Surgical Management of Hepatobiliary and Pancreatic Disorders, Second Edition (Oncologysurgery)

Citation preview

Second Edition Edited by Graeme J. Poston, Michael D’Angelica, and René Adam About the book Hepato-Pancreato-Biliary (HPB) surgery is now firmly established within the repertoire of modern general surgery. This new edition has been completely rewritten by world-leading surgeons to reflect the considerable advances made in the surgical management of HPB disorders since the highly successful first edition. This new edition includes:

• A comprehensive section on anatomy, imaging, and surgical technique • Over 20 new chapters, including a complete account of pediatric HPB disorders • Almost 300 high-resolution images, many in full color Surgical Management of Hepatobiliary and Pancreatic Disorders, Second Edition, comprehensively covers the full spectrum of common HPB diseases and associated surgical techniques to assist not only the general surgeon in regular practice, but also surgical trainees and those in related specialties of oncology, radiology, gastroenterology, and anesthesia.

About the Editors Graeme j. Poston, MS, FRCS (Eng), FRCS (Ed), is Director of Surgery and Hepatobiliary Surgeon, University Hospital Aintree, Liverpool, UK. He is the President of the Association of Upper Gastrointestinal Surgeons of Great Britain and Ireland (AUGIS), PresidentElect of the European Society of Surgical Oncology (ESSO), Past President of the British Association of Surgical Oncology (BASO), and author of numerous publications and national/international guidelines relating to the practice of HPB surgery. Michael D’Angelica, MD, is an Associate Attending at Memorial Sloan-Kettering Cancer Center and an Associate Professor at Cornell University/Weill Medical Center. He is currently the Program Chairman of the American Hepato-Pancreato-Biliary Association and a writing member of the National Comprehensive Cancer Network (NCCN) practice guidelines for hepatobiliary malignancy. René Adam, MD, PHD, is Hepatobiliary Surgeon and Professor of Surgery, Hôpital Paul Brousse, Université Paris-Sud, Villejuif, France.

This book demonstrates the wisdom of the new knowledge and technical skills of these diverse disciplines where cooperative efforts contribute toward the benefit of the patients with HPB disorders. Also Available Hepatocellular Carcinoma: A Practical Approach Edited by Bandar Al Knawy, K. Rajendra Reddy and Luigi Bolondi ISBN: 9780415480802 e-ISBN: 9780203092880

Improved Outcomes in Colon and Rectal Surgery Edited by Charles B. Whitlow, David E. Beck, David A. Margolin, Terry C. Hicks and Alan E. Timmcke ISBN: 9781420071528 e-ISBN: 9781420071535

Textbook of Surgical Oncology Edited by Graeme J. Poston, R. Daniel Beauchamp, and Theo J. M. Rogers ISBN: 9781841845074 e-ISBN: 9780203003220

Surgical Management of Hepatobiliary and Pancreatic Disorders

• An in-depth coverage of benign and malignant disorders of the liver, pancreas, and bile ducts and gallbladder

With a Foreword by Yuji Nimura, MD, President of the Aichi Cancer Center, Japan, and Past President of the IHPBA

Poston • D’Angelica • Adam

Surgical Management of Hepatobiliary and Pancreatic Disorders

Second Edition

Telephone House, 69-77 Paul Street, London EC2A 4LQ, UK 52 Vanderbilt Avenue, New York, NY 10017, USA

www.informahealthcare.com

Surgical Management of

Hepatobiliary and Pancreatic Disorders Second Edition

Edited by

Graeme J. Poston Michael D’Angelica René Adam

Surgical Management of Hepatobiliary and Pancreatic Disorders

Surgical Management of Hepatobiliary and Pancreatic Disorders Second Edition Edited by Graeme J. Poston MS, FRCS (ENG), FRCS (ED) Centre for Digestive Diseases University Hospital Aintree and Department of Surgery The Royal Liverpool University Hospitals Liverpool, UK

Michael D’Angelica MD Weill Medical College of Cornell University and Memorial Sloan-Kettering Cancer Center New York, New York, USA and

René Adam MD, PHD AP-HP Hôpital Paul Brousse Centre Hépato-Biliaire Villejuif, France

First published in 2003 by M. Dunitz Ltd, United Kingdom This edition published in 2010 by Informa Healthcare, Telephone House, 69-77 Paul Street, London EC2A 4LQ, UK. Simultaneously published in the USA by Informa Healthcare, 52 Vanderbilt Avenue, 7th floor, New York, NY 10017, USA. © 2011 Informa UK Ltd, except as otherwise indicated. No claim to original U.S. Government works. Reprinted material is quoted with permission. Although every effort has been made to ensure that all owners of copyright material have been acknowledged in this publication, we would be glad to acknowledge in subsequent reprints or editions any omissions brought to our attention. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, unless with the prior written permission of the publisher or in accordance with the provisions of the Copyright, Designs and Patents Act 1988 or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 0LP, UK, or the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA (http://www.copyright.com/ or telephone 978750-8400). Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. This book contains information from reputable sources and although reasonable efforts have been made to publish accurate information, the publisher makes no warranties (either express or implied) as to the accuracy or fitness for a particular purpose of the information or advice contained herein. The publisher wishes to make it clear that any views or opinions expressed in this book by individual authors or contributors are their personal views and opinions and do not necessarily reflect the views/opinions of the publisher. Any information or guidance contained in this book is intended for use solely by medical professionals strictly as a supplement to the medical professional’s own judgement, knowledge of the patient’s medical history, relevant manufacturer’s instructions and the appropriate best practice guidelines. Because of the rapid advances in medical science, any information or advice on dosages, procedures, or diagnoses should be independently verified. This book does not indicate whether a particular treatment is appropriate or suitable for a particular individual. Ultimately it is the sole responsibility of the medical professional to make his or her own professional judgements, so as appropriately to advise and treat patients. Save for death or personal injury caused by the publisher’s negligence and to the fullest extent otherwise permitted by law, neither the publisher nor any person engaged or employed by the publisher shall be responsible or liable for any loss, injury or damage caused to any person or property arising in any way from the use of this book. A CIP record for this book is available from the British Library. ISBN-13: 978-1-84184-693-4 Orders may be sent to: Informa Healthcare, Sheepen Place, Colchester, Essex CO3 3LP, UK Telephone: +44 (0)20 7017 5540 Email: [email protected] Website: http://informahealthcarebooks.com/ For corporate sales please contact: [email protected] For foreign rights please contact: [email protected] For reprint permissions please contact: [email protected]

Typeset by Exeter Premedia Services Printed and bound in the United Kingdom

Contents List of contributors Foreword Preface I

vii x xi

1 17

Margo Shoup and Jason W. Smith

3 Hepatic resection

166

C. Kahlert, R. DeMatteo, and J. Weitz

Robert Jones and Graeme J. Poston

2 Anatomy of the pancreas

154

Kaori Ito

17 Noncolorectal, nonneuroendocrine metastases

ANATOMY/IMAGING/SURGICAL TECHNIQUE 1 Surgical anatomy of the liver and bile ducts

16 Management of neuroendocrine tumor hepatic metastasis

18 Chemotherapy-associated hepatotoxicity

19 Thermal ablation of liver metastases 24

173

Martin Palavecino, Daria Zorzi, and Jean-Nicolas Vauthey

180

Samir Pathak and Graeme J. Poston

Ajay V. Maker and Michael D’Angelica

4 Ultrasound for HPB disorders

36

20 Resection for hepatocellular carcinoma

Duan Li and Lucy Hann

5 Liver surgery in elderly patients

46

Rajesh Satchidanand, Stephen W. Fenwick, and Hassan Z. Malik

53

21 Treatment of laparoscopically discovered gallbladder cancer

Gerardo Sarno and Graeme J. Poston

6 Small solitary hepatic metastases: when and how? David L. Bartlett and Yuman Fong

7 Managing complications of hepatectomy

63 73

Thilo Hackert, Moritz Wente, and Markus W. Büchler

9 Surgical complications of pancreatectomy

81

Steven C. Katz and Murray F. Brennan

10 Laparoscopy in HPB surgery

89

Nicholas O’Rourke and Richard Bryant

11 Cross-sectional imaging for HPB disorders (MRI and CT)

192

197

Jason K. Sicklick, David L. Bartlett, and Yuman Fong

Fenella K. S. Welsh, Timothy G. John, and Myrddin Rees

8 Pancreatic resection

ii. Primary

22 Liver transplantation for HCC: Asian perspectives Shin Hwang, Sung-Gyu Lee, Vanessa de Villa, and Chung Mao Lo

23 Non-surgical treatment of hepatocellular carcinoma

Lawrence H. Schwartz

216

Ghassan K. Abou-Alfa and Karen T. Brown

24 Resection of intrahepatic cholangiocarcinoma

223

Junichi Arita, Norihiro Kokudo, and Masatoshi Makuuchi

25 Transplantation for hilar cholangiocarcinoma 100

208

229

Julie K. Heimbach, Charles B. Rosen, and David M. Nagorney

26 Rare vascular liver tumors

233

Jan P. Lerut, Eliano Bonaccorsi-Riani, Giuseppe Orlando, Vincent Karam, René Adam, and the ELITA-ELTR Registry

II LIVER A. Malignant i. Metastases 12 Liver metastases: detection and imaging

109

Valérie Vilgrain, Ludovic Trinquart, and Bernard Van Beers

13 Surgery for metastatic colorectal cancer

27 Management of recurrent pyogenic cholangitis 118

René Adam and E. Hoti

14 Chemotherapy for metastatic colorectal cancer

135

Gerardo Sarno and Graeme J. Poston

242

W. Y. Lau and C. K. Leow

28 Liver abscess: amebic, pyogenic, and fungal

Derek G. Power and Nancy E. Kemeny

15 Multimodal approaches to the management of colorectal liver metastases

B. Benign

253

Purvi Y. Parikh and Henry A. Pitt

29 Benign solid tumors of the adult liver

261

Mark Duxbury and O. James Garden

148

30 Liver trauma

271

Timothy G. John, Myrddin Rees, and Fenella K. Welsh

v

CONTENTS 31 Portal hypertension

280

Michael D. Johnson and J. Michael Henderson

32 Liver transplantation for acute and chronic liver failure

A. Malignant 288

Vincent Kah Hume Wong and J. Peter A. Lodge

33 Benign cystic disease of the liver

301

Stephen W. Fenwick and Dowmitra Dasgupta

34 Management of hydatid disease of the liver

308

Adriano Tocchi

35 Surgical management of primary sclerosing cholangitis

324

401

Michael G. House and Keith D. Lillemoe

44 Cystic tumors of the pancreas

407

Peter J. Allen and Murray F. Brennan

414

Stephen N. Hochwald and Kevin Conlon

432

B. Benign 329

Hiromichi Ito and William R. Jarnagin

47 Acute pancreatitis

439

C. Ross Carter, A. Peter Wysocki, and Colin J. McKay

333

Yuji Nimura

48 Chronic pancreatitis

451

Jakob R. Izbicki, Oliver Mann, Asad Kutup, and Kai A. Bachmann

343

Nick Stern and Richard Sturgess

49 Pancreatic injury

463

Demetrios Demetriades, Beat Schnüriger, and Galinos Barmparas

B. Benign 39 Choledochal cyst detected in adulthood

43 Palliation of pancreas cancer

Jooyeun Chung, Lisa J. Harris, Hamid Abdollahi, and Charles J. Yeo

A. Malignant

38 Endoscopic management of malignant biliary obstruction

380

André L. Mihaljevic, Jörg Kleeff, and Helmut Friess

46 Rare tumors of the pancreas

III BILE DUCTS AND GALLBLADDER

37 Extrahepatic cholangiocarcinoma

42 Adenocarcinoma of the pancreas

45 Neuroendocrine pancreatic tumors

Jason A. Breaux and Steven A. Ahrendt

36 Management of advanced gallbladder cancer

IV PANCREAS

50 Pancreas transplantation 354

470

Khalid Khawaja

Bilal Al-Sarireh and Hassan Malik

40 Bile duct injuries and benign biliary strictures

360

Steven M. Strasberg

41 Gallstones and common bile duct stones—surgical and non-surgical approaches Matthew P. Dearing and Michael Rhodes

vi

V PEDIATRIC HPB DISORDERS 51 Pediatric HPB disorders

373

478

Maureen McEvoy and Michael P. La Quaglia

Index

489

List of contributors Ghassan K. Abou-Alfa MD Assistant Attending, Memorial Sloan-Kettering Cancer Center, and Assistant Professor, Weill Medical College at Cornell University, New York, New York, USA Hamid Abdollahi MD Senior Resident (General Surgery), Department of Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania, USA René Adam MD, PhD AP-HP Hôpital Paul Brousse, Centre Hépato-Biliaire, Inserm, Unité 785, and Université Paris-Sud, UMR-S 785, Villejuif, France Steven A. Ahrendt MD Associate Professor of Surgery, University of Pittsburgh Medical Center, UPMC Passavant Cancer Center, Pittsburgh, Pennsylvania, USA Peter J. Allen MD Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Bilal Al-Sarireh MBBCh, FRS, PhD Consultant Hepatopancreatobiliary and Laparoscopic Surgeon, Swansea University, and Department of Surgery, Morristown Hospital, Swansea, UK Junichi Arita MD, PhD Hepato-Biliary-Pancreatic Surgery Division, Artificial Organ and Transplantation Division, Department of Surgery, Graduate School of Medicine, University of Tokyo, Bunkyo-ku, Tokyo, Japan Kai A. Bachmann Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Galinos Barmparas Division of Trauma and Surgical Critical Care, University of Southern California, Los Angeles, California, USA David L. Bartlett Department of Surgery, University of Pittsburgh, Pittsburgh, Pennsylvania, and National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA Eliano Bonaccorsi-Riani Th. STARZL Abdominal Transplant Unit, Cliniques Universitaires St Luc Université catholique de Louvain, Department of Abdominal and Transplantation Surgery, Brussels, Belgium Jason A. Breaux MD Surgical Oncology Fellow, University of Pittsburgh Medical Center, UPMC Cancer Pavilion, Pittsburgh, Pennsylvania, USA Murray F. Brennan Benno C. Schmidt Clinical Chair in Oncology, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Karen T. Brown MD Attending Radiologist, Memorial Sloan-Kettering Cancer Center, and Professor of Clinical Radiology, Weill Medical College at Cornell University, New York, New York, USA

C. Ross Carter West of Scotland Pancreatic Unit, Glasgow Royal Infirmary, Glasgow, Scotland, UK Jooyeun Chung MD Department of Surgery, The Methodist Hospital, Houston, Texas, USA Kevin Conlon Professor of Surgery, University of Dublin, Trinity College Dublin, and Professorial Surgical Unit, Education Centre, AMNCH, Dublin, Ireland Michael D’Angelica MD Weill Medical College of Cornell University and Memorial Sloan-Kettering Cancer Center, New York, New York, USA Dowmitra Dasgupta MD, FRCS Consultant Hepato-Pancreatico-Biliary Surgeon, Department of Upper GI Surgery, Castle Hill Hospital, Cottingham, UK Matthew P. Dearing Department of Surgery, Norfolk & Norwich University Hospital, Norwich, UK R. DeMatteo Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Demetrios Demetriades Division of Trauma and Surgical Critical Care, University of Southern California, Los Angeles, California, USA Mark Duxbury Clinical Surgery, University of Edinburgh Royal Infirmary, Edinburgh, UK Stephen W. Fenwick MD, FRCS Consultant Hepatobiliary Surgeon, North Western Hepatobiliary Unit, University Hospital Aintree, Lower Lane, Liverpool, UK Yuman Fong MD Hepatobiliary Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Helmut Friess Chirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, Munich, Germany O. James Garden Regius Professor of Clinical Surgery, Clinical and Surgical Sciences (Surgery), University of Edinburgh, Royal Infirmary, Edinburgh, UK Thilo Hackert Department of Surgery, University of Heidelberg, Heidelberg, Germany Lisa J. Harris MD Senior Resident (General Surgery), Department of Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania, USA J. Michael Henderson Chief Quality Officer, Cleveland Clinic, Cleveland, Ohio, USA

Richard Bryant MBBS, FRACS Royal Brisbane Hospital, Brisbane, Queensland, Australia

Stephen N. Hochwald MD Chief, Division of Surgical Oncology, University of Florida, Gainesville, Florida, USA

Markus W. Büchler Department of General Surgery, University of Heidelberg, Heidelberg, Germany

Michael G. House MD Assistant Professor, Department of Surgery, Indiana University School of Medicine, Indianapolis, Indiana, USA

vii

LIST OF CONTRIBUTORS Lucy Hann MD Professor of Radiology, Weill Cornell Medical Center, and Director of Ultrasound Memorial Sloan-Kettering Cancer Center, New York, New York, USA Julie K. Heimbach Mayo Clinic, Rochester, Minnesota, USA Steven N. Hochwald University of Florida Medical School, Box 100286, Gainesville, FL 32610–0286, USA E. Hoti AP-HP Hôpital Paul Brousse, Centre Hépato-Biliaire, Villejuif, France, and Liver Transplant Unit, Saint Vincent’s University Hospital, Dublin, Ireland Shin Hwang Professor, Division of Hepatobiliary Surgery and Liver Transplantation, Department of Surgery, University of Ulsan College of Medicine, Seoul, Korea Hiromichi Ito MD Department of Surgery, Michigan State University, Lansing, Michigan, USA Kaori Ito MD Department of Surgery, Michigan State University, Lansing, Michigan, USA Jakob R. Izbicki FACS Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany William R. Jarnagin MD Hepatobiliary Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Timothy G. John MD, FRCSEd (Gen) Hepatobiliary Unit, Basingstoke and North Hampshire Hospitals NHS Foundation Trust, Basingstoke, UK Michael D. Johnson MD Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA Robert Jones MB, ChB, MRCS Clinical Fellow, North Western Hepatobiliary Centre, Aintree University Hospital, Liverpool, UK C. Kahlert Department of Surgery, University of Heidelberg, Heidelberg, Germany Vincent Karam Centre Hépatobiliaire, Hôpital Paul Brousse, Villejuif, France Steven C. Katz MD Director of Surgical Immunotherapy, Roger Williams Medical Center, Providence, Rhode Island, USA Khalid Khwaja MD Director of Kidney and Pancreas Transplantation, Senior Staff Surgeon, Lahey Clinic, Burlington, Massachusetts, USA Nancy E. Kemeny MD Memorial Sloan-Kettering Cancer Center, New York, New York, USA Jörg Kleeff Department of Surgery, Klinikum rechts der Isar, Technische Universität München, Munich, Germany Norihiro Kokudo MD, PhD Hepato-Biliary-Pancreatic Surgery Division, Artificial Organ and Transplantation Division, Department of Surgery, Graduate School of Medicine, University of Tokyo, Bunkyo-ku, Tokyo, Japan Asad Kutup Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

viii

W. Y. Lau Faculty of Medicine, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong, SAR C. K. Leow Mount Elizabeth Medical Centre, Singapore, Singapore Keith D. Lillemoe MD Jay L. Grosfeld Professor and Chairman, Department of Surgery, Indiana University School of Medicine, Indianapolis, Indiana, USA Sung-Gyu Lee Professor, Division of Hepatobiliary Surgery and Liver Transplantation, Department of Surgery, University of Ulsan College of Medicine, Seoul, Korea Michael P. La Quaglia MD Department of Surgery, Pediatric Surgery Service, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Jan P. Lerut MD, PhD, FACS Th. STARZL Abdominal Transplant Unit, Cliniques Universitaires St Luc Université catholique de Louvain, Department of Abdominal and Transplantation Surgery, Brussels, Belgium Duan Li MD Assistant Attending Radiologist, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Chung Mao Lo Professor, Department of Surgery, Queen Mary Hospital, The University of Hong Kong, Hong Kong, China J. Peter A. Lodge MD, FRCS Professor and Clinical Director, HPB & Transplant Unit, St. James’ University Hospital, Leeds, UK Ajay V. Maker MD Director of Surgical Oncology, Creticos Cancer Center–Advocate Illinois Masonic Medical Center; Departments of Surgery and Microbiology/Immunology, University of Illinois at Chicago, Chicago, Illinois, USA Masatoshi Makuuchi MD, PhD Hepato-Biliary-Pancreatic Surgery Division, Artificial Organ and Transplantation Division, Department of Surgery, Graduate School of Medicine, University of Tokyo, Bunkyo-ku, Tokyo, Japan Hassan Malik MD, FRCS Hepatobiliary Unit, Department of Surgery, University Hospital Aintree, Liverpool, UK Oliver Mann Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Maureen McEvoy MD Department of Surgery, Pediatric Surgery Service, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Colin J. McKay West of Scotland Pancreatic Unit, Glasgow Royal Infirmary, Glasgow, Scotland, UK André L. Mihaljevic Department of Surgery, Klinikum rechts der Isar, Technische Universität München, Munich, Germany David M. Nagorney Mayo Clinic, Rochester, Minnesota, USA Yuji Nimura MD President, Aichi Cancer Center, Chikusaku, Nagoya, Japan Giuseppe Orlando Th. STARZL Abdominal Transplant Unit, Cliniques Universitaires St Luc Université catholique de Louvain, Department of Abdominal and Transplantation Surgery, Brussels, Belgium

LIST OF CONTRIBUTORS Nicholas O’Rourke MBBS, FRACS Royal Brisbane Hospital, Brisbane, Queensland, Australia Martin Palavecino MD Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA Purvi Y. Parikh MD Department of Surgery, Albany Medical College, Albany, New York, USA Samir Pathak MD, ChB, MSC, MRCS Clinical Fellow, North Western Hepatobiliary Centre, Aintree University Hospital, Liverpool, UK Henry A. Pitt MD Indiana University, Indianapolis, Indiana, USA Graeme J. Poston MS, FRCS (Eng), FRCS (Ed) Centre for Digestive Diseases, University Hospital Aintree, and Department of Surgery, The Royal Liverpool University Hospitals, Liverpool, UK Derek G. Power MD Memorial Sloan-Kettering Cancer Center, New York, New York, USA Myrddin Rees MS, FRCS, FRCS (Ed) Hepatobiliary Unit, Basingstoke and North Hampshire Hospitals NHS Foundation Trust, Basingstoke, UK Michael Rhodes Department of Surgery, Norfolk & Norwich University Hospital, Norwich, UK Charles B. Rosen Mayo Clinic, Rochester, Minnesota, USA

Jason W. Smith MD Chief Resident, Department of Surgery, Loyola University Medical Center, Maywood, Illinois, USA Nick Stern Consultant Gastroenterologist, Digestive Diseases Department, University Hospital Aintree, Liverpool, UK Richard Sturgess Consultant Gastroenterologist and Clinical Director, Digestive Diseases Department, University Hospital Aintree, Liverpool, UK Adriano Tocchi Head of 1st Department of Surgery and Chief of the Gastro-intestinal and Hepato-biliary Surgical Service, University of Rome Sapienza Medical School, Rome, Italy Ludovic Trinquart Department of Radiology, Assistance-Publique Hôpitaux de Paris, Hôpital Beaujon, Clichy, France Bernard Van Beers Department of Radiology, Assistance-Publique Hôpitaux de Paris, Hôpital Beaujon, Clichy; Université Paris; and Centre de recherche biomédicale Bichat-Beaujon, Paris, France Jean-Nicolas Vauthey MD Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA Valérie Vilgrain Department of Radiology, Assistance-Publique Hôpitaux de Paris, Hôpital Beaujon, Clichy; Université Paris; and Centre de recherche biomédicale Bichat-Beaujon, Paris, France

Gerardo Sarno MD Clinical Fellow, North Western Hepatobiliary Centre, Aintree University Hospital, Liverpool, UK

Vanessa de Villa Assistant Professor, Department of Surgery, Queen Mary Hospital, The University of Hong Kong, Hong Kong, China

Rajesh Satchidanand MD, FRCS Clinical Fellow, North Western Hepatobiliary Centre, Aintree University Hospital, Liverpool, UK

J. Weitz MD Department of Surgery, University of Heidelberg, Heidelberg, Germany

Beat Schnüriger Division of Trauma and Surgical Critical Care, University of Southern California, Los Angeles, California, USA Lawrence H. Schwartz Department of Radiology, Columbia University College of Physicians and Surgeons, and Radiologist-in-Chief, New York–Presbyterian Hospital/ Columbia University Medical Center, New York, New York, USA Margo Shoup MD, FACS Chief, Division of Surgical Oncology, Department of Surgery, Loyola University Medical Center, Maywood, Illinois, USA Jason K. Sicklick Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA Steven M. Strasberg MD, FRCS(C), FACS, FRCS (Ed) Pruett Professor of Surgery and Head Hepato-Pancreato-Biliary and Gastrointestinal Surgery, Washington University in Saint Louis and Barnes-Jewish Hospital, Saint Louis, Missouri, USA

Fenella K. S. Welsh MA, MD, FRCS (Gen Surg) Hepatobiliary Unit, Basingstoke and North Hampshire Hospitals NHS Foundation Trust, Basingstoke, UK Moritz Wente Department of Surgery, University of Heidelberg, Heidelberg, Germany Vincent Kah Hume Wong MBCB, MRCS Research Fellow in Hepatopancreatobiliary & Transplant Surgery, HPB & Transplant Unit, St. James’ University Hospital, Leeds, UK A. Peter Wysocki Department of Surgery, Logan Hospital, Meadowbrook, Queensland, Australia Charles J. Yeo MD The Samuel D. Gross Professor and Chair, Department of Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Daria Zorzi MD Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA

ix

Foreword As recent progress in hepato-pancreato-biliary (HPB) surgery has been evident since the first edition of this book was published eight years ago, Dr. Graeme Poston, Dr. Mike D’Angelica, and Dr. René Adam, internationally recognized authorities in HPB surgery, have attempted to rewrite the second edition, joined by selected numerous worldwide specialists renowned as expert authors in each field to present a current view of the surgical and non-surgical management of benign and malignant HPB disorders. This book demonstrates the wisdom of the new knowledge and technical skills of these diverse disciplines where cooperative

x

efforts contribute toward the benefit of the patients with HPB disorders. The general surgeon will find this volume to be a useful source of current thoughts on how to manage the diverse HPB diseases. Yuji Nimura MD President, Aichi Cancer Center Professor Emeritus, Nagoya University Graduate School of Medicine Past President, International Hepato-Pancreato-Biliary Association (IHPBA)

Preface Hepato-pancreato-biliary (HPB) surgery is now firmly established within the repertoire of modern general surgery. Indeed, in many major tertiary centers there are now specific teams for both pancreatic and liver surgery. However, in most hospitals outside these major centers the day-to-day management and decision-making for patients with these disorders remains the remit of the general surgeon. Following the launch of the highly successful first edition of this book eight years ago there have been considerable advances in the surgical management of HPB disorders. Many of these relate to related specialties (radiology, oncology, gastroenterology, and anesthesia) and also directly to surgery (liver transplantation, caval bypass and replacement, laparoscopic surgery to name but a few). As such the second edition has been completely rewritten from scratch. As with the first edition, the purpose of this edition is twofold. First, it is intended to cover the spectrum of common

HPB diseases that will confront the general surgeon in his or her regular practice. Second, we hope that this work will be sufficiently comprehensive to cover the broad spectrum of HPB surgery for candidates coming to examinations at the completion of surgical training. We are indebted to the many international contributors for their perseverance and patience over the gestation of this project, which is greatly appreciated. Lastly, we are grateful to our publishers, Informa Healthcare, for their help during the preparation of this project. Graeme J. Poston Michael D’Angelica René Adam September 2010

xi

1

Surgical anatomy of the liver and bile ducts Robert Jones and Graeme J. Poston

The success of any surgical intervention on the liver and bile ducts is totally dependent on a thorough working knowledge of their anatomy. As the number of patients undergoing hepatobiliary surgery is increasing, good understanding of the anatomy of this area is increasingly important for any surgeon with an interest in the gastrointestinal tract. Command of this anatomy is also essential for the successful interpretation of functional imaging of hepatobiliary anatomy. When operating on the liver and biliary tree, the surgeon has to obey three basic tenets. ● ●

●

Remove all pathologically involved tissue. Preserve the maximal amount of functioning nonpathological liver tissue. Perform safe resection, while ensuring adequate blood supply to the remaining hepatic parenchyma.

Historically, the liver was described according to its morphological appearance (1,2). However, these three tenets have altered the approach to surgery, and the liver is now considered from a functional and therefore surgical perspective.

morphological anatomy Historically, when viewed at laparotomy, the liver appears divided into a larger “right” lobe, and a smaller “left” lobe by the umbilical fissure and falciform ligament (Figs. 1.1 and 1.2) (3). Situated on the inferior surface of the right lobe is the transverse hilar fissure, which constitutes the posterior limit of the right lobe. The “quadrate” lobe was defined as the portion of the right lobe lying anterior to this transverse hilar fissure and to the right of the umbilical fissure, its other margin being defined by the gallbladder fossa. The “caudate” lobe, which is anatomically and functionally separate from the rest of the liver, lies posterior to the hilum, between the portal vein and the inferior vena cava (IVC) (4). This historical anatomical approach does not consider the vasculature or biliary drainage of the liver and is of only limited use when planning surgical resection.

early application of the functional anatomy Isolated liver wounds, usually as a result of military action, had been successfully treated since the early seventeenth century (5,7), but the first attempt at resection of a liver tumor was not made until 1886, when the French surgeon Luis excised a solid liver tumor by ligating and cutting through a pedunculated left lobe “adenoma.” Attempts to suture the severed pedicle were unsuccessful, and the stump was returned to the peritoneal cavity. Not surprisingly, the patient succumbed some six hours later (8). In 1888, Rex reported a “new” arrangement of the right and left lobes of the liver and further refined our understanding of

lobar anatomy (2). The first successful elective liver resection was performed two years later by von Langenbuch, who excised a portion of the left lobe of the liver containing an adenoma in 1888 (9). He had to reopen the abdomen several hours after the operation because of reactionary hemorrhage, but was able to ligate the bleeding vessels and return the oversewn liver to the abdomen. Two years later in 1890, the Baltimore surgeon McLane Tiffany reported the successful removal of a benign liver tumor (10), and the following year Lucke described the successful resection of a cancerous growth of the liver (11). Surgery was now becoming a recognized treatment for liver pathology. Advances in surgery closely mirrored increased understanding of the functional anatomy of the liver (12–14). The first attempt to define the functional anatomy of the liver, which could possibly guide current surgical practice, was made by Cantlie in 1898, while working in Hong Kong. He dissected the livers of executed prisoners (15) and making vascular casts, he demonstrated that the main division between the right and left lobe in fact extended from approximately the gallbladder fossa, to the right side of the IVC, posterosuperiorly. Cantlie’s line, therefore, follow a line drawn from the gallbladder fossa, along the middle hepatic vein, to the IVC (Figs. 1.2 and 1.3) (3). In 1911, Wendel reported the first case of right lobectomy for a primary tumor (16), however this procedure did not follow the precise anatomical plane described by Cantlie. In 1939, while working in Paris, the Vietnamese surgeon Ton That Tung described the venous drainage of the liver in relation to the true lobar anatomy (Fig. 1.4) (17). The first anatomically correct description of a left lateral segmentectomy was made by Raven in 1948 while resecting metastatic colon cancer (18). Four years later, Lortat-Jacob and Robert finally described a similar approach to the true right hepatic lobectomy, based on the anatomical principles described by Cantlie (Fig. 1.6) (19). Healey and Schroy were the first to demonstrate in 1953 that the right lobe was further divided into an anterior and a posterior sector (20). They also showed that the left lobe was divided into a medial and lateral sector by the line of the falciform ligament and umbilical vein (Fig. 1.5). Understanding of the functional anatomy of the liver continued to develop, and in 1957, Goldsmith and Woodburne described a number of anatomical planes through the liver parenchyma that followed this functional anatomy. Their paper finally defined true right lobectomy (right hepatectomy), left lobectomy (left hepatectomy), and left lateral segmentectomy (Fig. 1.6) (21).

appreciation of segmental anatomy Probably the most important anatomical contribution to modern liver surgery comes from the work of the late Claude

1

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS IVC

Middle hepatic vein lying among Cantlie's line

IVC

Left lobe Right lobe

IVC

IVC Right free border of lesser omentum

Cantlie's line

Figure 1.1 Morphological anatomy.

Gallbladder

Figure 1.3 Cantlie’s line.

Umbilical fissure

Cantlie's line

Gallbladder

Quadrate lobe

supply (inflow and outflow), and therefore viability, to the remaining hepatic parenchyma. The description of Couinaud is the most complete and exact, and also the most useful for the operating surgeon, and therefore it is this description that will be used throughout this book.

segmental anatomy of the liver Transverse hilar fissure Gastrohepatic omentum

Common bile duct, hepatic artery and portal vein

IVC Caudate lobe

Figure 1.2 Anatomical features.

Couinaud, who in 1957 produced a huge number of vasculobiliary casts of the liver (23,24). Couinaud was able to demonstrate that the liver appeared to consist of eight anatomical segments, each of which could potentially be separately resected without affecting the physiological viability of the other segments. Couinaud redefined the caudate lobe as segment 1 and Goldsmith and Woodburne’s left lobe as segments 2 and 3. The quadrate lobe was termed segment 4, and more recently has been subdivided by further studies of its portal blood supply into 4A (superiorly) and 4B (inferiorly). The right liver consists of segments 5 (anteroinferiorly), 6 (posteroinferiorly), 7 (posterosuperiorly), and 8 (anterosuperiorly) (Fig. 1.7). Couinaud later suggested a further clarification, in which the caudate lobe to the left of the IVC remained segment 1, with that to the right being redefined as segment 9 (25). Resections based on these anatomical segments enable the surgeon to safely operate following the three central tenets described above; remove all pathologically involved tissue, preserve the maximal amount of nonpathological liver tissue, and perform safe resection, while ensuring an adequate blood

2

These anatomical studies of the functional anatomy of the liver allow us to define hepatic segments based upon both the distribution of the portal pedicles and the drainage of the hepatic veins (Fig. 1.5). The three main hepatic veins (right, middle, and left) divide the liver into four sectors, each of which receives a portal pedicle containing branches of the hepatic artery, hepatic duct, and portal vein; thus producing an alternation between hepatic veins and portal pedicles. These four sectors, demarcated by the hepatic veins, are the portal sectors, each sector therefore receiving an independent portal supply. For the same reason, the scissurae containing the hepatic veins are termed the portal scissurae while the scissurae containing portal pedicles are the hepatic scissurae (Fig. 1.5). Thus, the liver is divided by the main portal scissura along the line of the middle hepatic vein into two discrete hemilivers, along the line previously described by Cantlie (15). We therefore refer to these hemilivers as right and left livers, rather than right and left lobes, to avoid confusion with the anatomical lobes, particularly since there is no visible surface marking that permits individualization of the “true” lobes. As described by Cantlie, the main portal scissura runs posteriorly from the middle of the gallbladder fossa to the right side of the IVC (Fig. 1.5). Therefore, the right and left livers, demarcated by the main portal scissura, are independent in terms of their portal and arterial vascularization and their biliary drainage. These right and left livers are both further divided into two by the other two portal scissurae, delineated by the right and left hepatic veins. Goldsmith and Woodburne refer to these further divisions as “segments” (21), but for the rest of this book, we will use the more generally accepted nomenclature of Couinaud, which refers to these divisions as “sectors” (23). The

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS Right liver

Left liver

IVC

Middle hepatic vein (usually enters left vein before IVC)

Left heptic vein

Right hepatic vein

Caudate hepatic veins (variable)

Right inferior hepatic vein (variable) IVC Gallbladder, note that the middle vein may lie superficially in the gallbladder fossa Figure 1.4 Venous drainage of the liver.

IVC

Middle hepatic vein in main portal scissura following Cantlie's line

7

2

Right hepatic vein in right portal scissura

8

Left hepatic vein in left portal scissura Lateral segment of left lobe

3 1 4

Right posterior sector

Falciform ligament 6

5 Medial segment of left lobe Right anterior sector Right liver

Portal vein Left liver

Figure 1.5 Functional sectoral anatomy and relationship to hepatic scissurae.

right liver is divided by the right portal scissura (right portal vein) into an anteromedial (or anterior) sector containing segments 5 inferiorly and 8 superiorly, and a posterolateral (or posterior) sector containing segments 6 inferiorly and 7 superiorly (Fig. 1.5). When the liver lies in its normal position within the upper abdominal cavity, the right posterolateral sector lies directly behind the right anteromedial sector, and this scissura is therefore almost in the coronal plane. Therefore in the clinical setting (particularly when imaging the liver), it is better to speak of these anterior and posterior sectors (Fig. 1.5). The exact location of the right portal scissura is imprecise, because it has no external landmarks. According to Couinaud (23), it extends from the edge of the liver at the middle point between the back of the liver and the right side of the

gallbladder bed along the right hepatic vein posteriorly to the confluence of the right hepatic vein and the IVC (26–28). The venous drainage of the right liver is variable in that, in addition to the right and middle hepatic veins, there are often a number of smaller hepatic veins draining directly into the IVC from segments 6 and 7. Not infrequently (63–68%) segment 6 drains directly into the IVC through a distinct inferior right hepatic vein, larger than these other venous tributaries to the IVC, which can be a significant bonus in the preservation of residual hepatic function when undertaking extended left hepatectomies (Fig. 1.4) (29,30). The left portal scissura, along the left hepatic vein, divides the left liver into two sectors: an anterior sector containing segments 3 and 4 and a posterior sector containing segment 2

3

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

(A)

(B)

(C)

(D)

(E) Figure 1.6 Formal hepatectomies: (A) right hepatectomy; (B) left hepatectomy; (C) left lateral segmentectomy; (D) extended left hepatectomy; (E) extended right hepatectomy.

2

8 8

7

7

1

2

3 1 5

4

3

4 5

6 6

(A)

(B)

Figure 1.7 Functional division of the liver and of the liver segments according to Couinaud’s nomenclature (A) as seen in the patient and (B) in the ex vivo position.

4

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS (Fig. 1.5). It is important to note that the left portal scissura does not follow the umbilical fissure; this portal scissura contains a hepatic vein and the umbilical fissure contains a portal pedicle. Therefore the left portal scissura lies posterior to the ligamentum teres, inside the left lobe of the liver (Fig. 1.5). The middle hepatic vein (defining the main portal scissura) usually enters the left hepatic vein some 1 to 2 cm before the left hepatic vein joins the IVC (Fig. 1.4) (30). Occasionally the middle and left hepatic veins enter the IVC separately, and in 2 out of 34 of Couinaud’s casts, the middle vein and left veins joined at more than 2.5 cm from the IVC (30). Such an anomaly must be detected and excluded during isolated resection of segment 4, since if it is not seen, and the last 2 cm of the left vein is damaged, segments 2 and 3 will be needlessly sacrificed (and in the case of extended right hepatectomy, threaten future remnant liver viability). The caudate lobe (segments 1 and 9) is the dorsal portion of the liver, lying posteriorly and surrounding the retrohepatic IVC. It lies directly between the portal vein (anteriorly) and the IVC (posteriorly). The main bulk of the caudate lobe lies to the left of the IVC, with its left and inferior margins being free in the lesser omental bursa (Fig. 1.2). The gastrohepatic (lesser) omentum separates the caudate from segments 2 and 3 of the left liver. The left portion of the caudate lobe lies inferior to the right between the left portal vein and the IVC, as the caudate process. This process then fuses inferiorly with segment 6 of the right liver. The amount of caudate lobe that lies on the right side is variable, but usually small. The anterior surface of the caudate lobe lies within the hepatic parenchyma against the posterior intrahepatic surface of segment 4, demarcated by an oblique plane slanting from the left portal vein to the left hepatic vein. The caudate lobe must be considered functionally as an isolated autonomous segment, since its vascularization is independent of the portal division and of the three main hepatic veins. It receives a variable arterial and portal blood supply from both the right and left portal structures, although the right caudate lobe consistently receives an arterial supply from the right posterior artery. Biliary drainage is likewise into both the right and left hepatic ducts. However, the left dorsal duct can also join the segment 2 duct. The small hepatic veins of the caudate lobe drain directly into the IVC. This independent functional isolation of the caudate lobe is clinically important in Budd–Chiari syndrome; if all three main hepatic veins are obliterated, the only functioning hepatic venous drainage is through the caudate lobe, which therefore undergoes compensatory hyperplasia.

anatomical classification of hepatectomies Hepatic resections can be classified as “anatomical” and “nonanatomical.” Anatomical hepatectomies (hepatectomies reglees) are defined by resection of a portion of liver parenchyma defined by the functional anatomy. These resections are called left or right hepatectomies, sectorectomies, and segmentectomies. Nonanatomical hepatectomies involve resection of a portion of hepatic parenchyma not limited by anatomical scissurae. Such resections are usually inappropriate,

as they will leave behind devascularized residual liver and will also probably not adequately excise all the pathologically involved parenchyma. The usual anatomical hepatectomies can be considered in two groups: right and left hepatectomies in which the line of transection is the main portal scissura separating the right and left livers along the middle hepatic vein, and right and left hepatectomies in which the line of transection commences in the umbilical fissure. For some time the latter definition, initially proposed by Goldsmith and Woodburne (21), has been the accepted convention. We would encourage the use of the former definition, as segment 4 (quadrate lobe) is anatomically part of the left liver (Fig. 1.9), and this convention was adopted universally at the 2000 Brisbane Congress of the IHPBA (Brisbane Convention), and will be used hereafter in this book. Using this functional approach to liver anatomy, we can define numerous potential liver resections based upon the order (first, second, third) of the hepatic divisions (main portal scissura, anterior and posterior right portal scissurae, left portal scissura) (28). With regard to the first order division, right hepatectomy or hemihepatectomy (removal of the right liver/hemiliver) therefore consists of the resection of segments 5 to 8 (stipulating ± segment 1). Left hepatectomy or hemihepatectomy (removal of the left hemiliver or liver) is the removal of segments 2–4 (stipulating ± segment 1) (Fig. 1.6). In certain pathologies (multiple liver metastases or large tumors transgressing the main portal scissura) hepatectomies can be extended to include adjacent segments and sectors of the other liver. Therefore extended right hepatectomy (right trisegmentectomy or extended right hemihepatectomy) will also include resection of segment 4 (stipulating ±segment 1), taking portal structures to the right of the falciform ligament (Fig. 1.6). Similarly, extended left hepatectomy (left trisegmentectomy or extended left hemihepatectomy) would include resection of segments 5 and 8 en bloc with segments 2 to 4 (stipulating ± segment 1) (Fig. 1.6). When discussing second order divisions, individual sectors can be resected in isolation or in adjacent pairs depending upon the distribution of pathology. Therefore right anterior sectionectomy refers to the en bloc resection of segments 5 and 8 (between the main portal scissura (middle hepatic vein) and right portal scissura (right portal vein) on their pedicle of the anterior division of the right portal vein). Right posterior sectionectomy (previously referred to as right posterior or lateral sectorectomy) is the contiguous resection of segments 6 and 7, posterior to the right portal scissura (on the pedicle of the posterior division of the right portal vein) (Fig. 1.8). On the left side, isolated excision of segment 4 can be described as left median sectionectomy, although it is also legitimate to refer to it as resection segment 4 or segmentectomy 4. One area of confusion in these definitions of hepatectomies comes in the simultaneous resection of segments 2 and 3 (Fig. 1.10). Goldsmith and Woodburne originally described this procedure as a left hepatic lobectomy (21). Describing this as left lateral segmentectomy is technically wrong since the true left lateral segment (and sector) comprises no more than segment 2 (excision of which in isolation can therefore be

5

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS described as left lateral or posterior sectorectomy). It is now accepted convention that resection of segments 2 and 3 is regarded as a left lateral sectionectomy (but can also legitimately be referred to as bisegmentectomy 2–3). With regard to the third order divisions, resection is now at the level of the individual hepatic segment(s). Therefore these resections are referred to as segmentectomy (classified according to the segment being removed: 1–9). Similarly, segments 5 and 6 can be resected en bloc (and this used to be described as a right inferior hepatectomy) and this should now be described as bisegmentectomy 5–6. If there is a significant right inferior hepatic vein draining segments 5 and 6, then segments 7 and 8 can be resected with the right hepatic vein (bisegmentectomy 7–8) (Fig. 1.8).

surgical approach to the caudate lobe This resection (segmentectomy 1 or 9, or 1 and 9 en bloc) is initially achieved by dissection of the coronary ligament up to the right of the IVC, being careful to avoid the right hepatic vein. The falciform ligament is then dissected to the IVC, the

lesser omentum being incised close to the liver. Opening the left coronary ligament allows ligation of the inferior phrenic vein. The caudate veins running directly to the IVC are now exposed and can be divided between ligatures as they run up the back of the caudate lobe. After the hilar plate is lowered to expose the right and left portal pedicles, the portal inflow to both the right and left caudate segments can be identified, ligated, and divided. The caudate lobe is now isolated and the main portal fissure is divided to separate segments 4, 7, and 8. Note that the caudate segment 1s not defined macroscopically from segment 6.

the biliary tract Accurate biliary exposure and precise dissection are the two most important steps in any biliary operative procedure and are both totally dependent on a thorough anatomical understanding of these structures. Several authors have described the anatomy of the biliary tract (17,22,23), but unfortunately the surgical implications have been incompletely described and continue to be misunderstood by many surgeons.

(A)

(B)

(C)

(D)

(E) Figure 1.8 Other hepatic sectorectomies: (A) right posterior sectorectomy; (B) right anterior sectorectomy; (C) left medial sectorectomy (segments 4A and 4B); (D) right inferior hepatectomy; (E) right superior hepatectomy.

6

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS

intrahepatic biliary anatomy The right liver and left liver are respectively drained by the right and the left hepatic ducts. The caudate lobe (segments 1 and 9) is drained by several ducts joining both the right and left hepatic ducts (20). The intrahepatic ducts are tributaries of the corresponding hepatic ducts, which form part of the major portal tracts invaginating Glisson’s capsule at the hilus and penetrating the liver parenchyma (Fig. 1.11). There is variation in the anatomy of all three components of the portal triad structures (hepatic ducts, hepatic arteries, and portal vein), but it is the portal vein that shows the least anatomical variability. In particular, the left portal vein tends to be consistent in location (23). Bile ducts are usually located above the portal vein whereas the corresponding artery will lie below. Each branch of the intrahepatic portal vein corresponds to one or two intrahepatic bile ducts, which converge outside the liver to form the right and left hepatic ducts, in turn joining to form the common hepatic duct. The left hepatic duct drains segments 2, 3, and 4, which constitute the left liver. The duct draining segment 3 is found a little behind the left horn of the umbilical recess, from where it passes directly posteriorly to join the segment 2 duct to the left

of the main portal branch to segment 2. At this point, the left branch of the portal vein turns forward and caudally in the recessus of Rex (23) (Figs. 1.12 and 1.13). As the duct draining segment 3 begins its posterior course it lies superficially in the umbilical fissure, often immediately under Glisson’s capsule. As such it is usually easily accessible at surgery to allow a biliary– enteric (segment 3 hepaticojejunostomy) anastomosis for biliary drainage if such access is not possible at the porta hepatis. The left hepatic duct then passes beneath the left liver at the posterior base of segment 4, lying just above and behind the left branch of the portal vein. After the left duct crosses the anterior edge of that vein it joins the right hepatic duct to form the common duct at the hepatic ductal confluence. In this transverse portion, where it lies below the liver parenchyma, it receives one to three small branches from segment 4 (23). The right hepatic duct (Fig. 1.14) drains segments 5 to 8 and arises from the convergence of the two main sectoral (anterior 5 and 8, and posterior 6 and 7) tributaries. The right posterior sectoral duct runs almost horizontally (26) and comprises the confluence of the ducts from segments 6 and 7 (Fig. 1.15). The right posterior duct joins the right anterior sectoral duct (formed by the confluence of the ducts from segments 5 and 8)

Figure 1.9 Completion of segment 4 resection with portal bifurcation lying inferiorly in front of the inferior vena cava.

Figure 1.11 Exposing the hilar plate by raising the inferior surface of segment 4B, thus demonstrating the condensation of Glisson’s capsule, which will cover the extra hepatic confluence of the right and left hepatic ducts.

Figure 1.10 Left lateral segmentectomy immediately prior to division of the portal structure lying inferiorly and the left hepatic vein lying superiorly.

Figure 1.12 Exposing the recessus of Rex by distraction of the falciform ligament to demonstrate the bifurcation of segment 3 and segment 4 bile ducts.

7

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS 4 RPV

2

RHD RHA

4 (ant.) 3

CHD

PV

HA

Recessus of Rex

Figure 1.13 Biliary and vascular anatomy of the left liver. Note the position of segment 3 duct above the corresponding vein and its relationship to the recessus of Rex.

as it descends vertically (26). This right anterior sectoral duct lies to the left of the right anterior sectoral branch of the intrahepatic portal vein as it ascends within the parenchyma (Fig. 1.15). The junction of the two main right biliary ducts usually occurs immediately above the right branch of the portal vein (23). The right hepatic duct is considerably shorter than its counterpart on the left, which it joins to form the common hepatic duct in front of the right portal vein (Fig. 1.15). The caudate lobe (segments 1 and 9) has its own separate biliary drainage. This segment comprises two anatomically and functionally distinct portions, a caudate lobe proper (which consists of a right and left part) located at the posterior aspect of the liver, and a caudate process passing behind the portal structures to fuse with segment 6 of the right liver. In nearly half of individuals, three separate bile ducts drain these distinct parts, while in a quarter of individuals, there is a common biliary duct between the right portion of the caudate lobe proper and the caudate process, while the left part of the caudate lobe is drained by an independent duct. However, the site of drainage of these ducts is variable. Most authors advocate en bloc resection of the caudate lobe during resection of hilar cholangiocarcinoma (31), since the tumor usually infiltrates these ducts draining the caudate lobe. Certainly these authors have demonstrated that in 88% of cases of hilar cholangiocarcinoma coming to resection there is histological evidence of tumor infiltration of the caudate lobe along these ducts.

extrahepatic biliary anatomy

Figure 1.14 Demonstration of the right hepatic duct lying within the gallbladder fossa.

The detail of this section will be confined to the upper part of the extrahepatic biliary tree, above the common bile duct, since the common bile duct is also covered in chapter 2. The right and left hepatic ducts converge at the right of the hilum of the liver, anterior to the portal venous bifurcation and overlying the origin of the right portal vein. The biliary confluence

Anterior sectoral duct

8

5 7 Posterior sectoral duct

LHD LPV 6 LHA CHD

HA PV

Figure 1.15 Biliary and vascular anatomy of the right liver. Note the horizontal course of the posterior sectoral duct and the vertical course of the anterior sectoral duct.

8

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS is separated from the posterior aspect of the base of segment 4 by a fusion of connective tissue investing from Glisson’s capsule to form the fibrous hilar plate. This hilar plate has no vascular interposition and, when opened behind the posterior aspect of the base of segment 4, will display the extrahepatic confluence of the right and left hepatic ducts (Fig. 1.16). The main bile duct is divided into its upper part, the common hepatic duct, and lower part, the common bile duct, by the entry of the cystic duct from the gallbladder. This point of confluence of hepatic and cystic ducts to form the common bile duct is widely variable, and any surgeon performing the operation of cholecystectomy has a duty of care to their patient to be fully aware of this anatomic variability (lest they mistake the common bile duct, or less frequently the common or right hepatic ducts for the cystic duct, resulting in catastrophic consequences for the patient). The main bile duct normally has a diameter of up to 6 mm and passes downward anterior to the portal vein in the right free border of the lesser omentum. The bile duct is closely related to the hepatic artery as it runs upwards on its left side before dividing into its left and right branches, the right hepatic artery usually passing posteriorly to the bile duct. The cystic artery, which usually arises from the right hepatic artery, crosses the common hepatic duct as frequently anteriorly as it does posteriorly (Figs. 1.17 and 1.18). Calot’s triangle was originally defined by the common hepatic duct lying medially, inferiorly by the cystic duct and superiorly by the cystic artery (32). However, the usually accepted surgical definition of this triangle has been modified to that of the “cholecystectomy” triangle, which defines the

upper border as the inferior surface of the liver (and therefore contains the cystic artery) (33). The junction of the cystic duct and common hepatic duct varies widely and may even occur behind the pancreas. The retropancreatic portion of the bile duct approaches the duodenum obliquely, accompanied by the terminal part of the duct of Wirsung (see chap. 2). These two ducts join to enter the duodenum through the sphincter of Oddi at the papilla of Vater (34,35).

gallbladder and cystic duct The gallbladder lies within the cystic fossa on the underside of the liver in the main liver scissura, thereby defining the junction between the right and left hemilivers. It is separated from the hepatic parenchyma by the cystic plate, which is an extension of connective tissue from the hilar plate (described previously). The anatomical relationship of the gallbladder to the liver ranges from hanging by a loose peritoneal reflection to being deeply embedded within the liver parenchyma. The gallbladder varies in size and consists of a neck, body, and fundus, which usually reaches the free edge of the liver, still closely applied to the cystic plate. Large gallstones impacting within the neck of the gallbladder may create a Hartmann’s pouch (33), and inflammation secondary to this can obscure the anatomical plane between the gallbladder and the common hepatic duct (thus obliterating the cholecystectomy triangle). This degree of inflammation can make dissection during cholecystectomy difficult, increasing the risk of damage to the common hepatic duct (36). Other structures similarly threatened during this dissection as part of cholecystectomy for

Segment 4 Glisson's capsule Lig.teres

RHD RHA

LPV LHD LHA

RPV Cystic artery Cystic duct Gallbladder CHD

Umbilical fissure

HA

Line of incision of hilar plate to expose left hepatic duct CBD

Retroduodenal artery Gastroduodenal artery Splenic vein

Cystic plate

Hilar plate

Figure 1.16 Demonstration of the relationship between the posterior aspect of the base of segment 4 and the biliary confluence. Note the extension of Glisson’s capsule to invest the portal structures at the hilum (hilar plate) and extending over the hepatic surface of the gallbladder (cystic plate). Exposure of the extrahepatic left hepatic duct is achieved by incising the hilar plate at the base of segment 4 medially as far as the umbilical fissure.

Superior mesenteric artery and vein Figure 1.17 Anterior aspect of biliary anatomy. Note the hepatic duct confluence anterior to the right hepatic artery and origin of the right portal vein. Note also the course of the cystic artery, arising from the right hepatic artery and passing posteriorly to the common hepatic duct.

9

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS chronic cholecystitis include the right hepatic artery (in up to 50% of cholecystectomy bile duct injuries, so rendering the upper bile duct ischemic with ramifications for the timing of bile duct repair), the right hepatic duct, and in exceptional circumstances, a low-lying middle hepatic vein lying superficially just below the gallbladder fossa.

(A)

(B)

(D)

(E)

(G)

biliary anomalies

(C)

(F)

(H)

Figure 1.18 The eight most common variations in the anatomy of the arterial supply (cystic artery) to the gallbladder.

(A)

The cystic duct arises from the neck of the gallbladder and in 80% of people descends to join the common hepatic duct in its supraduodenal course. Its length varies widely but its luminal diameter is usually between 1 and 3 mm. The mucosa of the cystic duct is arranged in spiral folds (valves of Heister) (33). In a small number of cases, the cystic duct joins the right hepatic duct or occasionally a right hepatic sectoral duct. The gallbladder receives its blood supply by the cystic artery, the anatomy of which varies widely (Fig. 1.18). The most common variant arises directly from the right hepatic artery, then dividing into an anterior and posterior branch. The venous drainage of the gallbladder is directly through the gallbladder fossa to the portal vein in segment 5 (Fig. 1.19).

The biliary anatomy described above, comprising a right and left hepatic duct joining to form a common hepatic duct occurs in between 57% (23) and 72% (8) of cases. This variance may be explained by Couinaud’s (23) description of a triple confluence of right posterior sectoral duct, right anterior sectoral duct, and left hepatic duct in 12% of cases, which Healey and Schroy do not describe. There are many other abnormalities in biliary anatomy. Couinaud described a right sectoral duct joining the main bile duct in 20% of individuals (right anterior sectoral in 16%, right posterior sectoral in 4%). In addition, a right sectoral duct (posterior in 5%, anterior in 1%) may join the left hepatic duct in 6% of cases. In 3% of cases, there is an absence of a defined hepatic duct confluence with all the sectoral ducts joining separately and in 2% the right posterior sectoral duct may join the neck of the gallbladder or be entered by the cystic duct (23) (Fig. 1.20). Similarly, there are common variations of the intrahepatic biliary anatomy. Healey and Schroy (20) describe the classical intrahepatic biliary arrangement outlined above in 67% of

(B) Figure 1.19 (A) Venous drainage of the gallbladder. (B) The lymphatic drainage of the gallbladder towards the coeliac axis.

10

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS cases, with ectopic drainage of segment 5 in 9%, segment 6 in 14%, and segment 8 in 20% of the cases. In addition, they describe a subvesical duct in 20% to 50% of the cases (8,37). This subvesical duct may lie deeply embedded in the cystic plate and can join either the common or right hepatic ducts. This duct does not drain any specific area of the liver and never communicates with the gallbladder, but may be damaged during cholecystectomy and therefore contribute to postoperative biliary leak. On the left side, the commonest anomaly is a common union of ducts of segments 3 and 4 (25% of cases), and in only 2% does the segment 4 duct independently join the common hepatic duct (Fig. 1.21). Gross described a number of anomalies of the accessory biliary apparatus in 1936 (38). These include bilobed and

ra

ra

rp

Ih

Ih

rp

12%

57% (A)

(B) ra

ra

Ih rp

Ih rp

20%

4%

16% C2

C1 (C) ra

ra rp

rp

Ih

Ih

1%

5%

6%

D2

D1 (D)

4 4

ra

3

ra

rp 2

2

rp

1 3%

3

1

2%

1%

E1

E2

(E)

ra rp Ih

2% (F) Figure 1.20 Main variations of the hepatic duct confluence.

duplicated gallbladder (39,40), septum and diverticulum of the gallbladder, and variations in cystic duct anatomy including a double cystic duct (41). More rare is agenesis of the gallbladder (42,43) (Fig. 1.22). Furthermore, the gallbladder may be abnormally positioned, either lying deep within the liver parenchyma or lying under the left liver (44). The union of the cystic duct with the common hepatic duct may be angular, parallel, or spiral. The most frequent union is angular (75%) (45), while the cystic duct may run parallel with the hepatic duct in 20%, both encased in connective tissue. In 5% of cases, the cystic duct may approach the hepatic duct in a spiral fashion, usually passing posteriorly to the common hepatic duct before entering on its left side (Fig. 1.23).

the arterial blood supply of the bile ducts The hepatic artery usually arises as one of the three named branches of the coeliac trunk, along with the left gastric and splenic arteries (Fig. 1.24). The first named branch of the hepatic artery is the gastroduodenal artery and either of these arteries may then give rise to the right gastric and retroduodenal arteries (Fig. 1.24). The hepatic artery then divides into right (giving rise to the cystic artery) and left hepatic arteries. This arrangement holds true for 50% of cases. In nearly 25% of cases, the right hepatic artery arises separately from the superior mesenteric artery, indicative of the joint fore- and mid-gut origin of the liver (Fig. 1.25). In the remaining 25% of cases, the left hepatic artery arises from the left gastric artery. Occasionally, other variations will occur. These variations will be readily apparent to an experienced surgeon at operation. The authors do not advocate preoperative angiography to delineate these anomalies prior to routine hepatectomy. The extrahepatic biliary system receives a rich arterial blood supply (46), which is divided into three sections. The hilar section receives arterioles directly from their related hepatic arteries and these form a rich plexus with arterioles from the supraduodenal section. The blood supply of the supraduodenal section is predominantly axial. Most vessels to this section arise from the retroduodenal, right hepatic, cystic, gastroduodenal, and retroportal artery. Usually, eight small arteries, each 0.3 mm in diameter, supply the supraduodenal section. The most important of these vessels run along the lateral borders of the duct and are referred to as the 3 o’clock and 9 o’clock arteries. Of the arteries supplying the supraduodenal section, 60% run upward from the major inferior vessels while 38% run downward from the right hepatic artery. Only 2% are nonaxial, arising directly from the main trunk of the hepatic artery as it runs parallel to the bile duct. The retropancreatic section of the bile duct receives its blood supply from the retroduodenal artery. The veins draining the bile duct mirror the arteries and also drain the gallbladder. This venous drainage does not enter the portal vein directly but seems to have its own portal venous pathway to the liver parenchyma (47). It has been proposed that arterial damage during cholecystectomy may result in ischemia leading to postoperative stricture of the bile duct (47), although it seems unlikely that ischemia is the major mechanism in the causation of bile duct stricture after cholecystectomy.

11

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS 7

8

7

6 91% 7

8 7

8

7

8

86%

5

7

7 8

6 5%

8

8

6

4%

5

(A) seg V

5

5 10%

2%

2%

(B) seg VI

7 6

80%

3

3

6 5

2

2

7

5

20%

a 67%

b 1%

2

(C) seg VIII 3 c 1% 2

2 3

3

d 25%

e 1% 2

3 f 1% 2

3 g 4% (D) seg IV Figure 1.21 Variations of the intrahepatic biliary anatomy.

the anatomy of biliary exposure Although intraoperative ultrasound has made easier the location of dilated intrahepatic biliary radicals, surgical exposure of the extrahepatic biliary confluence and the segment 3 duct demands knowledge of precise anatomical landmarks. Biliary– enteric anastomosis necessitates precise bile duct exposure to facilitate the construction of a mucosa to mucosa apposition (36,48–50). To expose the extrahepatic biliary confluence, the base of the quadrate lobe (segment 4) is lifted upward and Glisson’s capsule is incised at its base (see Fig. 1.16) (51). This technique is also sometimes referred to as “lowering the hilar plate.” In only 1% of cases is this made difficult by any vascular imposition between the hilar plate and the inferior aspect of the liver. This maneuver will expose considerably more of the left hepatic duct than the right, which runs a shorter extrahepatic course.

12

Contraindications to this approach include patients with a very deep hilum, which is displaced upward and rotated laterally (36), and those patients who have undergone removal or atrophy of either the right or left livers resulting in hilar rotation. In this situation, the bile duct may come to lie behind the portal vein. When approaching the segment 3 duct (segment 3 hepaticojejunostomy), follow the round ligament (in which runs the remnant of the obliterated umbilical veins) through the umbilical fissure to the point where it connects with the left branch of the portal vein within the recessus of Rex. This junction may sometimes be deeply embedded within the parenchyma of the fissure. The bile ducts of the left liver are located above the left branch of the portal vein, whereas the corresponding arteries lie below the portal vein. Dissection of the round ligament on its left side allows exposure of either

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS

(A)

(B)

1

2

1

(D)

2

1

liver split to the left of the umbilical fissure in order to widen the fissure to achieve adequate access to the biliary system. Access to the right liver system is less readily achieved than to the left as the anatomy is more imprecise. However, intraoperative ultrasonography greatly enhances the ability of the surgeon to locat e these ducts at surgery. The ideal approach on the right side is to the segment 5 duct (52), which runs on the left side of its corresponding portal vein (23). The duct is exposed by splitting the liver over a short distance to the right of the gallbladder fossa, commencing at the right side of the porta hepatis. The segment 5 duct should lie relatively superficially on the left aspect of the portal vein to that segment.

3

1

(C)

2

Figure 1.22 Main variations in gallbladder and cystic duct anatomy: (A) bilobed gallbladder; (B) septum of gallbladder; (C) diverticulum of gallbladder; (D) variations in cystic duct anatomy.

(A) 75%

(B) 20%

(C) 5%

Figure 1.23 Different types of union of the cystic duct and common hepatic duct: (A) angular (75%); (B) parallel (20%); (C) spiral (5%).

the pedicle or anterior branch of the duct from segment 3. This dissection is achieved by mobilizing the round ligament and pulling it downwards, thereby freeing it from the depths of the umbilical fissure. This procedure usually requires the preliminary division of the bridge of liver tissue that runs between the inferior parts of segments 3 and 4. The umbilical fissure is then opened and with downward traction of the ligamentum teres an anterior branch of the segment 3 duct is exposed on its left side. Sometimes it may be necessary to perform a superficial liver split to gain access to this duct. In the usual situation of chronic biliary obstruction with dilatation of the intrahepatic bile ducts, the segment 3 duct is generally easily located above the left branch of the portal vein. However, in the situation of left liver hypertrophy, it may be necessary to perform a more extensive

radiological anatomy of the liver Accurate preoperative localization of liver pathology using radiological techniques is of increasing importance, as any potential resection depends largely on the segmental localization. Imaging is generally performed using ultrasound, computed tomography (CT), and magnetic resonance (MR). Ultrasound is excellent for imaging bile ducts, cysts, abscesses, and tumors. Hepatic circulation can also be accurately assessed using a Doppler technique. Ultrasound is also the imaging modality of choice for the biliary tree. However, the accuracy of ultrasound imaging is very operator dependent, and fine detail can be limited. Examination is limited by body habitus, and can be restricted by overlying bowel gas. CT scanning is an excellent method of assessing the liver parenchyma. It is able to identify a variety of different pathologies, and CT with IV contrast is the most commonly used method of imaging liver metastases. MR is excellent for the imaging and characterizing primary liver tumors, and is useful for the identification of hemangiomas, which can resemble metastases on CT scanning. Methods for defining segmental anatomy on ultrasound, CT, and MR images follow the anatomical landmarks previously described (53). These methods generally involve using three vertical planes along the lines of the main hepatic veins to divide the liver into its four sectors, with a transverse scissura along the portal vein further subdividing these four sectors to give the eight Couinaud segments. These anatomical landmarks are generally easily identifiable on standard imaging. The middle hepatic vein, left hepatic vein, and ligamentum teres provide good landmarks for dividing the left liver into its four segments. The right hepatic vein can usually be clearly seen dividing the right liver into its two sectors.

hepatic veins In an oblique ultrasonic view, the three hepatic veins join the IVC to form a characteristic W, with its base on the IVC. A similar view can be seen on CT scan. These veins are usually easily seen: the left hepatic vein separating segment 2 from segments 3 and 4, the middle hepatic vein separating segment 4 from 5 and 8, and the right hepatic vein separating 5 and 8 from 6 and 7.

portal system The portal supply to the left lobe, when viewed obliquely, can be seen as a side-on “H,” with the left portal vein giving its

13

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

Left branch of the hepatic artery Right branch of the hepatic artery Hepatic artery

3 o'clock artery 9 o'clock artery Common hepatic artery Retroduodenal artery

Gastroduodenal artery

(A)

M.H. artery L.H. artery R.H. artery

Left gastric

Cystic

Aorta

Proper hepatic

Celiac trunk

Right gastric Splenic

Supraduodenal

Common hepatic

Gastroduodenal (B)

Figure 1.24 (A) The biliary duct blood supply; (B) conventional arterial anatomy of the liver (50%).

branch to segment 2, before dividing into the terminal branches to 3 and 4. The portal supply to the right lobe also demonstrates a sideon “H” in the oblique view. The right branch of the portal vein forms the cross bar of the H, with the branches to segment 5 to 8 forming the arms.

gallbladder, ligamentum venosum, and falciform ligament Radiological landmarks of these structures are fallible (Figs. 1.26–1.28). Significant variations in intrahepatic vascular anatomy may result in incorrect identification of lesion location. A study by Rieker et al. looked at CT scans of patients who underwent liver resection. The location of the lesion was

14

identified using the landmarks outlined above. The scans were then reviewed, with the lesion being attributed to the nearest portal branch. Sixteen percent of lesions had a different segmental location if the portal branch was used instead of the conventional technique (Fig. 1.29) (54).

key points ●

●

A full understanding of the lobar, sectoral, and segmental anatomy of the liver and biliary system is an essential prerequisite for successful liver surgery. The surgeon must appreciate the wide variation in extrahepatic biliary anatomy.

SURGICAL ANATOMY OF THE LIVER AND BILE DUCTS

2

4a

8 (A)

IVC

(B)

7

(C)

(D)

(E)

(F)

Figure 1.27 CT scan of upper liver in venous phase showing the left, middle and right hepatic veins draining into the inferior vena cava (IVC).

Figure 1.25 Variations in anatomy of hepatic arterial supply.

Figure 1.28 CT scan of the liver in portal phase showing the left portal vein passing anteriorly between segments 3 and 4 within the recessus of Rex.

RAPV LPV

RPPV

Figure 1.26 Portal phase CT scan through porta hepatis showing the left portal vein (L) lying centrally and the anterior (RA) and posterior (RP) divisions of the right portal vein (R).

MPV

Figure 1.29 Percutaneous direct portogram showing the relationships of the anterior (RAPV) and posterior (RPPV) to the main (MPV) and left (LPV) portal veins.

15

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

references 1. Glisson F. Anatomia Hepatis. London: Typ. Du-Gardianis, 1654. 2. Rex 1888. Cited in Hobsley M. The anatomical basis of partial hepatectomy. Proc R Soc Med Engl 1964; 57: 550–4. 3. Schwartz SI. Historical Background. In: McDermott WV Jr, ed. Surgery of the liver. Boston, MA: Blackwell Scientific, 1989: 3–12. 4. McIndoe AH, Counsellor VX. A report on the bilaterality of the liver. Arch Surg 1927; 15: 589. 5. Lau WY. The history of liver surgery. J R Coll Surg Edin 1997; 42: 303–9. 6. Mikesky WE, Howard JM, DeBakey ME. Injuries of the liver in three hundred consecutive cases. Int Abstr Surg 1956; 103: 323–4. 7. Dalton HC. Gunshot wound of the stomach and liver treated by laparotomy and suture of the visceral wounds. Ann Surg 1888; 8: 81–100. 8. Luis A. Di un adenoma del fegato. Centralblatt fur chirg 1887; 5: 99. Abstract from Ganzy, delle cliniche 1886, 23, No 15. 9. Langenbuch C. Ein Fall von Resektion eines linksseitigen Schnurlappens der Leber. Berl Klin Wosch 1888; 25: 37–8. 10. Tiffany L. The removal of a solid tumor from the liver by laparotomy. Maryland Med J 1890; 23: 531. 11. Lucke F. Entfernung der linken Krebsiten Leber Lappens. Cantrallbl Chir 1891: 6: 115. 12. Cattell RB. Successful removal of liver metastasis from carcinoma of the rectum. Lehey Clin Bull 1940; 2: 7–11. 13. Wangensteen OH. The surgical resection of gastric cancer with special reference to: (1) the closed method of gastric resection; (2) coincidental hepatic resection; and (3) preoperative and postoperative management. Arch Surg 1943; 46: 879–906. 14. Keen WW. Report of a case of resection of the liver for the removal of a neoplasm with a table of seventy six cases of resection of the liver for hepatic tumor. Ann Surg 1899; 30: 267–83. 15. Cantlie J. On a new arrangement of the right and left lobes of the liver. J Anat Physiol (Lond) 1898; 32:4–9. 16. Wendel W. Beitrage zur Chirurgie der Leber. Arch Klin Chir Berlin 1911; 95: 887–94. 17. Ton That Tung. La vascularisation veineuse du foie et ses applications aux resections hepatiques. These, Hanoi, 1939. 18. Raven RW. Partial hepatectomy. Br J Surg 1948; 36: 397–401. 19. Lortat-Jacob JL, Robert HG. Hepatectomie droite regle. Presse Med 1952; 60: 549–50. 20. Healey JE Jr, Schroy PC. Anatomy of the biliary ducts within the human liver. Arch Surg 1953; 66: 599–616. 21. Goldsmith NA, Woodburne RT. Surgical anatomy pertaining to liver resection. Surg Gynaecol Obstet 1957; 195: 310–18. 22. Hjortsjo CH. The topography of the intrahepatic duct systems. Acta Anat 1951; 11: 599–615. 23. Couinaud C. Le foie. Etudes anatomiques et chirurgicales. Paris: Masson, 1957. 24. Couinaud C. Lobes et segments hepatiques. Note sur l’architecture anatomiques et chirurgicales du foie. Presse Med 1952; 62: 709–12. 25. Couinaud C. Anatomy of the dorsal sector of the liver. In: Couinaud C, ed. New Considerations on Liver Anatomy. Paris: Couinaud, 1998: 39–61. 26. Ton That Tung. Les Resections Majeures et Mineures Du Foie. Paris: Masson, 1979. 27. Caprio G. Un caso de extirpacion die lobulo izquierdo die hegado. Bull Soc Cir Urag Montevideo 1931; 2: 159. 28. Bismuth H, Houssin D, Castaing D. Major and minor segmentectomies “reglees” in liver surgery. World J Surg 1982; 6: 10–24. 29. Mancuso M, Nataline E, Del Grande G. Contributo alla conoscenza della struttura segmentaria del fegato in rapportto al problema della resezione epatica. Policlinico, Sez Chir 1955; 62: 259–93.

16

30. Couinaud C. Surgical anatomy of the liver revisited. C Couinaud, 15 rue Spontini, Paris, 1989. 31. Mizumoto R, Kawarada Y, Suzuki H. Surgical treatment of hilar carcinoma of the bile duct. Surg Gynecol Obstet 1986; 162: 153–8. 32. Rocko JM, Swan KG, Di Gioia JM. Calot’s triangle revisited. Surg Gynecol Obstet 1981; 153: 410–14. 33. Wood D. Eponyms in biliary tract surgery. Am J Surg 1979; 138: 746–54. 34. Byden EA. The anatomy of the choledochaoduodenal junction in man. Surg Gynecol Obstet 1957; 104: 641–52. 35. Delmont J. Le sphincter d’Oddi: anatomie traditionelle et fonctionelle. Gastroenterol Clin Biol 1979; 3: 157–65. 36. Bismuth H, Lazorthes F. Les Traumatismes Operatoires de la Voie Biliaire Principale. Paris: Masson, Vol 1, 1981. 37. Champetier J, Davin JL, Yver R, Vigneau B, Letoublon C. Aberrant biliary ducts (vasa aberrantia): surgical implications. Anat Clin 1982; 4: 137–45. 38. Gross RE. Congenital anomalies of the gallbladder. A review of a hundred and forty-eight cases with report of a double gallbladder. Arch Surg 1936; 32: 131–62. 39. Hobby JAE. Bilobed gallbladder. Br J Surg 1979; 57: 870–2. 40. Rachad-Mohassel MA, Baghieri F, Maghsoudi H, Nik Akhtar B. Duplication de la vesicule biliaire. Arch Francais des Maladies de l’Appareil Digestif 1973; 62: 679–83. 41. Perelman H. Cystic duct duplication. J Am Med Assoc 1961; 175: 710–11. 42. Boyden EA. The accessory gallbladder. An embryological and comparative study of aberrant biliary vesicles occurring in man and the domestic mammals. Am J Anat 1926; 38: 177–231. 43. Rogers HI, Crews RD, Kalser MH. Congenital absence of the gallbladder with choledocholithiasis. Literature review and discussion of mechanisms. Gastroenterology 1975; 48: 524–9. 44. Newcombe JF, Henley FA. Left sided gallbladder. A review of the literature and a report of a case associated with hepatic duct carcinoma. Arch Surg 1964; 88: 494–7. 45. Kune GA. The influence of structure and function in the surgery of the biliary tract. Ann R Coll Surg Engl 1970; 47: 78–91. 46. Northover JMA, Terblanche J. A new look at the arterial blood supply of the bile duct in man and its surgical implications. Br J Surg 1979; 66: 379–84. 47. Northover JMA, Terblanche J. Applied surgical anatomy of the biliary tree. In: Blumgart LH, ed. Biliary Tract, Vol 5. Edinburgh: Churchill Livingstone, 1982. 48. Bismuth H, Franco D, Corlette NB, Hepp J. Long term results of Roux-enY hepaticojejunostomy. Surg Gynecol Obstet 1978; 146: 161–7. 49. Voyles CR, Blumgart LH. A technique for construction of high biliary enteric anastomoses. Surg Gynecol Obstet 1982; 154: 885–7. 50. Blumgart LH, Kelley CJ. Hepaticojejunostomy in benign and malignant bile duct stricture: approaches to the left hepatic ducts. Br J Surg 1984; 71: 257–61. 51. Hepp J, Couinaud C, L’abord et L’utilisation du canal hepatique gauche dans le reparations de la voie biliaire principale. Presse Med 1956; 64: 947–8. 52. Smadja C, Blumgart LH. The biliary tract and the anatomy of biliary exposure. In: Blumgart LH, ed. Surgery of the Liver and Biliary Tract, 2nd edn. Edinburgh: Churchill Livingstone, 1994: 11–24. 53. Strunck H, Stuckmann G, Textor J et al. Limitations and pitfalls of Couinauds segmentation of the liver in transaxial imaging. Eur Radiol 2003; 13: 2472–82. 54. Rieker O, Mildenberger P, Hintze C et al. Segmentanatomie der Leber in der Computertomographie: Lokalisieren wir die Lasionen richtig. Rofo 2000; 171: 147–52.

2

Anatomy of the pancreas Margo Shoup and Jason W. Smith

topography of the pancreas The shape and size of the pancreas are highly variable but in general it has a roughly trapezoidal shape and lies in the retroperitoneum of the upper abdomen (1). It is a finely lobular structure with a tan to dull yellow color that reaches from the medial concavity of the duodenum up and to the left terminating at the hilum of the spleen. The volume of the pancreas increases rapidly during childhood, plateaus from 20 to 60 years, and then steadily decreases; however, the percentage of parenchyma versus fat in the pancreas continues to increase during life slowly replacing functional tissue (2) (Fig. 2.1). The pancreas is divided into three major regions, the head and uncinate, the neck, and the body and tail (3). The head is the most medial portion of the gland. It is the widest and thickest part, having the most globular ultrastructure and is cradled in the concavity of the duodenum lying just to the right of the second lumbar vertebra (1). There is an inferior projection to the head of the pancreas that lies posterior to the superior mesenteric vessels, which makes up the uncinate process. The head and uncinate are intimately associated with the duodenum, sharing an abundant network of anastomosing vessels. The posterior surface of the head of the pancreas is in apposition to the inferior vena cava, aorta, right spermatic and ovarian vessels, and right renal vessels and separated from them by the avascular fusion fascia of Treitz (4). The anterior surface is covered by the transverse colon and its mesentery (5,6). The neck of the pancreas is 2 to 3 cm in length and overlies the confluence of the superior mesenteric vein (SMV) and splenic vein by which it is grooved. It is related superiorly to the pylorus and first portion of the duodenum (3,4). The body of the pancreas extends from body of the second lumbar vertebra over the left kidney and begins to taper into the tail as it reaches the hilum of the spleen. The prismatic shape of the pancreas flattens in the tail. The splenic vein runs the length of the pancreas on the posterior surface, while the artery courses along the superior edge of the body. The body of the pancreas lies over the aorta and the left renal pedicle and kidney and is separated from these structures by the fusion fascia of Toldt (4). Inferiorly, it abuts the mesentery of the transverse mesocolon, while superiorly and anteriorly it abuts the stomach but is separated from it by the posterior parietal peritoneum (7).

ductal anatomy of the pancreas There are numerous configurations of the ducts of the pancreas and their relationships to each other, the duodenum and the common bile duct. The significance of the pancreas became understood only after the discovery of the main pancreatic duct by Wirsung in 1643. He noted that there was a duct that traversed the length of the organ with numerous

tributary ducts coming off at near right angles and that this duct opened into the duodenum, and he saw that there were occasionally two ducts in the gland (1). It was Santorini who finally concluded that, in the normal condition, there existed two ducts with the smaller of the two emptying into the duodenum by way of a small papilla approximately 2 cm nearer to the stomach than the major duct and this smaller duct bears his name (5). The smaller duct is patent all the way to the duodenum in only 60% of specimens and the duct of Wirsung represents the larger of the two; however, in about 10% of specimens, the duct of Santorini is the main drainage for the pancreas. Also in about 10% of cases, the two ducts are not in communication with each other (1) (Fig. 2.2). The parenchyma of the pancreas consists of small lobules divided by connective tissue. These lobules are centered around the main tributary ducts that run to the main pancreatic duct. Smaller branches off of these tributaries define further septated regions within the lobules of pancreatic tissue. The main branches of the pancreatic duct tend to meet the main duct on its superior and inferior aspect. The diameter of the main pancreatic duct is reported to be between 2.6 and 4.8 mm in the head, 2.0 and 4.0 mm in the body, and 0.9 and 2.4 mm in the tail (3). The duct runs in a relatively superficial position in the tail and after traversing the neck of the pancreas it dives deep into the parenchyma as it crosses the head and is near the dorsal surface of the pancreas as it nears the confluence with the common bile duct (CBD) and the duodenum (1). The lower portion of the CBD lies in contact with the head of the pancreas for between 2 and 7 cm and 40% of the time it lies in a groove between the surface of the pancreas and the duodenum. In the remainder of cases, it lies within the parenchyma of the pancreas (7). During embryological development, the lower duct of Wirsung arises in the ventral pancreatic bud adjacent to the early hepatic duct. Therefore, the association of the duct of Wirsung with the CBD is a consistent feature of the ductal anatomy of the pancreas (1). The duct of Wirsung and the CBD unite 6 to 8 mm within the papilla and form a common channel, which is slightly dilated and referred to as the ampulla of Vater. In just over 10% of cases, the two ducts do not form a short common channel and instead enter the duodenum independently on the papilla (5).

arterial anatomy of the pancreas The pancreas enjoys an abundant arterial blood supply that draws from both the celiac axis and the superior mesenteric artery (SMA). The pancreas is supplied from the celiac axis by the superior pancreaticoduodenal artery from the gastroduodenal artery (GDA), and the dorsal pancreatic and pancreatica magna arteries from the splenic artery (Fig. 2.3). The distal and inferior borders of the pancreas are supplied by the caudal and inferior pancreatic arteries, which are formed by

17

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

Figure 2.1 Overview of the relationship of the pancreas to other important structures in the upper abdomen. Plate 1098, From Anatomy of the Human Body, Henry Gray 1918.

(A)

(B)

(C)

Figure 2.2 (A) Duct of Santorini is patent all the way to the duodenum. (B) Duct of Santorini is the main drainage. (C) The two ducts are not in communication with each other.

ramifications with the dorsal pancreatic, pancreatica magna, and splenic arteries. The SMA gives rise to the more variable inferior pancreaticoduodenal (IPD) artery, which divides into branches to form both an anterior and posterior anastomotic arcade with branches from the superior pancreaticoduodenal artery (8). Superior Pancreaticoduodenal Artery The superior pancreaticoduodenal is a short branch of the GDA that arises after the takeoff of the right gastroepiploic artery (Fig. 2.4). It is angiographically identifiable in about 10% of specimens and is generally about 8 mm in length (9). Although rare, it is reported to occasionally arise from the left hepatic artery. When present, the superior pancreaticoduodenal artery divides into anterior and posterior branches, which anastomose with the inferior branches from the SMA. In the remaining cases, the posterior superior pancreaticoduodenal

18

(PSPD) artery is seen arising from the GDA prior to the right gastroepiploic takeoff. The anterior superior pancreaticoduodenal (ASPD) artery has a caliber between 1 and 3 mm and is considered the most important blood supply to the head of the pancreas. In the majority of cases, it is a terminal branch of the GDA after it has given off the PSPD and the right gastroepiploic arteries. The ASPD can be duplicated in up to 7% of cases and rarely is absent. Case reports of extremely rare anomalies exist, reporting the origin of this artery from almost all of the major branches of the celiac and SMAs (9). Posterior Superior Pancreaticoduodenal Artery This artery forms the superior portion of the posterior arcade that forms anastomoses with the posterior branch of the IPD artery. The PSPD artery is most commonly found as a branch of the GDA 1 to 2 cm after the takeoff of the hepatic artery (10). Up to 10% of cases may see the PSPD arise from the superior

ANATOMY OF THE PANCREAS

Cystic artery

Probe passed through epiploie foramen

a S

C r e a t o r

t

o

c

h

m

O x e n t e

Figure 2.3 Arterial anatomy of the pancreas, the celiac axis and its major branches. Plate 532, From Anatomy of the Human Body, Henry Gray 1918.

pancreaticoduodenal and in rare instances may arise from any of the hepatic arteries. The most common course of the PSPD after it leaves the GDA posteriorly is it runs over the portal vein (PV) and the anterior edge of the top of the pancreas where it enters the gland and finds the common bile duct and makes a right-handed spiral around the duct passing posterior to it just above the ampulla. It then runs deep in the parenchyma of the pancreas to find its connection with the posterior inferior artery. The PSPD gives off collateral branches to form the blood supply to the intrapancreatic portion of the common bile duct, it generally gives off the supraduodenal artery and occasionally the retroduodenal artery, rarely it may give a branch to the gallbladder or an accessory right hepatic artery (10). Inferior Pancreaticoduodenal Artery The IPD artery is present in about 70% of cases and is the common trunk that gives rise to the anterior and posterior inferior pancreaticoduodenal (AIPD and PIPD) arteries that form the anastomotic arcades supplying the head of the pancreas (11). In the remaining 30% of cases, the AIPD and PIPD arise directly from the SMA. The IPD may arise directly from the SMA as the first collateral branch from 2 to 5 cm distal to the origin and take a short course from its posterior takeoff into the inferior edge of the pancreatic parenchyma, or alternatively, it may arise as a common trunk with the first jejunal

branch, the pancreaticoduodenaljejunal (PDJ) trunk in which case it takes a longer course to the pancreas. The IPD crosses posterior to the SMV and the posterior surface of the pancreas and does not give off any branches prior to dividing into its anterior and posterior termini (11). Anterior and Posterior Inferior Pancreaticoduodenal Arteries These arteries supply the inferior part of the anastomotic arches that supply the head of the pancreas. They arise most often from a common IPD artery. They may also originate directly from the SMA or less commonly directly from the first jejunal artery or from a replaced hepatic artery. The main course of the AIPD is to follow the inferior curve of the pancreas and find its partner the ASPD (12). It may give off a branch to the duodenal–jejunal flexure or to form a transverse pancreatic artery. The PIPD runs more posterior and cephalad than the AIPD and ultimately finds the PSPD or alternatively terminates as small end arteries. It may supply a collateral branch to the transverse pancreatic artery when present (12). Dorsal Pancreatic Artery The main blood supply to the neck and body of the pancreas is the dorsal pancreatic (DP) artery. It most commonly arises from the splenic artery near its origin at the celiac axis (13). It may also take its origin from the celiac trunk itself, the

19

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

S t o m a c h

Figure 2.4 Arterial anatomy of the pancreas, demonstrating the gastroduodenal and its branches of the anterior and posterior pancreaticoduodenal arteries forming the anastomotic arcades with the branches from the superior mesenteric artery. Plate 533, From Anatomy of the Human Body, Henry Gray 1918.

common hepatic or the GDA. Alternatively, the DP may arise from the SMA. The course of the DP artery is usually in the form of an inverted “T” with a right and left branch that form after a short 1 to 3 cm course. When the artery arises from the splenic artery, it tends to angle back to the right, if it takes off from the celiac, hepatic, or GDA, then it transverses the neck in a leftward direction. When coming from the SMA it comes up from the bottom of the pancreas. The right branch of the DP forms an anastomosis with left anastomotic pancreatic artery from the ASPD. The left branch becomes the transverse pancreatic artery (13). Caudal and Great Pancreatic Arteries The great pancreatic artery is often present and is given off from the splenic artery at the junction of the body and tail. It collateralizes with the transverse pancreatic artery. The caudal pancreatic artery takes its origin from the left gastroepiploic, the distal splenic artery or a branch from the splenic hilum and forms anastomotic connections with the great pancreatic and transverse pancreatic arteries (3). The arterial blood supply to the pancreas is rich and complex. Most of the primary arterial conduits form some

20

anastomotic connection and this shared blood supply is one of the challenges of pancreatic surgery. When operating in the deepest recesses of the abdomen, having an intimate knowledge of the standard arterial anatomy as well as the most common alternatives will allow the pancreatic surgeon to maximize patient safety. That same surgeon must keep in mind that the arterial anatomy in this area is subject to wide variation and that one must always be prepared to address the aberrant anatomy. To that end, having good preoperative imaging to establish before the operation what the arterial anatomy is can be a valuable aid whether by angiography or by computed tomography (CT) angiography. Venous Drainage of the Pancreas The veins of the pancreas follow the course of the corresponding arteries in most cases. They are generally more superficially located than the arteries and depending on the location in the pancreas drain into the PV, SMV, the inferior mesenteric vein, or the splenic vein. In the head of the pancreas, there is a venous arcade that mirrors the arterial anastomoses and of the four main veins all, but the PSPD vein, which empties directly into the PV, find their way to the SMV. In addition, there are

ANATOMY OF THE PANCREAS

12

12p1

12b1

12a1

1

10 9

12p2

7

11

12a2 3 8a

5

16 8p

10

14a

2 6 4

14b

13a 17a

14c

18

14d

17b 14V

13b

15

14d

Figure 2.5 Lymphatic drainage of the pancreas.

numerous small bridging veins between the head of the pancreas and the SMV and PV as they course behind the pancreas, which must be carefully ligated during a resection. The fact that there are rarely venous branches that enter the SMV or PV on their anterior surfaces makes the dissection along the plane anterior to these vessels possible during pancreaticduodenectomy. Two large veins drain the body and tail of the pancreas, the splenic vein, which courses along the superior edge of the pancreas and the transverse pancreatic vein along the inferior margin. The portal vein is formed on the posterior surface of the neck of the pancreas by the confluence of the splenic vein and the SMV. The inferior mesenteric vein may join at this point as well, but more commonly joins the splenic vein or SMV proximal to the confluence (Fig. 2.4).

lympatic drainage of the pancreas The lymphatic drainage of the pancreas is rich and drains each lobular division with frequent anastomotic connections and the ultrastructure is similar to that in other solid organs of the abdomen(14) (Fig. 2.5). These lobular lymphatics coalesce to form several trunks that empty into the primary lymph node basins for the pancreas before quickly reaching the thoracic duct (15). The drainage of the pancreas can be roughly divided into right and left side based on the ventral and dorsal anlage of the primordial pancreas. The left side of the system drains the upper portion of the head, the neck, and body and tail,

while the right side drains the lower portion of the head, which developed from the ventral bud and constitutes the retroportal lymphatics (15,16). The superior pancreatic nodes drain the upper half of the neck, body and tail of the pancreas, and a portion of the head. They primarily lie along the superior border of the gland or in the gastropancreatic fold and gastrohepatic ligament (17). The inferior pancreatic nodes similarly drain the inferior half of the gland and lie along the inferior border as well as draining into the superior mesenteric nodes or the periaortic nodes. The anterior nodes are located along the surface of the pancreas that lies adjacent to the duodenum and are called the infrapyloric lymph nodes and the pancreaticoduodenal nodes. These anterior nodes may also drain into nodes along the root of the transverse colonic mesentery that is adjacent to the head of the pancreas. The posterior nodes run along the posterior pancreaticoduodenal border and include the nodes along the lower portion of the common bile duct, portal vein and nodes at the origin of the SMA. The tail of the pancreas forms several lymphatic trunks that reach out into the hilum of the spleen and form the superior and inferior lymph nodes (3,16). This simplified lymphatic mapping system is that adapted by the International Union against Cancer (UICC). A more comprehensive and clinically useful system was developed by the Japanese Research System, which divides lymph node stations into 18 different designations and rates them according to the likelihood of metastatic spread. Nodal stations 13 and 17 are the most likely to harbor disease with

21

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Right gastroepiploic vein Splenic vein

Portal vein

PSPD-V Gastrocolic trunk

Superior mesenteric vein AIPD-V PIPD-V

First jejunal tributary

ASPD-V Figure 2.6 Major venous drainage for the pancreas.

47% and 29%, respectively (18,19) (Fig. 2.6). Classification of lymphatic involvement will become increasingly important as increasing numbers of targeted therapies become available in pancreatic cancer.

innervation of the pancreas The pancreas receives fibers from both the sympathetic and parasympathetic nervous systems. The sympathetic innervation is via the splanchnic nerves, which carry both afferent fibers and efferent fibers, while the parasympathetic innervation is via the vagus nerve, which also has afferent and efferent supply to the pancreas. Parasympathetic innervation provides stimulatory signals to the islet cells to increase insulin secretion in response to food intake, while increased sympathetic tone suppresses insulin secretion and stimulates the secretion of glucagon (20,21). Efferent pain fibers are found in both the splanchnic and vagal nerves and localization of these fibers has been a difficult clinical problem in the management of pain in both inflammatory and malignant diseases of the pancreas. The right, and more prominently the left, celiac ganglion provide the majority of the direct innervation to the posterior head, body, and tail of the pancreas via fibers that course along the splenic artery (16). Neural ganglia around the common hepatic artery also provide fibers that course along the GDA to the head and uncinate process of the pancreas (22). Recently, it has been shown that the celiac ganglion bearing the splanchnic efferent fibers can be identified by endoscopic ultrasound and precise localization of neurolytic therapies can be applied to improve the success of these approaches (23,24). Enteropancreatic nervous connections have also been demonstrated from both the stomach and proximal duodenum to the pancreas (24–26). These connections suggest that there is crosstalk directly from the gastrointestinal tract to the pancreas coordinating exocrine and/or endocrine secretions with gut function.

22

The pancreas lies in the recesses of the upper abdomen and remains one of the most challenging organs to manage from a clinical or operative standpoint. Its rich blood supply, close associations with major vascular structures, intimate relation to the common bile duct, and the attachments to the duodenum and spleen all contribute to the complexity of surgical intervention in both malignant and benign disease (7). A thorough understanding of the three dimensional relationship of the arterial blood supply and major veins in proximity to the pancreas make approaching pancreatic resection possible. As we move into an era of minimally invasive surgery, being able to recognize the anatomy and its variations with minimal cues from adjacent structures will become increasingly important and continued study of these complex relationships allows the mind to know, so that the eye may see.

references 1. Opie EL. Anatomy of the Pancreas and its Variations. Disease of the Pancreas: Its Cause and Nature, 1st edn. Philadelphia, PA: J.B. Lippincott Company, 1903: 359. 2. Saisho Y, Butler AE, Meier JJ, et al. Pancreas volumes in humans from birth to age one hundred taking into account sex, obesity, and presence of type-2 diabetes. Clin Anat 2007; 20: 933–42. 3. Skandalakis LJ, Colborn GL, Skandalakis JE. Surgical anatomy of the pancreas. In: Baker RJ, Fischer JE, eds. Mastery of Surgery, Vol. 2, 4th edn. Philadelphia, PA: Lippincott Williams & Wilkins, 2001: 2448. 4. Kuroda A, Nagai H. Surgical anatomy of the pancreas. In: Howard J, Idezuki Y, Ihse I, Prinz R, eds. Surgical Diseases of the Pancreas, 3rd edn. Baltimore, MD: Lippincott Williams & Wilkins, 1998: 869. 5. Cattell RB, Warren KW. The anatomy and physiology of the pancreas. In: Cattell RB, Warren KW, eds. Surgery of the Pancreas. Philadelphia, PA: Saunders, 1953. 6. Hollinshead WH. The thorax, abdomen and pelvis. In: Hollinshead WH, ed. Anatomy for Surgeons. Vol. 2. New York: Medical Department, Harper and Row Publishers, 1971: 430. 7. Anson BJ, McVay CB, Callander CL. The Abdomen. Surgical Anatomy. Philadelphia, PA: Saunders, 1971. 8. Woodburne RT, Olsen LL. The arteries of the pancreas. Anat Rec 1951; 111: 255–70. 9. Bertelli E, Di Gregorio F, Bertelli L, Mosca S. The arterial blood supply of the pancreas: A review. I. The superior pancreaticoduodenal and the anterior superior pancreaticoduodenal arteries. An anatomical and radiological study. Surg Radiol Anat 1995; 17: 97–106, 101–3. 10. Bertelli E, Di Gregorio F, Bertelli L, Civeli L, Mosca S. The arterial blood supply of the pancreas: A review. II. The posterior superior pancreaticoduodenal artery. An anatomical and radiological study. Surg Radiol Anat 1996; 18: 1–9. 11. Bertelli E, Di Gregorio F, Bertelli L, Civeli L, Mosca S. The arterial blood supply of the pancreas: A review. III. The inferior pancreaticoduodenal artery. An anatomical review and a radiological study. Surg Radiol Anat 1996; 18: 67–74. 12. Bertelli E, Di Gregorio F, Bertelli L, Orazioli D, Bastianini A. The arterial blood supply of the pancreas: A review. IV. The anterior inferior and posterior pancreaticoduodenal aa., and minor sources of blood supply for the head of the pancreas. An anatomical review and radiologic study. Surg Radiol Anat 1997; 19: 203–12. 13. Bertelli E, Di Gregorio F, Mosca S, Bastianini A. The arterial blood supply of the pancreas: A review. V. The dorsal pancreatic artery. An anatomic review and a radiologic study. Surg Radiol Anat 1998; 20: 445–52. 14. Navas V, O’Morchoe PJ, O’Morchoe CC. Lymphatic system of the rat pancreas. Lymphology 1995; 28: 4–20. 15. Pissas A. Anatomoclinical and anatomosurgical essay on the lymphatic circulation of the pancreas. Anat Clin 1984; 6: 255–80. 16. Donatini B, Hidden G. Routes of lymphatic drainage from the pancreas: A suggested segmentation. Surg Radiol Anat. 1992; 14: 35–42.

ANATOMY OF THE PANCREAS 17. Hartley M, Finch-Jones M. Anatomy of the pancreas. In: Poston G, Blumgart L, eds. Surgical Management of Hepatobiliary and Pancreatic Disorders, 1st edn. London: Martin Dunitz, 2002: 19–28. 18. Bogoevski D, Yekebas EF, Schurr P, et al. Mode of spread in the early phase of lymphatic metastasis in pancreatic ductal adenocarcinoma: Prognostic significance of nodal microinvolvement. Ann Surg 2004; 240: 993–1000, discussion 1000–1. 19. Sakai M, Nakao A, Kaneko T, et al. Para-aortic lymph node metastasis in carcinoma of the head of the pancreas. Surgery 2005; 137: 606–11. 20. Benthem L, Mundinger TO, Taborsky GJ, Jr. Parasympathetic inhibition of sympathetic neural activity to the pancreas. Am J Physiol Endocrinol Metab 2001; 280: E378–81. 21. Jarhult J, Falck B, Ingemansson S, Nobin A. The functional importance of sympathetic nerves to the liver and endocrine pancreas. Ann Surg 1979; 189: 96–100.

22. Yoshioka H, Wakabayashi T. Therapeutic neurotomy on head of pancreas for relief of pain due to chronic pancreatitis; a new technical procedure and its results. AMA Arch Surg 1958; 76: 546–54. 23. Levy MJ, Topazian MD, Wiersema MJ, et al. Initial evaluation of the efficacy and safety of endoscopic ultrasound-guided direct Ganglia neurolysis and block. Am J Gastroenterol 2008; 103: 98–103. 24. Kirchgessner AL, Liu MT, Gershon MD. In situ identification and visualization of neurons that mediate enteric and enteropancreatic reflexes. J Comp Neurol 1996; 371: 270–86. 25. Holst JJ, Schwartz TW, Knuhtsen S, Jensen SL, Nielsen OV. Autonomic nervous control of the endocrine secretion from the isolated, perfused pig pancreas. J Auton Nerv Syst 1986; 17: 71–84. 26. Kirchgessner AL, Gershon MD. Innervation and regulation of the pancreas by neurons in the gut. Z Gastroenterol Verh 1991; 26: 230–33.

23

3

Hepatic resection Ajay V. Maker and Michael D’Angelica

introduction Though liver anatomy and physiology have been studied for centuries, liver surgery still is a relatively young field. Just 30 years ago, the mortality of major hepatic resection neared 25%. This high mortality limited its utility and deterred patients and referring physicians from considering surgery. The current generation of hepatobiliary surgeons has an increased understanding of the segmental anatomy of the organ and has seen a dramatic decrease in the mortality of liver surgery to nearly 1% largely due to a dramatic decrease in blood loss (1). This chapter will address the basic principles and techniques to safely approach liver resection.

basic principles Surgical Indications: Benign vs. Malignant Disease Though this chapter focuses on the technical aspects of hepatic resection, an understanding of when liver resection is indicated is of paramount importance. Due to advances in modern imaging techniques and an increased knowledge of the natural history of liver lesions, tumors that may have been resected in the past for diagnostic uncertainty are now often observed. Similarly, malignant lesions that were not resected in the past but referred for nonsurgical therapy are now being treated with resection. Indications for specific benign and malignant processes are outlined in other chapters; however, the general principles are mentioned here. Benign Disease Partial hepatectomy for benign conditions should be parenchymal preserving and reserved for lesions that are symptomatic, have premalignant potential, or carry an unclear diagnosis. Wide margins are not necessary, therefore in some cases, for example, focal nodular hyperplasia (FNH) or hemangiomas, enucleation may be safely performed, although in some instances an anatomic segmental resection may be the safest approach (2–4). This is addressed at the end of the chapter and detailed in other chapters. Malignant Disease Partial hepatectomy for malignant conditions must obtain a clear surgical margin, and is suitable for well-selected patients with both primary and metastatic cancer. We have found increased patient survival with margins of at least 1 cm in patients undergoing resection for metastatic colorectal cancer (5–13), though other series suggest that a negative margin, regardless of the distance, is sufficient (14,15). The exception may be in slow-growing tumors with multiple liver metastases, such as neuroendocrine tumors, where tumor debulking may be of value. As long as the functional remnant liver is adequate, usually about 25% liver volume in otherwise

24

healthy individuals, excision of tumor can prolong life and in some cases provide long-term disease-free survival. Patient Selection Proper patient selection is critical to both the safety and efficacy of hepatic resection. One should evaluate the patient’s general state of health, the condition of the liver, and the volume of the future liver remnant to properly assess the risk of general anesthesia, major abdominal surgery, and liver resection. Subcostal and upper abdominal incisions are painful and may result in respiratory splinting and increased pulmonary complications compared to other incisions (16). For this reason, assessment of the patient’s ability to mobilize early and ambulate postoperatively must not be underestimated. Though there are many algorithms to evaluate liver function in patients with chronic liver disease, the Child-Pugh classification is a useful preoperative indicator, and patients with a Pugh score of B or C should generally not undergo liver resection. Hepatic resection in cirrhotic patients is particularly difficult with operative mortality increasing with advanced Child classification. Hepatic resection in the setting of portal hypertension is generally not recommended (17), as this condition predisposes the liver to higher portal pressures and diminished ability to increase portal flow to the liver remnant postoperatively, thereby inhibiting normal liver regeneration and increasing the risk of life-threatening bleeding. The cirrhotic liver has decreased regenerative capacity and impairment in liver function is greater, lasts longer, and can result in permanent liver failure. A low platelet count, splenomegaly, ascites, or evidence of varices on preoperative radiography may be the only findings to alert the surgeon to hepatic dysfunction. Many noncirrhotic patients that present for hepatic resection have abnormal liver function due to chemotherapy, diabetes, or obesity. These diseased livers carry an increased risk of functional impairment with large resections and may also have impaired function despite retained volume (18,19). In these livers, careful preoperative planning must be done to achieve a parenchymal sparing resection. Biopsy, if performed, can give clues to the fat content of the liver, as can preoperative imaging (20). Early data suggest that MRI spectroscopy can also accurately quantify hepatic fat content, and this may prove to be a useful tool in preoperative liver assessment and operative planning (20,21). In cases where liver function may be impaired, or where extended resection is necessary to gain tumor-free margins, portal vein embolization is being employed to induce hypertrophy of the proposed liver remnant (22,23). No absolute guidelines for embolization can be made; however, preoperatively induced liver hypertrophy is a valuable tool in planning and executing major liver resections (24). Furthermore, chronic biliary obstruction inhibits liver function and, thus patients with

HEPATIC RESECTION hilar cholangiocarcinomas are also at increased risk of liver failure postoperatively. The functional residual liver volume should be calculated to insure adequate liver function postresection. A healthy, noncirrhotic individual requires a functional hepatic reserve of at least 20% of the original nontumoral liver volume. The regenerative capacity of the liver should enable full functional compensation within weeks of resection; once greater than 70% of liver volume is resected, however, there is a risk of clinically significant liver insufficiency. This risk is minimal if specimen volume has been replaced with tumor, in which case compensatory hypertrophy will have already occurred. Preoperative Imaging (See Also Chapters 3, 4, and 11) Fine-cut triphasic helical computed tomography (CT) with CT angiogram is the single most useful study in preoperative evaluation of liver tumors. When the study includes the chest, abdomen, and pelvis, preoperative staging is reliable and can identify areas outside of the liver that may need further evaluation or confirm nonoperative candidates. CT can define the vascular anatomy, identify anatomical variants, determine resectability, estimate the functional liver residual volume, and identify preoperative biliary drainage strategies, thereby obviating the need for further radiographic studies. CT angiography in particular has almost prevented the need for traditional angiography. 3-D reconstruction of the vasculature is particularly helpful in identifying vascular anomalies quickly and temporally. Furthermore, 3-D reconstruction of the vascular anatomy may lead to more accurate visualization of tumor– vessel relationships and may be a more accurate study to predetermine the operative line of transection (25). Magnetic resonance imaging can also provide high-quality vascular and volumetric assessments of the liver but its principal role is in characterizing liver tumors of unclear etiology. In experienced hands, ultrasound is a fast, inexpensive, and noninvasive modality that can quickly obtain information regarding tumor size and the amount of liver involvement, particularly in gallbladder and biliary tumors. It is especially helpful in distinguishing cysts from solid tumors and should be used in addition to CT to evaluate cysts for the presence of septations or mural thickening, which would suggest cystadenoma or a cystadenocarcinoma. Duplex ultrasound is also particularly helpful as a dynamic study to identify vasculature in relation to tumor masses. Anesthetic Techniques Operative and perioperative morbidity and mortality have been decreased in part due to changes in anesthetic practices over the evolution of hepatic resection. A focus on maintaining low central venous pressure (CVP) can greatly reduce blood loss and keep the operative field clean for proper visualization of the biliary and vascular anatomy during parenchymal transaction. This is accomplished by positioning the patient in mild Trendelenberg and minimizing intravenous fluid to maintain systolic blood pressures above 90 mmHg and urine output to about 25 mL/h. If the IVC is still distended after mobilization of the liver, parenchymal transection can wait until central venous pressure is decreased through use of

narcotics, vasodilatory inhalation agents, or direct vasodilators. A central venous pressure of less than 5 mmHg can be maintained during the periods of liver mobilization and parenchymal transection. Though a cental venous catheter is a useful tool to follow the CVP, the surgeon can also look for a nondistended IVC and for blood coursing through flat intrahepatic veins. If transection is performed under Pringle control, bleeding is generally from hepatic veins, therefore, with a low hepatic venous pressure, even large tears in hepatic veins can be visualized to allow ligation or repair without massive hemorrhage. By Poiseuille’s law, blood flow is exponentially proportional to the radius of the vessel; therefore, even minor decreases in venous distention can decrease blood loss exponentially. With these techniques, the risk of postoperative renal failure has not been shown to be significant, nor has the risk of air embolism, which can be minimized, regardless, by keeping the patient in about 15° of Trendelenberg (26,27). Normal resuscitation is performed after the resection is completed and hemostasis has been achieved.

basic techniques Positioning, Skin Incision, and Exposure The patient should be positioned supine with the arms extended at right angles to the body. Any self-retaining retractor can be utilized, however, we prefer the Goligher retractor to elevate the costal margin, and this crossbar can be fitted to the table to form a 45° angle from top of the crossbar to the xyphoid. The patient should be prepped from the mid-chest to below the umbilicus, and draped to expose the right chest in the event a right thoracotomy is necessary to gain additional exposure. Though some groups routinely make a J-shaped thoracoabdominal incision, in our experience a thoracoabdominal incision was rarely necessary in over 1800 cases (2). We employ selective use of diagnostic laparoscopy based on the risk of unresectable disease (28), and conform the type of incision to the expected resection. For access to both lobes of the liver, a bilateral subcostal incision can be used with or without vertical midline extension. For the great majority of liver resections, we employ a “hockey stick” incision, which includes a right subcostal incision with vertical midline extension to the xyphoid. These incisions, when combined with the Goligher retractor, provide good exposure of the suprahepatic IVC, even with large right-sided tumors. We have found a higher rate of incisional hernia with a “Mercedes” incision compared to a “hockey stick” incision (29). For left-sided resections, a midline incision may suffice. Occasionally, when there is severe right-sided hepatic atrophy or exposure to the suprahepatic IVC is necessary for safety, extension into the right chest can be helpful (Fig. 3.1). Mobilization The ligamentum teres is divided between clamps and ligated, leaving a long secure ligature that is used as a handle to further expose the porta hepatis. The thin veil of the falciform ligament is incised along its length to free it from the anterior abdominal wall and expose the ligamentum teres. In obese individuals, the area where the falciform is fused to the anterior abdominal wall may be invested within a large fat pad. This fat pad can be removed with diathermy from beneath

25

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS both sides of the exposed fascia, improving exposure and aiding with fascial reapproximation at the end of the case. The falciform ligament is divided up to the suprahepatic IVC (Fig. 3.2). Bimanual palpation of the liver should be performed to assess the extent of hepatic disease. Segment 4 should be carefully retracted cephalad to expose the clear veil of lesser omentum anterior to the caudate lobe and attaching to the ligamentum venosum. This is incised, allowing palpation of

A

F

D

C

B

E

Figure 3.1 Incisions for liver resection. B-D, initial upper midline exploration. A-B-C, ideal for exposure of the whole liver (hockey stick). C-D-E, the classic chevron incision with A-D (Mercedes) extension. C-D, right subcostal incision. F, thoracoabdominal extension. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

the caudate and celiac axis, and providing access through the foramen of Winslow to the porta hepatis. Intraoperative ultrasound is used at this point to define the extent of disease, vascular relationships, and to confirm resectability. To mobilize the right liver, the leaf of the right coronary ligament is dissected from the falciform ligament and carefully incised over the IVC and territory of the right hepatic vein. This should be done sharply with downward traction on the liver and superior traction on the diaphragm. Once the right hepatic vein is identified, the right coronary ligament is taken close to the liver surface to its furthest extent laterally and the right triangular ligament is divided. To complete the mobilization, the right liver must be freed inferiorly. Omental and peritoneal attachments to the liver and gallbladder are divided to expose the inferior extent of the right triangular ligament. The retroperitoneal attachments are incised off the right adrenal gland and the liver can then be rotated medially to expose the retrohepatic IVC. If the right liver is to be resected or control of the right hepatic vein is needed, the multiple small venous branches from the IVC to the posterior liver must be individually dissected, controlled, and divided. Large accessory inferior right hepatic veins are common and may require division with a vascular stapler or control with vascular clamps and ligatures. It is critical for the surgeon on the left side of the table to retract the right liver medially to expose these branches and prevent injury to the cava. When all of these branches are ligated and divided, all that is left to expose the right hepatic vein will be a fibrous band of tissue that runs lateral to the vein, encircles the IVC, and courses posteriorly to the left and posterior border of the caudate, known as the caval ligament (Fig. 3.3). A tunnel can be safely created medial to this ligament and lateral to the right hepatic vein with a Kelly clamp or renal pedicle clamp in order to allow either a ligature or a

Figure 3.2 Mobilization of the liver begins with downward traction on the liver and division of the falciform ligament to the inferior vena cava. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

26

HEPATIC RESECTION vascular load endo-GIA staple fire. Once this is divided, the right liver is mobile and the lateral aspect of the right hepatic vein is exposed. The left liver is mobilized similarly, however since it does not lie on the vena cava, an extensive caval dissection is not necessary. Sharp and blunt dissection over the suprahepatic IVC will expose the groove between the right vein and the common trunk of the middle and left and middle hepatic veins. Downward traction on the liver and cephalad traction on the diaphragm help expose the left coronary ligament. The groove between the left and middle hepatic veins can be exposed with sharp dissection if there is no long intrahepatic common channel (Fig. 3.4). Care must be taken here to identify the phrenic vein as it courses on the underside of the diaphragm to enter the IVC, as it can be inadvertently injured if the triangular ligament is not properly exposed or not divided close to the liver surface (Fig. 3.5). As the left lateral segment is released

from its peritoneal attachments, it is also useful to place a hand or laparotomy pad under the left lateral segment and anterior to the caudate to provide traction and to protect the stomach, bowel, and spleen from diathermic injury. Further mobilization of the left liver can be accomplished by dividing the lesser omentum as well as the ligamentum venosum either at the left portal vein or left hepatic vein insertions to expose these vessels and the underlying caudate lobe.

vascular isolation Once the liver is mobilized, there are essentially three steps to safely perform a hepatectomy. These involve vascular inflow control, vascular outflow control, and parenchymal transection. Inflow Control All major hepatic resections require control of the vascular inflow to be accomplished safely. Furthermore, adequate

Figure 3.3 Multiple small venous branches from the IVC to the posterior liver must be individually dissected and divided. When all of these branches are controlled, all that is left to expose the right hepatic vein will be a fibrous band of tissue, the caval ligament. A tunnel can be safely created behind this ligament and above the right hepatic vein with a Kelly clamp. Once this is divided, the entire right liver is mobile and the venous outflow can be encircled and controlled. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

27

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS hepatic arterial and portal venous inflow must be maintained to the remnant liver. Selective inflow control may be achieved extrahepatically (30), intrahepatically during parenchymal transection (1,31), or by intrahepatic pedicle control via hepatotomies (32,33). In the extrahepatic approach, the hepatic artery and portal vein branches are dissected at the porta hepatis and controlled outside of the liver. In this approach, the individual artery and portal vein have to be separately identified and ligated since they have not yet entered the liver as a portal pedicle. The advantages of this approach are early vascular control prior to transection and demarcation of the liver on its surface. The disadvantages are a somewhat tedious dissection and the potential for injury to contralateral structures. The presence of tumor abutting the hilum may mandate extrahepatic inflow control. The right hepatic artery usually courses posterior to the common hepatic bile duct and can be dissected from the right side of the porta hepatis and controlled. Once divided, the proximal artery stump can be retracted anteriorly exposing the underlying portal vein. All branches must be carefully dissected and identified prior to division to insure that there is no compromise of flow to the future liver remnant, a potentially fatal complication. There is typically a small branch to the caudate process coming off the right portal vein proximally that may have to be controlled. As opposed to the short extrahepatic course from the hilum to the right liver, the vascular inflow to the left liver can be controlled in the umbilical fissure (34). The left portal vein and duct are mobilized by lowering the hilar plate. Here the left hepatic artery is typically found running cephalad along the left side of the porta hepatis anteriorly. It is prudent to insure that one has not inadvertently ligated the artery proximal to the right hepatic artery takeoff by confirming a pulse to the right liver. Once the left hepatic artery is divided, the underlying left portal vein can be dissected behind it. A branch to caudate lobe is very constant and should be preserved if the caudate is not going to be resected. Proximal dissection and identification of the right portal vein from the left side is worthwhile to confirm anatomy. Unless mandated by tumor proximity, we prefer to transect the bile duct (left or right) intrahepatically during parenchymal transection to absolutely avoid contralateral injury. This is especially important on the left side where there is often variant drainage of the major right sectoral ducts to the left hepatic duct. An alternative to extrahepatic inflow control at the hilum is intrahepatic control using a pedicle ligation technique. This technique is most appropriate for right-sided tumors not encroaching on the hilus. The portal triads carry Glisson’s capsule with them into the liver substance forming a sturdy pedicular sheath that can be dissected and clamped. Exposure of the pedicles can be accomplished by parenchymal transection down to the sheaths or by hepatotomies in the liver substance above the pedicle. For exposure of the right-sided inflow pedicle(s), hepatotomies are typically made along the inferior part of the gallbladder fossa and the caudate process and a large clamp is used to encircle the inflow structures. The whole right pedicle can be controlled this way or the right anterior and posterior sectoral pedicles can be encircled separately. The approach is rapid and avoids dissection of the contralateral

28

structures in the hilum, but risks injury to the pedicle before encircling the triad, or hemorrhage from coursing veins, which commonly run close to the pedicles. Though total vascular isolation has been employed by some groups (35–38), we have found that total vascular isolation techniques were not necessary in more than 1800 consecutive liver resections (2). Outflow Control Though there are multiple small veins that drain the right lobe and segment I directly into the retrohepatic vena cava, the majority of hepatic blood flow drains into the inferior vena cava (IVC) via the left, middle, and right hepatic veins. In

Figure 3.4 Sharp and blunt dissection over the suprahepatic IVC exposes the right, middle and left hepatic veins. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

Left phrenic vein

Diaphragm

Left triangular ligament

Diathermy

Figure 3.5 Downward traction on the liver and cephalad traction on the diaphragm help expose the left coronary ligament. Care must be taken here to identify the phrenic vein as it courses on the underside of the diaphragm to enter the IVC, as it can be inadvertently injured if the triangular ligament is not properly exposed or not divided close to the liver surface. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

HEPATIC RESECTION major hepatectomy, extrahepatic control of these vessels is preferred. Standard anatomy consists of a single right hepatic vein entering the vena cava, and a left and middle hepatic vein that is joined and entering the cava as a single trunk. Autopsy studies of the left and middle hepatic venous trunk have elucidated at least five types of hepatic vein trunk variants (39). The right hepatic vein is typically encircled after the dissection of the vena cava and caval ligament has been carried out as described earlier. The base of the right hepatic vein should be dissected sharply and once exposed, a clamp can be passed between the right and middle hepatic veins. Exposure of the left and middle hepatic vein extraheaptically can be challenging. The groove between the right and middle hepatic veins is initially developed from above the liver. The left liver is mobilized and the ligamentum venosum is divided just before its insertion into the left hepatic vein. Here a tunnel is carefully developed underneath the middle and left hepatic vein and they are encircled (Fig. 3.6). It is often difficult to individually encircle the left or middle hepatic vein extrahepatically but this depends on the anatomy of the common trunk. It is important to identify the hepatic venous anatomy on preoperative imaging and recognize variations in the branching patterns, since bleeding in this area can be difficult to control. Ligation of the hepatic venous outflow of the liver can also be accomplished during parenchymal transection with careful exposure of the cava and the origin of these veins once the liver has been transected to expose them. The exposures for specific resections are discussed later in the chapter. Parenchymal Transection Once vascular inflow and outflow to the lobe or segment has been controlled, all that remains is division of the liver parenchyma. There are many techniques to accomplish this. The instruments used are left to the surgeon’s preference, but it is imperative that the vessels and ducts divided be identified and dissected before division. Transection of the liver should be a deliberate dissection of intrahepatic structures rather than simply coagulation of liver tissue. In addition to the ability to confidently ligate each branch on the transection

(A)

line, it allows one to identify the venous drainage and pedicle inflow to the remnant liver. Moreover, in cases where the tumor margin is adjacent to major hepatic veins and portal pedicles, it allows precise extirpation of the tumor. For these reasons, we prefer a simple crushing technique. Glisson’s capsule is scored with diathermy along the transection line and a Kelly clamp is used to crush the liver tissue and expose the vessels and ducts for clipping, ligation, or bipolar energy sealing. Larger pedicles are suture ligated or stapled (40). The operative surgeon crushes the tissue in small linear planes, the assistant clips or seals the vessels, and the surgeon divides the structures. In this fashion, the transection line is quickly and efficiently completed. Though not always necessary, inflow occlusion with a Pringle maneuver may be used to decrease blood loss, and an entire lobe can often be transected with three to four sessions of 10–15 minutes on Pringle with 5 minutes off. After removal of the specimen, the raw surface is carefully inspected for bile leaks, which are suture ligated or clipped. Some groups advocate injection of dye or intralipid via the cystic duct to identify open biliary tributaries for ligation. Since drainage is associated with prolonged hospital stay, increased infection, and no change in a need for interventional radiology directed drainage, we do not routinely place drains after hepatic resection in the absence of biliary reconstruction (41).

major hepatic resection: definitions and specific considerations Multiple descriptions of liver anatomy and resections by anatomists and surgeons have resulted in terminologies that can be confusing and imprecise. A recent consensus conference in Brisbane, Australia, with the American Hepato-PancreatoBiliary Association has published new guidelines to clarify this nomenclature. When unclear, or if there is confusion about the description of a resection, one should revert to naming the numerical segments involved. The right liver is comprised of segments V–VIII and the left liver is comprised of segments II–IV. Appropriate terms for resection of the right or left liver would be “hepatectomy” or “hemi-hepatectomy.”

(B)

Figure 3.6 (A) Medial retraction of the left lateral segment exposes the ligamentum venosum. (B) The ligamentum venosum is divided sharply where it is tethered to the left hepatic vein, releasing the vein and enabling a tunnel to be dissected under the middle and left hepatic veins and anterior to the IVC. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

29

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Extending a right hepatectomy to include segment IV or a left hepatectomy to include segments V and VIII would be described as a “right/left trisectionectomy” or “trisegmentectomy.” Resection of segments II and III is often referred to as a “left lateral segmentectomy” or “sectionectomy.” There are essentially five types of major resection. The nomenclature of these resections is based on the anatomical classification (Table 3.1) (Fig. 3.7) (42–45). Right Hemihepatectomy (Right Hepatectomy, Right Hepatic Lobectomy) A right hemihepatectomy involves excision of segments V–VIII. The right lobe is completely mobilized and the right hepatic vein is isolated. The peritoneum overlying the common bile duct and extending into Calot’s triangle is incised to expose the cystic artery and duct. These are ligated and divided. A long tie is left on the proximal cystic stump and used as a retractor to help expose the common bile duct and dissect the vasculature. The hilar plate is lowered to protect the left hepatic duct. We typically do not dissect the right hepatic duct extrahepatically, but address it during parenchymal transection to absolutely avoid any potential for injury to the left hepatic duct. The right hepatic artery usually passes posterior to the common bile duct (Fig. 3.8) and is sharply dissected, ligated, and divided to the right of the common duct. Superior traction on the right hepatic artery stump will help expose the portal vein. The portal bifurcation is approached laterally and posteriorly. When dissecting the right portal vein, care should be taken to identify the first posterior branch to the right side of the caudate. Circumferential control of the right portal vein should not be attempted until this branch is identified and dissected as bleeding from this vein can be troublesome. Once a few centimeters of right portal vein are fully exposed and the left portal vein has been visualized, it is encircled and divided. Clamping of the right portal vein at this point should confirm demarcation of the right liver. Occasionally, the right anterior and posterior sectoral portal vein branches arise independently from the portal vein. In this instance, they must be individually dissected and ligated after confirming flow to the left liver. The right hepatic vein is isolated and divided as described previously. It is important that all the retrohepatic veins are first controlled and divided, that the dissection extends to the

left of the IVC, and that the right hepatic vein is skeletonized completely right at the liver surface. It is especially important to gain extrahepatic control of the vein with large tumors near the hepatic venous confluence or in the posterior sector near the vena cava, where it can be difficult to obtain tumor clearance without excessive traction on the vein. Alternatively, the right hepatic vein can also be controlled from within the liver during parenchymal transection, however, this usually forces the hepatic transection to the right of the true principal midliver plane. After inflow ligation, a line of demarcation becomes evident. Figure-of-eight stay sutures are placed to either side of this line and parenchymal transection can begin safely. The surgeon’s left hand lifts the left lobe from above the IVC carefully as the transection plane is deepened. This will expose the middle hepatic vein, and division of the specimen can proceed to the right or left of the vein depending on tumor clearance. As the dissection proceeds superiorly, the segment V and then VIII hepatic veins are divided along the middle hepatic vein. The main right portal pedicle is exposed and divided with the endo-GIA stapler. This will control the right hepatic duct if it was not controlled extrahepatically. Alternatively, an anterior approach can be used to resect the right lobe of the liver. This approach is advantageous when the right lobe cannot be mobilized due to a large right-sided tumor, or there is a large mass adherent to the diaphragm or IVC (46). In this approach, after extrahepatic inflow division, the liver is transected without mobilization. It is then freed from its venous and ligamentous attachments to the IVC and peritoneum. The parenchyma is transected from the anterior liver surface to the IVC along the line of demarcation, and venous tributaries are controlled from the front, including the right hepatic vein (47,48). To help control bleeding in the deeper parenchymal plane, the “hanging maneuver” may be employed (49). In this maneuver, the anterior plane of the IVC is dissected from the liver undersurface. The most inferior veins draining the caudate are ligated and divided, and a tunnel is carefully created anterior to the IVC to the space between the right and middle hepatic veins with a Kelly clamp. This is a blind tunnel of 4 to 6 cm. A tape is passed that can then be used to elevate the liver away from the anterior surface of the IVC, helping to define the plane of transection and facilitating exposure of the deeper tissues. In this technique, the right portal pedicle

Table 3.1 Anatomy and Classification of Major Hepatic Resections Anatomic Classification Couinaud

Goldsmith and Woodburne

Brisbane

Segments resected

Right hepatectomy Right lobectomya Left hepatectomy Extended left hepatectomya Left lobectomy

Right hepatic lobectomy Extended right hepatic lobectomy Left hepatic lobectomy Extended left lobectomy Left lateral segmentectomy

Right hemihepatectomy Right trisectionectomy Left hemihepatectomy Left trisectionectomy Left lateral sectionectomy

V, VI, VII, VIII IV,V,VI, VII, VIIIb II, III, IV II, III, IV, V, VIIIb II, III

a

Often referred to as trisegmentectomy. May also include segment I.

b

30

HEPATIC RESECTION is divided, parenchymal transection is completed to the IVC, the lateral venous attachments to the IVC are ligated and divided, the right hepatic vein is stapled, the coronary and triangular ligaments are divided, and the specimen is removed. Right Trisectionectomy (Right Lobectomy, Extended Right Lobectomy, Right Trisegmentectomy) A right trisectionectomy is a right hemihepatectomy extended to include segment IV. The liver is mobilized as described for a right hepatectomy. To approach the inflow and outflow of segment IV, the ligamentum teres is elevated to expose the umbilical fissure. If a bridge of tissue between segments III and IV is present concealing the fissure, this should be divided with diathermy. Here, the ligamentum teres can be traced to its embryologic origin at the left portal vein. Incising the fibrous tissue that tethers the left main pedicle to the base of the umbilical fissure releases the left-sided structures from the

undersurface of segment IV, and it opens up the fissure. To safely perform a right trisectionectomy, the left hepatic duct should be freed clear of the proposed plane of transection. This is accomplished by lowering the hilar plate as previously described. The inflow and outflow to the right liver are controlled and divided as previously described. Once the right hepatic vein has been divided, the middle vein can usually be encircled. The liver tissue is divided to the right of the falciform ligament and the pedicles feeding segments IVa and IVb are ligated and divided as they come off the main left pedicle (Fig. 3.9). Unless tumor mandates, a deliberate dissection within the umbilical fissure is usually not necessary. As the plane of transection is deepened toward the IVC superiorly, the middle hepatic vein is encountered, dissected, and divided with a stapler. It is absolutely critical to protect the left hepatic vein as narrowing or transection of this vein will likely result in liver failure or massive hemorrhage secondary to a lack of other venous return from the liver.

(A)

(B)

(C)

(D)

(E) Figure 3.7 The anatomy and classification of major hepatic resections. (A) right hepatectomy, (B) left hepatectomy, (C) left lobectomy, (D) extended left hepatectomy, (E) right lobectomy.

31

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

Figure 3.9 To expose and control the portal pedicles to segment IV, the liver tissue is divided to the right of the falciform ligament and the pedicles feeding segments IVA and IVB are ligated and divided as they come off the main left pedicle. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

Figure 3.8 During right hepatectomy, the right hepatic artery usually passes posterior to the common bile duct and is sharply dissected, ligated, and divided to the right of the duct. After cholecystectomy, retraction of the cystic duct will expose the underlying artery. Source: Blumgart; Surgery of the Liver, Biliary Tract and Pancreas, 4th Edition; Chapter 80; copyright Elsevier.

Left Hemihepatectomy (Left Hepatectomy, Left Hepatic Lobectomy) A left hemihepatectomy involves excision of segments II– IV. The left lobe of the liver is mobilized, the umbilical fissure is exposed, and the hilar plate is lowered as previously described. The gastrohepatic ligament is entirely divided, with care taken to identify any accessory or replaced left hepatic arteries not identified on preoperative imaging. The left hepatic artery is dissected at the base of the umbilical fissure and divided. The caudate branch of the portal vein is identified before the left main portal vein enters the umbilical fissure. If the caudate lobe is to be spared, the portal vein is ligated and divided distal to this vein. A line of demarcation marking the right-sided border of segment IV corresponds with a plane that usually extends from the IVC to the base of the gallbladder fossa (“Cantlie’s line”). This “principle plane” is the same as that seen in a right hemihepatectomy. Segments II and III are reflected medially and the middle and left hepatic veins are identified, encircled and divided extrahepatically as described earlier. The left hepatic vein is often not amenable to circumferential extrahepatic exposure initially but can be exposed after splitting the liver back to its origin. Parenchymal transection completes the excision. Left Trisectionectomy (Extended Left Hepatectomy, Extended Left Lobectomy, Left Trisegmentectomy) A left trisectionectomy involves removal of segments II, III, IV, V, and VIII. The entire liver is mobilized. The inflow and outflow to the left lobe are controlled as previously

32

described for a left hemihepatectomy. The inflow to segments V and VII can be addressed in a few ways. The anterior sectoral pedicle can be encircled intrahepatically either through hepatotomies or after transection in the right scissura to the left of the right hepatic vein. The pedicle can be encircled and clamped confirming flow the posterior sector. Alternatively, an extensive hilar dissection can be carried out to identify and divide the arterial and portal branches to the right anterior sector. It is critical that preoperative imaging is reviewed for anatomic variations in the inflow and outflow to the right liver. Once the anterior sectoral inflow is divided, a near horizontal line of demarcation becomes evident anterior to the right hepatic vein and dividing the right anterior and posterior sectors. Parenchymal transection continues anterior to the right hepatic vein and the specimen is removed. The middle hepatic vein is necessarily taken as part of this resection and is addressed as described earlier. A left trisectionectomy is a challenging operation that requires significant experience with major hepatic resections. Left Lateral Sectionectomy (Left Lobectomy, Left Lateral Segmentectomy) A left lateral sectionectomy involves removal of segments II and III. The left lobe of the liver is mobilized and the hilar plate is lowered as previously described. Just to the left of the umbilical fissure, the portal pedicles to segments II and III are identified and divided. These can be identified and controlled through multiple hepatotomies or during parenchymal transaction in a plane just to the left of the falciform ligament. A deliberate dissection in the umbilical fissure is usually not necessary. The left hepatic vein is usually divided after parenchymal resection back to its origin but can also be controlled extrahepatically as described in the “outflow control” section of the chapter.

HEPATIC RESECTION

wedge vs. segmental resection The segmental anatomy of the liver, as defined by Couinaud, divides the liver into eight independent segments, each with its own inflow and biliary drainage (see chapter 1) (42,50). As a result, each segment can be individually resected without affecting the inflow or outflow to the rest of the liver. Segment-oriented hepatectomy spares normal parenchyma and is particularly useful when bilateral noncontiguous segments are involved or in patients with chronic liver disease. Nonanatomic wedge resections can be useful for small peripheral tumors that are not close to major inflow pedicles or venous branches for which adequate tumor margins can be obtained. Though some groups have shown that anatomical resection is not superior to wedge resection for tumor clearance, pattern of recurrence, or survival (51), in our experience anatomic segmental resection resulted in improved tumor clearance and patient survival compared to wedge resection (52). Wedge excision may risk fracturing the plane between the tumor and normal liver, margin positivity, and intraoperative hemorrhage (12,53). Anatomic resection may provide better visibility, decrease the risk of major hemorrhage, and in many cases provide a wider margin of resection. Segmentectomy I (Caudate Resection) The caudate lobe is often resected with a right or left hemihepatectomy, however, isolated caudate resection may be performed for solitary tumors in segment I. The anatomy of the caudate lobe between the IVC, portal triad, and hepatic veins can make resection tedious and challenging. The caudate lobe straddles both hepatic lobes and therefore receives vascular inflow from both the right and left portal pedicles (54). Venous drainage is directly into the IVC via one to nine short hepatic veins (55). The left edge of the caudate fuses with the IVC via a fibrous band of tissue that encircles the IVC and attaches to segment VII. In many patients, this caval ligament may be composed of liver parenchyma. Dissection at the base of the umbilical fissure exposes the caudate branches of the left portal vein and hepatic artery for ligation and division. Segments II and III of the liver are mobilized and reflected to the right, exposing the caudate where it lies on the IVC. The left lateral attachments of the caudate to the IVC are divided (56). Exposure and division of the left caval ligament can be challenging and care should be taken to avoid injury to the cava inferiomedially and the base of the left and middle hepatic veins superiorly. With anterior traction on the caudate, the short hepatic veins draining into the IVC on the posterior aspect of the caudate can be visualized and controlled. If there is a bulky tumor in the caudate or anterior traction of the lobe is difficult, the retrohepatic veins can be approached from the right side, by mobilizing the right lobe and turning it to the left, then dissecting and dividing all the veins starting below the caudate and continuing onto the anterior surface of the IVC (57). The caudate branch from the right portal vein should also be identified and ligated. To complete the resection, the tissue joining the caudate to segment VII must be transected. Anteriorly and superiorly,

care must be taken to avoid injury to the middle and left hepatic veins. Segments II or III The approach to excising either segment II or III is the same as that for a left lateral segmentectomy, except the plane between the segments needs to be defined. This plane is identified by the course of the left hepatic vein, segment 3 being anterior and segment 2 being posterior. Inflow control to either segment is achieved in the umbilical fissure. Ligation of the portal pedicle will guide resection along the plane of demarcation. Care must be taken to divide the branches of the left hepatic vein draining the excised segment, but to leave the main left hepatic vein intact to drain the remnant liver. Segment IV As described in a right trisectionectomy, the inflow to segment IV is found to the right of the umbilical fissure. The hilar plate is lowered to protect the left bile duct and to provide access to the multiple pedicles to segment IV. Ligation of these pedicles will provide a line of demarcation along Cantlie’s line. During parenchymal transection, the venous drainage of segments IVa and IVb are divided sequentially to the left of the middle vein on the lateral border of the segment, and along the umbilical fissure on the medial border of the segment, where the umbilical vein often courses. The middle hepatic vein can be sacrificed in this operation if necessary with adequate drainage of the right liver and segments 2 and 3. Segments V and VIII (Anterior Sector) The inflow to segments V and VIII are from the right anterior sectoral pedicle. This can be approached and controlled extrahepatically or intrahepatically as described for a left trisectionectomy. If the anterior and posterior sectoral pedicles branch within the liver parenchyma, a hepatotomy over the anterior pedicle is necessary. Alternatively, the liver can be transected in the principal plane down to the base of the anterior sector where its origin can be controlled, typically posterior to terminal middle hepatic vein branches. The anterior sector lies between the right and middle hepatic veins, i.e., between Cantlie’s line and a transverse plane anterior to segments VI and VII. This horizontal plane of transection can be better defined by clamping the anterior pedicle to demarcate the right, left, and posterior borders. The transection line between V and VIII is demarcated and defined intrahepatically when control of the isolated segmental inflow is obtained. The middle hepatic vein can usually be safely divided in this operation if necessary, but in the absence of a large accessory right hepatic vein, the right hepatic vein must be preserved for adequate drainage of the posterior sector. Segments VI and VII (Posterior Sector) The inflow to segments VI and VII are from the posterior sectoral pedicle. This can often be approached and controlled in the fissure of Ganz, though the anatomy of anterior and posterior pedicles can be highly variable. If the anterior and posterior sectoral pedicles branch within the liver parenchyma,

33

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS the portal pedicles must be approached during parenchymal transection. The medial plane of transection can be better defined by clamping the posterior pedicle to demarcate the border. The classic description of a single pedicle from the posterior sectoral pedicle feeding either segment VI or VII is the exception rather than the rule (58), therefore, careful parenchymal dissection, preoperative study of the CT, and intraoperative ultrasound are critical to these resections. If a posterior sectorectomy is to be performed, the right hepatic vein can be sacrificed since the anterior sector drains into the middle hepatic vein. Central Hepatectomy (Segments IV, V, and VIII) A central hepatectomy with various amount of extension into any of the three segments can be performed combining the techniques described above. Typically this requires dividing the middle hepatic vein intrahepatically near its origin. The techniques of a segment IV resection and anterior sectorectomy are essentially combined. This is a challenging operation that requires a substantial surface of liver to be transected but can be very useful to spare parenchyma while removing centrally placed tumors.

enucleation of benign tumors ₍see chapters 28, 32, and 33₎ When indicated, hepatectomy for benign conditions should be parenchymal preserving. Though anatomic resection along segmental planes is sometimes necessary, some benign tumors may be enucleated, for example, adenomas, fibronodular hyperplasia, metastatic neuroendocrine tumors, and hemangiomas (2,4). Hemangiomas in particular push liver tissue away as they grow, and create a fibrolamellar plane of tissue that defines the border between cavernous tissue and normal liver parenchyma (2). The arterial supply to the hemangioma can be determined from preoperative imaging and is clamped, allowing the tumor to decompress via the venous outflow. The hepatic tissue over the mass is then incised to enter an avascular plane surrounding the tumor. Small vessels that traverse this plane are ligated and divided. The majority of the dissection can be done with the surgeon’s finger, and the mass is shelled out. This approach preserves normal parenchyma, eliminates the need for hepatic venous outflow control, limits blood loss, and has fewer complications than lobectomy (3,4). Management of benign lesions is covered in further detail in other chapters.

conclusion Major hepatic resections for benign and malignant tumors can be accomplished safely and efficaciously. Proper patient selection, precise preoperative imaging, specific anesthetic techniques, and knowledge of the principal complications are essential. Study of each patient’s segmental anatomy will allow inflow and outflow control and the ability to tailor the resection needed for each individual.

references 1. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: Analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002;236(4):397–407.

34

2. Baer HU, Dennison AR, Mouton W, et al. Enucleation of giant hemangiomas of the liver: Technical and pathologic aspects of a neglected procedure. Ann Surg 1992;216(6):673–6. 3. Gedaly R, Pomposelli JJ, Pomfret EA, Lewis WD, Jenkins RL. Cavernous hemangioma of the liver: anatomic resection vs. enucleation. Arch Surg 1999 Apr;134(4):407–11. 4. Yoon SS, Charny CK, Fong Y, et al. Diagnosis, management, and outcomes of 115 patients with hepatic hemangioma. J Am Coll Surg 2003;197(3): 392–402. 5. Are C, Gonen M, Zazzali K, et al. The impact of margins on outcome after hepatic resection for colorectal metastasis. Ann Surg 2007 Aug;246(2): 295–300. 6. Cady B, Jenkins RL, Steele Jr GD, et al. Surgical margin in hepatic resection for colorectal metastasis: A critical and improvable determinant of outcome. Ann Surg 1998;227(4):566–71. 7. Cady B, Stone MD, McDermott Jr WV, et al. Technical and biological factors in disease-free survival after hepatic resection for colorectal cancer metastases. Arch Surg 1992;127(5):561–9. 8. Ekberg H, Tranberg KG, Andersson R. Determinants of survival in liver resection for colorectal secondaries. Br J Surg 1986;73(9):727–31. 9. Elias D, Cavalcanti A, Sabourin JC, et al. Resection of liver metastases from colorectal cancer: The real impact of the surgical margin. Eur J Surg Oncol 1998;24(3):174–9. 10. Kato T, Yasui K, Hirai T, et al. Therapeutic results for hepatic metastasis of colorectal cancer with special reference to effectiveness of hepatectomy: Analysis of prognostic factors for 763 cases recorded at 18 institutions. Dis Colon Rectum 2003;46:522–31. 11. Ohlsson B, Stenram U, Tranberg KG. Resection of colorectal liver metastases: 25-year experience. World J Surg 1998;22(3):268–77. 12. Scheele J, Stang R, Altendorf-Hofmann A, Paul M. Resection of colorectal liver metastases. World J Surg 1995;19(1):59–71. 13. Shirabe K, Takenaka K, Gion T, et al. Analysis of prognostic risk factors in hepatic resection for metastatic colorectal carcinoma with special reference to the surgical margin. Br J Surg 1997;84(8):1077–80. 14. Hamady ZZR, Cameron IC, Wyatt J, et al. Resection margin in patients undergoing hepatectomy for colorectal liver metastasis: A critical appraisal of the 1 cm rule. Eur J Surg Oncol 2006;32(5):557–63. 15. Pawlik TM, Scoggins CR, Zorzi D, et al. Effect of surgical margin status on survival and site of recurrence after hepatic resection for colorectal metastases. Ann Surg 2005;241(5):715–24. 16. Mimica Z, Pogorelic Z, Perko Z, et al. Effect of surgical incision on pain and respiratory function after abdominal surgery: a randomized clinical trial. Hepatogastroenterology 2007;54(80):2216–20. 17. Llovet JM, Fuster J, Bruix J. The Barcelona approach: diagnosis, staging, and treatment of hepatocellular carcinoma. Liver Transpl 2004;10 (2 Suppl 1):S115–20. 18. Zorzi D, Laurent A, Pawlik TM, et al. Chemotherapy-associated hepatotoxicity and surgery for colorectal liver metastases. Br J Surg. 2007;94(3): 274–86. 19. Kooby DA, Fong Y, Suriawinata A, et al. Impact of steatosis on perioperative outcome following hepatic resection. J Gastrointestinal Surg 2003;7(8):1034–44. 20. Siegelman ES, Rosen MA. Imaging of hepatic steatosis. Semin Liver Dis 2001;21(1):71–80. 21. Orlacchio A, Bolacchi F, Cadioli M, et al. Evaluation of the severity of chronic hepatitis C with 3-T1H-MR spectroscopy. AJR Am J Roentgenol 2008;190(5):1331–9. 22. Abulkhir A, Limongelli P, Healey AJ, et al. Preoperative portal vein embolization for major liver resection: a meta-analysis. Ann Surg 2008;247(1):49–57. 23. Covey AM, Brown KT, Jarnagin WR, et al. Combined portal vein embolization and neoadjuvant chemotherapy as a treatment strategy for resectable hepatic colorectal metastases. Ann Surg 2008;247(3):451–5. 24. Belghiti J. Arguments for a selective approach of preoperative portal vein embolization before major hepatic resection. J Hepatobiliary Pancreat Surg 2004;11(1):21–4. 25. Lamade W, Glombitza G, Fischer L, et al. The impact of 3-dimensional reconstructions on operation planning in liver surgery. Arch Surg 2000;135(11):1256–61.

HEPATIC RESECTION 26. Cunningham JD, Fong Y, Shriver C, et al. One hundred consecutive hepatic resections: Blood loss, transfusion, and operative technique. Arch Surg 1994;129(10):1050–6. 27. Melendez JA, Arslan V, Fischer ME, et al. Perioperative outcomes of major hepatic resections under low central venous pressure anesthesia: Blood loss, blood transfusion, and the risk of postoperative renal dysfunction. J Am Coll Surg 1998;187(6):620–5. 28. D’Angelica M, Fong Y, Weber S, et al. The role of staging laparoscopy in hepatobiliary malignancy: prospective analysis of 401 cases. Ann Surg Oncol 2003 Mar;10(2):183–9. 29. D’Angelica M, Maddineni S, Fong Y, et al. Optimal abdominal incision for partial hepatectomy: increased late complications with Mercedes-type incisions compared to extended right subcostal incisions. World J Surg 2006;30(3):410–8. 30. Blumgart L. Hepatic resection. In: Dudley HAF, Rob C, Smith of Marlow RS, Pories WJ, eds. Rob & Smith’s operative surgery, 4th edn. London, Boston: Butterworth Scientific, 1982:v. 31. Tung TT. Les Re?sections Majeures et Mineures du Foie. 1979. 32. Launois B, Jamieson GG. The posterior intrahepatic approach for hepatectomy or removal of segments of the liver. Surg Gynecol Obstetrics 1992;174(2):155–8. 33. Launois B, Jamieson GG. The importance of Glisson’s capsule and its sheaths in the intrahepatic approach to resection of the liver. Surg Gynecol Obstetrics 1992;174(1):7–10. 34. Hepp J, Couinaud C. [Approach to and use of the left hepatic duct in reparation of the common bile duct.]. Presse Med. 1956;64(41):947–8. 35. Delva E, Camus Y, Nordlinger B, et al. Vascular occlusions for liver resections. Operative management and tolerance to hepatic ischemia: 142 cases. Ann Surg 1989;209(2):211–8. 36. Emond J, Wachs ME, Renz JF, et al. Total vascular exclusion for major hepatectomy in patients with abnormal liver parenchyma. Arch Surg 1995;130(8):824–30; discussion 30–1. 37. Emre S, Schwartz ME, Katz E, Miller CM. Liver resection under total vascular isolation. Variations on a theme. Ann Surg 1993;217(1):15–19. 38. Hannoun L, Borie D, Delva E, et al. Liver resection with normothermic ischaemia exceeding 1 h. Br J Surg. 1993;80(9):1161–5. 39. Nakamura S, Tsuzuki T. Surgical anatomy of the hepatic veins and the inferior vena cava. Surgery Gynecol Obstetrics 1981;152(1):43–50. 40. Fong Y, Blumgart LH. Useful stapling techniques in liver surgery. J Am Coll Surgeons 1997;185(1):93–100. 41. Fong Y, Brennan MF, Brown K, Heffernan N, Blumgart LH. Drainage is unnecessary after elective liver resection. Am J Surg 1996;171(1):158–62.

42. Couinaud C. Le foie. Etudes anatomiques et chirurgicales. Le Foie: Etudes Anatomiques et Chirurgicales. 1957. 43. Goldsmith NA, Woodburne RT. The surgical anatomy pertaining to liver resection. Surg Gynecol Obstet 1957;105:310–8. 44. Starzl TE, Iwatsuki S, Shaw BW, Jr, et al. Left hepatic trisegmentectomy. Surg Gynecol Obstet 1982;155(1):21–7. 45. Starzl TE, Koep LJ, Weil R, 3rd, et al. Right trisegmentectomy for hepatic neoplasms. Surg Gynecol Obstet 1980;150(2):208–14. 46. Chik BH, Liu CL, Fan ST, et al. Tumor size and operative risks of extended right-sided hepatic resection for hepatocellular carcinoma: implication for preoperative portal vein embolization. Arch Surg 2007;142(1):63–9; discussion 9. 47. Lai EC, Fan ST, Lo CM, Chu KM, Liu CL. Anterior approach for difficult major right hepatectomy. World J Surg 1996;20(3):314–7; discussion 8. 48. Lai ECS, Fan ST, Lo CM, et al. Hepatic resection for hepatocellular carcinoma: An audit of 343 patients. Ann Surg 1995;221(3):291–8. 49. Belghiti J, Guevara OA, Noun R, Saldinger PF, Kianmanesh R. Liver hanging maneuver: a safe approach to right hepatectomy without liver mobilization. J Am Coll Surg 2001;193(1):109–11. 50. Couinaud C. Bases anatomiques des hepatectomies gauche et droite reglees. J Chir 1954;70:933–66. 51. Zorzi D, Mullen JT, Abdalla EK, et al. Comparison between hepatic wedge resection and anatomic resection for colorectal liver metastases. J Gastrointest Surg 2006;10(1):86–94. 52. DeMatteo RP, Palese C, Jarnagin WR, et al. Anatomic segmental hepatic resection is superior to wedge resection as an oncologic operation for colorectal liver metastases. J Gastrointest Surg 2000;4(2):178–84. 53. Polk W, Fong Y, Karpeh M, Blumgart LH. A technique for the use of cryosurgery to assist hepatic resection. J Am Coll Surg 1995;180(2):171–6. 54. Mizumoto R, Suzuki H. Surgical anatomy of the hepatic hilum with special reference to the caudate lobe. World J Surg 1988;12(1):2–10. 55. Heloury Y, Leborgne J, Rogez JM, et al. The caudate lobe of the liver. Surg Radiol Anat 1988;10(1):83–91. 56. Takayama T, Makuuchi M. Intraoperative ultrasonography and other techniques for segmental resections. Surg Oncol Clin N Am 1996;5(2):261–9. 57. Lerut J, Gruwez JA, Blumgart LH. Resection of the caudate lobe of the liver. Surgery Gynecol Obstetrics 1990;171(2):160–2. 58. Hata F, Hirata K, Murakami G, Mukaiya M. Identification of segments VI and VII of the liver based on the ramification patterns of the intrahepatic portal and hepatic veins. Clin Anat 1999;12(4):229–44.

35

4

Ultrasound for HPB disorders Duan Li and Lucy Hann

introduction Ultrasound is the initial study of choice in most clinical situations due to the lack of ionizing radiation, relatively low cost, and accessibility in varied settings such as at the bedside or in the operating suite. Ultrasound differs from other crosssectional imaging techniques in that it uses sound propagation and reflection from interfaces within tissue for imaging. Images are generated by piezoelectric material within the transducer that transmits and receives the sound signal. Higher frequency transducers provide the best resolution, but high frequencies are attenuated more rapidly in tissue. For that reason, transducer frequency is selected for the application. Superficial structures are evaluated at frequencies in the range of 6 to 18 MHz and transabdominal ultrasound, which requires better penetration, typically uses frequencies ranging from 3 to 6 MHz. Doppler is a unique feature of ultrasound for imaging vessels and blood flow. When moving blood is insonated, the frequency of the returning signal is proportional to blood velocity. A cursor is placed over a specific blood vessel and images are obtained in both gray scale and Doppler (termed “Duplex scanning”). The Doppler information can then be displayed in three different formats: (1) spectral Doppler, (2) color Doppler, and (3) power Doppler. Spectral Doppler shows a waveform with velocity changes and flow direction over time. Color Doppler displays mean velocities and direction of flow within vessels. The color codes assigned for velocities are usually displayed in the upper left aspect of the image. Power Doppler gives the amplitude of the Doppler signal without direction or frequency information; since it is not angledependent, it is very useful for imaging low flow and tortuous vessels. Ultrasound contrast agents further improve applications for vascular imaging. Current contrast agents use microbubbles encapsulated within thin lipid spheres. After intravenous injection, the microbubbles remain intravascular and do not diffuse into the interstitium as do MRI and CT contrast agents. After a low-power ultrasound signal is applied, the microbubbles oscillate (expand and contract) at harmonic frequencies that are detected by the transducer (1,2). With these ultrasound contrast agents, it is now possible to image tumor vasculature in exquisite detail (3–7) (Fig. 4.1). Despite the versatility of ultrasound, there are limitations. Sound is reflected at bone and air interfaces so scans are obtained from different positions to avoid intestinal air or rib artifact. This lack of standardized perspective compared to axial imaging format of CT and MRI may present difficulty for referring clinicians who are unfamiliar with the technique. To lessen bowel gas interference, 6-hour fast is recommended to improve visualization of the pancreas and liver and to provide sufficient gallbladder distension. Another significant limitation

36

is that ultrasound is operator-dependent; skilled technologists and radiologists are essential since diagnosis is made at image acquisition. For best results, the surgeon should communicate to the radiologist the specific clinical questions so that appropriate targeted images can be obtained at the time of the examination. This chapter will discuss ultrasound applications for diagnosis of hepatic, gallbladder, biliary, and pancreatic abnormalities. The role of specialized ultrasound techniques such as endoscopic ultrasound and intraoperative ultrasound will also be addressed.

liver Anatomically the liver is divided into sectors that are defined by the scissurae that contain the hepatic veins; these sectors are then subdivided into individual hepatic segments that each contain intact portal and arterial inflow and hepatic venous outflow and draining bile ducts (8,9). Ultrasound hepatic anatomy is shown in Fig. 4.2. Diffuse Liver Disease Diffuse liver abnormalities include fatty infiltration, hepatitis, and cirrhosis. Hepatic steatosis is present in 17% to 33% of the general population and in 70% of overweight individuals (10). On ultrasound, the liver has diffusely increased echogenicity and in advanced cases significant sound beam attenuation obscures the deep liver. Areas of focal sparring may be seen anterior to the portal confluence and adjacent to the gallbladder. Hepatic steatosis impacts perioperative outcome and accurate preoperative diagnosis would be useful (11). Fatty infiltration increases liver stiffness, which can be measured by tissue displacement in response to the transmitted ultrasound wave. These elastography techniques hold promise for diagnosis of diffuse infiltrative liver diseases such as hepatic steatosis and early-stage hepatic fibrosis (12–14). Focal Hepatic Lesions Cystic lesions Ultrasound is the best modality to differentiate cystic from solid liver masses and to determine the internal architecture of cystic lesions. Simple cysts, found in 2% to 3% of patients (15), have thin wall, no internal echoes, and bright posterior enhancement. Even if the cyst is lobulated or has thin septation, benign diagnosis can be made (16). Symptomatic large simple cysts may be treated with ultrasound-guided aspiration and sclerosis (15,17), but it is extremely important to assess the cyst wall. Mural nodularity, thick tumor rim, and internal vascularity may indicate neoplasm such as biliary cystadenoma and these lesions should not be unroofed or aspirated since complete surgical resection is required. Cystic liver metastases present as complex cysts often with solid or

ULTRASOUND FOR HPB DISORDERS

(A)

(B)

Figure 4.1 Microbubble contrast enhanced ultrasound image of a hypervascular liver mass. (A) Contrast enhanced image shows the intense hypervascularity of this liver lesion (arrow) that proved to be focal nodular hyperplasia. (B) The lesion (arrows) is subtle on the corresponding grayscale image. (Complements of Siemens Medical Solutions, Ultrasound Division. Malvern, PA.)

irregular rim. These are typically from sarcoma, cystadenocarcinomas of the ovary and pancreas, and mucinous colon carcinoma primaries (16,18). Ovarian metastases are characteristically peripheral implants. Squamous cell tumors with necrosis appear as cystic masses and other metastases may cavitate in response to chemotherapy. The appearance of cyst contents on ultrasound can be used for differential diagnosis. Pyogenic abscess initially may be echogenic and later liquified with debris, fluid-fluid levels, and irregular wall (Fig. 4.3). Echogenic reflections with reverberations, seen in 20% to 30% of cases, suggest air within the abscess (18). The classic echinococcal cyst is a complex cyst with well-defined wall, containing double echogenic lines. Multiple, internal echogenic foci, “snowstorm signs” settle in the dependent portions of the cyst. Localized splits in the cyst wall, with floating, undulating membranes, are also characteristic and the cyst wall may calcify (19,20). Hematomas in the acute stage may be echogenic and then they have layering lowlevel echoes from blood, and later become honeycombed with septation. When a preexisting cyst becomes hemorrhagic, internal septation may be thick and irregular, but they float freely in real-time and are not rigid. Solid Liver Lesions Solid liver lesions are further characterized by lesion echogenicity, vascularity, and peripheral halo. Definitive diagnosis of benignity can be made for hemangiomas, focal fatty infiltration, and focal fatty sparing because of their classic ultrasound features. Benign focal nodular hyperplasia can also be identified when the characteristic “spokewheel” vascular pattern, tortuous feeding artery, and marked hypervascularity are seen on contrast-enhanced ultrasound or Doppler images (Fig. 4.1). Hypoechoic liver masses and lesions with a peripheral halo are suspicious for malignancy. Although CT and MRI are used for tumor staging, there can be added benefit from ultrasound to

(1) assess lesions that are “too small to characterize” by CT, (2) define the relationship of tumor to bile ducts, and (3) evaluate vascular encasement and tumor margin (Fig. 4.4). Typical hemangioma, seen in 70% to 80% of cases, is a uniformly echogenic mass with sharp margin (21) (Fig. 4.5A). The multiple vascular interfaces within the hemangioma cause the increased echogenicity and margin is well-demarcated since histopathologically hemangiomas lack a capsule. Hemangiomas have absent or minimal flow on Doppler imaging; they are never hypervascular. Another common appearance of hemangioma is a mass with thin peripheral echogenic rim with mixed central echogenicity (Fig. 4.5B). Giant hemangiomas > 5 cm often lack these characteristic ultrasound features because of central fibrosis, necrosis, and myxomatous degeneration. A study of 213 patients with typical hemangioma appearance and without risk for hepatic malignancy found only one patient with malignancy on long term follow-up and concluded that typical hemangiomas in low-risk patients do not require follow-up (22). This rule does not apply to patients with cirrhosis, hepatitis, or chronic liver disease that places them at increased risk for hepatocellular carcinoma, nor does it apply to patients who already have malignancies, and particularly not to those with primary tumors that exhibit echogenic metastases. Caturelli et al. (23) studied 2,000 patients with cirrhosis. Of these, 44 had hemangioma-like lesions. On follow-up, half proved to be hepatocellular carcinomas and half hemangiomas. Thus, in patients at risk for hepatocellular carcinoma, any echogenic lesion merits further evaluation or follow-up. Other benign conditions such as focal fatty infiltration and focal sparing are diagnosed by geographic margins and typical location in segment 4 anterior to the portal vein bifurcation or less commonly, adjacent to the gallbladder. Focal fat appears echogenic relative to normal liver and areas of focal sparing are less echogenic than fatty infiltrated liver. A useful finding on Doppler evaluation is that vessels cross undisturbed

37

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS without displacement through areas of focal fat or focal sparring (24,25). A hypoechoic halo around a liver lesion indicates a clinically significant mass, suspicious for malignancy, including hepatocellular carcinoma and hepatic adenoma and metastases from colorectal, gastrointestinal, neuroendocrine, renal cell, choriocarcinoma, and vascular primaries such as Kaposi sarcoma (Fig. 4.4). Pathologically, the halo is caused by proliferating malignant cells, compression of the liver parenchyma, and dilated sinusoids. The hypoechoic halo sign has a 95% positive predictive value and an 87% negative value for differentiating

metastases from hemangioma (26,27). A hypoechoic halo may be seen even in small lesions 1, 5. size > 5 cm, 6. positive margin, and 7. extrahepatic disease. From this, we formulated a clinical risk score (CRS) based on the first five of these factors for use in patient selection for surgery and for stratification of patients for clinical studies. Using one point for each criterion, a summed score of 0–2 puts patients in a low-risk group and is a strong indication for hepatectomy. In the patients with small tumors, a maximum score of 4 is possible. The 5-year survival of patients with small tumors and 0–2 points on the CRS is 47% and the median survival is 56 months (33). Patients with a score of 3–4 are in a high-risk group, with a median survival of 32 months and 5-year survival of 24% (Fig. 6.1). In these high-risk patients, a period of observation with no therapy or systemic chemotherapy allowing for the extent of metastases to declare themselves is reasonable. Improved imaging techniques such as fluorodeoxyglucose positron emission tomography (FDG PET) scanning should be considered and may help discover extrahepatic disease noninvasively in these patients at high risk for additional cancer (34). Finally, these patients should be considered for clinical studies of aggressive adjuvant chemotherapy after liver resection. 1.0

Survival

0.8

0.6

0.4

0.2

0.0 0

12

24

36

48

60

Months Figure 6.1 Prediction of long-term outcome for small (5 cm and (5) CEA > 200 ng/dl. For score = 0–2 (N = 236) (open box), the median survival was 56 months and the 5-year survival 47%. For score = 3–4 (N = 57) (filled triangles), the median survival was 32 months and the 5-year survival 24%.

56

Neuroendocrine Tumors Patients with symptomatic neuroendocrine tumors should be considered for resection or ablation. For the small tumor, symptoms are most likely derived from hormonal secretion by the tumors, and such hormone levels will also provide a marker for effectiveness of the ablation or resection. For asymptomatic tumors, a period of observation to allow assessment of the pace and aggressiveness of the tumors is reasonable when the tumors are small. At the first signs of progression, resection or ablation should be considered. Noncolorectal, Nonneuroendocrine Tumors Harrison et al. defined prognostic factors involved in the resection of noncolorectal, nonneuroendocrine hepatic metastases (26). In this study, 96 patients underwent liver resection. The prognostic factors of significance on multivariate analysis included the disease-free interval (>36 months), curative resection (versus palliative incomplete resection), and primary tumor type. Their conclusions would suggest that regardless of histology, with a long disease-free interval patients may benefit from surgical resection.

resection techniques For small solitary metastases to the liver, the goal of resection is to completely excise the tumor while preserving the maximum normal hepatic parenchyma. Preserving parenchyma facilitates postoperative recovery and also provides flexibility for further resections should intrahepatic recurrences occur (35). Small surface-oriented metastases can be excised using a nonanatomic wedge resection, whereas deeper lesions require formal segmentectomies or sectorectomies. A goal of at least a 1 cm margin is reasonable (36). The use of intraoperative ultrasound is important to rule out other small hepatic metastases, which may not be evident on preoperative scans and in defining the intersegmental planes for designing the approach to segmentectomy. Even for wedge resections, ultrasound is beneficial in defining the vascular anatomy around the lesion, which may help minimize blood loss. Wedge Resections Wedge resections must be performed meticulously to avoid inadvertently leaving a positive margin. Large chromic liver sutures can be placed and used for retraction during dissection. The parenchymal dissection should be performed along the lines used for other forms of liver resection. We prefer the Kelly clamp technique where the clamp is used to crush the normal parenchyma, exposing vessels that are then clipped, tied, suture ligated, or stapled using a vascular stapling device (37). The Pringle maneuver is used intermittently for 5 minutes at a time followed by reperfusion of the parenchyma, during which time the argon beam coagulator is used to coagulate small bleeding vessels on the surface. This technique is superior to the simple use of electrocautery for the dissection, which is often attempted for what seems to be routine wedge resections. The char effect of the electrocautery prevents adequate visualization of the anatomy, making it quite easy to stray into large vessels or into the tumor.

SMALL SOLITARY HEPATIC METASTASES: WHEN AND HOW? The most difficult margin in performing a wedge resection is the deep margin of dissection. Using intraoperative ultrasound, the depth of dissection should be measured prior to the initiation of parenchymal dissection, including at least a 1-cm margin deep to the tumor. The dissection should be carried down perpendicular to the liver surface to the predetermined depth. At this point, the tumor can be lifted up and dissection can proceed horizontally across the base of the wedge. The tendency to resect with a “V-shaped approach” is more likely to be complicated by a positive deep margin. At the end of the dissection, the Pringle maneuver is removed and the argon beam coagulator is used to control bleeding vessels. Careful examination is made for any evidence of a bile leak, which is controlled with suture ligature. For larger lesions where it is especially difficult to achieve the deep margin safely, a cryoassisted wedge resection can be performed (38). The cryotherapy probe is inserted into the tumor and freezing is begun with real time ultrasound imaging. When the zone of freezing is confirmed by ultrasound to be at least 1 cm beyond the tumor, wedge resection is performed using the freeze margin as the margin of resection. The cryotherapy probe makes a ready retracting device and the parenchyma is usually easy to dissect at the margin of the ice-ball. Freezing must continue intermittently during dissection to ensure that the ice-ball does not retract and expose the tumor.

clamped at its junction with the vena cava during parenchymal transection to further minimize blood loss. When the solitary metastases lie near an intersegmental plane, two segments can be removed. This is most easily done as a formal sector such as the left lateral sectorectomy (segments II and III) and right posterior sectorectomy (segments VI and VII). The caudate lobe (segment I) can be resected as an isolated segmentectomy when the tumor is confined to this lobe (42). This requires a more extensive dissection, including complete division of all the perforating caudate veins draining directly into the vena cava as well as the numerous small portal triads extending off the main left pedicle at the base of the umbilical fissure. Figure 6.2 demonstrates a case of a small, solitary segment of hepatic metastasis for colorectal cancer, which was detected on an MRI scan used for screening because of a rising CEA. Although this was a surface lesion, intraoperative ultrasound revealed the segment VI triad immediately adjacent to the tumor. The segment VI triad was located by ultrasound and ligated at its origin with minimal parenchymal dissection. The intersegmental planes were then marked by electrocautery and a formal segmentectomy was performed with negative margins. While an aggressive resection was indicated and performed, the patient can still undergo a formal left or right hepatic lobectomy in the future if indicated. No dissection of the vena cava or porta hepatis was required.

Segmental Resections For all but the most superficial lesions, we prefer a segmental approach for the resection of tumor (39). Segmental resections have a significantly lower rate of pathologic positive margins, and this translates into improved long-term survival (40). Small, deep solitary metastases and surface lesions adjacent to major vascular structures lend themselves particularly well to segmentectomies or sectorectomies. The intersegmental planes can be identified intraoperatively using vascular landmarks with the aid of intraoperative ultrasound. Using these planes for parenchymal dissection will minimize blood loss and help ensure a safe margin. Inflow occlusion for the segment can almost always be performed first, thereby producing demarcation of the segmental planes to further enhance the dissection. The portal triad to segments II, III, and IV can be identified and controlled within the umbilical fissure with little parenchymal dissection (37). The right posterior sectoral pedicle can be found by dividing the parenchyma along a horizontal cleft (fissure of Gans) present on the inferior surface of the right lobe of the liver. The pedicle can be traced to its bifurcation to segments VI and VII for control of the individual segmental portal triads. The anterior sectoral pedicle can be dissected from an inferior or anterior approach. The major hepatic veins lie within the intersegmental planes and can be a source of significant blood loss during the parenchymal transection phase of a segmentectomy. The use of low central venous pressure (0–5 mmHg) during parenchymal dissection can decrease back bleeding in these veins (41). Extrahepatic control of the left, middle, and right hepatic veins can also be achieved and the vein of concern temporarily

Morbidity and Mortality The mortality rates for major hepatic resection have decreased significantly over time to a common reporting of mortality in the 1% to 4% range (43). These values are even lower for wedge resections and segmentectomies. In a recent report of 270 wedge or segmental resections, the operative mortality was 0.5% (40). This low mortality is not surprising considering that the main cause of death in studies of liver resection is liver failure secondary to inadequate residual normal parenchyma, an unlikely event for resection of small solitary hepatic metastases where minimal normal parenchyma is sacrificed. While mortality rates are low, the complication rate for major hepatic resection is still relatively high, ranging from 20% to 50% (5). Bile leaks, perihepatic abscess, hemorrhage, cardiopulmonary complications, pleural effusions, pneumonia, and pulmonary embolism are among the most common complications (43). Many of these could be expected after segmentectomy and wedge resections as well as major hepatic resections. Even though these complications do not translate into a high mortality rate, they may affect recovery time and quality of life. While this is not a significant issue for patients expected to undergo a long-term disease-free interval or cure, it may be significant for patients whose survival is expected to be of the order of months. For those patients with aggressive tumors who are likely to fail outside the liver in the near future, less invasive techniques which are associated with a lower complication rate and quicker recovery time are more appealing. Ablative Techniques Other minimally invasive techniques include local ablative therapies such as laparoscopically directed cryotherapy (44) or

57

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS

(A)

(B)

(C)

(D)

Figure 6.2 An example of a small, solitary colorectal metastasis to segment VI. (A) MRI reveals subtle abnormality not seen on CT scan. (B) Intraoperative ultrasound reveals the tumor and adjacent segment VI portal vein. (C) Intersegmental planes have been marked on the liver capsule with electrocautery and parenchymal dissection begun. (D) Resected segment with tumor (microscopic negative margins). (Special thanks to Dr Peter Choyke for MRI scan.)

radiofrequency ablation (45). These techniques will be discussed further in chapter 8. They provide ideal alternatives to laparotomy and major liver resection for the treatment of small solitary hepatic metastases, since the small tumor is the most likely to be completely treated by ablation techniques. Furthermore, treatment by ablative techniques does not preclude future resection. Percutaneous approaches to tumor ablation are even more attractive than laparoscopic procedures. Local injection of toxic agents such as ethanol has been shown to be effective for hepatocellular cancers, however these agents have not been proven for other histologies and are known to be poorly effective for colorectal cancer (2). Radiofrequency ablation can be performed percutaneously under ultrasound guidance with local anesthesia. Figure 6.3 demonstrates a case of a metastatic pancreatic cancer 2 years after a dramatic primary response to gemcitabine and radiation therapy. Because the patient will likely begin to fail in multiple sites in the near future with limited survival potential, a laparotomy and hepatic resection was not considered reasonable. She was treated with percutaneous radiofrequency ablation, achieving a good zone of necrosis encompassing the mass, and she spent only one day in the hospital with very minimal discomfort. How such procedures, which have low morbidity and which maintain quality of life, will factor

58

in the treatment of patients with small hepatic metastases must be addressed by studies with sufficient follow-up to define the local recurrence rate.

adjuvant chemotherapy The role for adjuvant systemic chemotherapy after the removal of small solitary hepatic metastases is not well defined. Even for hepatic colorectal metastases, which are commonly treated with surgery, data on adjuvant chemotherapy after liver resection is sparse. Two retrospective studies have suggested a benefit of adjuvant systemic chemotherapy after metastasectomy, but others have not supported this (6,46–48). Use of systemic chemotherapy after resection of hepatic colorectal metastases is based mainly on data demonstrating adjuvant 5-fluorouracil (5-FU) and levamisol or 5-FU and leucovorin to decrease recurrence rate and improve survival when used after resection of the primary tumor (49). It is hoped that a similar benefit will be seen when 5-FUbased chemotherapy is used after metastasectomy. Current practice is to offer adjuvant 5-FU-based chemotherapy after hepatic resection to patients who have had no previous chemotherapy. There are currently no data to support the use of irinotecan and oxaliplatin in an adjuvant setting, although studies are in progress.

SMALL SOLITARY HEPATIC METASTASES: WHEN AND HOW?

(A)

(B)

(C) Figure 6.3 An example of a small, solitary pancreatic cancer metastasis treated with percutaneous radiofrequency ablation. (A) Pretreatment CT scan reveals hypodense 3 cm right lobe liver metastasis. (B) Ultrasound hoto with radiofrequency probe inserted into tumor. (C) Post-treatment scan (3 weeks) reveals large zone of necrosis replacing prior tumor. (Special thanks to Dr Thomas Shawker for ultrasound photo.)

For patients with hepatic colorectal metastases, the most common site of tumor recurrence after liver resection is the remnant liver (50). In the treatment of patients with small hepatic metastases, there is particular concern that even smaller undetected metastases may subsequently present as a liver tumor recurrence. Regional chemotherapy to treat the liver site is therefore a theoretically attractive option for adjuvant care. Data addressing the utility for such hepatic arterial infusional (HAI) chemotherapy had been sparse, consisting only of four small single-arm studies (51–53) and a single, small, randomized trial consisting of 36 patients (54). These preliminary studies demonstrated safety of such an approach, but efficacy data were insufficient to support the routine use of adjuvant intraarterial chemotherapy. Two large randomized trials examining adjuvant HAI have been completed. In the first trial (55), 224 patients from 25 centers were randomized to either no adjuvant therapy or adjuvant HAI 5-FU + systemic folinic acid. Although no difference was found between the groups, technical factors compromised this study such that only 34 of the 114 patients randomized to chemotherapy completed the adjuvant treatments. In another study, Kemeny et al. randomized 156 patients to either systemic 5-FU + leucovorin or HAI floxuridine (FUDR) + systemic 5-FU after complete resection of tumor (56). There was a significant survival advantage to HAI that is most likely related to local liver tumor control. We believe HAI chemotherapy is effective and

should be considered as an adjuvant to resection of hepatic colorectal metastases. For noncolorectal, nonneuroendocrine histologies metastatic to the liver, the most likely cause of death will be related to the disease outside the liver, regardless of how the liver is managed. For patients who are likely to develop systemic metastases in the near future, it may be reasonable to offer chemotherapy prior to resection. If the tumor responds, then a resection will be performed with confidence that other micrometastatic disease may be effectively treated with chemotherapy. If the tumor does not respond and the liver remains the only site of metastatic disease, resection is performed with increased confidence conferred by the longer period of observation. If the patient advances systemically during chemotherapy, then it is very unlikely that a resection would have been of benefit and the patient will have avoided the potential morbidity, pain, discomfort, and recovery time of an hepatic resection. That patient can go on to obtain second-line chemotherapy, investigational chemotherapy, or have no additional treatment.

conclusions Algorithms for the management of small solitary hepatic metastases are shown in Figure 6.4. Both patient and tumor characteristics must be considered in making management decisions. The most important tumor-related characteristic is

59

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Colorectal metastases

High CRS (3–4)

Low CRS (0–2)

Observation or chemotherapy

Resection

No extrahepatic progression

Extrahepatic progression

Resection or ablation

Chemotherapy

Ablation

Resection

Adjuvant therapy protocol

Adjuvant therapy protocol

(A) Neuroendocrine metastases

Symptomatic

Non-colorectal non-neuroendocrine Asymptomatic Long disease-free interval

Resection

Ablation

Short disease-free interval

Observation Resection Progression

No progression

Resection or ablation

Observation

(B)

Effective chemotherapy (>20% response)

No effective chemotherapy

Trial of chemotherapy

Ablation vs observation

(C)

Figure 6.4 Algorithms for the management of small hepatic metastases. (A) Algorithm for colorectal metastases (CRS, clinical risk score). (B) Algorithm for neuroendocrine metastases. (C) Algorithm for non-colorectal, non-neuroendocrine metastases.

histology. For patients with colorectal cancer (Fig. 6.4A), the prognostic factors for tumor recurrence after resection are well defined. Using the clinical risk score (CRS) as selection criterion, patients with CRS = 0–2 are ideal candidates for resection. Those with CRS = 3–4 should consider observation or chemotherapy prior to a definitive hepatic procedure. Immediate ablation or resection should be performed in the setting of a clinical trial, and most appropriately a trial examining adjuvant therapy. For neuroendocrine cancers (Fig. 6.4B), symptomatic tumors should be treated with resection and/or ablation when possible. When the cancer is found in an asymptomatic patient, a period of observation is not unreasonable because of the often indolent nature of these tumors. At resection, the principle should be to leave as much normal liver behind in order to minimize the risk of liver failure and in order to allow for

60

repeat anatomic liver resections in the future for recurrent disease. Enucleation with positive margins is acceptable for treatment of this histology because resection is almost never curative, and such cytoreduction can provide significant and durable palliation with minimum risk. For patients with small, solitary, noncolorectal nonneuroendocrine tumors, the most significant factor in terms of prognosis seems to be the disease-free interval (Fig. 6.4C). For patients with a long disease-free interval from primary resection a curative surgical resection is indicated as the most effective means of therapy. While it may be still unlikely that these patients can be cured, they must be given the benefit of the doubt and the most optimal procedure performed. The definition of “long” has been arbitrarily set at 36 months by Harrison et al. (26), but in reality it must vary according to histology. For gastric cancer, 12–24 months would be

SMALL SOLITARY HEPATIC METASTASES: WHEN AND HOW? considered long, whereas for ocular melanoma, 3–5 years would be more reasonable. Patients with a short disease-free interval from a tumor with a poor prognosis should undergo a trial of chemotherapy if there is a known effective agent. If no effective agent exists (as is the case for most solid malignancies), then these patients are ideal for an experimental, minimally invasive, local ablative therapy. This provides an advantage to observation alone, given the low but definite risk of the metastases spreading during the observation period. It will be psychologically more comforting to the patient to know that the lesion has been ablated, and risk, pain, and recovery duration are minimal. Observation alone is also quite reasonable, but it is often not accepted by patients. Patient-related factors must also be taken into consideration. Patients who have concomitant illnesses that make them poor operative candidates may be better served with a minimally invasive or percutaneous technique, even in the case of potentially curable metastases from colorectal cancer. Because of improvements in diagnostic techniques and the routine use of serum tumor markers, the detection of small solitary hepatic metastases from various tumors will likely increase in the future. A uniform approach to these patients such as that which is outlined in the treatment algorithm should be considered.

key points Factors that determine management ● ● ● ●

Natural history of tumor type Expected cure rate after surgical treatment Effectiveness of alternative treatment strategies Morbidity of surgical resection

Survival rates following hepatic resection ●

●

Good evidence for long-term survival Colorectal metastases Neuroendocrine metastases Survival possible in highly selected cases Breast cancer Sarcoma (especially gastrointestinal stromal tumors) Melanoma

Patient selection factors in colorectal metastases ●

●

Contraindications Extrahepatic disease (except solitary pulmonary metastases) Positive hilar lymph nodes Relative contraindications Presentation within 12 months of resection of primary tumor CEA >200 ng/dl >1 liver tumor Tumor >5 cm in size Positive resection margin

references 1. Nuzzo G, Giuliante F, Giovannini I, Tebala GD, Clemente G, Vellone M. Resection of hepatic metastases from colorectal cancer. Hepato-gastroenterology 1997; 44: 751–9. 2. Alexander HR, Bartlett DL, Fraker DL, Libutti SK. Regional treatment strategies for unresectable primary or metastatic cancer confined to the liver. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology. Philadelphia, PA: JB Lippincott, 1996: 1–19. 3. Wagner JS, Adson MA, van Heerden JA, Adson MD, Ilstrup DM. The natural history of hepatic metastases from colorectal cancer. A comparison with resective treatment. Ann Surg 1984; 199: 502–8. 4. Wood CB, Gillis CR, Blumgart LH. A retrospective study of the natural history of patients with liver metastases from colorectal cancer. Clin Oncol 1976; 2: 285–8. 5. Fong Y, Blumgart LH, Cohen AM. Surgical treatment of colorectal metastases to the liver. CA Cancer J Clin 1995; 45: 50–62. 6. Hughes KS, Simon R, Songhorabodi S. Resection of the liver for colorectal carcinoma metastases: a multi-institutional study of indications for resection. Surgery 1988; 103: 278–88. 7. Rosen CB, Nagorney DM, Taswell HF et al. Perioperative blood transfusion and determinants of survival after liver resection for metastatic colorectal carcinoma. Ann Surg 1992; 216: 493–505. 8. Scheele J, Stang R, Altendort-Hofmann A, Paul M. Resection of colorectal liver metastases. World J Surg 1995; 19: 59–71. 9. Taylor M, Forster J, Langer B, Taylor BR, Greig PD, Mahut C. A study of prognostic factors for hepatic resection for colorectal metastases. Am J Surg 1997; 173: 467–71. 10. Fong Y, Cohen AM, Fortner JG et al. Liver resection for colorectal metastases. J Clin Oncol 1997; 15: 938–46. 11. Moertel CG. An odyssey in the land of small tumors. J Clin Oncol 1987; 5: 1503–22. 12. Thompson GB, van Heerden JA, Grant CS, Carney JA, Ilstrup DM. Islet cell carcinomas of the pancreas: a twenty-year experience. Surgery 1988; 104: 1011–7. 13. Declore R, Friesen SR. Gastrointestinal neuroendocrine tumors. J Am Coll Surg 1994; 178: 188–211. 14. Godwin JD. Carcinoid tumors: an analysis of 2837 cases. Cancer 1975; 36: 560–9. 15. Sjoblom SM. Clinical presentation and prognosis of gastrointestinal carcinoid tumours. Scand J Gastroenterol 1988; 23: 779–87. 16. McEntee GP, Nagourney DMN, Kvols LK, Moertel CG, Grant CS. Cytoreductive hepatic surgery for neuroendocrine tumors. Surgery 1990; 108: 1091. 17. Chen H, Hardacre JM, Uzra A, Cameron JL, Choti MA. Isolated liver metastases from neuroendocrine tumors: does resection prolong survival? J Am Coll Surg 1998; 187: 88–92. 18. Ihse I, Persson B, Tibblin S. Neuroendocrine metastases of the liver. World J Surg 1995; 19: 76–82. 19. Raab R, Nussbaum KT, Behrend M, Weimann A. Liver metastases of breast cancer: results of liver resection. Anticancer Res 1998; 18: 2231–3. 20. Elias D, Lasser PH, Montrucolli D, Bonvallot S, Spielmann M. Hepatectomy for liver metastases from breast cancer. Eur J Surg Oncol 1995; 21: 510–3. 21. Jaques DP, Coit DG, Casper ES, Brennan MF. Hepatic metastases from soft-tissue sarcoma. Ann Surg 1995; 221: 392–7. 22. Schwartz SI. Hepatic resection for noncolorectal nonneuroendocrine metastases. World J Surg 1995; 19: 72–5. 23. Salmon RJ, Levy C, Plancer C, et al. Treatment of liver metastases from uveal melanoma by combined surgery-chemotherapy. Eur J Surg Oncol 1998; 24: 127–30. 24. Ochiai T, Sasako M, Mizuno S. Hepatic resection for metastatic tumours from gastric cancer: analysis of prognostic factors. Br J Surg 1994; 81: 1175–8. 25. Bines SD, England G, Deziel DJ, Witt TR, Doolas A, Roseman DL. Synchronous, metachronous and multiple hepatic resections of liver tumors originating from primary gastric tumors. Surgery 1993; 114: 799–805. 26. Harrison LE, Brennan MF, Newman E et al. Hepatic resection for noncolorectal, nonneuroendocrine metastases: a fifteen-year experience with ninety-six patients. Surgery 1997; 121: 625–32.

61

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS 27. Fujisaki S, Takayama T, Shimada K, et al. Hepatectomy for metastatic renal cell carcinoma. Hepato-gastroenterology 1997; 44: 817–9. 28. Iwatsuki S, Shaw BW, Starzl TE. Experience with 150 liver resections. Ann Surg 1983; 197: 247. 29. Talmadge JE, Fidler IJ. Enhanced metastatic potential of tumor cells harvested from spontaneous metastases of heterogeneous murine tumors. J Natl Cancer Inst 1982; 69: 975–80. 30. Elias D, Saric J, Jaeck D et al. Prospective study of microscopic lymph node involvement of the hepatic pedicle during curative hepatectomy for colorectal metastases. Br J Surg 1996; 83: 942–5. 31. Nanko M, Shimada H, Yamaoka H et al. Micrometastatic colorectal cancer lesions in the liver. Jpn J Surg 1998; 28: 707–13. 32. Hughes KS, Simon R, Songhorabodi S, Adson MA. Resection of the liver for colorectal carcinoma metastases: a multi-institutional study of patterns of recurrence. Surgery 1986; 100: 278–84. 33. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999; 230(3): 309–18. 34. Delbeke D, Vitola JV, Sandler MP et al. Staging recurrent metastatic colorectal carcinoma with PET. J Nucl Med 1997; 38: 1196–201. 35. FernE1ndez-Trigo V, Shamsa F, Sugarbaker PH. Repeat liver resections from colorectal metastasis. Surgery 1995; 117: 296–304. 36. Shirabe K, Takenaka K, Gion T et al. Analysis of prognostic risk factors in hepatic resection for metastatic colorectal carcinoma with special reference to the surgical margin. Br J Surg 1997; 84: 1077–80. 37. Blumgart LH. Liver resection—liver and biliary tumours. In: Blumgart, LH ed. Surgery of the Liver and Biliary Tract. New York: Churchill Livingstone, 1994: 1495–538. 38. Polk W, Fong Y, Karpeh M, Blumgart LH. A technique for the use of cryosurgery to assist hepatic resection. J Am Coll Surg 1995; 180: 171–6. 39. Billingsley KG, Jarnagin WR, Fong Y, Blumgart LH. Segment-oriented hepatic resection in the management of malignant neoplasms of the liver. J Am Coll Surg 1999; 187: 471–81. 40. DeMatteo RP, Palese C, Jarnagin WJ, Sun RL, Blumgart LH, Fong Y. Anatomic segmental hepatic resection is superior to wedge resection as an oncologic operation for colorectal liver metastases. J Gastrointest Surg 2000; 4(2): 178–84. 41. Cunningham JD, Fong Y, Shriver C. One hundred consecutive hepatic resections: blood loss, transfusion and operative technique. Arch Surg 1994; 129: 1050–6. 42. Bartlett D, Fong Y, Blumgart LH. Complete resection of the caudate lobe of the liver: technique and results. Br J Surg 1996; 83: 1076–81.

62

43. Fong Y, Blumgart LH. Hepatic colorectal metastasis: current status of surgical therapy. Oncology 1998; 12: 1489–94. 44. Lezoche E, Paganini AM, Feliciotti F, et al. Ultrasound-guided laparoscopic cryoablation of hepatic tumors: preliminary report. World J Surg 1998; 22: 829–36. 45. Siperstein AE, Rogers SJ, Hansen PD, Gitomersky A. Laparoscopic thermal ablation of hepatic neuroendocrine tumor metastases. Surgery 1997; 122: 1147–55. 46. Fortner JG, Silva JS, Golbey RB. Multivariate analysis of a personal series of 247 consecutive patients with liver metastases from colorectal cancer: I. Treatment by hepatic resection. Ann Surg 1984; 196: 306–16. 47. Butler J, Attiyeh FF, Daly JM. Hepatic resection for metastases of the colon and rectum. Surg Gynecol Obstet 1986; 162: 109–13. 48. Pagana TJ. A new technique for hepatic infusional chemotherapy. Semin Surg Oncol 1986; 2: 99–102. 49. Moertel CG, Fleming TR, Macdonald JS et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 1990; 322: 352–8. 50. Blumbart LH, Fong Y. Surgical management of colorectal metastases to the liver. Curr Prob Surg 1995; 5: 333–428. 51. Goodie DB, Horton MD, Morris RW, Nagy LS, Morris DL. Anaesthetic experience with cryotherapy for treatment of hepatic malignancy. Anaes Int Care 1992; 20: 491–6. 52. Moriya Y, Sugihara K, Hojo K, Makuuchi M. Adjuvant hepatic intra-arterial chemotherapy after potentially curative hepatectomy for liver metastases from colorectal cancer: a pilot study. Eur J Surg Oncol 1991; 17: 519–25. 53. Curley SA, Roh MS, Chase JL, Hohn DC. Adjuvant hepatic artery infusion chemotherapy after curative resection of colorectal liver metastases. Am J Surg 1993; 166: 743–8. 54. Kemeny MM, Goldberg D, Beatty D et al. Results of a prospective randomized trial of continuous regional chemotherapy and hepatic resection as treatment of hepatic metastases from colorectal primaries. Cancer 1986; 57: 492–8. 55. Lorenz M, Muller HH, Schramm H et al. Randomized trial of surgery versus surgery followed by adjuvant hepatic arterial infusion with 5-fluorouracil and folinic acid for liver metastases of colorectal cancer. Ann Surg 1998; 228: 756–62. 56. Kemeny N, Huang Y, Cohen AM et al. Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer. N Engl Med 1999; 341(27): 2039–48. 57. Que FG, Nagorney DM, Batts KP, Linz LJ, Kvols LK. Hepatic resection for metastatic neuroendocrine carcinomas. Am J Surg 1995; 169: 36–43.

7

Managing complications of hepatectomy Fenella K. S.Welsh, Timothy G. John, and Myrddin Rees

introduction The safety of elective liver surgery has improved dramatically in the past 30 years. A multicenter American series comprising 621 liver resections published in the late 1970s reported a 13% mortality (1). By contrast, recent large published series describe posthepatectomy mortality rates of 0% to 4.4%, with 19.6% to 45% morbidity (2–8) (Table 7.1). Furthermore, individual units have demonstrated a significant reduction in morbidity and mortality over time, despite ever-widening the indications for hepatectomy (6,8). This dramatic improvement in immediate postoperative outcome can be explained by increased specialization of liver surgery in high-volume centers (9), better selection of patients in terms of hepatic functional reserve and comorbid conditions, advances in surgical technique, including greater understanding of hepatic segmental anatomy and improved instrumentation for the parenchymal transection. Furthermore, anesthesia and critical care has improved enormously, the routine use of low central venous pressure (CVP) anesthesia being a particular advance. However, even a 20% complication rate remains significant, particularly if the indication for hepatectomy is for livingdonor transplantation. Furthermore, postoperative morbidity can also adversely affect disease-specific and disease-free survival (10–12). Thus the short- and long-term consequences of postoperative morbidity, coupled with increasing litigation, and limited health care resources, has renewed the drive to further improve the immediate outcome from liver resection, with emphasis on prevention of and improved management of complications, when they occur. The precise definitions of the specific complications such as bleeding, bile leak, and hepatic insufficiency are still without consensus. Moreover, the stratification of the severity of each complication is still unclear. Standardized definitions, grading, and reporting of the complications of hepatectomy are needed to allow an objective, quality assessment of outcome data from different units and further improve results. The system proposed and validated by Clavien, focusing on the therapeutic consequences of complications in order to rank their severity, is currently the best available (13). However, it is still not universally adopted within the surgical community. A number of studies have attempted to identify the risk factors associated with complications and death from hepatectomy, three of which are detailed in Table 7.2. From these studies, there is consensus that the estimated blood loss or blood transfusion rate, the extent of hepatic resection, and an additional extrahepatic procedure are all independent predictors of morbidity and mortality. In addition, medical comorbidity, an elevated preoperative creatinine, preoperative thrombocytopenia, or hypoalbuminemia also appear to increase the operative risk. However, in the Hong Kong study (8), while cirrhosis per se was associated with increased

postoperative morbidity and mortality on an initial univariate analysis, it failed to independently predict outcome on subsequent multivariate analysis. Similarly, Belghiti’s group found that the in-hospital mortality rate was significantly higher in those patients with cirrhosis (8.7%) compared to those without underlying liver disease (1%, p < 0.001), but this was not subjected to multivariate analysis (7). Thus while liver resection in cirrhotic patients is technically more challenging than resecting normal liver, with a higher incidence of bleeding, septic complications, and postoperative liver failure (14), these two studies would suggest that in experienced high-volume centers, liver resection can be safely performed in patients with early cirrhosis. The common complications of hepatectomy may be classified as specific to the procedure or of a more general nature (Table 7.3). This chapter will deal with these complications in turn, focusing on their definition, incidence, predisposing factors, prevention, presentation, investigation, and treatment.

bleeding Incidence Bleeding is the most feared complication of hepatectomy, both on the operating table and in the immediate aftermath of surgery. In the 1960s and 1970s, it was the cause of major morbidity and mortality. The 1974 Liver Tumor Survey was a multicenter series of 621 hepatic resections performed in 98 U.S. centers, published in 1977. It reported a 13% mortality, with 15 of the 82 deaths (18%) due to exsanguinating hemorrhage in the operating room and bleeding being the documented primary cause of death in 26 of the 76 patients (34%), where the cause of death could be determined (1). However, bleeding is now relatively rare, with the median estimated blood loss for an elective hepatectomy being 345 to 600 ml (3,6) and the need for perioperative blood transfusion now being the exception rather than the rule. Indeed, the incidence of major hemorrhagic complications is rare, 0.7% (7/1005) in our own series (3). Of these seven cases, there were no on-table deaths, five patients were treated nonoperatively and two underwent reexploration for bleeding from a hepaticojejunal anastomosis and a left caudate branch of the portal vein respectively. In the Sloan-Kettering series of 1803 patients, the incidence is similar (1%) (6). Prevention Prevention remains the key to the management of bleeding. In the preoperative assessment, a careful drug history should be taken. If the patient is on drugs such as aspirin, clopidogrel, or warfarin, the indication for the treatment should be reviewed, and the drugs stopped where possible. Patients on warfarin as prophylaxis for thromboembolic events can be managed with an inferior vena cava (IVC) filter, placed preoperatively. It is

63

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Table 7.1 Morbidity and Mortality from Hepatic Resection in Recent Large Case-Series Reference

Years of study

No of centers

No of resections

Imamura et al. Rees et al. Wei et al. Malik et al. Jarnagin et al.

1994–2002

1

1056

1987–2005 1992–2002 1993–2006 1991–2001

1 2 1 1

1005 423 687 1803

Belghiti et al.

1990–1997

1

747

Poon et al.

1989–2003

1

1222

Case-mix 50% HCC 29% cirrhotic 100% CRLM 100% CRLM 100% CRLM 62% CRLM 10% HCC Elective & emergency. 35% benign 28% HCC 17% CRLM 32% cirrhotic 60% HCC 33% cirrhotic

Mortality

Morbidity

0%

39%

1.5% 1.7% 3.0% 3.1%

25.9% 19.6% 29.5% 45.0%

4.4% all 3.9% elective 8.7% cirrhotic 25.0% emergency 4.9%

22.0%

32.4%

Abbreviations: CRLM, colorectal liver metastases; HCC, hepatocellular carcinoma.

Table 7.2 Three Studies Reporting the Independent Predictors of Morbidity and Mortality after Hepatic Resection Reference Jarnagin et al.

Years of study 1991–2001

No of resections 1803

Belghiti et al.

1990–1997

478 elective resections, no cirrhotics

Poon et al.

1989–2003

1222

Predictors of morbidity Estimated blood loss Extent of resection + EH procedure ↑ preoperative creatinine Hypoalbuminemia Medical comorbidity Male gender ASA score Extent of resection Steatosis Blood transfusion + EH procedure Thrombocytopenia Blood transfusion + EH procedure

Predictors of mortality Estimated blood loss Extent of resection + EH procedure ↑ preoperative bilirubin Thrombocyt openia Age + EH procedure (in patients with malignancy)

Hypoalbuminemia Thrombocytopenia ↑ preoperative creatinine Major resection Blood transfusion

Abbreviation:+EH Procedure, additional extra-hepatic procedure.

important to identify patients with tricuspid regurgitation or right heart disease, where the anesthetist may encounter difficulties in lowering the CVP, as this may influence the extent of resection. Careful evaluation and correction of coagulation abnormalities should be performed pre- and perioperatively, particularly in the cirrhotic patient. Two key maneuvers are used to prevent bleeding during hepatic transection: portal triad clamping and low CVP anesthesia. Portal triad clamping, first described by Pringle in 1908, reduces hepatic arterial and portal venous bleeding (15,16). Although a European survey demonstrated that the use of inflow occlusion is not universal, it did confirm that most hepatic surgeons resort to it in difficult cases and that experienced surgeons are more likely to use it routinely (17). However, a recent systematic review and meta-analysis of the effect of inflow occlusion on postoperative morbidity and mortality failed to demonstrate any significant outcome

64

benefit (18). This is confirmed by another systematic review published in 2009, which compared 166 patients with vascular occlusion to 165 patients with no vascular occlusion (19). However, despite the small numbers involved, this later study showed that blood loss was significantly lower in those patients who had vascular occlusion. A low CVP reduces back bleeding from hepatic veins during the transection (20–22) and is now accepted practice during liver resection worldwide. Indeed, following the introduction of low CVP anesthesia in our own unit, the mean blood loss was significantly reduced from 2116 to 426 ml (3). However, these techniques can test the patients’ cardiovascular reserve. Obstructing the portal blood flow causes venous congestion of bowel and in combination with warm ischemic liver injury, releasing a flush of anerobic metabolites and cytokines back into the circulation on release of the clamp (23). Low CVP anesthesia relies on patients being maintained in a

MANAGING COMPLICATIONS OF HEPATECTOMY Table 7.3 Complications of Hepatectomy General complications Immediate (on table) Early (days)

Late (weeks/ months)

Specific complications

• Hypothermia

• Bleeding

• Respiratory— atelectasis, pleural effusion, pneumonia • Cardiovascular— DVT, PE, MI, arrhythmias, CVA • Renal failure • Wound infection • Pain • Incisional hernia

• • • •

Bleeding Bile leak Hepatic insufficiency Intra-abdominal abscess

Investigation and Treatment The hemoglobin concentration and clotting screen should be performed urgently, ensuring that the patient has an up-todate cross match. Any coagulopathy should be corrected. If the patient remains shocked, appropriate investigations may include endoscopy and mesenteric arteriography. Ultimately, as in our own series, small number of patients may need to return to the operating theatre for surgical control of hemorrhage.

biliary complications • Biliary stricture

Abbreviations: DVT, deep vein thrombosis; PE, pulmonary embolus; MI, myocardial infarction; CVA, cerebrovascular accident.

hypovolemic state until liver resection has been completed (20,21). This is in contrast to most other major surgical procedures, where patients have large volumes of crystalloid and colloid perioperatively. While a recent meta-analysis has confirmed that the use of the antifibrinolytic agent aprotinin can significantly reduce transfusion requirements during liver transplantation (24), there is no evidence for its routine use during liver resection (25). In contrast, a prospective double-blind randomized trial of tranexamic acid, another antifibrinolytic agent, has shown that its use perioperatively significantly reduced the blood loss and transfusion requirements in elective liver resection (26). Two prospective randomized controlled trials have failed to show any benefit of using recombinant factor VIIa in either noncirrhotic (27) or cirrhotic (28) patients undergoing hepatectomy. Since the early 1990s, the use of fibrin sealants has become a popular aid hemostasis at the hepatic parenchymal transection site. Two early randomized trials suggested some benefit in achieving hemostasis (29) and reducing postoperative blood loss (30), although the numbers involved were small. A more recent trial of a carrier-bound fibrin sealant (TachoSil®) suggested it was quicker and more effective hemostasis compared to argon beam coagulation (31). However, the numbers involved were again small (1000, or distal pancreatectomy (12). The magnitude of the pancreatic leak may also depend on whether the source is the main duct or parenchyma (19). Those patients who do develop pancreatic leaks are more likely to suffer from other complications or death, and this risk is exacerbated by superimposed infection (12). Numerous strategies have been attempted to minimize the chances of pancreatic leakage and these are discussed below.

Delayed Gastric Emptying The incidence of delayed gastric emptying (DGE) following PD ranges from 4% to 29% (5,11,13) and is associated with other intraabdominal complications (Table 9.1). While DGE is not associated with an increased risk of death, it does prolong hospitalization time (5,20). Parameters used to define DGE include the volume of nasogastric tube output, the length of time before tolerance of oral feeding, and results of scintigraphic studies. At our institution, DGE is defined as failure to achieve oral intake sufficient to maintain adequate hydration by postoperative day 10 (9). DGE is thought to be due to numerous factors, including management of the pylorus, extent of retroperitoneal dissection, intraabdominal fluid collections, and decreased motilin activity (21). Early reports indicated that pylorus-preserving pancreaticoduodenectomy (PPPD) increased the risk of DGE (22,23) but subsequent studies have failed to confirm this (Table 9.4) (24,25). Radical resection or extended retroperitoneal dissection may also be associated with DGE (26). It is unclear if more extensive dissection has a direct effect or if higher rates of pancreatic leak, sepsis, or hemorrhage predispose to DGE (25). Expeditious management of fluid collections, infection, or bleeding may limit gastric dysmotility. An additional contributing factor to DGE may be reduced levels of circulating motilin following PD (27). In a randomized control trial (RCT) including 118 patients undergoing PD, erythromycin, the motilin analogue, reduced the DGE rate from 30% to 19% compared to placebo (21). By contrast, routine nasogastric decompression or withholding of oral feeding has not been shown to affect the rate of DGE. Based upon data from RCTs involving patients subjected to gastrectomy, routine nasogastric tube placement following pancreatic surgery is unnecessary (28,29). Furthermore, early oral feeding should be considered following major abdominal procedures (30,31). Postpancreatectomy Hemorrhage Postpancreatectomy hemorrhage (PPH) occurs in 2% to 9% of cases and the consequences may be severe (8,32–37). The initial evidence of hemorrhage may be the “sentinel bleed,” which is present in 30% to 100% of patients prior to massive PPH (38–41). Risk of PPH is related to inadequate intraoperative hemostasis, bile leak, pancreatic leak, intraabdominal infection, and sepsis (39,40,42–44). The presence of jaundice at the time of pancreatic resection may increase the risk of PPH, but this is not lessened by preoperative biliary drainage (37). The implications and management of PPH vary depending on the time of onset and source (4). Postoperative bleeding within the first 24 hours is most often due to a technical failure and requires reoperation if severe (35). The most appropriate course of action for PPH occurring beyond the immediate postoperative period will

81

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Table 9.1 Complications Following Pancreaticoduodenectomy Author Balladur (42) Bottger (10) Gouma (11) Balcom (5) Muscari (7) Winter (13) Vin (12) House (49) Baker (47)

N

Pancreatic Fistula or Leak

Delayed Gastric Emptying

Hemorrhage

Bile Leak

Overall Complications

223 228 300 489 300 1423 680 356 440

13% 8% 7% 13% 17% 9% 18% 15% 16%

NR NR 29% 12% NR 15% NR 4% 7%

9% NR 5% NR 6% NR NR NR 2%

NR 3) or stones above the cystic duct confluence. To perform laparoscopic choledochotomy, the anterior surface of the common bile duct is dissected just sufficiently to confidently identify the anatomy. A 1.5-cm vertical incision is made in the common bile duct below the cystic duct confluence. Filling of the duct system with saline via the transcystic catheter distends the collapsed duct and helps prevent injury to the posterior duct wall when the anterior wall is incised. Another method is to gently lift up the anterior wall with a suitable small atraumatic grasper (such as a “dolphin-nose”) introduced via the right subcostal port. This will create a small transverse ridge of the anterior duct wall, which can then be cut using scissors introduced via the epigastric port, thus creating a vertical incision that can then be extended with the scissors. A similar effect can also be created using stay sutures. Initial flushing via the choledochotomy with the suckeraspirator and massaging of the duct may remove the stones. A choledochoscope can be introduced, this time from the epigastric port. A 3 or 5 mm flexible scope may be employed, or even a rigid ureteroscope if the orientation is suitable. (On the rare occasions that a rigid ureteroscope is required, it can

LAPAROSCOPY IN HPB SURGERY sometimes be introduced transcystically via the epigastric port if the orientation is suitable.) With the choledochoscope, the stones can be removed under direct vision, and the flexible choledochoscope can be maneuvered into both the upper ducts and lower CBD to confirm clearance of all calculi. On rare occasions, hydraulic lithotripsy may be required to break up impacted stones (27). Where there is confidence about stone clearance and biliary drainage, choledochotomy can be simply closed by suturing (25). If there is any doubt about biliary drainage or duct clearance, then choledochotomy should be closed after passage of an antegrade biliary stent, or a T-tube inserted. Choledochoduodenostomy For the elderly patient with a suspected benign stricture, and a reasonable stone load, laparoscopic choledochoduodenostomy is a good option (23). As in open surgery, a common duct diameter of greater than 10 mm is preferable. A continuous absorbable suture is used, and the operation mimics the open procedure with anastamosis of the choledochotomy to a longitudinal opening in the duodenum.

pancreatic pseudocyst Pancreatic pseudocysts can be managed endoscopically with gastrotomy and stenting (perhaps the only current valid indication for NOTES [Natural Orifice Transabdominal Endoscopic Surgery]). Pancreatic pseudocysts can also be drained internally via a laparoscopic approach (38–46). Most commonly the pseudocyst is located in the lesser sac and the appropriate procedure is a cyst-gastrostomy. An anterior gastrotomy is made. The cyst can be seen bulging forward, adherent to the posterior stomach wall, which is incised with diathermy or the harmonic scalpel to enter the cyst. Cyst fluid will come flowing out under pressure at this point, and is important to have an instrument ready to pass into the cyst so that the point of communication is not lost. The cyst fluid is aspirated with the sucker and the cyst emptied. A linear stapler is then introduced into the small cyst-gastrotomy to create a wide cystgastrostomy, and the residual unstapled edges are sutured together. The pseudocyst can be entered with the laparoscope and inspected, and any debris removed. The anterior gastrotomy is then closed with sutures or a stapling device. In some cases, the position of the pseudocyst will require a side-to-side cyst-gastrostomy or Roux-en-Y cyst-enterostomy. Published reports suggest that laparoscopic cyst-gastrostomy has a higher initial success rate and lower recurrence rate than endoscopic cyst-gastrostomy (42,47). As the cyst-gastrostomy created via the endoscopic approach is only small, any large debris is unable to exit the cyst. However, endoscopic approaches can be improved with using balloon dilatation and multiple stents to maintain better drainage, endoscopic ultrasound to guide the procedure and avoid vessels (48,49), and with the development of stapling instrumentation for natural orifice surgery (50).

imaging studies, but be found to have locally advanced disease or small liver or peritoneal metastases (imaging-occult metastases) that render the disease inoperable (Fig. 10.3). Staging laparoscopy can identify these patients and therefore spare the patient a laparotomy. Staging laparoscopy in its simplest form involves visual inspection of the peritoneal and liver surfaces, but may also include laparoscopic ultrasound, trial dissection, or peritoneal washing for cytology. Staging laparoscopy is preferable to a nontherapeutic laparotomy to identify unresectability. The hospital stay is shorter (51,52), and the patient is able to start chemotherapy sooner (53). The risks of a staging laparoscopy are low, with morbidity reported at 0% to 4% and mortality 0% to 0.15% (54). Port-site recurrences are uncommon, between 0% and 2% (54), and usually occur in patients with extensive peritoneal carcinomatosis. Staging laparoscopy may be performed as a prelude to resection in the same procedure or as a separate procedure prior to planned resection—there can be significant scheduling issues depending on the institution if an aborted procedure means allocated theater time is unable to be utilized. The yield of staging laparoscopy depends on many factors. The type and stage of the malignancy affects the likely presence of imaging-occult metastases, as does the quality and type of the imaging performed. The extent of the staging procedure is also important—whether laparoscopic ultrasound, peritoneal washings or trial dissection is included. It is also obviously influenced by what findings are considered to contraindicate resection; for example, localized peritoneal disease or porta hepatis nodes for colorectal liver metastases, or involvement of the portal vein requiring vein resection and grafting in pancreas cancer may not be considered contraindications to resection. The value of a positive staging laparoscopy also depends on whether any required palliative procedures, such as biliary or gastric bypass in carcinoma of the head of the pancreas, can be performed laparoscopically. In adenocarcinoma of the pancreas, after high-quality CT scanning, staging laparoscopy has been shown to identify

laparoscopic staging A proportion of patients with hepatobiliary and pancreatic malignancies will appear to be resectable on noninvasive

Figure 10.3 Peritoneal metastases at staging laparoscopy and carcinoma of the head of the pancreas.

91

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS unresectability in 15% to 51% of patients, and spare 10% to 31% of patients an unnecessary laparotomy (51,55–61). Laparoscopic ultrasound has been shown to add information in 12% to 14% of patients (62–64). Patients with tumors larger than 3 cm are more likely to have unsuspected metastases at exploration (65), as are patients with a Ca 19.9 level greater than 150 kU/L (66,67). Positive peritoneal lavage has been found in 3% to 51% of patients (57,68–76), and is more likely in locally advanced or metastatic tumors (77), larger tumors, and tumors of the body or tail (70,78). Positive peritoneal cytology, which has the same prognosis as metastatic disease (79), is the only marker of unresectability in 1% to 14% of patients (57,69,70,76). Tumors of the body and tail of the pancreas are twice as likely as pancreatic head lesions to have imaging-occult metastases (57,69). Imaging-occult metastases are uncommon in nonpancreatic periampullary tumors (60,80,81) and routine laparoscopy in these patients is probably not indicated. Patients who on imaging have locally advanced, unresectable pancreatic cancer should also be considered for staging laparoscopy, as those without metastatic disease can be considered for chemoradiotherapy regimens aimed at local control or even downstaging followed by resection, regimens which would incur unnecessary treatmentrelated morbidity for those with metastatic disease (69,77,82). In colorectal liver metastases, laparoscopy will identify unresectable disease in 10% to 38% of patients, with a sensitivity of 39% to 75% (83–90). Laparoscopy is more likely to be positive in patients with a higher clinical risk score (83,86,91). In noncolorectal, nonneuroendocrine liver metastases, laparoscopy has been reported to identify unresectable disease in 25% of patients, with a sensitivity of 66% (92). Staging laparoscopy is useful for patients with primary biliary malignancies. For patients with suspected resectable gall bladder carcinoma on imaging, the yield for detecting unresectable disease is 56% to 62% (93,94), though the yield is less for intrahepatic cholangiocarcinoma (93) at 36% and hilar cholangiocarcinoma (93–95) at 25%. The yield for hilar cholangiocarcinomas is higher for T2 or T3 lesions than for T1 lesions (94) (36% vs. 9%). In hepatocellular carcinoma that is considered suitable for curative resection, peritoneal dissemination is uncommon, and standard laparoscopy is unlikely to add much information. Laparoscopy with laparoscopic ultrasound, however, can identify the extent of the primary tumor, additional imagingoccult tumors, portal or hepatic venous tumor thrombus or an inadequate hepatic remnant, with a yield for unresectability of 10% to 36% and a sensitivity of 63% to 96% (96–100). The results obtained will depend on the type and quality of preoperative imaging and the level of experience with laparoscopic ultrasound.

laparoscopic palliative bypass In patients with inoperable periampullary tumors, there is often biliary and/or gastric obstruction that requires relief. The traditional teaching in open surgery was to perform both a biliary and gastric bypass whether or not the patient was symptomatic. If at laparoscopy the tumor is found to be unresectable, in the absence of actual or impending biliary or

92

gastric outlet obstruction, it is not necessary to perform a palliative biliary or gastric bypass (101). In many instances, the endoscopic approach is effective to relieve obstruction. Duodenal stenting is safer and provides a better quality of life than laparoscopic gastrojejunostomy in the short term (102), although laparoscopic gastrojejunostomy may provide a more durable result for patients with a longer life expectancy (103). ERCP with placement of a plastic biliary stent has a lower morbidity than traditional open surgical bypass, although plastic biliary stents have a tendency to occlude, resulting in recurrent biliary obstruction requiring a repeat procedure (104). Metallic stents, however, have a much higher patency rate in the longer term, and can serve many patients for the remainder of their survival (104–106). In some cases, however, stenting fails for technical reasons or due to inability to access the ampulla. In these cases, a laparoscopic bypass is a useful option (107–116), with the potential for lower morbidity and shorter hospital stay than an open surgical procedure (113,115). A laparoscopic biliary bypass is most easily performed as a stapled or sutured side-to-side cholecystojejunostomy. The main limitation of this approach is that the confluence of the cystic and common hepatic ducts must be well above the tumor to prevent recurrent biliary obstruction (117). This can be confirmed at the procedure by cholangiography via the fundus of the Gall bladder—a Verres needle with large syringe attached is used to empty the gall bladder of bile, which is then filled with contrast to confirm that the cystic duct confluence is more than 1 cm above the level of the tumor. If this is not the case, an hepaticojejunostomy is constructed. A gastrojejunostomy is typically fashioned in an antecolic, isoperistaltic stapled side-to-side manner.

laparoscopic pancreatectomy Distal pancreatectomy is well suited to a laparoscopic approach. The usual indication is a solid or cystic tumor of the tail of the pancreas that is not clearly benign on preoperative imaging. The procedure may involve en-bloc resection of the spleen and splenic vessels; preservation of the spleen with preservation of the splenic vessels; or preservation of the spleen without preservation of the splenic vessels with the spleen supplied from the short gastric and gastroepiploic vessels (the Warshaw technique (118)). For lesions close to the spleen, when splenectomy is necessary, the approach can be similar to laparoscopic splenectomy, with the patient left side up, and the spleen and distal pancreas mobilized from lateral to medial. After division of the short gastric vessels and the gastrocolic omentum, the pancreas can be divided en bloc with the splenic vessels using a linear stapler. For a medial to lateral approach, the pancreatic neck is divided, either with a stapling device or with the harmonic scalpel with subsequent suture closure of the pancreatic stump. Where the splenic vessels are being resected, the splenic vein is divided with a stapling device and the splenic artery divided with a stapling device or locking clips. If the splenic vessels are to be preserved, then the tail of the pancreas is dissected carefully from them with control of the small vessels with clips and/or the harmonic scalpel or

LAPAROSCOPY IN HPB SURGERY electrosurgical sealing device. Otherwise the dissection continues in the relatively avascular plane behind the splenic vein. At this point, if the spleen is to be preserved with the Warshaw technique, then the splenic hilum is divided with a stapling device taking care to preserve the short gastric vessels, and the gastroepiploic arcade. Otherwise if the spleen is to be resected, the dissection continues in this plane behind the splenic vein to complete the mobilization of the spleen and complete the resection. The specimen is retrieved in a bag and a closed suction drain is placed. Laparoscopic distal pancreatectomy has been shown to be a safe procedure, with a shorter hospital stay and overall morbidity that is less than the open procedure (119–124). The main complication is a pancreatic fistula occurring in about 15% of patients, though this occurs at no greater rate than with an open resection (120,124). The application of fibrin glue to the stump (125) and the use of staple line mesh reinforcement (126) have both shown some benefit in small studies in reducing this rate, and in open surgery the placement of a transampullary stent (127) has shown some benefit, as has identification and direct suture of the main pancreatic duct (128), although the optimal management of the pancreatic stump is still to be determined. Preservation of the spleen by the Warshaw technique can be complicated by infarction of the lower pole of the spleen (129,130). Laparoscopic central pancreatectomy has been reported in the literature (131) and successfully been performed twice by one of the authors. The indication of a central tumor where diabetes is a risk postoperatively is not common. Laparoscopic enucleation of insulinomas has been reported in small series but is associated with a significant rate of pancreatic fistula (129,132,133). Intraoperative ultrasound is essential to ensure that the main pancreatic duct is not close to the resection line. Laparoscopic pancreaticoduodenectomy has been reported in small numbers (134–137). The procedure is feasible but prolonged and difficult, and the potential role for this procedure remains to be determined.

laparoscopic liver resection The laparoscopic approach to liver resections presents certain technical challenges. It is a heavy solid organ that can be cumbersome to mobilize and manipulate, parenchymal transection requires the identification and control of large vessels with the potential for significant bleeding, and the paucity of external anatomical markers can make the maintenance of surgical orientation to ensure a satisfactory oncologic clearance difficult. Laparoscopic liver resection was initially reported in 1995 by Rau (138), Cuesta (139), and Hashizume (140). Anatomic resections in the form of left lateral sectionectomy were reported in 1996 by Azagra (141) and Kaneko (142), formal hemihepatectomies were reported in 1998 by Huscher (143), and Cherqui (144) reported the first significant series of 30 patients in 2000. The largest series were recently reported by Koffron (145) and Buell (146). Dr. Joe Buell organized the first international consensus meeting on laparoscopic liver resection, held in Louisville,

Kentucky in November 2008. Agreed definitions of laparoscopic liver surgery include the following: ●

●

●

●

Pure laparoscopic: where the liver resection is completed laparoscopically and the specimen removed via a remote incision; Hand assisted: where the surgeon operates with his nondominant hand inside the abdomen, placed via an airtight device, through which the specimen is removed; Hybrid liver resection (145): where the liver is mobilized laparoscopically and most of the resection is done through a smaller than usual right upper quadrant incision; Conversion: where the surgeon changes to an open operation from one of the above. One can also convert from pure laparoscopic to hand assist or hybrid.

The most suitable cases for a laparoscopic approach are solitary small (40 yrs, jaundice to encephalopathy time >7 days, serum bilirubin >300 µmol/l, PT >50 sec or INR >3.5 Category 7: Etiology: Acute presentation of Wilson’s disease, or Budd–Chiari syndrome. A combination of coagulopathy and any grade of encephalopathy Category 8: Hepatic artery thrombosis on days 0–21 after liver transplantation Category 9: Early graft dysfunction on days 0–7 after liver transplantation with at least two of the following: AST >10,000 IU/l, INR >3.0, serum lactate >3 mmol/l, absence of bile production Category 10: Any patient who has been a live liver donor who develops severe liver failure within 4 wks of the donor operation AOD – Acetaminophen overdose, PT – Prothrombin time, INR – International Normalized Ratio, ICP – Intracranial Pressure, FiO2 – inspired oxygen concentration

account for the majority of the LT indications in HIV patients. In addition to the selection criteria mentioned previously, criteria specific to HIV in the United Kingdom (32) include 1. CD4 counts >200 cells/µl or >100 cells/µl in the presence of portal hypertension 2. Absence of viremia 3. Absence of AIDS defining illness after immune reconstitution 4. Anti-retroviral therapeutic options available if HIV disease reactivates.

selection of donor The “ideal” deceased donor profile is as follows: age 50 years of age with cerebrovascular disease as cause of death (34,35). Expanded Criteria Donor The impact of changing deceased donor characteristics and widening gap between the donor pool and the waiting list means that deceased donors with features deviating from the donor profile are increasingly utilized (36,37). The term “extended or expanded criteria donor” (ECD) has been coined for such donors (Table 32.3), which suggests a higher risk of graft failure and decreased survival. Rather than a clear “good or bad” liver graft, ECD represents a spectrum of cumulative donor risks which should be taken into consideration (33,38,39). Steatosis and Abnormal Liver Function A recent international consensus meeting on ECD grafts (40) recommended against using liver allografts with severe steatosis (>60%) and from elderly donors in HCV-infected recipients. Abnormal liver function tests are not contraindications but careful assessment of other donor factors is essential, especially if there is a marked rise in gamma glutamyl transpeptidase level (>200 UI/L). Elderly Donors Patients with liver transplant from elderly donors have shown comparable survival, provided that there are no additional risk factors (41,42). While there is no clear age limit in utilizing an

289

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Table 32.3 Definition for Expanded Criteria Liver Donors (ECD)a Elevated risk for transmission of a disease (viral, eg, Hepatitis C or B) or bacterial infection (especially recovery after sepsis with bacteriemia or donor malignancy, etc.) Acute hemodynamic deterioration with the risk of organ loss Donor age >65 yrs Donor BMI >30 kg/m2 Bilirubin (total) >3 mg/dl or 51 µmol/l ASAT (GOT) or ALAT (GPT) above three times the upper reference threshold Serum sodium >165 mmol/l Hospitalization in ICU >7 days Hepatic steatosis >40% a There is no internationally agreed definition of ECD as yet. Above is the German Medical Association definition of ECD donors for liver donation (154).

elderly donor liver, there are two caveats: (i) liver grafts of advanced age have reduced regenerative capacity and synthetic function, and are more susceptible to cold ischemic injury and (ii) elderly liver grafts should not be used for HCV-positive patients because of the high risk of severe HCV recurrence (43) and reduced graft and patient survival (44). Donors with Infection The use of donors with bacteremia or bacterial meningitis seems safe as the risk transmission is low (∼4%) (40,45), especially when appropriate prophylactic antibiotic therapy is instituted (46). However, donors with systemic sepsis, multiorgan failure, tuberculosis, or infected with multi-resistant organisms should be avoided (46,47). Donors with Hepatitis B Core Antibody Positive The use of hepatitis B core antibody (anti-HBc) positive donor livers and post-operative management remain controversial. Recipients who receive anti-HBc-positive livers are at a greater risk of developing de novo HBV with reported incidences of 72% (48) to 78% (49), compared to candidates who had antiHBc-negative livers (0.5%) (49). This is due to detectable levels of HBV DNA in serum and/or liver tissues of anti-HBc-positive donors, despite no other serological markers of HBV being present (50,51). Interestingly, studies have shown that recipients who are anti-HBc and/or anti-HBs-positive are protected with a much lower risk of de novo HBV from an anti-HBcpositive liver (48,52). Management strategies to prevent de novo HBV in recipients of anti-HBc-positive livers involve using lamivudine and/ or hepatitis B immunoglobulin (HBIG) depending on the serological status of the donor and recipient. Recently, adefovir dipivoxil was proposed as an alternative to HBIG (23). There is no consensus as yet on the prophylactic strategies with some units preferring HBIG monotherapy (53) and others using lamivudine only (54). Our unit’s protocol comprises of long-term lamivudine for recipients of anti-HBc-positive livers. For patients receiving hepatitis B surface antigen-positive livers, long-term lamivudine and adefovir regimens are used.

290

Donors with Hepatitis C Infection Livers from donors infected with HCV represent 2% to 6% of the donor pool. As the risk of HCV transmission is high, HCVinfected liver grafts should be limited to HCV-positive recipients only. Ideally, a biopsy of the HCV-positive graft should be obtained prior to transplantation to assess fibrosis (47). Using HCV-positive grafts with minimal inflammation and fibrosis does not negatively impact outcome. Several studies (55–58) have shown comparable 5-year graft and patient survival between HCV-positive and HCV-negative liver grafts in HCVinfected patients. Donors with Malignancy As the age of donors increases, the risk of cadaveric donors with previous malignancy increases. Currently, the incidence of donors having either an active or treated malignancy is 3% (59) and the risk of transmitted cancer is between 0.02% (60) and 0.2% (61). The most common cancer in an organ donor is skin malignancy followed by brain tumor (59). The risk of donor transmitted malignancy increases with tumor grade (Table 32.4) and any donors with a history of metastatic cancer should be excluded. Split-Liver Split-liver transplantation (SLT) in adults is associated with increased graft failure and recipient morbidity (33,62,63). Current UNOS criteria for potential splittable donors are donor age 400 ml) then multiple procedures can be attempted. The catheter should be left in place and the procedure repeated according to daily cyst drainage. Complications from the procedure are mild pain, transient hyperpyrexia, nausea and vomiting. To reduce these symptoms, patients can be premedicated with an opiate analgesic and an anti-emetic, and a local anesthetic, such as lignocaine can be instilled into the cyst prior to the injection of the

alcohol. Severe pain can be a sign of alcohol leaking into the peritoneal cavity. If this occurs the procedure should be abandoned and repeated at a later date. Significant but uncommon complications relate to the needle puncture. These include pleural effusion and hemothorax (32). There are also reports of a transient elevation of blood alcohol level, and for this reason patients should be advised not to drive or operate machinery following the procedure. Aspiration sclerotherapy does have several advantages over other more invasive approaches. It is a relatively simple technique which does not require general anesthesia and can be performed on an out-patient basis. However, it does require experienced radiologists with expertise in the interpretation of hepatobiliary imaging and with competence in interventional hepatobiliary procedures. For these reasons the procedure should only be performed where such expertise is available. The published results of aspiration sclerotherapy appear encouraging. Recurrence rates are reported at between 0 and 30% (11,26–28,31–37), although there is significant disparity between series. The number of patients is often small; the indications are variable with polycystic patients often enrolled alongside patients with simple cysts. The techniques reported vary in terms of sclerosant concentration, volume administered, duration of sclerosis, and whether single or multiple procedures were performed. The length of patient follow-up is variable and reported outcome measures include symptomatic relief or a reduction in cyst volume or the complete ablation of the cyst. Recently, one group has reported similar results when comparing prolonged negative-pressure catheter drainage with single session alcohol sclerotherapy (38). Clearly further studies with greater patient numbers and standardized outcome measures, including quality of life assessment, are required to fully assess this technique.

surgical treatment Figure 33.1 Multidetector computed tomography (MDCT) image with iodinated intravenous contrast media. A large simple cyst is seen in segments 6 and 7. A small cyst is noted on the periphery of segment 2.

(A)

There are three surgical techniques that have been employed in the management of simple benign liver cysts. These are fenestration (deroofing of the cyst), local excision of the cyst (cystectomy), and anatomic or non-anatomic liver resection.

(B)

Figure 33.2 (A) T1-weighted MRI image through the upper abdomen showing multiple cysts throughout the liver. Fluid is dark on T1-weighting (B) T2-weighted coronal view of the abdomen of the same patient as in Figure 33.2A. The cysts are clearly seen as bright throughout the liver. This section is taken through the vascular pedicle of the liver.

302

BENIGN CYSTIC DISEASE OF THE LIVER Fenestration has emerged as the most popular technique with fewer complications than the more radical procedures. The concept of cyst fenestration was first described by Lin et al. in 1968 (39). The principle is to form a permanent communication between the cyst cavity and the peritoneum so that any fluid subsequently produced by the cyst lining can be re-absorbed by the peritoneum. The technique involves aspiration of the cyst content, which can be inspected and sent for cytology, microbiological culture, and tumor marker analysis, followed by a wide excision of the extra-hepatic portion of the cyst wall. The cavity can be inspected and biopsies taken from any area of concern. The excised cyst wall should be subjected to histopathological assessment. Hemostasis can be achieved by oversewing the edge of the cyst wall.

laparoscopic surgery In recent years, the technique of laparoscopic cyst fenestration has developed and should now be the first-line surgical approach for patients with symptomatic simple cysts. Laparoscopic cyst fenestration has all the advantages of minimally invasive surgery including reduced postoperative pain, shorter hospital stay, and quicker recovery. The laparoscopic procedure was initially recommended for cysts in segments II, III, IVb, and V. However, with increasing experience, location should not be considered a contraindication to laparoscopic surgery (40). A standard 4 port laparoscopic technique is used with the patient in a supine position, although for right-sided posteriorly placed cysts a left lateral position can give superior access. An angled laparoscope (30 or 45 degrees) gives superior views of the cyst cavity. The cyst can be punctured by insertion of a trocar through the exposed wall. The contents are sent for analysis. A wide excision of the cyst wall is then made, taking great care not to venture into the hepatic parenchyma. Various techniques have been employed to achieve hemostasis of the remnant cyst wall including diathermy, over sewing of the remnant cyst wall edge and using a laparoscopic linear cutting vascular stapler to excise the cyst wall. The authors’ preference is to use harmonic shears. The resected cyst wall is removed in an endoscopic retrieval bag, and the remnant cyst wall is inspected. Although unusual, if a bile leak is identified it must be controlled either with a suture or with a laparoscopic clip. Some authors advocate obliteration of the residual cyst wall with electrocautery. If this is attempted it should be done without breaching the cyst wall as major vascular structures, distorted by the cyst, can lie just underneath. The argon beam coagulator is probably best suited to the task (41) as the burn is very superficial. There is the potential risk of gas embolus and careful control of intraperitoneal pressure must be maintained. To prevent recurrence of the cyst, particularly when the exposed cyst cavity will come to lie against the abdominal wall, an omentoplasty should be performed. A pedicle of omentum is mobilized from the transverse colon and advanced into the cyst cavity. It can be secured either with sutures or with laparoscopic clips. A cholecystectomy is not routinely performed unless there is evidence of cholecystolithiasis, or where the cyst drainage leaves the gallbladder excessively mobile and at risk of subsequent torsion.

The results of laparoscopic cyst fenestration are encouraging though most reported studies have few patients with short follow-up periods. Three recent studies have reported outcomes in patients treated with laparoscopic cyst fenestration with a mean follow-up greater than 4 years (40,42,43). There is no reported mortality, with asymptomatic recurrence seen in up to 50% (40) but symptomatic recurrence noted in only 4.5% (43).

open surgery With the development of the laparoscopic technique, the procedure of open cyst fenestration for benign simple cysts should be reserved for those patients with recurrent disease or with extensive abdominal adhesions that preclude the laparoscopic technique. More radical procedures, including cystectomy and liver resection, carry a greater morbidity and mortality than cyst fenestration (44). Cystectomy can be particularly dangerous as the adjacent parenchyma is compressed and major vascular structures are often within the walls of the cyst. The main indication for resection is when there is suspicion that the cyst may be neoplastic in nature.

polycystic liver disease Adult Polycystic Liver Disease (PCLD) is a hereditary condition characterized by the development of multiple macroscopic and microscopic cysts throughout the liver. They are histopathologically similar to simple biliary cysts. There are two main forms of the disease and both show an autosomal dominant pattern of inheritance. The most common form of PCLD develops in association with autosomal-dominant polycystic kidney disease (ADPKD) (Figs. 33.3 and 33.4). This is one of the most common inherited diseases with a prevalence of 1 in 400 to 1 in 1000 and accounts for 8% to 10% of all cases of end stage renal failure (45). Factors associated with more extensive hepatic involvement are increasing age, female gender, severity of renal disease, and severity of renal dysfunction (46). The prevalence of liver cysts in ADPKD rises from 20% in the third to 70% in the seventh decade of life (47). The severity of disease in females correlates with the number of pregnancies and the use of exogenous female hormones (46), and may be due to the stimulatory effects of estrogen (48). The much rarer hereditary form of PCLD, known as Autosomal Dominant Polycystic Liver Disease (ADPLD), occurs with liver-only involvement (49).

clinical presentation The majority of patients with PCLD are asymptomatic and, as with simple cysts, the diagnosis is made during routine investigation (50). Laboratory tests of liver function including bilirubin, alkaline phosphatase, alanine aminotransferase, and prothrombin time are usually normal. Symptoms tend to occur in patients with longstanding disease and are related to liver enlargement and compression of adjacent organs. Patients may complain of an increase in abdominal girth, upper abdominal pain, early satiety, nausea, respiratory compromise, and limitation in physical ability. More significant complications include hemorrhage into a cyst, infection within a cyst and compression of vascular and

303

SURGICAL MANAGEMENT OF HEPATOBILIARY AND PANCREATIC DISORDERS Table 33.1 Gigot Classification of Adult Polycystic Liver Disease Classification Type I Type II

Type III

Figure 33.3 Portal venous phase MDCT image showing multiple large cysts throughout the liver. Cysts are also noted within the kidneys.

Features Limited number (