. J. Urol. 1995; 154:714. 59. Husmann, D. A. Renal dysplasia: the risks and consequences of leaving dysplastic tissue in situ. Urology 1998; 52(4):533–536. 60. Javadpour, N., Chelouhy, E., Moncada, L. et al. Hypertension in a child caused by a multicystic kidney. J. Urol. 1970; 104:918– 921. 61. Chen, Y.-H., Stapleton, B., Roy, S. et al. Neonatal hypertension from a unilateral multicystic dysplastic kidney. J. Urol. 1985; 133:664–665. 62. Susskin, M. R., Kim, K. S., & King, L. R. Hypertension and multicystic kidney. Urology 1989; 34:362. 63. Angermeier, K. W., Kay, R., & Levin, H. Hypertension as a complication of multicystic dysplastic kidney. Urology 1992; 39:55. 64. Andretta, E., Aragona, F., Talenti, E., & Passerini Glazel, G. Il rene multicistico. Urol. Pediatr. 1995; 1:31–38. 65. Pettersson, B. A. & Klauber, G. T. Prognosis for children born with multicystic dysplastic kidneys (abstract). Br. J. Urol. 1996; 77(Suppl.1):18. 66. Rudnik-Sch¨oneborn, S., John, U., Deget, F. et al. Clinical features of unilateral multicystic renal dysplasia in children. Eur. J. Pediatr. 1998; 157:666–672. 67. Snodgrass, W. T., Wacksman, J., & Homsy, Y. Hypertension associated with multicystic dysplastic kidney in children. J. Urol. 2000; 164(2):472. 68. Emmert, G. K. & King, L. R. The risk of hypertension is underestimated in the multicystic dysplastic kidney: a personal perspective. Urology 1994; 44:404–405. 69. Taylor, R. C., Azmy, A. F., & Young, D. G. Long-term follow-up of surgical renal hypertension. J. Pediatr. Surg. 1997;22:228. 70. Hendren, W. H., Kim, S. H., Herrin, J. T. et al. Surgically correctable hypertension of renal origin in childhood. Am. J. Surg. 1982; 143:432. 71. Konda, R., Sato, H., Ito, S. et al. Renin containing cells are present predominantly in scarred areas but not in dysplastic regions in multicystic dysplastic kidney. J. Urol. 2001; 166:1910. 72. Lapis, H., Doshi, R. H., Watson, M. A. et al. Reduced renin expression and altered gene transcript profiles in multicystic dysplastic kidneys. J. Urol. 2002; 168(4 Pt 2):1816. 73. LaSalle, M. D., Stock, J. A., & Hanna, M. K. Insurability of children with congenital urological anomalies. J. Urol. 1997; 158:1312– 1315. 74. Perez, L. M., Naidu, S. I., & Joseph, D. B. Outcome and cost analysis of operative versus nonoperative management of neonatal multicystic dysplastic kidneys. J. Urol. 1998; 160:1207–1211.

55 Urolithiasis Bartley G. Cilento, Jr,1 Gerald C. Mingin,2 and Hiep T. Nguyen1 2

Introduction There is a considerable body of literature dedicated to adult stone disease but very little written about pediatric stone disease. While nephrolithiasis is less prevalent in children, it can be associated with significant morbidity. This chapter will touch upon the pathophysiology, the surgical management, and the current state of evidence base data, particularly long-term outcomes of pediatric stone disease.

Epidemiology Childhood urolithiasis varies widely around the world both in the composition and location of the stones. For example, it is endemic in the Far East and Turkey where the stones are primarily ammonium uric acid stones. In European countries, the majority of stones found in children are composed of struvite and organic matrix that are frequently the result of urinary tract infections from bacteria containing urease. In the USA, infection-related stones are rare. Stone composition in the USA is mostly calcium-based. In North America and other developed countries, nephrolithiasis in children has remained stable over the last 20 years. In Europe and the United States, the incidence of stone disease is similar, ranging from 0.13 to 1.0 cases per 1000 hospital admissions.1,2 The disease always varies geographically within countries. For example, the incidence tends to be higher in the warm climates of the southeastern United States and California.3 The incidence is also higher in areas with immigrants from countries where stone disease is endemic. The ratio of boys to girls affected by stone disease is 1.2–1.7:1 in the USA and 1.9–3:1 in the UK.4,5 In

1 Department of Urology, MA Children’s Hospital, Boston, Department of Urology, Denver Children’s Hospital, Denver, CO, USA

these countries the majority of children who form stones are Caucasian, while less than 10% are African American.2 In developing countries, stone disease is still endemic, e.g., in countries such as Turkey and parts of North Africa. Forty years ago, 50% of the stones seen in these children were bladder stones, associated with an infectious etiology. Currently, 75%–80% of these children have upper tract stones and 5% appear to be caused by infection.6 This pattern closely resembles what is seen in North America and in parts of Western Europe, where metabolic disorders are responsible for 50% of stones. This shift in the composition and location of the stone is likely to be due to improved living standards with better health care, and improved nutrition with increased protein consumption.7 Recent studies suggest a shift in the age of patients presenting with a stone, as well as in the size of the stone. In a retrospective review of 103 patients, children presenting between 1988 and 1995 had a mean age of 13.9 years and an average stone size of 10.7 mm, while those presenting after 1995 had a mean age of 10.9 years, and a stone size of 7.5 mm.8 Once a child develops urolithiasis, there is a 6%–54% chance of recurrence within 6 years. Those children identified with a metabolic abnormality have a fivefold chance of recurrence in comparison to those children without an identifiable metabolic abnormality. In children, the manifestations of stone can be less dramatic than in adults. In terms of presenting symptoms, abdominal pain occurred in 50%, hematuria in 33%, and infection in 11% of children. Urolithiasis was an incidental finding in 15% of children undergoing radiographic imaging. In the USA, the vast majority of stones are located in the renal pelvis (75%). Ten percent of stones are found in

Pediatric Surgery and Urology: Long-term Outcomes, Mark Stringer, Keith Oldham, Pierre Mouriquand. C Cambridge University Press, 2006. Published by Cambridge University Press.


695

696

B. G. Cilento, Jr, G. C. Mingin and H. T. Nguyen

Table 55.1. Metabolic etiology of stones Calcium lithiasis Hypercalciuria (>4 mg/kg per day) Normocalcemic (8.7–10.1 mg/dl) Hypercalcemic (>10.1 mg/dl)

Hyperoxaluria (>40 mg/1.73 m2 per day)

Hyperuricosuria (>815 mg/1.73 m2 per day)

Hypocitraturia (5.5 mg/dl)

Cystine lithiasis (>75 mg/g Cr) Misc. inborn errors of metabolism

Idiopathic – absorptive, renal leak RTA Associated with bone resorption – Immobilization, hyperparathyroidism GI hyperabsorption – Vitamin D intoxication – Sarcoidosis – Phosphorus depletion – Hyper- or hypothyroidism – Bartter’s syndrome – Malignancy Hereditary type I and II Enteric – malabsorption, IBD, bowel resection Dietary (increased oxalate consumption, vitamin C) Leukemia, lymphoma Idiopathic, familial Inborn errors of metabolism (glycogen storage disease) Metabolic acidosis Distal RTA GI disease Potassium depletion Familial hyperuricemia Blood disorders Inborn errors of metabolism (Lesch–Nyhan syndrome) Defect in COLA reabsorption Xanthinuria Orotic aciduria Adenine phosphoribosyltransferase deficiency

the ureter and another 10% are located within the bladder. Stones larger than 5 mm are unlikely to pass spontaneously. Bladder stones are being found with increasing frequency in patients who have undergone bladder augmentation.

Etiology Recurrent childhood nephrolithiasis has a number of documented causes, including metabolic abnormalities, infection, structural anomalies, idiopathic, and iatrogenic causes. A metabolic abnormality has been found in more

than 50% of children with stones. These metabolic abnormalities include hypercalciuria, hyperoxaluria, hyperuricosuria, hypocitrauria, cystinuria, and renal tubular acidosis (RTA) (Table 55.1). Hypercalciuria is the most common of the metabolic disorders seen in children. Fifty percent or more of these stones are composed of calcium oxalate or calcium phosphate. The causes of hypercalciuria include idiopathic, reabsorptive, iatrogenic, and granulomatous disease (Table 55.2). Idiopathic hypercalciuria is divided into Type I absorptive (diet independent) with increased calcium absorption on both a normal and increased calcium diet and Type II (diet dependent) with decreased calcium absorption on a low calcium diet. The second type of idiopathic hypercalciuria is renal leak hypercalciuria, where impairment in the tubular reabsorption of calcium results in a subsequent increase in PTH, and 1, 25-OH Vitamin D-mediated calcium absorption. Primary hyperparathyroidism is caused by an adenoma or hyperplasia of the parathyroid gland. Primary hyperparathyroidism can be sporadic, familial, recessive, or dominant, and is part of the multiple endocrine neoplasia (MEN) syndromes. Hypercalciuria may be due to iatrogenic causes, such as immobilization of children recovering from surgery and the use of calcium wasting diuretics (furosemide).3 Rarely, hypercalciuria is seen in patients with sarcoidosis. Another metabolic cause of nephrolithiasis is hyperoxaluria. Primary hyperoxaluria is responsible for 50% of oxalate-related stones.9 Children have nephrocalcinosis and recurrent stone formation, which will progress to renal insufficiency. Secondary hyperoxaluria occurs in the setting of intestinal malabsorption, resulting in an increase in the colonic permeability of oxalate, as well as less calcium for oxalate binding. Cystine stones are the least common type of stones in children. Cystinuria is a recessive disorder with an abnormality in the renal and intestinal transport of the dibasic amino acids: cystine, ornithine, lysine, and arginine. Cystine stones may be seen in infancy, where males and females are equally affected.10 Cystine stones can also be seen in patients with associated metabolic abnormalities (such as hypercalciuria, hyperuricosuria and hypocitraturia), hemophilia, retinitis pigmentosa, muscular dystrophy, trisomy 21, and hereditary pancreatitis.1,3,10 Five percent of pediatric stone formers will have a uric acid stone.1,4,11 Elevated urine uric acid levels are caused by a high purine diet, metabolic errors, myeloproliferative and seizure disorders, hemolysis, and malignancy. Low urine pH or an increased ammonium concentration will cause uric acid stones to precipitate. Increased uric acid

Urolithiasis

Table 55.2. Major forms of hypercalciuria

Absorptive hypercalciuria Diet dependent and independent Increased Ca absorption

Renal hypercalciuria

Increased serum Ca

PTH

Urine Ca

Renal leak Urine Ca

Serum Ca

Primary hyperparathyroidism

PTH

1,25(OH)2D

Ca absorption

Adenoma PTH

Bone reabsorption

1,25(OH)2D

levels are seen in children with primary gout, Lesch–Nyhan syndrome, glycogen storage disease, pernicious anemia, polycythemia, and renal tubular defects. Drugs such as salicylates, probenicid, phenobarbital, valproic acid, sulfinpyrazone and alcohol are also predisposing factors.12 Urinary citrate is a natural inhibitor of stone formation. Hypocitrauria can lead to stone formation and 19%–63% of children diagnosed with a stone have hypocitrauria. Metabolic acidosis can accentuate the problem by leading to a decrease in urine citrate excretion. For example, low citrate levels are seen in distal renal tubular acidosis and in severe dehydration with electrolyte imbalance. Short gut or inflammatory bowel disease, cystic fibrosis, diuretic therapy, and urinary tract infections are also frequent causes of hypocitrauria. A high protein diet and excessive intake of red meat results in an increased acid load and have been shown to decrease urine citrate.13 In the USA, infectious stones comprise 24% of total stone formers.4,11 The majority are boys less than 5 years old with upper tract calculi. Struvite stones are the most common and are due to the urea splitting bacteria such as Proteus,

Serum Ca

Urine Ca

Ca absorption

Klebsiella, Pseudomonas, Providencia, Enterococci and the non-urea splitting E. coli. Functional obstruction with urinary stasis can predispose children to stone formation. In fact, 30% or more of children with urinary tract stones have a concomitant anatomic abnormality that predisposes to urine stasis. These abnormalities include ureteropelvic and ureterovesical obstruction as well as obstructed megaureter, single system ureterocele, and neuropathic bladder.

Clinical presentation In children, the manifestations of stone can be less dramatic than in adults. In terms of presenting symptoms, abdominal pain occurred in 50%, hematuria in 33%, and infection in 11% of children with stones. Urolithiasis was an incidental finding in 15% of children undergoing radiographic imaging. The clinical presentation of a child with nephrolithiasis can vary with age. Hematuria is the most common

697

698

B. G. Cilento, Jr, G. C. Mingin and H. T. Nguyen

RBUS Normal

Questionable stone

stone

Consider CT scan Obstructed

Stent or percutaneous nephrostomy

CT scan w/o contrast Stone

No stone

Retrograde pylogram

ESWL, percutanous nephrolithotomy, ureteroscopy, open surgery

Fig. 55.1. Evaluation for a suspected stone.

presenting sign in children under 5 years old. Gross or microscopic hematuria occurs in 30% to 90% of children with stones, while pain is a more common finding in older children. Unlike in adults, the pain is rarely localized to the flank with radiation to the groin.12 Children with stones may experience abdominal pain, which can mimic appendicitis or gastroenteritis. Children can also present with infection or other urinary symptoms such as dysuria, frequency, or incontinence. A urinary tract infection may be the only presenting sign of a stone in infants.

Evaluation of a child with a suspected stone The first step in the evaluation of infants and children with a suspected stone is directed radiologic imaging (Fig. 55.1). The goal of the radiological evaluation is to confirm the diagnosis, assess the stone burden, location, and the degree urinary obstruction. In children, the initial imaging modality of choice is the renal/bladder ultrasound (RBUS) because it is non-invasive. Experienced pediatric ultrasonographers and current imaging technology have dramatically increased the diagnostic accuracy of this modality such that it is uncommon for the authors to need further imaging such as the intravenous pyelogram (IVP) or computerized tomography (CT). Even ureteral stones can be seen with RBUS but are certainly more difficult to visualize than stones located within the kidney. If clinical suspicion is high and the RBUS has not demonstrated a stone, further evaluation with a spiral CT scan (with and without) contrast may be helpful in accessing the stone size, location of the stone, and the amount of kidney function. Because of its high speed and sensitivity, CT scan has eliminated the need for intravenous pyelography. Rarely, a retrograde pyelogram is needed to further clarify

the stone position or urinary tract anatomy. In addition to the radiographic studies, a complete blood count (CBC), serum electrolytes, BUN, creatinine, serum calcium, and phosphorus should be obtained to accurately assess the child’s condition. The urine should also be cultured for bacteria and analyzed for pH, protein, and white blood cells. The urine should also be examined for stone crystals. The clinical picture and radiographic findings will determine the treatment approach: observation or surgical intervention. Children should receive narcotics sufficient to relieve pain and be adequately hydrated, usually 1.5 to 2 times maintenance fluid. Urinary tract infections should be treated with antibiotics. Children who are afebrile, have a stone less then 4 mm in size, and have no associated anatomical abnormalities, may be observed. Asymptomatic children with stones greater than 4 mm in size will likely require intervention since these stones are not likely to pass spontaneously. If the child is obstructed, immediate drainage of the affected kidney is required. Uncontrolled pain, nausea, and emesis even in the presence of a small stone are indications for definitive intervention.

Surgical treatment in children with a first-time stone In children with upper tract stones, extracorporeal shock wave lithotripsy (ESWL) is the preferred first-line therapy particularly for small to moderate calculi (4–10 mm) and no anatomic abnormalities (Fig. 55.2). There are no set guidelines for the use of ESWL in children but it has been shown to be safe and effective. Jayanthi et al. reported success with stones up to 1.5 cm in size in children age 17 months to 14 years.14 Most centers report a stone-free rate between 70% and 80% for medium size stones,15,16 with even higher stone-free rates using second-generation lithotripters (Table 55.3). Use of ESWL in staghorn or partial staghorn calculi is controversial, with reports of stonefree rates in excess of 70%; however, 50% of these patients required one or more treatments.17 The advantages of ESWL in children include its noninvasiveness and infrequent need to place a ureteral stent. Most of the newer ESWL machines use ultrasound localization and thereby avoid exposure to ionizing radiation. Either intravenous sedation and general anesthesia are used during the ESWL procedure and the choice is generally determined by surgeon, anesthesiologist, and institutional preferences. In addition, because of their small size, children have decreased shock wave attenuation, limiting trauma to the surrounding organs. Newer machines like the lithostar, which uses piezoelectric energy to deliver

Urolithiasis

Treatment of stones

Surgery

Observation

Stone size > 4 mm Symptoms (e.g. pain) Infection Associated anatomic abnormalities

Stone size < 4 mm Min. symptoms (e.g., pain) No infection No associated anatomic abnormalities

Ureteroscopy

ESWL

Ureteral stone > 4 mm Ureteral stone with obstruction Older children

Stone size 4–10 mm No associated anatomic abnormalities Located in kidney or proximal ureter Able to be visualized on fluoro. or RPG Larger and older children (> 2 years old) No Hx of cystine stone

Open surgery PCNL

Large stone burden Need to correct associated anatomic abnormalities

Stone size > 10–15 mm Associated anatomic abnormalities (UPJ obstruction) Stone not able to be visualized well Failed ESWL Cystine or dense calcium stone

Fig. 55.2. Treatment of pediatric stones.

the shock precisely to a smaller focal zone, minimize damage even further.17 Long-term studies indicate that ESWL is safe in children. Renal function and linear growth are not affected, and there is no increased risk of hypertension or scarring.14 The disadvantages of ESWL include the need for fluoroscopic localization of the stone in some cases, the potential development of urosepsis, steinstrasse (accumulation of multiple stone fragments within the ureter), hematuria, and flank pain. However, these complications occur infrequently with second-generation lithotripters.18 Another disadvantage is the higher retreatment rate and the number of shocks needed with the second-generation piezoelectric lithotripters. The energy delivered to the stone is actually lower because of the small focal volume, necessitating further treatments when compared to the standard electrohydraulic units.19 In children with a large upper tract (1.5 cm) stone, percutaneous nephrolithotomy (PCNL) may be considered. Woodside performed the first pediatric PCNL with a stone free rate of 100% and no complications.20 Similar series report comparable findings in children of all ages using standard adult techniques (Table 55.3).21,22 Like ESWL, there are no absolute indications for PCNL. Relative indications include: a large stone burden of 1.5 cm or larger; associated urinary tract anatomical abnormalities; inability to visualize the stone; failed ESWL and evidence of a dense

calcium monohydrate or cystine stone. The advantages of PCNL treatment include definitive removal of large stones. The disadvantages of PCNL include renal scarring,23 a high retreatment rate with the mini-percutanous technique,24 and higher complication as compared with ESWL. The overall complication rate of PCNL is 4%.25 However, with appropriate patient selection and the use of laser lithotripsy, retreatment can be minimized. Distal ureteral stones can be managed with ESWL or ureteroscopy. Children with a ureteral stone of 4 mm or larger that cannot be visualized with ESWL should undergo ureteroscopic stone extraction. The success rate of ureteroscopy varies between 80% and 100%. An EHL or Holmium–Yag probe can be passed through a 6.9 or 7.5 Fr. pediatric ureteroscope. After stone fragmentation, ureteral stents are placed for a period of several days to a week, depending on the amount of manipulation. The stents are then removed cystoscopically. In older children, the stent can be attached to a string, which is brought out through the urethra and removed without a second procedure. Since the incidences of vesicoureteral reflux after ureteroscopy is extremely low,26 a VCUG is usually not needed after stent removal. Open stone surgery in children is rarely indicated. A large stone burden that would require multiple endoscopic procedures or the presence of an associated congenital abnormality would be an indication to proceed with open surgery.

699

700

B. G. Cilento, Jr, G. C. Mingin and H. T. Nguyen

Table 55.3. Reported outcomes of pediatric stone therapy Outcome – Follow-up in Modality/author Year Patients stone free (%) Retreatment months ESWL Choong 31 Elsobky 19 Orsola 17 Jayanthi 14 Tekin 32 Al busaidy 33 Nazli 34 Picramenos 15 Gschwend 35 Hasanglou 36 Cranidis 37 Van Horn 16 Myers 38 Longo 39

2000 23 2000 148 1999 15 1999 24 1998 59 1998 55 1998 67 1996 96 1996 27 1996 103 1996 28 1995 32 1995 238 1995 70

PCNL Choong 31 Sahin 40 Badawy 22 Al-Shammari 21 Jackman 24 Mor 41 Kurzrock 42

2000 2000 1999 1998 1998 1997 1996

Ureteroscopy Van Savage 43 Wollin 44 Jayanthi 14 Al Busaidy 33 Minevich 45 Kurzrock 42 Smith 46 Scarpa 47 Shroff 48

2000 1999 1999 1997 1996 1996 1996 1995 1995

91 86 73 70 71 87 87 82 92 63 55 68 77 99

13 64 36 NR NR 36 42 28 NR NR NR NR 14 29

36 3 60 1–48 3 3 24 33 46 6 NR 67 3 3

30 14 60 8 7 25 8

90 100 83 88 85 92 100

10 36 NR 12.5 43 12 25

36 1–16 3–72 3–6 3 2–66 5–96

33 19 12 43 7 17 11 7 13

100 84 92 94 85 100 82 100 77

0 16 NR 7 0 0 18 0 NR

24 .5–12 1–48 NR NR 5–96 .5–22 NR NR

Consideration should be given to the fact that children tolerate a flank or dorsal lumbotomy incision well. The advantages of open surgery include rapid and effective stone clearance with a single procedure; however, open surgery especially anatrophic nephrolithotomy can be associated with functional renal deterioration.27

Evaluation of a child with a recurrent stone Children with recurrent stones should be thoroughly evaluated in order to identify the etiology of the stone disease and possibly prevent recurrence (Fig 55.3). Twenty to

History: medications Malignancy, metabolic disorders

Spot urine for ca/cre ratio (Table 55.5)

Stone analysis

PE: height/weight/BP Signs of a congenital anomaly

Medical treatment (Table 55.6)

UTI

VCUG

24 h urine

Cyanide nitroprusside test

Fig. 55.3. Work-up for a repeat stone former.

50% of children with urolithiasis have a family history of stone disease and may point to a metabolic etiology or a tendency toward recurrence.11 A careful inquiry should be made for a history of metabolic disorders, distal RTA, and malignancy. A review of medications (Table 55.4) such as diuretics, steroids, vitamin D analogues, or chemotherapeutic medications is also important. Whenever possible, a chemical analysis of the stone is essential. The physical examination should assess weight and blood pressure, since these abnormalities may suggest an underlying metabolic, systemic, or anatomic defect. Signs of a congenital anomaly such as myelomeningocele should also be noted. In children with a proven urinary tract infection, a voiding cystourethrography is suggested to rule out associated anatomical abnormalities such as vesicoureteral reflux or posterior urethral valves. Laboratory studies should include a serum calcium level and spot urine for calcium, uric acid and oxalate level, relative to creatinine (Table 55.5). These are helpful screening tests and for follow-up in younger children, where obtaining multiple samples may prove difficult.12 If the above results are normal, the work-up is complete. Abnormalities noted on the spot urine test require a 24 h urine collection to characterize the problem better; however, the correlation between the two studies is high. An elevation in calcium, uric acid, or oxalate noted on a 24 h urine collection, can tailor the work-up toward identifying the specific metabolic defect. The amount of urinary calcium varies with age. In children less than 1 year old, the calcium to creatinine ratio varies between 0.81 and 0.42 (Table 55.5). Hypercalciuria can be confirmed on a 24 h urine collection (≥4 mg/kg per day). It has been shown that preterm and term infants on breast milk have a higher amount of calcium in their urine.28 If calcium and sodium are both elevated on urine collection, a repeat urine collection after several weeks of

Urolithiasis

Table 55.4. Medications frequently associated with urolithiasis Calcium stones

Oxalate stones

Uric acid stones

Acetazolamide Calcium channel blockers Furosemide Vitamin D Theophylline Hyperalimentation Triamterene

Vitamin C Phosphate binding antacids

Probenecid sulfinpyrazone

Table 55.5. Urinary solute to creatinine ratio Urinary solute: creatinine ratio Calcium (mg/dl) Neonates 34 weeks Children Oxalate (mg/dl) Infants 0–6 months Children 4 years Uric acid (mg/dl) Timed collections Calcium (mg/kg per 24 h) Oxalate (mg/1.73 m2 per 24 h) (mmol/1.73 m2 per 24 h) Uric acid (mg/1.73 m2 per 24 h) Cystine (mg/24 h) Citrate (mg/g creatine/24 h) Male Female Magnesium (mg/kg per 24 h)

Normal values