C o n s u l t a n t Pat h o l o g y
liver pathology
Consultant Pathology Series David E. Elder, MB, ChB, FRCPA Series Editor Tumorigenic Melanocytic Proliferations David E. Elder Brain Tumors Richard Prayson, Bette Kleinschmidt-DeMasters, and Mark L. Cohen Head and Neck Pathology Leon Barnes, Raja Seethala, and Simion Chiosea Liver Pathology Linda D. Ferrell and Sanjay Kakar
Forthcoming Volumes in the Series Thyroid Papillary Lesions Virginia A. LiVolsi and Jennifer L. Hunt Urinary Bladder Diagnosis Robert O. Petersen
C o n s u l t a n t Pat h o l o g y Volume 4
liver pathology EDITORS Linda D. Ferrell, MD Professor and Vice Chair of Clinical Affairs Director of Surgical Pathology Department of Pathology University of California San Francisco San Francisco, California Sanjay Kakar, MD Associate Professor and Vice Chair of Pathology University of California, San Francisco Chief of Pathology San Francisco VA Medical Center San Francisco, California
New York
Acquisitions Editor: Richard Winters Cover Design: Joe Tenerelli Compositor: S4Carlisle Publishing Services Printer: SCI Visit our website at www.demosmedpub.com © 2011 Demos Medical Publishing, LLC. All rights reserved. This book is protected by copyright. No part of it may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Medicine is an ever-changing science. Research and clinical experience are continually expanding our knowledge, in particular our understanding of proper treatment and drug therapy. The authors, editors, and publisher have made every effort to ensure that all information in this book is in accordance with the state of knowledge at the time of production of the book. Nevertheless, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application of the information in this book and make no warranty, expressed or implied, with respect to the contents of the publication. Every reader should examine carefully the package inserts accompanying each drug and should carefully check whether the dosage schedules mentioned therein or the contraindications stated by the manufacturer differ from the statements made in this book. Such examination is particularly important with drugs that are either rarely used or have been newly released on the market. Library of Congress Cataloging-in-Publication Data Liver pathology / Linda D. Ferrell and Sanjay Kakar, editors. p. ; cm. — (Consultant pathology series ; 4) Includes bibliographical references. ISBN 978-1-933864-93-8 1. Liver — Diseases. I. Ferrell, Linda D. II. Kakar, Sanjay.
III. Series: Consultant pathology series ; 4.
[DNLM: 1. Liver Diseases — pathology. 2. Liver — pathology. WI 700] RC846.9.L582 2011 616.3'623071 — dc22
2011001173
Special discounts on bulk quantities of Demos Medical Publishing books are available to corporations, professional associations, pharmaceutical companies, health care organizations, and other qualifying groups. For details, please contact: Special Sales Department Demos Medical Publishing 11 W. 42nd Street, 15th Floor New York, NY 10036 Phone: 800–532–8663 or 212–683–0072 Fax: 212–941–7842 E-mail:
[email protected] Made in the United States of America 11 12 13 14 15
5 4 3 2 1
My parents, Savita and Ramesh, for guiding me to make erudite choices and inculcating a simple uncomplicated view of life. My late uncle Vinni, for being a treasure trove of inspiration and an unending source of the zeal to excel. My sister, Shalini, for always considering me an unconditional hero capable of achieving anything. My mentors, Larry and Linda, for shepherding me in the world of pathology, and for sharing their penchant of making the correct diagnosis, both in life and in liver biopsies. My patients, whose condition I hope this exercise will help to ameliorate in some small way. My wife Shalini and son Soham for making everything worthwhile. SANJAY KAKAR
My husband, Rick, for his support, even when I was working late and at home on the book. Roddy and Peter, for helping me start along the pathway of liver. All the co-authors, who helped make this book possible. LINDA D. FERRELL
This page intentionally left blank
C on ten ts
Series Foreword Preface Acknowledgments Contributors
xv xvii xix xxi
1. Acute Hepatitis
1
Sanjay Kakar 1.1 Acute Hepatitis With Inflammation-Dominant Pattern 1.2 Acute Hepatitis With Bridging Necrosis 1.3 Resolving Hepatitis 1.4 Nonspecific Reactive Hepatitis
2. Acute Liver Failure
7 9 11 13
15
Rageshree Ramachandran and Sanjay Kakar 2.1
Acute Liver Failure With Necrosis-Dominant Injury Pattern Sanjay Kakar
3. Autoimmune Hepatitis/Overlap Syndromes
18
21
Kay Washington 3.1 3.2 3.3 3.4 3.5
Autoimmune Hepatitis With Bile Duct Injury Versus Primary Biliary Cirrhosis Autoimmune Hepatitis–Primary Biliary Cirrhosis Overlap Syndrome Autoimmune Hepatitis–Primary Sclerosing Cholangitis Overlap Syndrome Chronic Hepatitis C With Autoantibodies Versus Autoimmune Hepatitis Syncytial Giant Cell Hepatitis
4. Fatty Liver Disease
26 30 32 34 36
39
Matthew M. Yeh and Elizabeth M. Brunt 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8
Steatosis With Inflammation Versus Steatohepatitis Steatohepatitis With Minimal Ballooning Steatohepatitis Without Activity Nonalcoholic Steatohepatitis With Moderate/Marked Portal Inflammation Steatohepatitis With Elevated Serum Iron Indices and Siderosis Alcoholic Steatohepatitis Drugs and NAFLD Microvesicular Steatosis
vii
41 43 45 47 52 54 56 57
viii
CONTENTS
4.9 4.10 4.11
Pediatric Fatty Liver Disease Cynthia Behling Chemotherapy-Associated Steatohepatitis Due to Irinotecan Vikram Deshpande and Gregory Y. Lauwers Subacute Steatohepatitis Linda D. Ferrell
5. Hepatic Granulomas and Granulomatous Hepatitis
58 61 64
67
Laura W. Lamps 5.1 5.2 5.3 5.4 5.5 5.6
Necrotizing Epithelioid Granulomas Sarcoidosis Fibrin Ring Granulomas Schistosomiasis Cat Scratch Disease Chronic Granulomatous Disease David E. Kleiner
6. Cholestasis
70 73 77 79 82 85
89
Jay H. Lefkowitch 6.1 6.2 6.3 6.4 6.5 6.6 6.7
Drug-Induced Pure Cholestasis Bile Ductular Cholestasis Associated With Sepsis Crohn Disease With Primary Sclerosing Cholangitis Chronic Large Bile Duct Obstruction of Uncertain Cause Primary Sclerosing Cholangitis, Exclude Cholangiocarcinoma Immunoglobin G4–Associated Cholangitis Versus Primary Sclerosing Cholangitis Vikram Deshpande Hepatolithiasis Polly W. Y. Lam and Wilson M. S. Tsui
7. Bile Duct Damage and Ductopenia
96 99 102 104 107 110 113
117
Kay Washington 7.1 7.2 7.3 7.4
Primary Biliary Cirrhosis With Nonspecific Changes and Positive Antimitochondrial Antibody Antimitochondrial Antibody-Negative Primary Biliary Cirrhosis Primary Biliary Cirrhosis With Ductopenia Primary Biliary Cirrhosis With Cirrhosis
8. Ductal Plate Malformations and Cystic Diseases of the Liver
120 123 125 128
131
Barton Kenney and Dhanpat Jain 8.1 8.2 8.3 8.4 8.5
Congenital Hepatic Fibrosis Versus Cirrhosis Caroli Disease/Syndrome Versus Other Cystic Disease Adult Polycystic Liver Disease Versus Caroli Disease Choledochal Cyst Solitary Hepatic Cyst Versus Hydatid Cyst
143 145 147 149 151
CONTENTS
9. Hereditary Hyperbilirubinemias
ix
153
Sarangarajan Ranganathan 9.1
Dubin-Johnson Syndrome
10. Neonatal Cholestatic Liver Disease
157
159
Grace E. Kim and Linda D. Ferrell 10.1 Biliary Atresia 10.2 Neonatal Hepatitis With Hypopituitarism 10.3 Paucity of Intrahepatic Bile Ducts 10.4 Neonatal Hepatitis Due to Alpha-1-Antitrypsin Deficiency
11. Sinusoidal Dilatation and Congestion
164 167 170 172
175
Sanjay Kakar 11.1 Budd-Chiari Syndrome Versus Biliary Disease 11.2 Sinusoidal Obstruction 11.3 Veno-Occlusive Disease (Sinusoidal Obstruction Syndrome) 11.4 Amyloidosis
12. Peliosis Hepatis
177 180 182 184
187
Sandra Fischer and Maha Guindi 12.1 Alcoholic Lipopeliosis 12.2 Lipopeliosis in Transplanted Donor Livers
13. Portal Hypertension Without Cirrhosis
188 189
191
Sandra Fischer and Maha Guindi 13.1 Hepatoportal Sclerosis 13.2 Portal Vein Thrombosis 13.3 Budd-Chiari Syndrome 13.4 Regressed Cirrhosis Case
14. Clinical and Morphological Spectrum of Liver Diseases in Pregnancy
191 193 195 199
203
Andrew Kenneth Burroughs and Amar Paul Dhillon 14.1 14.2 14.3
Recurrent Cholestasis of Pregnancy Acute Fatty Liver of Pregnancy Toxemia/HELLP Syndrome
15. Drug-Induced Liver Injury
205 207 210
213
David E. Kleiner 15.1 15.2 15.3 15.4 15.5
Acetaminophen-Induced Fulminant Liver Failure Statin-Associated Acute Hepatotoxicity Drug-Induced Autoimmune Hepatitis Drug-Induced Cholestatic Hepatitis Drug-Induced Ductopenia
217 220 223 226 229
x
CONTENTS
15.6 15.7 15.8 15.9
Methotrexate-Induced Chronic Liver Disease Liver Injury Due to Total Parenteral Nutrition Amiodarone-Induced Phospholipidosis Drug-Induced Microvesicular Steatosis
16. Cytoplasmic Globules
232 235 237 240
243
Elaine S. Chan and Matthew M. Yeh 16.1 16.2
Alpha-1-Antitrypsin Deficiency Alpha-1-Antichymotrypsin Deficiency
17. Glycogenic Abnormalities on Liver Biopsy
245 249
251
Michael Torbenson 17.1 17.2 17.3 17.4 17.5 17.6
Glycogenic Hepatopathy Glycogen in the Liver: Abnormal Versus Normal Glycogenic Hepatopathy, Cause Uncertain Glycogenic Hepatopathy, Type II Diabetes Glycogen Pseudo–Ground-Glass Smooth Endoplasmic Reticulum Proliferation
18. Macrophage Infiltrate
255 257 258 260 261 263
265
Ryan M. Gill, Sanjay Kakar, and Linda D. Ferrell 18.1 Gaucher Disease 18.2 Niemann-Pick Disease
19. Approach to Liver Biopsy With Minimal or Nonspecific Histologic Findings
271 273
275
Dhanpat Jain and Sanjay Kakar 19.1
Mild Hepatic Steatosis Versus Ito Cell Lipidosis Katharine van Patten, Sanjay Kakar, and Dhanpat Jain
20. Interpreting Iron in Liver Specimens
278
281
Michael Torbenson 20.1 20.2 20.3 20.4 20.5 20.6
Genetic Hemochromatosis Grading Iron Hepatic Iron Index Marked Hepatic Iron but No Genetic Mutation Iron in the Setting of Chronic Hepatitis C Neonatal Hemochromatosis Linda D. Ferrell
21. Wilson Disease
284 289 291 293 294 296
299
Linda D. Ferrell 21.1 21.2 21.3
Fulminant Form of Wilson Disease Chronic Hepatitis Due to Wilson Disease Cirrhosis With Chronic Hepatitis Consistent With Wilson Disease
301 304 305
CONTENTS
22. Liver Transplant Pathology
xi
307
Oyedele Adeyi 22.1 22.2 22.3 22.4 22.5 22.6 22.7 22.8 22.9 22.10
Acute Cellular Rejection Recurrent Hepatitis C Acute Cellular Rejection Versus Recurrent Hepatitis C Late Cellular Rejection Versus Autoimmune Hepatitis Versus Recurrent Hepatitis C Fibrosing Cholestatic Hepatitis C Versus Biliary Obstruction Versus Adverse Reaction to Medication Mechanical Biliary Obstruction Versus Chronic Rejection Chronic Rejection Versus Recurrent Primary Sclerosing Cholangitis Versus Non-PSC Stricture Zone 3 (Centrilobular) Necrosis Cytomegalovirus Hepatitis Graft Versus Host Disease
23. Benign Hepatocellular Lesions
309 311 313 315 319 323 326 330 337 340
343
Valérie Paradis 23.1 23.2 23.3 23.4 23.5 23.6
Atypical Focal Nodular Hyperplasias on Imaging Focal Nodular Hyperplasia Versus Inflammatory/Telangiectatic Hepatocellular Adenoma Hepatocellular Adenoma Subtyping: Inflammatory/Telangiectatic Versus Steatotic Adenoma Hepatocellular Adenoma Subtyping: Associated Liver Nodules Hepatocellular Adenoma Subtyping: Inflammatory/ Telangiectatic Adenoma Hepatocellular Adenoma Subtyping: Adenoma With Atypical Features
24. Biliary Neoplasms
347 349 351 354 356 359
361
Kisha Mitchell and Dhanpat Jain 24.1 24.2 24.3 24.4 24.5 24.6 24.7
Bile Duct Adenoma Versus Biliary Hamartoma Intrahepatic Cholangiocarcinoma Versus Hepatocellular Carcinoma Cholangiocarcinoma in Association With Von Meyenburg Complexes Diagnosis of Hilar/Extrahepatic Cholangiocarcinoma Bile Duct Cystadenoma/Carcinoma Versus Foregut Cyst Biliary Adenofibroma Vikram Deshpande and Gregory Y. Lauwers Biliary Papillomatosis/Intraductal Cholangiocarcinoma Wilson M. S. Tsui
25. Hepatocellular Carcinoma
365 370 373 375 378 381 383
387
Prodromos Hytiroglou 25.1 25.2 25.3
Well-Differentiated Hepatocellular Carcinoma Poorly Differentiated Hepatocellular Carcinoma Early Hepatocellular Carcinoma
394 398 403
xii
CONTENTS
26. Hepatocellular Carcinoma Variants
407
Shriram Jakate and Deborah Giusto 26.1
Pseudoglandular Hepatocellular Carcinoma Versus Cholangiocarcinoma and Metastatic Adenocarcinoma 26.2 Hepatocellular Versus Neuroendocrine Carcinoma 26.3 Hepatocellular Carcinoma, Clear Cell Variant 26.4 Scirrhous Hepatocellular Carcinoma 26.5 Combined Hepatocellular-Cholangiocarcinoma 26.6 Diffuse Cirrhosis-Like Hepatocellular Carcinoma 26.7 Spectrum of Cytoplasmic Contents in Hepatocellular Carcinoma 26.8 Poor Differentiation and Vascular Invasion in Hepatocellular Carcinoma 26.9 Pedunculated Hepatocellular Carcinoma 26.10 Ablated Hepatocellular Carcinoma
27. Metastatic Tumors: Illustration of Immunohistochemical Workup
409 412 414 417 419 421 423 425 427 428
431
Rageshree Ramachandran and Sanjay Kakar 27.1 27.2
Hepatocellular Carcinoma Versus Metastatic Adenocarcinoma Hepatocellular Carcinoma Versus Metastatic Polygonal Cell Tumor
28. Hepatoblastoma
436 438
441
Sarangarajan Ranganathan 28.1 28.2 28.3 28.4
Biopsy Diagnosis of Hepatoblastoma Macrotrabecular Hepatoblastoma Versus Hepatocellular Carcinoma Small Cell Hepatoblastoma Versus Other Small Round Cell Tumors Teratoid Hepatoblastoma Versus Malignant Teratoma/Yolk Sac Tumor
29. Vascular Tumors 29.1 29.2 29.3 29.4
Cavernous Hemangioma Variants Linda D. Ferrell Epithelioid Hemangioendothelioma Hala R. Makhlouf and Zachary D. Goodman Hepatic Angiosarcoma Hala R. Makhlouf and Zachary D. Goodman Infantile Hemangioma Michael Torbenson
30. Hematopoietic Tumors of the Liver
447 449 452 455
457 457 459 462 465
469
Patrick A. Treseler and John P. Higgins 30.1 30.2 30.3 30.4 30.5 30.6 30.7
Dense Small B-Cell Infiltrate With Reactive Follicles Diffuse Large B-Cell Infiltrate Portal Infiltrate With Reed-Sternberg Cells Polymorphic Lymphoid Infiltrate in a Transplant Patient Sinusoidal T-Cell Infiltrate Portal and Lobular Infiltration by Blasts Portal and Lobular Infiltration by Mature Granulocytes
476 480 482 485 487 490 492
CONTENTS
31. Other Infiltrative Neoplasms of Liver
xiii
495
Lawrence Burgart 31.1
Portal-Based Infiltrative Neoplasm Versus Biliary Disease
32. Mesenchymal Tumors of the Liver 32.1 32.2 32.3 32.4 32.5
Mesenchymal Hamartoma Wendy L. Frankel and Xiaoping Zhou Embryonal Sarcoma Wendy L. Frankel and Xiaoping Zhou Angiomyolipoma Cherise Marie Cortese and Raouf E. Nakhleh Angiomyolipoma, Inflammatory Variant Linda D. Ferrell Malignant Angiomyolipoma—Malignant Perivascular Epithelioid Cell Tumor William A. Ahrens
496
499 499 502 505 508 510
Appendices A. Adequacy of Needle Biopsy B. Grading and Staging C. Special Stains in Liver Biopsy Pathology
515 515 517 521
Index
523
This page intentionally left blank
S e r i e s Forewo rd
D
iagnostic surgical pathology remains the gold standard for diagnosis of most tumors and many inflammatory conditions in most, if not all, organ systems. The power of the morphologic method is such that, in many instances, a glance at a thin section of tissue stained with two vegetable dyes is sufficient to determine with absolute certainty whether a patient should undergo a major procedure or not, or whether a patient is likely to live a healthy life or die of an inoperable tumor. In such cases, the diagnostic process is one of “gestalt,” a form of almost instantaneous pattern recognition that is similar to the recognition of faces, different brands of automobiles, or breeds of dogs. In other “difficult” cases, the diagnosis is not so obvious. In many of these cases, a diagnosis may be possible, but may be outside of the experience of the routine practitioner. In such a circumstance, it may be possible for a practitioner with more experience—a consultant—to make a diagnosis rather readily. In other cases, the problem may really not be suited to the histologic method. In these cases as well, a consultant may be invaluable in determining that it is simply not possible to make a reliable diagnosis with the materials available. In yet other cases, the diagnosis may be ambiguous, and again a consultant’s opinion can be important in establishing a differential diagnosis that may guide clinical investigation. There are many fine consultants available to the practicing surgical pathology community. Many of them have authored textbooks, and many of them give presentations at national meetings. However, these materials can offer only a superficial insight into the vast amount of knowledge that is
embedded in these individuals’ cerebral cortices—and in their filing cabinets. This series represents an effort to enable the dissemination of this hitherto-inaccessible knowledge to the wider community. Our authors are individuals who have accumulated large collections of difficult cases and are willing to share their material and their knowledge. The cases are based on actual consultations, and the indications for the consultation, when available, are presented, because these are the records of the manner in which these cases presented themselves as being problematic. We have asked the consultants, when possible, to present their consultation letters in much the same form (albeit edited to some degree) as that in which they were first presented, because these represent the true records of the clinical encounter. In addition, we asked the authors to amplify upon these descriptions, with brief reference to the literature, and to richly illustrate the case reports with high-quality digital images. Images from books in the series, as well as additional images to amplify the presentation of the cases, will be made available on a website for downloading, study, and use in education. These images, in some cases, have been derived from virtual slides, which also may be made available in the future from a digital repository for their additional educational value. David E. Elder, MB, ChB, FRCPA Professor of Pathology and Laboratory Medicine Hospital of the University of Pennsylvania Philadelphia, Pennsylvania
xv
This page intentionally left blank
Preface
T
he Hippocratic Oath enshrines the idea of medicine as more of art than science. This is particularly true of liver pathology, where the blend of the myriad H&E patterns, the intricate maze of reticulin network, the rich hues of a trichrome, the deep azure of the Prussian blue, and the blithe shades of PASD can inspire even the most insipid of imaginations. This text is an attempt to amalgamate the art and science of liver pathology using histologic patterns as a template to embark upon the diagnostic exercise. This book is part of the Consultant Pathology Series and will focus on diagnostic problems in liver pathology through case examples. The book is divided into two sections: non-neoplastic and neoplastic liver diseases. Liver biopsy interpretation in non-neoplastic liver diseases involves two steps: recognition of the morphological pattern of injury and identification of the disease processes that led to the identified pattern. The first step is essentially a morphological exercise and since the liver has a limited repertoire of response to injury, there is considerable overlap of morphological patterns across diverse disease processes. The second step requires correlation of the morphological pattern with clinical presentation, radiological findings, and laboratory data to establish an etiological diagnosis. It can be valuable to examine the biopsy without knowledge of clinical information for an unbiased appraisal of histologic abnormalities and perhaps as an act of showmanship for an uninitiated trainee or clinical colleague, but any attempt to establish the final liver biopsy diagnosis without the requisite clinical information is an act of brashness that is avoided by the astute pathologist. The non-neoplastic diseases section follows a patternbased approach comprising an introduction that defines the clinical and morphological features of the pattern and the relevant differential diagnoses. The individual disease entities and the exercise of reaching an etiological diagnosis are illustrated through case examples. The examples have been chosen from cases referred to experts in the field and highlight
the wide spectrum of problems in liver biopsy diagnosis. The format includes a case history, relevant laboratory data, imaging information and histologic features, followed by a discussion. The latter focuses on how the diagnosis was established based on the provided clinical and histologic features, and how the relevant differential diagnoses were excluded. The initial chapters cover patterns encountered in the most hepatitic and biliary disorders, followed by developmental, genetic, and pediatric diseases. Adverse drug reaction is one of the most common causes of liver injury and a systematic evidencebased approach is presented in chapter 15, embellished with numerous case examples. Both academic and community pathologists are increasingly presented with allograft-related problems. Chapter 22 outlines a pragmatic approach to determining the etiology of allograft liver dysfunction with emphasis on differential diagnosis. The neoplastic diseases section also follows a case-based approach with emphasis on the morphological and immunohistochemical features that establish the diagnosis and exclude other close mimics. Since limited tissue is available in needle biopsies, judicious use of immunohistochemistry is critical for diagnosis and a practical approach is outlined in chapter 27. In addition to typical hepatocellular and biliary tumors and biliary neoplasms, the approach to challenging situations like histological variants, hematopoietic disorders, and mesenchymal tumors is also illustrated through case examples. We hope that the style of this text and of this series in general will provide the requisite ammunition to the pathologist for assimilation of clinical and histological data to solve diagnostic conundrums, judiciously employ the immunohistochemical armamentarium, and generate pathology reports that provide the missing pieces of the jigsaw puzzle to the hepatologist.
xvii
Linda D. Ferrell, MD Sanjay Kakar, MD
This page intentionally left blank
Ac k n ow ledg m en ts
We would like to acknowledge the help of Caren Hale for organizing the manuscript and images. We appreciate the efforts of our publisher, Demos Medical Publishing, in their meticulous scrutiny of the written material and molding the text into its final shape. Special mention is due to Rich Winters, Executive Editor, for his unstinting effort and support throughout the gestation of this text as it evolved from an idea to its present final form.
xix
This page intentionally left blank
C o ntribu to rs
Oyedele Adeyi, MB, BS, FCAP, FRCP(C) Staff Pathologist Laboratory Medicine Program University Health Network Assistant Professor Laboratory Medicine and Pathobiology University of Toronto Toronto, Ontario, Canada William A. Ahrens, MD Adjunct Assistant Professor Department of Pathology University of North Carolina, Chapel Hill Chapel Hill, North Carolina Director of Gastrointestinal and Hepatobiliary Pathology Department of Pathology Carolinas Medical Center Charlotte, North Carolina Cynthia Behling, MD, PhD Staff Pathologist Department of Pathology Pacific Rim Pathology Group Sharp Memorial Hospital San Diego, California Elizabeth M. Brunt, MD Professor Department of Pathology and Immunology Washington University St. Louis, Missouri Lawrence Burgart, MD Clinical Professor of Pathology Department of Pathology University of Minnesota College of Medicine Minneapolis, Minnesota
Andrew Kenneth Burroughs, MBChB, Hons, FEBG, FRCP, FMedSci Consultant Physician and Hepatologist Royal Free and University College Medical School Professor of Hepatology University College London London, United Kingdom
Sandra Fischer, MD Assistant Professor of Pathology Laboratory Medicine Program Department of Laboratory Medicine and Pathobiology University Health Network University of Toronto Toronto, Ontario, Canada
Elaine S. Chan, MD Resident Department of Pathology University of Washington Seattle, Washington
Wendy L. Frankel, MD Vice Chair and Director of Anatomic Pathology Department of Pathology The Ohio State University Columbus, Ohio
Cherise Marie Cortese, MD Assistant Professor and Consultant Department of Laboratory Medicine and Pathology Mayo Clinic, Florida Jacksonville, Florida Vikram Deshpande, MD Assistant Professor of Pathology Department of Pathology Massachusetts General Hospital Assistant Professor of Pathology Department of Pathology Harvard Medical School Boston, Massachusetts Amar Paul Dhillon, MD, FRCP, FRCPath Professor of Histopathology Department of Cellular Pathology University College of London Medical School London, United Kingdom Linda D. Ferrell, MD Professor and Vice Chair of Clinical Affairs Director of Surgical Pathology Department of Pathology University of California, San Francisco San Francisco, California
xxi
Ryan M. Gill, MD, PhD Clinical Instructor Department of Pathology University of California San Francisco, California Deborah Giusto, MD Associate Pathologist Department of Pathology 4path Pathology Services Justice, Illinois Zachary D. Goodman, MD, PhD Director, Liver Pathology Research Department of Pathology Center for Liver Diseases Inova Fairfax Hospital Falls Church, Virginia Maha Guindi, FRCPC Associate Professor of Pathology Laboratory Medicine Program Department of Laboratory Medicine and Pathobiology University Health Network University of Toronto Toronto, Ontario, Canada
xxii
John P. Higgins, MD Associate Professor of Pathology Department of Pathology Stanford University Stanford, California Prodomos Hytiroglou, MD Professor of Pathology Department of Pathology Aristotle University Medical School Thessaloniki, Greece Dhanpat Jain, MBBS, MD Director of Gastrointestinal Pathology Associate Professor Department of Anatomic Pathology Yale University School of Medicine New Haven, Connecticut Shriram Jakate, MD, FRCPath Professor of Pathology, Gastroenterology, and Hepatology Department of Pathology Rush University Medical Center Chicago, Illinois Sanjay Kakar, MD Associate Professor and Vice Chair of Pathology University of California, San Francisco Chief of Pathology San Francisco VA Medical Center San Francisco, California Barton Kenney, MD Assistant Professor Department of Anatomic Pathology Yale University School of Medicine New Haven, Connecticut Grace E. Kim, MD Professor of Pathology Associate Director of Surgical Pathology Director of Pediatric Pathology Department of Anatomic Pathology University of California, San Francisco San Francisco, California David E. Kleiner, MD, PhD Director, Clinical Operations Laboratory of Pathology National Cancer Institute National Institutes of Health Bethesda, Maryland Polly W. Y. Lam, FRCPA, FHKCPath Senior Medical Officer Department of Pathology Queen Elizabeth Hospital Hong Kong, China
CONTRIBUTORS
Laura W. Lamps, MD Professor and Vice Chair Director, Diagnostic Laboratories Department of Pathology University of Arkansas for Medical Sciences Little Rock, Arkansas Gregory Y. Lauwers, MD Vice Chairman Department of Pathology Massachusetts General Hospital Chief, Gastrointestinal Pathology Service Department of Pathology Massachusetts General Hospital Professor of Pathology Department of Pathology Harvard Medical School Boston, Massachusetts Jay H. Lefkowitch, MD Professor of Clinical Pathology Department of Pathology College of Physicians and Surgeons Columbia University New York, New York Hala R. Makhlouf, MD, PhD Chief, Division of Hepatic Pathology Department of Hepatic and Gastrointestinal Pathology Armed Forces Institute of Pathology Washington, DC Kisha Mitchell, MD Assistant Professor Department of Anatomic Pathology Yale University School of Medicine New Haven, Connecticut Raouf E. Nakhleh, MD Professor Laboratory Medicine and Pathology Mayo Clinic, Florida Jacksonville, Florida Valérie Paradis, PD, PhD Professor Department of Pathology Beaujon Hospital INSERM U773 Beaujon Hospital Clichy, France Rageshree Ramachandran, MD, PhD Department of Pathology University of California San Francisco, California
Sarangarajan Ranganathan, MD Director, Anatomic Pathology Pediatric Pathology Division Associate Professor Department of Pathology Children’s Hospital University of Pittsburgh Medical Center Pittsburgh, Pennysylvania Michael Torbenson, MD Associate Professor Department of Pathology The Johns Hopkins School of Medicine Baltimore, Maryland Patrick A. Treseler, MD, PhD Professor of Pathology Department of Pathology University of California, San Francisco San Francisco, California Wilson M. S. Tsui, FRCPath, FHKCPath Consultant Pathologist Department of Pathology Caritas Medical Centre Hong Kong, China Katharine van Patten, MD Clinical Instructor Department of Anatomic Pathology Yale University School of Medicine New Haven, Connecticut Kay Washington, MD, PhD Professor Department of Pathology Vanderbilt University Medical Center Nashville, Tennessee Matthew M. Yeh, MD, PhD Associate Professor Department of Pathology University of Washington School of Medicine Seattle, Washington Xiaoping Zhou, MD, PhD Fellow, Gastrointestinal and Liver Pathology Department of Pathology The Ohio State University Columbus, Ohio
1 Acute Hepatitis SANJAY KAKAR
D E F I N I T I ON
4. Absence of fibrosis: Fibrosis is the morphological hallmark of chronicity in liver disease. By definition, there is no fibrosis in acute hepatitis. In clinical terms, chronicity is defined by persistence of clinical, biochemical, or serological evidence of liver dysfunction for more than 6 months.
Acute hepatitis is clinically defined by elevation of alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST) (at least twice normal). Markers of biliary disease like alkaline phosphatase (ALP) are either normal or the ratio of serum activity of ALT to serum activity of ALP is greater than or equal to 5 (1). By definition, there is no prior history of chronic liver disease. Histologically, acute hepatitis is defined by lobular hepatocellular injury. This pattern of injury typically comprises the following four features:
RO LE O F T H E PAT H O LO GIST
The diagnosis of acute hepatitis can be established in most cases from clinical and serological data. When liver biopsy is performed, the role of the pathologist is two-fold: (i) to confirm the diagnosis of acute hepatitis and exclude other etiologies like biliary disease, (ii) to help in establishing the etiology of acute hepatitis based on the morphological features. Several subpatterns can be recognized within the acute hepatitis pattern of injury. These subpatterns can be very helpful in narrowing the differential diagnosis.
1. Hepatocellular injury with or without necrosis: This is the hallmark of acute hepatitis and may be observed in the form of hepatocellular swelling and/or hepatocellular dropout (acidophil bodies/Councilman bodies/apoptosis). In severe cases, there may be necrosis of groups of hepatocytes (confluent necrosis). The necrosis may be random and nonzonal, or it may show predilection for certain zones such as around central veins. Bridging necrosis with centro-portal, centro-central, or less commonly portoportal bridging can be present, and may carry higher risk of progression to chronic hepatitis. When confluent necrosis is widespread (multiacinar confluent necrosis or panacinar necrosis), it constitutes the morphologic counterpart of fulminant hepatitis or acute liver failure (Chapter 2). The terms massive and submassive hepatic necrosis have also been used for these situations.
SUBPAT T ER NS O F ACUT E H EPAT IT IS Inflammation-Dominant Acute Hepatitis Defining features
1. Hepatocellular damage: hepatocellular swelling, acidophil bodies, necrosis (Figures 1.1 and 1.2). 2. Lobular inflammation: lymphocyte-predominant in most cases, variable numbers of other inflammatory cells can be present. 3. Portal inflammation may be present, but is not sufficient or necessary for the diagnosis.
Regenerative features like binucleate hepatocytes, mitoses, and thicker cell plates are common. Prominent Kupffer cells are often present in the sinusoids and can form small aggregates (microgranulomas). If the hepatocellular injury is severe enough to interfere with bile secretion, cholestasis can be present (cholestatic hepatitis). 2. Inflammation: The inflammatory infiltrate typically involves the hepatic parenchyma with variable involvement of portal tracts. Interface inflammation, characteristic of chronic hepatitis, can be present in some cases and is prominent in some etiologies like acute hepatitis A. The infiltrate is predominantly composed of lymphocytes, but depending on the underlying etiology and duration of illness, other types of cells like macrophages, plasma cells, and eosinophils can be present. 3. Absence of bile duct injury: Injury to interlobular bile ducts is absent or minimal. Bile ductular reaction can occur in the areas of hepatocellular necrosis and should not be misinterpreted as biliary injury. Neutrophils are commonly present in association with ductules. This phenomenon has been referred to as pericholangitis, but it should not be mistaken for acute cholangitis.
F I G U R E 1 . 1 Inflammation-dominant pattern of injury highlighting portal and lobular inflammation. The bile duct is intact. Interface activity is present. Hepatocellular swelling are not seen in the lobule.
1
2
CHAPTER
1:
ACUTE
FIGURE 1. 2 Same case as in Figure 1.1 emphasizing the hepato-
cellular swelling, dropout, lobular inflammation, and prominent Kupffer cells.
H E PAT I T I S
F I G U R E 1 . 4 Hepatocellular ballooning with Mallory-Denk bodies helps in the identification of steatohepatitic pattern of injury.
FIGURE 1. 3 Lobular inflammation and hepatocellular swelling in acute alcoholic hepatitis can mimic inflammation-dominant pattern of acute hepatitis when fat is absent or minimal.
F I G U R E 1 . 5 Same case as Figure 1.3. Pericellular fibrosis also points
Differential diagnosis
present clinically as acute hepatitis. The presence of fibrosis is helpful in establishing chronicity in these cases. The leading differential diagnoses of inflammationdominant acute hepatitis pattern are acute viral hepatitis, adverse drug reaction, autoimmune hepatitis, and Wilson disease. The etiology remains undetermined in 10% to 15% of cases.
Distinction from biliary disease is generally not a problem as the predominant pattern of hepatocellular injury is obvious and there is hepatitic pattern of liver enzyme elevations. Acute alcoholic hepatitis can clinically present as acute hepatitis. The histological pattern of injury is steatohepatitis (steatosis, hepatocellular swelling, Mallory-Denk bodies with or without pericellular fibrosis) and is usually different from the inflammation-dominant pattern of acute hepatitis. In some cases, the steatosis is minimal or absent and the hepatocellular swelling may be mistaken for acute hepatitis pattern of injury. The presence of Mallory-Denk bodies and pericellular fibrosis, along with history of alcohol intake help in establishing the correct diagnosis (Figures 1.3–1.5). In some instances, acute exacerbation of previously unknown chronic disease can
toward a steatohepatitic etiology.
1. Acute viral hepatitis: diagnosis rests on serological tests. Since these assays can occasionally yield false-negative results, polymerase chain reaction (PCR) for viral RNA/ DNA can be obtained. 2. Adverse drug reaction: diagnosis rests on history of exposure to drugs known to cause the pattern of injury observed on liver biopsy (see Chapter 15). Herbal supplements and over-the-counter medications should be
CHAPTER
1:
ACUTE
specifically sought. The temporal profile of disease onset and drug exposure is important in establishing causality. The disease usually subsides after cessation of the offending agent, but it can persist for weeks to months in some cases. Prominent eosinophilic infiltrate, granulomas, relatively sparse portal inflammatory component, sharply defined perivenular necrosis with minimal inflammation, and cholestasis out of proportion to hepatocellular injury (with or without bile duct damage) favor adverse drug reaction, but all these features lack specificity and sensitivity (see Chapter 15). 3. Autoimmune hepatitis (AIH) is a form of chronic hepatitis but can present clinically as acute hepatitis. Most cases with acute presentation also have fibrosis on liver biopsy and are classified as chronic hepatitis. However, around 10% of AIH lack fibrosis at initial presentation and need to be distinguished from other causes of acute hepatitis (2,3). High necroinflammatory activity and numerous plasma cells are typical of AIH, but are not specific. These features are seen in the active phase of the disease and are less prominent during the quiescent phase. Plasma cells can be observed in other causes of acute hepatitis and in biliary diseases like primary biliary cirrhosis (PBC). In some cases of AIH, plasma cells may be inconspicuous or absent. Hence, AIH does not have any specific histological features. The diagnosis is dependent on a scoring system based on clinical, laboratory, and histological findings (see Chapter 3). 4. Wilson disease can show steatosis or a steatohepatitis-like picture with glycogenated nuclei, acute hepatitis or fulminant hepatic failure, or chronic hepatitis or cirrhosis (see Chapter 21). Hepatic disease usually manifests earlier than neurological disease (mean age 8–12 years), and onset of liver disease after 50 years is rare. Low ceruloplasmin level and elevated 24-hour urinary copper can be obtained. Ceruloplasmin is an acute phase reactant and can be normal in 10% of cases. Histochemical stains for copper (rubeanic acid, rhodanine) or copper binding protein (orcein) are unreliable (4,5). Quantitative determination of copper by spectrophotometry using the tissue from the paraffin block is a reliable indicator of hepatic copper in Wilson disease. Normal hepatic copper levels are 15 to 55 μg/g of dry liver tissue. Patients with untreated Wilson disease invariably have levels exceeding 250 μg/g, and often exceeding 1000 μg/g. Normal hepatic copper excludes the possibility of untreated Wilson disease. Since copper is excreted in the bile, conditions with chronic cholestasis (like PBC and primary sclerosing cholangitis) also show elevated hepatic copper. However, these conditions rarely enter the clinical or histological differential diagnosis of Wilson disease. 5. Celiac disease is primarily considered a gastrointestinal disease, but can involve other organ systems. The most frequent manifestation of liver involvement by celiac disease is mild elevations of transaminases. This can be observed in 40% to 50% of untreated patients (6,7). Histologically, these patients show nonspecific reactive
H E PAT I T I S
3
hepatitis and revert to normal within 6 to 12 months of gluten-free diet. Other histological manifestations include acute hepatitis, chronic hepatitis, nodular regenerative hyperplasia and, rarely, cirrhosis. In addition, coexisting celiac disease is seen in 3% to 6% of AIH, PBC, and primary sclerosing cholangitis. The relationship of celiac disease and autoimmune liver disease is not clear. Autoimmune liver dysfunction often does not respond to gluten-free diet. Serological tests for celiac disease should be done in all cases of unexplained liver dysfunction. Cholestatic Hepatitis Pathologic features
This pattern shows features of inflammation-dominant acute hepatitis accompanied by cholestasis. This pattern is different from other acute hepatitis patterns both clinically and histologically. Clinically, a mixed hepatitis/cholestatic pattern of liver enzymes is often present with ALT and/or AST elevated (at least twice normal), ALP elevated (typically less than 5 times normal), and ratio of ALT activity to ALP activity is less than 5. Histologically, the picture can be dominated by cholestasis with only mild hepatocellular injury. Differential diagnosis
Most cases are drug related (Case 15.4). Any etiologies listed under inflammation-dominant acute hepatitis pattern can potentially lead to cholestatic hepatitis. Viral hepatitis A is well known to be associated with cholestasis (8,9). Acute Hepatitis, “Toxic” Pattern Pathologic features
The histologic picture is dominated by necrosis, whereas the inflammation is minimal or absent. Differential diagnosis
1. Dose-dependent drug toxicity: Most drugs lead to inflammation-dominant pattern of liver injury as the liver damage is immune mediated (idiosyncratic). In few instances (acetaminophen, halothane), the damage is due to direct dose-dependent toxicity of the drug or its metabolite (Case 15.1). 2. Toxins: organic solvents, mushroom poisoning, herbal medications. 3. Nonhepatotropic viruses: herpes simplex, adenovirus (Chapter 2). 4. Vascular etiologies: ischemia, acute onset of venous outflow obstruction as in Budd-Chiari syndrome. Vascular events superimposed on underlying chronic liver disease or cirrhosis can clinically present as acute hepatitis (Figure 1.6). Clinical or serological data and presence of fibrosis in areas unaffected by ischemic injury help in identifying the underlying chronic liver disease.
4
CHAPTER
1:
ACUTE
H E PAT I T I S
FIGURE 1. 6 Necrosis-dominant pattern of liver injury evidenced
F I G U R E 1 . 7 Acute hepatitis with centrizonal necrosis related
by parenchymal extinction and minimal inflammation. This was the result of an ischemic event in the setting of chronic hepatitis C and cirrhosis. Without this history, this would be interpreted as parenchymal extinction due to acute/subacute event since there is no definite fibrosis.
to Chaparral leaf, a herbal drug.
Acute Hepatitis With Bridging Necrosis Pathologic features
The morphology is similar to inflammation-dominant acute hepatitis pattern with the additional presence of bridging necrosis. Regenerative nodules can also be present. These features can be mistaken for bridging fibrosis and interpreted as chronic hepatitis or even cirrhosis. The distinction between bridging necrosis and fibrosis can be challenging on needle biopsies, but can be accomplished in most cases by trichrome and elastic stains (Case 1.2). Differential diagnosis
Same as inflammation-dominant acute hepatitis pattern.
F I G U R E 1 . 8 Same case as in Figure 1.7. Trichrome stain is helpful in highlighting the centrizonal necrosis. The pale staining cells around the central vein are macrophages. There is no fibrosis.
Acute Hepatitis With Centrizonal Necrosis Pathologic features
Inflammation-dominant acute hepatitis pattern with varying degrees of hepatocellular injury and inflammation accompanied by necrosis around the central vein (Figures 1.7 and 1.8). Differential diagnosis
Most cases are associated with adverse drug reaction, especially if the centrizonal necrosis is sharply circumscribed and is out of proportion to the inflammation. Prominent centrizonal necrosis can also occur in AIH (10).
Resolving Hepatitis Pathologic features
Mild portal and lobular inflammation with scattered pigmentladen macrophages in the sinusoids, often more prominent around the central vein. Hepatocellular damage is typically mild. Periodic acid-Schiff diastase (PASd) stain can highlight the macrophages. The PASd-positive macrophages are the most characteristic feature of this pattern, but their numbers can vary depending on the duration of the disease.
CHAPTER
1:
ACUTE
H E PAT I T I S
5
Differential diagnosis
A vast majority of cases with this pattern represent adverse drug reaction, but the resolving phase of any acute hepatitis can lead to this picture. Mild Hepatitis Pattern Pathologic features
The features are similar to inflammation-dominant acute hepatitis pattern, but are mild. Hepatocellular swelling or dropout can be minimal, and lobular inflammation can be focal. This makes it challenging to recognize this as acute hepatitis pattern and distinguish it from biliary disease. A combination of clinical and biopsy features are necessary for this distinction: 1. Liver enzymes: Elevated ALT and/or AST, with normal or near-normal ALP favors hepatitic disease. If ALP is elevated, the ratio of serum activity of ALT to serum activity of ALP 5 favors a hepatitic process. 2. Autoantibodies: Elevated IgG, smooth muscle antibodies, and liver-kidney microsomal antibodies are typical of AIH, whereas elevated IgM and antimitochondrial antibodies favor PBC. 3. Biopsy: Since only mild hepatocellular injury can be seen in biliary disease, it does not establish hepatitic etiology (Figure 1.9). Both portal and lobular inflammation can also be seen in biliary diseases. On the other hand, nondestructive inflammatory involvement of the bile ducts can be seen as a minor finding in hepatitic disease (Figure 1.10), but destructive bile duct features or ductopenia are not seen. Hence liver enzymes and autoantibodies play an important role in distinguishing mild hepatitic pattern from biliary disease. 4. Copper: Since copper is secreted in the bile, it accumulates in periportal hepatocytes in biliary diseases.
F I G U R E 1 . 1 0 Lymphocytic infiltration of bile duct in chronic hepatitis C. Destructive bile duct lesions or granulomatous cholangitis is not seen.
Demonstration of copper in periportal hepatocytes by histochemical stains (rhodanine, rubeanic acid, or orcein) favors biliary disease (11,12). Negative copper stain is not informative. Differential diagnosis
1. All the 4 etiologies listed above for inflammationdominant acute hepatitis pattern can also lead to mild hepatitis pattern of injury, although AIH is less likely with mild injury. 2. Nonspecific reactive hepatitis is a term used to describe portal and lobular inflammation, with or without mild hepatocellular damage that can occur in systemic inflammatory disorders (like systemic lupus erythematosus [SLE], rheumatoid arthritis, celiac disease) and infections, especially in the abdomen (see Chapter 1.4). 3. Chronic hepatitis. Mild hepatitis pattern without fibrosis can be seen early in the course of chronic hepatitis, especially chronic hepatitis C. Screening for anti-HCV antibodies is carried out by enzyme immunoassays. False-positive and less commonly false-negative results can occur (13). False negativity typically occurs in recent infections, immunocompromised states, dialysis, and cryoglobulinemia (14). Recombinant immunoblot assays or direct testing for viral RNA by PCR are considered confirmatory tests. Giant Cell Hepatitis Pathologic features
FIGURE 1. 9 Prominent lobular inflammation in PBC.
The features are similar to inflammation-dominant acute hepatitis pattern with additional presence of multinucleated hepatocytes. This feature is well described in neonatal hepatitis.
6
CHAPTER
1:
ACUTE
Small multinucleated cells can occur in acute hepatitis due to any cause, but true syncytial giant cells are uncommon in older children and adults. Differential diagnosis
The majority of these cases represent AIH, and autoantibodies are present in a significant number of cases (see Chapter 3.5). The overlap syndrome of AIH and primary sclerosing cholangitis has also been linked to giant cell hepatitis. Viral infections like measles, Epstein-Barr virus, paramyxovirus, HIV, and hepatotropic viruses (A, B, and C) have been associated with giant cell hepatitis. Certain drugs like 6-mercaptopurine and methotrexate have been implicated but the evidence is not strong. PAT H OL OG Y R E P ORT
The diagnosis should encompass the morphologic subpattern of acute hepatitis followed by a discussion of etiological diagnosis in the comment. The term “active hepatitis” can be used in place of “acute hepatitis” in pathology diagnosis. This term connotes active hepatocellular injury without specifying the acute or chronic nature of the process. The 2 scenarios where the term “active hepatitis” can be useful are: 1. If necrosis is present in the liver biopsy, areas of liver parenchymal collapse can be extremely difficult to distinguish from fibrosis. Trichrome and elastic stains can be helpful in this distinction. If this distinction cannot be made based on the histologic features, the use of “active hepatitis” as the diagnostic term avoids definite categorization of the disease process as acute or chronic. 2. By definition, acute hepatitis resolves within 6 months. In the absence of complete clinical information, the term “active hepatitis” can be used.
H E PAT I T I S
References 1. Watkins PB, Seeff LB. Drug-induced liver injury: summary of a single topic clinical research conference. Hepatology. 2006;43:618–631. 2. Burgart LJ, Batts KP, Ludwig J, Nikias GA, Czaja AJ. Recent-onset autoimmune hepatitis. Biopsy findings and clinical correlations. Am J Surg Pathol. 1995;19:699–708. 3. Fujiwara K, Fukuda Y, Yokosuka O. Precise histological evaluation of liver biopsy specimen is indispensable for diagnosis and treatment of acute-onset autoimmune hepatitis. J Gastroenterol. 2008;43:951–958. 4. Gollan JL, Gollan TJ. Wilson disease in 1998: genetic, diagnostic and therapeutic aspects. J Hepatol. 1998;28(suppl 1):28–36. 5. Davies SE, Williams R, Portmann B. Hepatic morphology and histochemistry of Wilson’s disease presenting as fulminant hepatic failure: a study of 11 cases. Histopathology. 1989;15:385–394. 6. Duggan JM, Duggan AE. Systematic review: the liver in coeliac disease. Aliment Pharmacol Ther. 2005;21:515–518. 7. Volta U. Pathogenesis and clinical significance of liver injury in celiac disease. Clin Rev Allergy Immunol. 2009;36:62–70. 8. Gordon SC, Reddy KR, Schiff L, Schiff ER. Prolonged intrahepatic cholestasis secondary to acute hepatitis A. Ann Intern Med. 1984;101:635–637. 9. Glikson M, Galun E, Oren R, Tur-Kaspa R, Shouval D. Relapsing hepatitis A. Review of 14 cases and literature survey. Medicine (Baltimore). 1992;71:14–23. 10. Hofer H, Oesterreicher C, Wrba F, Ferenci P, Penner E. Centrilobular necrosis in autoimmune hepatitis: a histological feature associated with acute clinical presentation. J Clin Pathol. 2006;59:246–249. 11. Lefkowitch JH. Special stains in diagnostic liver pathology. Semin Diagn Pathol. 2006;23:190–198. 12. Nemolato S, Serra S, Saccani S, Faa G. Deparaffination time: a crucial point in histochemical detection of tissue copper. Eur J Histochem. 2008;52:175–178. 13. Kesli R, Ozdemir M, Kurtoglu MG, Baykan M, Baysal B. Evaluation and comparison of three different anti-hepatitis C virus antibody tests based on chemiluminescence and enzyme-linked immunosorbent assay methods used in the diagnosis of hepatitis C infections in Turkey. J Int Med Res. 2009;37:1420–1429. 14. Chronic hepatitis C. Current disease management. National Institute of Health. http://digestive.niddk.nih.gov/ddiseases/pubs/chronichepc/ chronichepc.pdf
Case 1.1
Acute Hepatitis With Inflammation-Dominant Pattern SANJAY KAKAR
C L I N IC AL I N F OR M AT I ON
A 38-year-old man presented with a 6-week history of weakness, abdominal pain, jaundice, dark urine, and light colored stools. On examination, there was mild hepatomegaly but no stigmata of chronic liver disease. Liver enzyme results showed ALT 569 U/L, AST 621 U/L, ALP 136 U/L, total bilirubin 12.7 mg/dl, direct bilirubin 7.1 mg/dl. Serological tests for hepatitis A, B, C, D, and E were negative. There were no autoantibodies, and ceruloplasmin levels, and serum IgG were normal. No significant drug history could be elicited. Radiology studies did not reveal any evidence of venous outflow obstruction. R E A SON F OR R E F E R R AL
The clinical and histologic impression was acute hepatitis, but the etiology could not be determined.
F I G U R E 1 . 1 . 2 The hepatic parenchyma shows foci of hepatocyte dropout and prominent Kupffer cells along the sinusoids.
PAT H OL OG I C F E AT U R E S
The liver biopsy showed mild portal inflammation comprising of lymphocytes, few plasma cells, and rare eosinophils (Figure 1.1.1). There was no significant interface inflammation, and the bile ducts were intact. The hepatic parenchyma demonstrated mild hepatocellular swelling as well as several foci of hepatocyte dropout (Figure 1.1.2). Scattered foci of lobular lymphocytic inflammation were present (Figure 1.1.3). Prominent Kupffer cells were seen in the sinusoids, some forming small clusters (Figure 1.1.2). In addition, there was mild steatosis with no evidence of steatohepatitis (Figure 1.1.2).
F I G U R E 1 . 1 . 3 There are foci of lobular inflammation and mild
steatosis.
There was no fibrosis on trichrome stain. PASd stain highlighted macrophages in the sinusoids (Figure 1.1.4). There were no globules to suggest alpha-1-antitrypsin deficiency.
DIAGNO SIS
Acute hepatitis C .
FIGURE 1. 1. 1 Mild predominantly lymphocytic portal inflammation
with minimal interface activity and normal bile duct.
7
8
CHAPTER
1:
ACUTE
FIGURE 1. 1. 4 PASd stain highlights the macrophages in the
sinusoids. D I SC U SSI ON
The histologic picture shows inflammation-dominant acute hepatitis pattern of injury. The typical differential diagnosis in this setting is acute viral hepatitis, adverse drug reaction, AIH, and Wilson disease. In this patient, serological tests for hepatitis A, B, and C were negative arguing against acute viral hepatitis. After thorough review of medication history, no offending drugs or herbal supplements were identified. The lack of prominent plasma cells and absence of high necroinflammatory activity were not characteristic of AIH, although this diagnosis cannot be excluded on histologic grounds. The lack of elevated IgG and autoantibodies did not favor this possibility. Workup for Wilson disease revealed normal serum ceruloplasmin and urinary copper. Celiac disease, a rare cause of acute hepatitis, was considered but serological test for tissue transglutaminase antibodies was negative, making this unlikely.
H E PAT I T I S
The cause of 10% to 15% of acute hepatitis cases cannot be determined after complete workup. On further exploration, it was found that the patient got a tattoo 2 months prior to biopsy. He was tested for hepatitis B virus DNA, hepatitis C virus RNA, and human immunodeficiency virus (HIV) RNA. The PCR results for hepatitis C virus RNA yielded a copy number of 105/ml; tests for hepatitis B and HIV were negative. The clinical presentation, history of exposure, and positive hepatitis C virus (HCV) RNA led to the diagnosis of acute hepatitis C. Most cases of acute hepatitis C are asymptomatic and remain clinically undetected (1–4). Symptomatic disease occurs in around 25% of cases (4). Rare cases of fulminant hepatitis have been reported (1). Symptomatic disease, history of recent exposure, and demonstration of HCV RNA help to establish the diagnosis in the setting of clinical presentation of acute hepatitis. The viral RNA appears in the serum within 1 to 2 weeks of the infection (4). However, seroconversion can take 2 to 6 months and hence serological testing for antiHCV antibodies alone may yield false-negative results (1–4). Majority of the patients (50%–80%) will progress to chronic hepatitis C (1–3), defined by persistence of viral RNA after 6 months. Spontaneous resolution is more likely to occur in patients younger than 40 years, women, and those with symptomatic disease, whereas immunosuppressed individuals are more likely to progress to chronic disease (4). There is no correlation between HCV genotype or viral copy numbers and disease clearance.
References 1. Thomas DL, Seeff LB. Natural history of hepatitis C. Clin Liver Dis. 2005;9:383–398. 2. Heller T, Rehermann B. Acute hepatitis C: a multifaceted disease. Semin Liver Dis. 2005;25:7–17. 3. Chung RT. Acute hepatitis C virus infection. Clin Infect Dis. 2005;41:S14–S17. 4. Kamal SM. Acute hepatitis C: a systematic review. Am J Gastroenterol. 2008;103:1283–1297.
Case 1.2
Acute Hepatitis With Bridging Necrosis SANJAY KAKAR
C L I N IC AL I N F OR M AT I ON
A 42-year-old man presented with abdominal pain, jaundice, and tender hepatomegaly of 4 weeks’ duration. ALT and AST levels were greater than 1000 U/L, and ALP was minimally elevated. Ultrasound showed multiple liver nodules raising the possibility of cirrhosis. There was no history of prior liver disease. Serological tests for hepatitis A, B, and C as well as autoantibodies were negative. Serum ceruloplasmin was normal, and urinary copper was not elevated. There was no significant medication history. R E A SON F OR R E F E R R AL
The biopsy was preliminarily interpreted as chronic hepatitis with marked activity and possible cirrhosis, etiology undetermined. F I G U R E 1 . 2 . 2 Confluent necrosis associated with florid ductular
reaction.
PAT H OL OG I C F E AT U R E S
The liver biopsy showed marked portal and panacinar lymphoplasmacytic infiltrate (Figure 1.2.1). The bile ducts were intact. The hepatic parenchyma showed hepatocellular swelling, dropout, and confluent necrosis associated with ductular reaction (Figure 1.2.2). A nodular architecture was seen in several areas in the biopsy (Figure 1.2.3). It was difficult to determine whether this represents bridging necrosis or bridging fibrosis. The areas of bridging showed pale staining on trichrome in contrast to dense staining of portal tract collagen (Figures 1.2.4 A and B). No elastic fibers were seen in the area of bridging (Figure 1.2.5). These features support an acute process with necrosis and parenchyma collapse.
F I G U R E 1 . 2 . 3 Several areas in the biopsy showed a nodular architecture raising the possibility of cirrhosis.
DIAGNO SIS
Acute hepatitis with bridging necrosis and regenerative nodules, etiology undetermined.
DISCUSSIO N
The histologic picture shows inflammation-dominant pattern of injury with bridging necrosis and regenerative nodules. The distinction between bridging necrosis and bridging fibrosis in this setting can be challenging. The differentiation of acute
1. 2. 1 Dense portal and lobular lymphoplasmacytic infiltrate. The bile duct is intact.
FIGURE
9
10
CHAPTER
1:
ACUTE
H E PAT I T I S
A F I G U R E 1 . 2 . 5 Orcein stain shows absence of elastic fibers in the areas of bridging supporting necrosis with parenchymal collapse rather than fibrosis. Elastic fibers are present in the portal tract (left).
B FIGURE 1. 2. 4 (A) The trichrome stain showed pale staining in the
areas of bridging indicating that this represents necrosis rather than fibrosis. The portal tract shows dense staining of collagen. (B) The trichrome stain showed pale staining in the areas of bridging indicating that this represents necrosis rather than fibrosis. The portal tract shows dense staining of collagen.
hepatitis with confluent necrosis versus acute exacerbation of chronic disease can have important therapeutic implications. Although this can be challenging on hematoxylin and eosin (HE) stain on liver biopsy, histochemical stains for trichrome, elastic fibers, and reticulin can be very helpful. The combination of these 3 stains can distinguish necrosis from fibrosis in nearly all cases (1,2). 1. Trichrome stain: Areas with necrosis with parenchymal collapse are characterized by loose stroma with spaces between collagen fibers. Congestion or hemorrhage is often present in these areas. The trichrome stain shows a characteristic two-toned appearance in acute hepatitis with necrosis. The portal areas and walls of central veins
show dark and dense staining that reflects the presence of normal collagen bundles. The areas of necrosis with parenchymal collapse show lighter staining. Fibrous septa in chronic liver disease also show darkly staining dense collagen. It is important to have a good trichrome stain to appreciate the two-toned appearance. Areas of necrosis can be easily interpreted as fibrosis in overstained trichrome stain. 2. Elastic stain: Elastic fibers are present in normal portal tracts as well as in fibrous septa of chronic liver disease. They get deposited 2 to 3 months after injury. Absence of elastic fibers on elastic stains (Verhoeff elastic or orcein stain) favors necrosis over fibrosis. 3. Reticulin stain: Necrosis is accompanied by collapse of liver architecture. In these areas, reticulin stain outlines the collapsed residual architecture, but shows dense indistinct staining in fibrotic areas. The differential diagnosis in this setting is acute viral hepatitis, adverse drug reaction, AIH, and Wilson disease. Rare cases of celiac disease can also present with acute hepatitis (3,4). The clinical and laboratory workup does not support any of these possibilities in this case. The etiology of acute hepatitis cannot be determined in 10% to 15% of cases.
References 1. Scheuer PJ, Maggi G. Hepatic fibrosis and collapse: histological distinction by orcein staining. Histopathology. 1980;4:487–490. 2. Ferrell LD, Greenberg MS. Special stains can distinguish hepatic necrosis with regenerative nodules from cirrhosis. Liver Int. 2007;27: 681–686. 3. Duggan JM, Duggan AE. Systematic review: the liver in coeliac disease. Aliment Pharmacol Ther. 2005;21:515–518. 4. Volta U. Pathogenesis and clinical significance of liver injury in celiac disease. Clin Rev Allergy Immunol. 2009;36:62–70.
Case 1.3
Resolving Hepatitis SANJAY KAKAR
C L I N IC AL I N F OR M AT I ON
A 54-year-old woman presented with abdominal pain and mild jaundice for 2 weeks. Serological tests for viral hepatitis and autoantibodies were negative. ALT and AST were 400 to 500 U/L, and ALP was normal. The patient had been on lisinopril for hypertension for 2 months. Based on the clinical features, a diagnosis of lisinopril-induced liver injury was made. The drug was discontinued, but the ALT and AST did not return to normal 2 months after withdrawal, and a liver biopsy was performed. R E A SON F OR R E F E R R AL
The biopsy showed mild hepatocellular injury, and was referred for a definite diagnosis. PAT H OL OG I C F E AT U R E S F I G U R E 1 . 3 . 1 (B) The hepatic parenchyma shows mild lobular inflammation and pigment-laden macrophages along the sinusoids, especially around the central veins.
The liver biopsy shows mild portal inflammation composed of lymphocytes. The bile ducts are intact and there is no interface inflammation. Mild hepatocellular swelling is seen. There are small foci of lobular inflammation and scattered pigmentladen macrophages along the sinusoids (Figures 1.3.1A and B). PASd stain highlights these macrophages, especially around the central veins (Figure 1.3.2). D I AG N OS I S
Resolving hepatitis likely related to lisinopril-induced liver injury.
F I G U R E 1 . 3 . 2 The pigmented macrophages are highlighted by the
PASd stain. DISCUSSIO N
The histologic picture shows the resolving hepatitis pattern of injury. Mild hepatocellular damage and PASd-positive macrophages in the sinusoids are the characteristic features of this pattern. Most cases are due to adverse drug reaction. Clinical and biochemical resolution of drug-induced acute hepatitis occurs promptly in most cases following cessation of the offending agent. In a few instances, the resolution can be delayed for several weeks to months and these cases are more likely to be biopsied.
FIGURE 1. 3. 1 (A) The hepatic parenchyma shows mild lobular
inflammation and pigment-laden macrophages along the sinusoids, especially around the central veins.
11
12
CHAPTER
1:
ACUTE
This pattern can be seen in the resolving phase of acute hepatitis of any etiology. Viral hepatitis serologies and workup for AIH and Wilson disease should be considered in these cases, especially if a drug-related etiology is not readily apparent. Since hepatitis C serological tests can be false negative, testing for HCV RNA by PCR should be considered to exclude the diagnosis with certainty. Similar histological findings can be observed in nonspecific reactive hepatitis, a term used to describe mild hepatocellular injury in systemic diseases (1). These diseases include autoimmune diseases like SLE, rheumatoid arthritis, and celiac disease, and systemic infectious or inflammatory disorders.
H E PAT I T I S
Abdominal infections are particularly prone to show nonspecific reactive hepatitis. The patient was on lisinopril, which has been related to acute and fulminant hepatitis, and helped to establish the diagnosis (2).
References 1. Burt AD. Nonspecific reactive hepatitis. In: Burt AD, Portmann BC, Ferrell LD, eds. MacSween’s Pathology of the Liver. 5th ed. Churchill Livingstone, Elsevier, Philadelphia, PA; 2007: 881–883. 2. Larrey D, Babany G, Bernuau J, et al. Fulminant hepatitis after lisinopril administration. Gastroenterology. 1990;99:1832–1833.
Case 1.4
Nonspecific Reactive Hepatitis SANJAY KAKAR
C L I N IC AL I N F OR M AT I ON
A 64-year-old woman presented with acute cholecystitis and underwent cholecystectomy. Mild dilatation of the common bile duct was seen in the intrahepatic cholangiogram. The liver enzymes at the time of surgery showed mild elevation of ALT and AST (150–200 U/L), and ALP was mildly elevated (less than twice normal). Serological tests for viral hepatitis were negative. Antinuclear antibodies (ANAs) were present (1:160); test for smooth muscle antibodies was negative. An intraoperative core-needle liver biopsy was performed. R E A SON F OR R E F E R R AL
The biopsy showed mild hepatocellular inflammation and no definite cholestatic features, but a definite etiological diagnosis was not apparent.
F I G U R E 1 . 4 . 2 Mild lobular inflammation with minimal hepatocel-
lular injury.
PAT H OL OG I C F E AT U R E S
DISCUSSIO N
The liver biopsy shows mild lymphocytic portal inflammation without significant interface inflammation (Figure 1.4.1). The bile ducts are intact and there is no ductular reaction. Small foci of lobular inflammation are present (Figure 1.4.2). There is no fibrosis.
The histologic picture shows a mild hepatitis pattern of injury characterized by mild inflammation and hepatocellular damage. The patient has dilated biliary tree raising the possibility of large duct obstruction. However, the absence of portal expansion, edema, and ductular reaction does not support this possibility. The mild hepatitis pattern seen in this biopsy does not reliably distinguish between a mild acute hepatitis pattern and biliary disease. The following steps are essential to achieve this distinction:
D I AG N OS I S
Nonspecific reactive hepatitis, possibly related to acute cholecystitis.
1. Pattern of liver enzyme elevation: The mild elevation of ALT and AST with normal ALP, in this case, favors hepatitic disease. 2. Autoantibodies: ANA are not useful in distinguishing hepatitic versus biliary disease. Antimitochondrial antibodies were subsequently obtained to exclude PBC. ALP level can be normal in early PBC, and the biopsy may show nonspecific changes resembling the mild hepatitis pattern. 3. Serum immunoglobulins: Elevated serum IgG is characteristic of AIH, whereas increase in IgM level is typical of PBC. Both IgG and IgM were normal in this case. 4. Copper stain: Since copper is secreted in the bile, it tends to accumulate in biliary disease in periportal hepatocytes. Presence of periportal copper favors biliary disease, but a negative result is not helpful in excluding biliary disease. The copper stain was negative in this case. This exercise establishes a hepatitic pattern of biliary disease. The differential diagnosis includes a mild form of the 4 conditions that cause inflammation-dominant acute hepatitis
FIGURE 1. 4. 1 Mild lymphocytic inflammation in the portal tract
without interface activity.
13
14
CHAPTER
1:
ACUTE
pattern of injury: viral hepatitis, adverse drug reaction, AIH, and Wilson disease. The positive ANA had originally raised the clinical question of AIH. However, ANA can be positive in up to 30% of the normal population using 1:40 as the cutoff and in 3% with 1:320 as the cutoff. The modest transaminase elevations, negative SMA, normal IgG, absence of prominent necroinflammatory activity, and paucity of plasma cells do not provide support for AIH (1,2). The clinical and laboratory data do not support the other three possibilities. Rare possibilities like celiac disease should be evaluated with tissue transglutaminase antibodies depending on the clinical situation. In this case, the patient was undergoing surgery for acute cholecystitis. Nonspecific liver involvement is very common
H E PAT I T I S
in abdominal infections and manifests as mild hepatocellular injury and inflammation. The overall clinical and histological information points toward the diagnosis of nonspecific reactive hepatitis (3).
References 1. Czaja AJ, Norman GL. Autoantibodies in the diagnosis and management of liver disease. J Clin Gastroenterol. 2003;37:315–329. 2. Czaja AJ. Behavior and significance of autoantibodies in type 1 autoimmune hepatitis. J Hepatol. 1999;30:394–401. 3. Burt AD. Nonspecific reactive hepatitis. In: Burt AD, Portmann BC, Ferrell LD, eds. MacSween’s Pathology of the Liver. 5th ed. Churchill Livingstone Elsevier, Philadelphia, PA; 2007: 881–883.
2 Acute Liver Failure RAGESHREE RAMACHANDRAN AND SANJAY KAKAR
In Asia, viral hepatitis is the most common cause of ALF. Hepatitis B is common in China, Hong Kong, and Taiwan, whereas hepatitis E accounts for many cases in India (9). Acute flares of chronic HBV may present clinically as ALF in previously asymptomatic chronic HBV carriers. The 2000–2006 data from the Pediatric Acute Liver Failure Study Group shows that the cause cannot be identified in approximately 50% of cases (10). In the population aged 0 to 3 years, acetaminophen accounts for only 3% of cases compared with 12% overall in children. Metabolic disease represents 10% to 15%, and viral causes comprise 6% to 8% of cases. Ischemia and autoimmune disease each account for less than 10% of pediatric cases. The outcome in the pediatric setting is more favorable than that of adults, with high rates of survival for acetaminophen toxicity and hepatitis A–induced ALF. Death is attributed to cerebral edema or sepsis in most cases. Morbidities include respiratory distress with mechanical ventilatory support, acute renal failure, and infection (7). In the United States, approximately 5% to 10% of liver transplantations are performed every year for ALF (11). Outcome is predicted by degree of encephalopathy and patient age. The role of underlying etiology and histopathology (including severity of confluent necrosis) in predicting outcome is unclear (12), though biopsy diagnosis can be useful in assigning etiology, as in the case of AIH (13).
D E FI N I T I ON A N D T E R M I N OL OG Y
Acute liver failure (ALF) is defined as the onset of hepatic encephalopathy and coagulopathy within 8 to 26 weeks of the onset of symptoms in patients without known underlying liver disease (1). It can progress rapidly to multiorgan failure within a matter of days, necessitating emergency liver transplantation. Overall mortality rate is high and can reach up to 85% (2). Other terms that have been used are fulminant hepatitis, massive necrosis, and submassive necrosis. Since necrosis may not be present in every case and some cases may not be due to hepatitic causes, the term ALF is more appropriate. C L IN I C AL F E AT U R E S
In the United States, drugs are the most common cause of ALF in adults (Table 2.1), accounting for 25% to 50% of cases (3–6). Of these, acetaminophen accounts for the majority of drug-related cases. Based on the Acute Liver Failure Study Group data collected from 1998 to 2007, the remaining causes in the adult population include hepatitis B virus (HBV) (7%), hepatitis A virus (3%), autoimmune hepatitis (AIH) (5%), ischemia (4%), and Wilson disease (2%). In 14% of cases, the cause of fulminant hepatitis cannot be identified (7). Rare instances of ALF have been reported with amyloidosis (8).
PAT H O LO GIC FEAT UR ES A ND DIFFER ENT IA L DIAGNO SIS
TA B LE 2. 1 Causes of acute liver failure Disease Category
Specific Etiologies
Infectious (viral)
Hepatotropic viruses Hepatitis A, B, less commonly D, E Nonhepatotropic viruses HSV-1, HSV-2, adenovirus
Drugs/toxins
Dose-dependent toxicity Acetaminophen Ecstasy, cocaine Idiosyncratic reaction Isoniazid, halothane, herbal medications Toxins Amanita phalloides (mushrooms)
Metabolic
Wilson disease, Reye syndrome, nonalcoholic steatohepatitis (rare)
Pregnancy-associated
Acute fatty liver of pregnancy, HELLP syndrome
Vascular
Budd-Chiari syndrome, veno-occlusive disease, shock/sepsis (ischemia)
Other
AIH, metastatic neoplasms, amyloidosis
Liver biopsy is often contraindicated due to the risk of severe bleeding. In cases where biopsy is performed, the histologic features can be a useful guide to the etiology of ALF. Based on the morphological features, ALF can be divided into 3 histological patterns (Table 2.2): TA BL E 2 . 2 Histological patterns of injury in acute liver failure
and their differential diagnosis
Abbreviations: AIH, autoimmune hepatitis; HSV, herpes simplex virus.
Histological Pattern
Common Disease Entities
Necrosis with prominent inflammatory activity
Acute viral hepatitis, idiosyncratic drug reaction, AIH, Wilson disease
Necrosis with little inflammation
Dose-dependent drug toxicity (acetaminophen), toxins, nonhepatotropic viruses (herpes simplex), vascular etiologies
Microvesicular steatosis
Drugs (tetracycline, zidovudine, valproic acid), acute fatty liver of pregnancy, rare metabolic conditions
Abbreviation: AIH, autoimmune hepatitis.
15
16
CHAPTER
2:
ACUTE
LIVER
FAILURE
2. Necrosis with little or no inflammation. The liver shows confluent necrosis, with negligible inflammation. The leading cause in this category is dose-dependent drug injury exemplified by acetaminophen. The anesthetic agent halothane, recreational drugs such as cocaine and 3,4-methylenedioxymethylamphetamine (MDMA, Ecstasy), industrial organic compounds such as carbon tetrachloride, and herbal agents such as pennyroyal, glue thistle, and germander can lead to the same pattern. Some cases of ALF due to Wilson disease can also have this morphology. Infections with nonhepatotropic viruses typically result in necrosis with negligible inflammation. Herpes simplex and adenoviral hepatitis are the most common in this category. Other less common causes of ALF with minimal inflammation include metastatic carcinoma, acute Budd-Chiari syndrome, and ischemic hepatitis. FIGURE 2. 1 Atorvastatin-induced ALF showing confluent necrosis, prominent inflammatory activity, and a small area of residual liver parenchyma (left).
3. Extensive microvesicular steatosis. The liver biopsy shows diffuse microvesicular steatosis with little inflammation. A variable degree of cholestasis and necrosis may be present. These morphological changes are a manifestation of mitochondrial injury and are often accompanied by lactic acidosis (16).
1. Necrosis with prominent inflammatory activity. The liver shows confluent necrosis accompanied by prominent portal and panacinar inflammation (Figure 2.1). The confluent necrosis often involves the majority of the liver parenchyma (massive/submassive hepatic necrosis) and can be associated with prominent ductular reaction. Cholestasis may be present. The inflammation is dominated by lymphocytes and plasma cells; neutrophils can be seen in association with ductular reaction. Necrosis can be distinguished from fibrosis on trichrome and elastic stains (see Case 1.2).
Most cases result from drug toxicity (see Cases 4.8 and 15.9) related to tetracycline (antibiotic), valproic acid (anticonvulsant), zidovudine (nucleoside analog used in human immunodeficiency virus [HIV]), L-asparaginase (chemotherapeutic drug), and amineptine (antidepressant) (17–23). Other settings that can lead to microvesicular steatosis include alcohol foamy degeneration, Reye syndrome, acute fatty liver of pregnancy, Jamaican vomiting sickness, and rare metabolic conditions like congenital deficiency in urea-cycle enzymes and carnitine deficiency.
The differential diagnosis of this pattern of injury includes viral hepatitis, AIH, idiosyncratic drug reaction, and Wilson disease. The distinction may not be possible on histologic grounds alone, and correlation with clinical and laboratory tests is necessary. Serological tests and/or viral titers are necessary for the diagnosis of hepatotropic viral hepatitides. Immunohistochemical stains for hepatitis B core antigen and hepatitis B surface antigen can be useful in cases of acute exacerbation of chronic hepatitis but are negative in acute hepatitis B. Travel history to areas where viral hepatitis E is endemic can be helpful. The presence of numerous plasma cells raises the possibility of AIH, but it is not a specific finding. The diagnosis can be supported by elevated immunoglobulin G (IgG) levels and serum autoantibodies smooth muscle antibody. If liver biopsy is undertaken after preemptive treatment with steroids, inflammation may not be marked and plasma cells may not be prominent. Commonly implicated drugs in ALF include isoniazid, other antimicrobials (sulfonamides, cotrimoxazole, ketoconazole), monoamine oxidase inhibitors, and anticonvulsants (phenytoin, valproate) (14,15). Any drug that causes acute hepatitis can potentially cause ALF. Wilson disease should always be considered in patients younger than 50 years and especially if there is associated hemolysis.
References 1. Schiødt FV, Lee WM. Fulminant liver disease. Clin Liver Dis. 2003;7: 331–349, vi. 2. Ichai P, Samuel D. Etiology and prognosis of fulminant hepatitis in adults. Liver Transpl. 2008;14(suppl 2):S67–S79. 3. Williams R. Classification, etiology, and considerations of outcome in acute liver failure. Semin Liver Dis. 1996;16:343–348. 4. Lee WM. Acute liver failure. Clin Perspect Gastroenterol. 2001;2:101–110. 5. Williams R. Changing clinical patterns in acute liver failure. J Hepatol. 2003;39:660–661. 6. Hanje AJ, Chalasani N. How common is chronic liver disease from acute drug-induced liver injury? Gastroenterology. 2007;132:2067–2068, discussion 2068–2069. 7. Lee WM, Squires RH Jr, Nyberg SL, Doo E, Hoofnagle JH. Acute liver failure: summary of a workshop. Hepatology. 2008;47:1401–1415. 8. Hung HH, Huang DF, Tzeng CH, et al. Systemic amyloidosis manifesting as a rare cause of hepatic failure. J Chin Med Assoc. 2010;73:161–165. 9. Cheng VC, Lo CM, Lau GK. Current issues and treatment of hepatic failure including transplantation in Hong Kong and the Far East. Semin Liver Dis. 2003;23:239–250. 10. Squires RH Jr, Shneider BL, Bucuvalas J, et al. Acute liver failure in children: the first 348 patients in the pediatric acute liver failure study group. J Pediatr. 2006;148:652–658. 11. Gotthardt D, Riediger C, Weiss KH, et al. Fulminant hepatic failure: etiology and indications for liver transplantation. Nephrol Dial Transplant. 2007;22(suppl 8):viii5–viii8.
CHAPTER
2:
ACUTE
12. Hanau C, Munoz SJ, Rubin R. Histopathological heterogeneity in fulminant hepatic failure. Hepatology. 1995;21:345–351. 13. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31:929–938. 14. Russo MW, Galanko JA, Shrestha R, Fried MW, Watkins P. Liver transplantation for acute liver failure from drug-induced liver injury in the United States. Liver Transpl. 2004;10:1018–1023. 15. Saukkonen JJ, Cohn DL, Jasmer RM, et al. An official ATS statement: hepatotoxicity of antituberculosis therapy. Am J Respir Crit Care Med. 2006;174:935–952. 16. Pessayre D, Mansouri A, Berson A, Fromenty B. Mitochondrial involvement in drug-induced liver injury. Handb Exp Pharmacol. 2010;196: 311–365. 17. De Bus L, Depuydt P, Libbrecht L, et al. Severe drug-induced liver injury associated with prolonged use of linezolid. J Med Toxicol. 2010;6: 322–326.
LIVER
FAILURE
17
18. Haas S, Rockstroh JK, Spengler U, Fischer HP. Nucleoside induced hepatopathy in HIV patients. Diagnostic value of liver biopsy assessment. Pathologe. 2004;25:406–411. 19. Koch RO, Graziadei IW, Zangerle R, Romani N, Maier H, Vogel W. Acute hepatic failure and lactate acidosis associated with antiretroviral treatment for HIV. Wien Klin Wochenschr. 2003;115:135–140. 20. Bodmer M, Sulz M, Stadlmann S, Droll A, Terracciano L, Krähenbühl S. Fatal liver failure in an adult patient with acute lymphoblastic leukemia following treatment with L-asparaginase. Digestion. 2006;74:28–32. 21. Huang YL, Hong HS, Wang ZW, Kuo TT. Fatal sodium valproateinduced hypersensitivity syndrome with lichenoid dermatitis and fulminant hepatitis. J Am Acad Dermatol. 2003;49:316–319. 22. Uchida T, Kao H, Quispe-Sjogren M, Peters RL. Alcoholic foamy degeneration—a pattern of acute alcoholic injury of the liver. Gastroenterology. 1983;84:683–692. 23. Lionte C. Lethal complications after poisoning with chloroform—case report and literature review. Hum Exp Toxicol. 2010;29:615–622.
Case 2.1
Acute Liver Failure With Necrosis-Dominant Injury Pattern SANJAY KAKAR
C L I N I C AL I N F OR M AT I ON
A previously healthy 65-year-old man presented with abrupt onset of fever and abdominal pain. Regular medications included aspirin and acetaminophen. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) was markedly elevated (>1500 U/L). Serological studies for hepatitis A, B, and C as well as autoantibodies were negative. A liver biopsy was obtained. The patient was presumptively treated for acetaminophen toxicity but developed renal failure and hepatic encephalopathy. R E A S ON F OR R E F E R R A L
To establish etiology of liver failure. PAT H OL OG I C F E AT U R E S
F I G U R E 2 . 1 . 2 The nuclei of the hepatocytes at the periphery of the necrotic areas show eosinophilic ground-glass appearance with smudging and margination of nuclear chromatin.
The liver biopsy showed widespread hemorrhage and confluent nonzonal hepatocellular necrosis (Figure 2.1.1). The inflammatory infiltrate was scant. At the periphery of the necrotic areas, the nuclei of viable hepatocytes showed ground-glass or smudged appearance with margination of nuclear chromatin (Figure 2.1.2). Some nuclei showed large eosinophilic inclusions with a surrounding halo (Figure 2.1.3). Occasional multinucleated cells were present. Macrovesicular steatosis was also present, perhaps as an incidental finding. Portal tracts were largely unremarkable with intact bile ducts. Immunohistochemistry for herpes simplex virus (HSV) was positive, confirming the diagnosis (Figure 2.1.4).
F I G U R E 2 . 1 . 3 Some hepatocytes show prominent eosinophilic intranuclear inclusions surrounded by a halo.
DIAGNO SIS
Herpes simplex virus hepatitis leading to acute liver failure.
DISCUSSIO N
The histologic pattern of necrosis with minimal inflammation is typical of HSV hepatitis. The presence of viral inclusions excludes other etiologies that can cause a similar pattern of
FIGURE 2. 1. 1 Widespread hemorrhagic nonzonal necrosis with no
significant inflammation.
18
CASE
2.1:
ACUTE
LIVER
FAILURE
FIGURE 2. 1. 4 Immunohistochemistry shows herpes simplex virus
inclusions in the nuclei of infected cells.
injury, such as dose-dependent adverse drug reaction (acetaminophen), toxin-mediated injury, and vascular etiologies. HSV hepatitis usually occurs in neonates, pregnant women, or immunocompromised patients. However, fulminant HSV infection can affect otherwise healthy people. Hepatic involvement is the result of disseminated infection and can occur with both HSV types 1 and 2. Symptoms are generally nonspecific but may include fever, headache, and abdominal or muscle pain. Marked elevation of transaminases in the absence of jaundice and hepatomegaly is typical. Mucocutaneous manifestations are observed in half of the cases (1,2). Since the disease can follow a rapidly progressive course, early diagnosis and treatment with acyclovir are imperative. The presence of mucocutaneous rash, genital symptoms, and anicteric hepatitis with prominent transaminase elevations helps in raising clinical suspicion for HSV hepatitis. In cases of ALF of unknown etiology, preemptive antiviral therapy (acyclovir) is recommended during evaluation for liver transplantation (3). Survival rate is greater than 40% after transplantation for HSV-induced ALF and patients are advised to stay on lifelong antiviral prophylaxis, usually acyclovir (4). HSV-1 and HSV-2 are synonymous with human herpes viruses 1 and 2 (HHV-1 and HHV-2). Varicella-zoster virus (VZV), Epstein-Barr virus (EBV), and cytomegalovirus (CMV), are also part of the human herpes virus family (HHV-3, -4, and -5, respectively) and have the potential to cause hepatitis. EBV causes an infectious mononucleosis-like syndrome, with frequent but mild liver involvement in most cases. Liver transaminases can be elevated but jaundice is rare. The liver usually shows a diffuse lymphocytic sinusoidal infiltrate. A few atypical lymphocytes and occasional nonnecrotizing granulomas can be present. Focal apoptotic bodies can be seen, but cholestasis or prominent hepatocellular damage is not present. In situ hybridization and/or polymerase chain
WITH
NECROSIS-DOMINANT
INJURY
19
reaction can be used to confirm EBV infection (5). However, these cases are rarely biopsied. Rare cases of ALF have been reported with EBV (6). CMV is a cause of neonatal hepatitis in infancy. In immunocompetent individuals, it causes an infectious mononucleosis-like syndrome similar to EBV. Nonnecrotizing granulomas can be present. Among the immunocompromised hosts, CMV is especially common in renal transplant recipients. In liver transplant recipients, CMV hepatitis has to be distinguished from acute rejection. Characteristic CMV inclusions can be identified in hepatocytes, but significant inflammation or necrosis seen in other forms of acute hepatitis is often not present. Small neutrophilic collections resembling microabscesses can be present around the hepatocytes with viral inclusions (7,8). HHV-6 can cause mild, nonspecific hepatitis. HHV-8 seropositivity is common in patients with cirrhosis (9), but its role in ALF is unclear. Adenoviral hepatitis is histologically similar to HSV hepatitis and usually occurs in immunocompromised hosts but may rarely involve healthy individuals (10). Immunohistochemistry or in situ hybridization is necessary to confirm the diagnosis. Other less-common viral infections that can cause acute hepatitis with widespread necrosis include yellow fever, dengue fever, and Ebola fever. Parvovirus B19 can cause fulminant hepatitis in children. Mild nonspecific liver inflammation can also be seen with Coxsackie virus, rubella, and measles.
References 1. Peters DJ, Greene WH, Ruggiero F, McGarrity TJ. Herpes simplexinduced fulminant hepatitis in adults: a call for empiric therapy. Dig Dis Sci. 2000;45:2399–2404. 2. Sharma S, Mosunjac M. Herpes simplex hepatitis in adults: a search for muco-cutaneous clues. J Clin Gastroenterol. 2004;38:697–704. 3. Norvell JP, Blei AT, Jovanovic BD, Levitsky J. Herpes simplex virus hepatitis: an analysis of the published literature and institutional cases. Liver Transpl. 2007:1428–1434. 4. Riediger C, Sauer P, Matevossian E, Müller MW, Büchler P, Friess H. Herpes simplex virus sepsis and acute liver failure. Clin Transplant. 2009;23(suppl 21):37–41. 5. Suh N, Liapis H, Misdraji J, Brunt EM, Wang HL. Epstein-Barr virus hepatitis: diagnostic value of in situ hybridization, polymerase chain reaction, and immunohistochemistry on liver biopsy from immunocompetent patients. Am J Surg Pathol. 2007;31:1403–1409. 6. Ader F, Chatellier D, Le Berre R, Morand P, Fourrier F. Fulminant Epstein-Barr virus (EBV) hepatitis in a young immunocompetent subject. Med Mal Infect. 2006;36:396–398. 7. Seehofer D, Rayes N, Tullius SG, et al. CMV hepatitis after liver transplantation: incidence, clinical course, and long-term follow-up. Liver Transpl. 2002;8:1138–1146. 8. Razonable RR. Cytomegalovirus infection after liver transplantation: current concepts and challenges. World J Gastroenterol. 2008;14: 4849–4860. 9. Chou AL, Huang WW, Tsao SM, Li CT, Su CC. Human herpesvirus type 8 in patients with cirrhosis: correlation with sex, alcoholism, hepatitis B virus, disease severity, and thrombocytopenia. Am J Clin Pathol. 2008;130:231–237. 10. Wang WH, Wang HL. Fulminant adenovirus hepatitis following bone marrow transplantation. A case report and brief review of the literature. Arch Pathol Lab Med. 2003;127:e246–e248.
This page intentionally left blank
3 Autoimmune Hepatitis/Overlap Syndromes KAY WASHINGTON
modified in 1999 (2,3) (Table 3.2). This system classifies cases as definite or probable AIH, based on weighted parameters. In multiple reports, this system has a high degree of sensitivity (97%–100%) for diagnosis of AIH. The system also effectively
AU TOI M M U N E H E PAT I T I S Definition
Autoimmune hepatitis (AIH) is characterized as an unresolving hepatitis usually associated with hypergammaglobulinemia and tissue-directed autoantibodies and responding in most cases to immunosuppressive therapy. The pathogenesis appears to be aberrant autoreactivity in genetically susceptible individuals (1), with molecular mimicry between viral and self-antigens the likely basis for the autoimmune response.
TA BL E 3 . 2 The revised international autoimmune hepatitis group
modified scoring system
Diagnosis and Scoring Criteria
AIH is diagnosed by consideration of the combination of clinical and laboratory features with exclusion of other causes of liver disease such as viral hepatitis, Wilson disease, alpha1-antitrypsin deficiency, primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), alcohol abuse, and drug reaction (Table 3.1). Essential features are elevated serum transaminase levels, usually with antinuclear antibody (ANA) and/or smooth muscle antibody (SMA) seropositivity; immunoglobulin G (IgG) levels greater than 1.5 times normal; and compatible liver histology. Serum bilirubin levels are variable, but alkaline phosphatase is usually only mildly elevated. Diagnosis of AIH is straightforward in 50% of cases and is aided by serum studies and the scoring system developed by the International Autoimmune Hepatitis Working Group and TA B LE 3. 1 Histologic differential diagnosis of autoimmune
hepatitis Disease
Distinguishing Features
AIH
Prominent plasma cells Often very active necroinflammatory activity May show centrilobular necrosis and inflammation early in disease course
Chronic viral hepatitis
Portal lymphoid aggregates (HCV); steatosis (HCV); ground glass hepatocytes (HBV)
Drug-induced hepatitis
No helpful distinguishing features; may mimic AIH
Wilson disease
Mallory’s hyaline; copper deposition; prominent glycogenated nuclei
Primary biliary cirrhosis
Florid duct lesions and ductopenia
Primary sclerosing cholangitis
Fibrous obliteration of bile ducts
Category
Score
Female sex
2
ALP:AST (or ALT) ratio 1.5 1.5–3.0 3.0
2 0 2
Serum globulins or IgG above normal 2.0 1.5–2.0 1.0–1.5 1.0
3 2 1 0
Autoantibodies (ANA, SMA, or LKM-1) 1:80 1:80 1:40 1:40
3 2 1 0
Hepatitis viral markers Positive Negative
3 3
Drug history Positive Negative
4 1
Average alcohol consumption Low (25 grams/day) High (60 grams/day)
2 2
Liver histology Interface hepatitis Lymphoplasmacytic infiltrate Hepatocyte rosette pattern of regeneration None of the above Biliary changes Other features Other autoimmune disorders in patients or first-degree relatives
Abbreviations: AIH, autoimmune hepatitis; HBV, hepatitis B virus; HCV, hepatitis C virus.
3 1 1 5 3 3
Comments
Lower titers are considered significant in children and should be scored at least 1.
The patient should be tested for markers of hepatitis A, B, and C infection; tests for other viruses such as EBV and CMV may be considered. Recent use of known or suspected hepatotoxic drugs.
“Biliary changes” refers to bile duct patterns of injury typical of PBC or PSC or with ductopenia in an adequate biopsy. “Other features” are any suggesting an alternative etiology, eg, nonalcoholic fatty liver disease.
2
(continued)
21
22
CHAPTER
3:
AU T O I M M U N E
H E PAT I T I S / O V E R L A P
TA B LE 3. 2 The revised international autoimmune hepatitis group
modified scoring system (continued) Category Optional parameters in patients who are seronegative for ANA, SMA, and LKM-1 Seropositivity for other defined autoantibodies HLA DR3 or DR4
Score
Comments
2
Other defined antibodies are those with published evidence of relevance to AIH and include p-ANCA, anti-LC1, anti-SLA, and anti-ASGPR.
1 2 3
Response to therapy Complete relapse Interpretation of aggregate scores Pretreatment Definite AIH Probable AIH
15 10–15
Post-treatment Definite AIH Probable AIH
17 12–17
SYNDROMES
system that is more widely applicable in routine clinical practice (4). In this simplified system, 3 categories for grading histology are defined: 1. Typical histology for AIH. Each of the following 3 features is required: i. Interface hepatitis, with lymphocytic/lymphoplasmacytic inflammation in portal tracts and extending into the lobule (Figures 3.1 and 3.2) ii. Emperipolesis (engulfment of lymphocytes by hepatocytes) (Figure 3.1) iii. Hepatocyte rosette formation (Figure 3.3) 2. Histology compatible with AIH. Chronic hepatitis pattern of injury with lymphocytic infiltration but lacking some of the features considered “typical.” 3. Atypical histology for AIH. Features suggestive of other diagnoses (eg, steatohepatitis) are present.
Abbreviations: AIH, autoimmune hepatitis; ALP, alkaline phosphatase; ALT, alanine aminotransferase; ANA, antinuclear antibody; ASGPR, asialoglycoprotein receptor; AST, aspartate aminotransferase; CMV, cytomegalovirus; EBV, Epstein-Barr virus; HLA, human leukocyte antigen; IgG, immunoglobulin G; SMA, smooth muscle antibody; LC1, liver cytosol 1 antibody; anti-LKM-1, anti–liver-kidney microsomal antibody; p-ANCA, protoplasmic antineutrophil cytoplasmic antibodies; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; SLA, soluble liver antigen. From Ref. 2.
TA B LE 3. 3 Simplified diagnostic criteria for AIH Feature
Cutoff
Points
ANA or SMA
1:40
1
ANA or SMA
1:80
2*
OR LKM
1:40
OR SLM
Positive
IgG
Liver histology
Absence of viral hepatitis
FIGURE 3.1 Prominent interface hepatitis with emperipolesis (engulfment of lymphocytes by hepatocytes, arrow) in autoimmune hepatitis.
upper limit of normal
1
1.10 times upper limit of normal
2
Compatible with AIH
1
Typical AIH
2
Yes
2 6: probably AIH 7: definite AIH
*Addition of points achieved for all antibodies (maximum, 2 points) Abbreviations: AIH, autoimmune hepatitis; ANA, antinuclear antibody; IgG, immunoglobulin G; LKM, liver-kidney microsomal antibody; SMA, smooth muscle antibody. From Ref. 4.
excludes AIH in patients with PSC and biliary disorders (96%–100% accuracy for exclusion of definite AIH). The overall diagnostic accuracy of the scoring system is roughly 90%. Recently, a simplified scoring system (Table 3.3) has been proposed by the same group, with the goal of producing a
F I G U R E 3 . 2 Lobular necroinflammatory activity with apoptotic hepatocyte (arrow) in autoimmune hepatitis.
CHAPTER
3:
AU T O I M M U N E
H E PAT I T I S / O V E R L A P
SYNDROMES
23
TA BL E 3 . 4 Helpful clinicopathologic features for distinguishing
autoimmune hepatitis, primary biliary cirrhosis, primary biliary cirrhosis, and hepatitis C virus
FIGURE 3. 3 Hepatocellular regeneration resulting in “rosettes” in autoimmune hepatitis.
Clinical and Immunologic Features of AIH Clinical features
AIH, like most autoimmune disorders, is more common in female patients, with a male:female ratio of roughly 1:4. It can present at any age, but younger patients appear to have a more severe form of the disease. Presentation varies widely, ranging from asymptomatic elevations of serum liver enzymes, to massive hepatic necrosis resulting in fulminant hepatic failure, to cirrhosis with portal hypertension. Approximately 30% of patients, usually younger patients, will have an acute presentation mimicking acute viral hepatitis. Roughly 20% of patients will be asymptomatic, with elevated transaminases identified on screening examination or during evaluation for amenorrhea, thyroid disease, arthralgia, or diabetes mellitus. Such patients tend to be older (mean age, 48 years) than patients symptomatic at diagnosis (mean age, 41 years). About 50% of patients with AIH will have concurrent autoimmune disorders, most commonly thyroid disease or rheumatoid arthritis. The presence of ulcerative colitis, however, should raise questions about the diagnosis of AIH; such patients are more likely to have PSC (Table 3.4). Immunologic features
The 3 most commonly reported antibodies in AIH are ANA, SMA, and anti-liver-kidney microsomal (LKM) antibodies (Table 3.5). In general, though useful for diagnosis, the autoantibody titer does not reliably reflect disease severity or outcome. Of the patients with AIH, 70% to 80% have ANA or SMA antibodies or both (1:40). The ANA react mainly with histones and DNA, yielding a homogeneous pattern, but other patterns also occur, with no apparent clinical significance (5–22). The SMA reacts with several cytoskeletal components, including F-actin. Overall, 3% to 4% of the patients (usually children) present with anti-LKM-1 antibodies, without ANA or SMA.
Disease
Most Useful Features
AIH
• “Definite AIH” or “probable AIH” on modified AIH scoring system • Prominent lobular hepatitis • Prominent interface hepatitis with numerous plasma cells • Negative viral studies, including HCV RNA
PBC
• • • •
PSC
• Abnormal cholangiogram • Concentric periductal fibrosis • Ductopenia
HCV
• HCV RNA in serum • Plasma cells usually not prominent • Lobular activity is usually mild; severely active chronic hepatitis favors AIH
Florid duct lesion Ductopenia High titer AMA High alkaline phosphatase with lower transaminase levels
Abbreviations: AIH, autoimmune hepatitis; AMA, antimitochondrial antibodies; HCV, hepatitis C virus; PBC, primary biliary cirrhosis; PBV, primary biliary cirrhosis; PSC, primary sclerosing cholangitis.
TA BL E 3 . 5 Autoantibodies in autoimmune hepatitis Antibody
Test
Comments
ANA
Usual test is indirect IF. Hep-2 cells give higher values than tissue sections; ELISAs not sufficiently standardized.
Specific IF pattern has no clinical significance. Other liver diseases associated with ANA include PBC, PSC, HCV, drug injury, NASH.
SMA
IF and ELISA
Directed against F-actin; found alone or in conjunction with ANA in up to 87% of patients.
SLA/LP
Immunoassay or Western blot; commercial ELISA assay now approved
High specificity for AIH but limited sensitivity.
LKM
Indirect IF
Patients with Type 2 AIH usually have only LKM antibodies. LKM antibodies may be found in other conditions, notably hepatitis C.
Abbreviations: AIH, autoimmune hepatitis; ANA, antinuclear antibody; ELISA, enzyme linked immunosorbent assay; F-actin, filamentous-actin; HCV, hepatitis C virus; IF, immunofluorescence; LKM, liver-kidney microsomal antibody; anti-LKM, anti–liver-kidney microsomal antibody; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; SLA/LP, soluble liver antigen/liver pancreas antigen; SMA, anti–smooth muscle antibody.
This serologic marker is more common in southern Europe, but the true prevalence is not known; less than 4% of patients in the United States have LKM-1 antibodies compared with 20% in Europe. Antibodies against the soluble liver antigen/ liver pancreas antigen (SLA/LP) are found in 10% to 30% with AIH; although testing for this autoantibody is not routine, it appears to have a global distribution. Protoplasmic
24
CHAPTER
3:
AU T O I M M U N E
H E PAT I T I S / O V E R L A P
antineutrophil cytoplasmic antibodies (p-ANCA), though not specific, are present in 60% to 90% of patients (23). AIH may be subclassified based on the autoantibody profile (Table 3.6); Type 1 AIH is defined as ANA and/or SMA positive and has a bimodal age distribution with peaks between 10 and 20 and 45 and 70 years of age. Type 2 AIH is characterized by the presence of anti-LKM-1 antibodies; the majority of these patients are young women with severe disease. However, because Type 1 AIH is much more common than Type 2, most young women with severe disease are Type 1 AIH. Patients with Type 2 AIH present between ages 2 and 4 years and may have ectodermal dysplasia, mucocutaneous candidiasis, and endocrinopathies suggesting autoimmune polyglandular syndrome Type 1 (24). Type 3 AIH, characterized by positivity for SLA/LP antibodies, is clinically indistinguishable from Type 1. Approximately 10% to 20% of AIH patients are initially seronegative for the conventional autoantibodies, which may appear after immunosuppressive therapy is begun. p-ANCA testing may be helpful in this subgroup of patients, and viral hepatitis B and C should be excluded prior to initiation of treatment. Natural History and Treatment of AIH
AIH is characterized by a fluctuating clinical course, with waxing and waning of necroinflammatory activity, but symptomatic AIH is considered a progressive disease that if untreated will usually result in cirrhosis. About one-third of patients are cirrhotic at presentation and have a less favorable outcome (62% vs 94% 10-year survival) (5). Rarely, patients present with acute liver failure (ALF). The mainstay of treatment for AIH is long-term immunosuppression, typically corticosteroid therapy with or without azathioprine. In most cases, the necroinflammatory process responds promptly to immunosuppressive therapy, although relapses with withdrawal of therapy are common and the disease can recur following liver transplantation. After a biochemical response is achieved, many patients can be maintained on azathioprine monotherapy. A second liver biopsy to determine treatment endpoints based on histologic response is not uniform medical practice.
SYNDROMES
F I G U R E 3 . 4 Normal histologic findings after resolution of autoimmune hepatitis with corticosteroid therapy.
It is generally felt that histologic improvement lags behind biochemical improvement by several months, and often biopsies show low-grade necroinflammatory activity as long as 6 months after normalization of transaminase levels. Some hepatologists require normal histology on liver biopsy before withdrawal of treatment (Figure 3.4). However, histologic resolution does not guarantee that the patient will remain in remission. Portal plasma cell infiltrates during immunosuppressant therapy are associated with relapse upon drug cessation (6,7). For patients who present with fulminant hepatic failure secondary to AIH, liver transplantation may be necessary. Recurrence of AIH in the hepatic allograft is relatively common, occurring in approximately one-third of adult patients (8), and is more common in children. Outcome in most cases appears to be favorable, with few patients requiring retransplantation (8–10). Hepatocellular carcinoma, a relatively common complication of chronic viral hepatitis, is rarely seen in AIH, occurring only in the setting of cirrhosis (11).
References TA B LE 3. 6 Serologic classification of autoimmune hepatitis Autoantibodies
Features
Type 1
ANA and/or SMA
Most common type
Type 2
LKM, ANA, and SMA negative
Young children in the United States; initially thought to have worse prognosis but responds well to treatment.
Type 3
SLA/LP
Indistinguishable clinically from Type 1; clinical course is controversial, possibly more likely to relapse after corticosteroids.
Abbreviations: AIH, autoimmune hepatitis; ANA, antinuclear antibody; LKM, liverkidney microsomal antibody; SMA, smooth muscle antibody; SLA/LP, soluble liver antigen/liver pancreas antigen.
1. Krawitt EL. Autoimmune hepatitis. N Engl J Med. 2006;354:54–66. 2. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31(5):929–938. 3. Johnson PJ, McFarlane IG. Meeting report: International Autoimmune Hepatitis Group. Hepatology. 1993;18:998–1005. 4. Hennes EM, Zeniya M, Czaja AJ, et al. Simplified criteria for the diagnosis of autoimmune hepatitis. Hepatology. 2008;48(1):169–176. 5. Feld JJ, Dinh H, Arenovich T, Marcus VA, Wanless IR, Heathcote EJ. Autoimmune hepatitis: effect of symptoms and cirrhosis on natural history and outcome. Hepatology. 2005;42(1):53–62. 6. Verma S, Gunuwan B, Mendler M, Govindrajan S, Redecker A. Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy. Am J Gastroenterol. 2004;99:1510–1516.
CHAPTER
3:
AU T O I M M U N E
H E PAT I T I S / O V E R L A P
7. Czaja AJ, Carpenter HA. Histological features associated with relapse after corticosteroid withdrawal in type 1 autoimmune hepatitis. Liver Int. 2003;23(2):116–123. 8. Campsen J, Zimmerman MA, Trotter JF, et al. Liver transplantation for autoimmune hepatitis and the success of aggressive corticosteroid withdrawal. Liver Transpl. 2008;14(9):1281–1286. 9. Gonzalez-Koch A, Czaja AJ, Carpenter HA, et al. Recurrent autoimmune hepatitis after orthotopic liver transplantation. Liver Transpl. 2001;7(4):302–310. 10. Vogel A, Heinrich E, Bahr MJ, et al. Long-term outcome of liver transplantation for autoimmune hepatitis. Clin Transplant. 2004;18(1):62–69. 11. Yeoman AD, Al-Chalabi T, Karani JB, et al. Evaluation of risk factors in the development of hepatocellular carcinoma in autoimmune hepatitis: implications for follow-up and screening. Hepatology. 2008;48(3): 863–870. 12. Batts KP, Ludwig J. Chronic hepatitis: an update on terminology and reporting. Am J Surg Pathol. 1995;19(12):1409–1417. 13. Scheuer PJ. Classification of chronic viral hepatitis: a need for reassessment. J Hepatol. 1991;13:372–374. 14. Czaja AJ, Carpenter HA. Sensitivity, specificity, and predictability of biopsy interpretations in chronic hepatitis. Gastroenterology. 1993;105:1824–1832. 15. Singh R, Nair S, Farr G, Mason A, Perrillo R. Acute autoimmune hepatitis presenting with centrizonal liver disease: case report and review of the literature. Am J Gastroenterol. 2002;97:2670–2673. 16. Pratt DS, Fawaz KA, Rabson A, Dellelis R, Kaplan MM. A novel histological lesion in glucocorticoid-responsive chronic hepatitis. Gastroenterology. 1997;113:664–668.
SYNDROMES
25
17. Czaja AJ, Carpenter HA. Autoimmune hepatitis with incidental features of bile duct injury. Hepatology. 2001;34:659–665. 18. Bach N, Thung SN, Schaffner F. The histological features of chronic hepatitis C and autoimmune chronic hepatitis: a comparative analysis. Hepatology. 1992;15:572–577. 19. Zen Y, Harada K, Sasaki M, et al. Are bile duct lesions of primary biliary cirrhosis distinguishable from those of autoimmune hepatitis? Interobserver histological agreement on trimmed bile ducts. J Gastroenterol. 2005;40:164–170. 20. Daniels JA, Torbenson M, Anders RA, Boitnott JK. Immunostaining of plasma cells in primary biliary cirrhosis. Am J Clin Pathol. 2009;131(2):243–249. 21. Umemura T, Zen Y, Hamano H, Kawa S, Nakanuma Y, Kiyosawa K. Immunoglobin G4-hepatopathy: association of immunoglobin G4bearing plasma cells in liver with autoimmune pancreatitis. Hepatology. 2007;46(2):463–471. 22. Czaja AJ, Cassani F, Cataleta M, Valentini P, Bianchi FB. Antinuclear antibodies and patterns of nuclear immunofluorescence in type 1 autoimmune hepatitis. Dig Dis Sci. 1997;42(8):1688–1696. 23. McFarlane IG. Autoimmune hepatitis: diagnostic criteria, subclassifications, and clinical features. Clin Liver Dis. 2002;6(3): 605–621. 24. Gregorio GV, Portmann B, Karani J, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology. 2001;33(3):544–553.
Case 3.1
Autoimmune Hepatitis With Bile Duct Injury Versus Primary Biliary Cirrhosis KAY WASHINGTON
C L I N I C AL I N F OR M AT I ON
A 35-year-old woman presented with fatigue. There was no history of medication use other than occasional nonsteroidal anti-inflammatory agents for headache, and the patient did not consume alcohol. Physical examination revealed mild icterus and no hepatomegaly or splenomegaly. There were no stigmata of chronic liver disease. Laboratory testing revealed elevated serum transaminases (10 times normal), mildly elevated alkaline phosphatase (twice normal), elevated total bilirubin (3 times normal), and serum IgG (2 times normal). Serologic studies for hepatitis A and B were negative; polymerase chain reaction testing for hepatitis C RNA was negative. R E A S ON F OR R E F E R R A L
The biopsy shows a hepatitic pattern of injury with lymphocytic infiltration of bile ducts but without bile duct loss. Do the findings represent autoimmune hepatitis (AIH) or PBC?
FIGURE
3.1.2
Prominent lobular hepatitis and sinusoidal
lymphocytosis.
PAT H OL OG I C F E AT U R E S
The liver biopsy shows a hepatitis pattern of injury, with portal and periportal lymphoplasmacytic infiltrates, moderate interface hepatitis (Figure 3.1.1), and spotty hepatocyte necrosis with acidophilic bodies in the lobule (Figure 3.1.2). Plasma cells are prominent in the portal infiltrate and are present in clusters both in the lobule (Figure 3.1.3) and in portal tracts. Although bile ducts are focally infiltrated by
F I G U R E 3 . 1 . 3 Cluster of lobular plasma cells.
lymphocytes and show reactive changes (Figure 3.1.1), there is minimal bile duct loss and no florid duct lesions are seen. No hepatocyte “rosettes” are noted, and there is minimal emperipolesis.
DIAGNO SIS
Autoimmune hepatitis with moderate activity, with minor bile duct injury.
FIGURE 3. 1. 1 Portal lymphoplasmacytic inflammatory infiltrate with interface hepatitis and infiltration of bile duct by lymphocytes.
26
CASE
3.1:
AU T O I M M U N E
H E PAT I T I S
27
D I S C U S S I ON
The pathologic features when considered in concert with the laboratory findings are strongly suggestive of AIH. Further serologic tests were positive for antinuclear and smooth muscle-directed antibodies at less than 1:80 and negative for antimitochondrial antibodies. Application of the modified scoring system for AIH (see Table 3.2 in Chapter 3) based upon the available information provided yields a pretreatment score in the “definite AIH” range. Based on the biopsy and serologic findings, PBC can be excluded, and in the absence of a significant drug history and of evidence of viral hepatitis, the diagnosis of AIH can be made. Grading and Staging
Although not specifically developed for AIH, grading systems developed for chronic hepatitis are often used to grade and stage the necroinflammatory activity and extent of fibrosis. Four-tiered systems such as those described by Batts and Ludwig (1) and Scheuer (2) are perhaps the most widely used in everyday practice. In the Batts and Ludwig system, lobular and interface hepatitis are graded as minimal, mild, moderate, or severe, with the more severe determining the overall grade if a discrepancy exists between portal and lobular necroinflammatory activity. Staging is based on fibrosis, with stage 1 characterized by portal expansion of portal tracts without periportal extension, stage 2 by periportal fibrosis, stage 3 by portal-portal and portal-central bridging fibrosis, and stage 4 by cirrhosis (1).
F I G U R E 3 . 1 . 4 Cirrhosis in autoimmune hepatitis often shows no distinguishing features; degree of activity ranges from minimal in “burned out” cases to severe.
Distinction of Chronic Hepatitis From Chronic Biliary Disease
Distinction of chronic hepatitis from chronic cholestatic disorders such as PBC and PSC is based upon clinical and laboratory findings and histologic features (see Table 3.4 in Chapter 3). AIH is characterized by a hepatitis pattern of injury, usually with prominent interface hepatitis. In the acute phase, lobular inflammation may predominate and fibrosis is minimal. Plasma cells are often prominent and are sometimes seen singly and in clusters in the lobule; however, roughly one-third of biopsies from patients with well-documented AIH will have few or no plasma cells (3). The severity of necroinflammatory activity is quite variable, ranging from mildly active hepatitis to bridging necrosis to massive hepatic necrosis. Hepatocyte regeneration may be prominent, with regenerating rosettelike structures. Ballooning degeneration, spotty hepatocyte necrosis, and apoptotic bodies are common but not specific. As the disease progresses, periportal fibrosis with bile ductular reaction, formation of portal-portal and portal-central bridges, and nodular regeneration result in cirrhosis (Figure 3.1.4) with variable necroinflammatory activity. In PBC, destruction of interlobular bile ducts by a nonsuppurative inflammatory process (Figure 3.1.5) eventually results in chronic cholestasis. In advanced stages of the disease, feathery degeneration of hepatocytes (cholate stasis), due to accumulation of bile salts within the cytoplasm,
F I G U R E 3 . 1 . 5 Bile duct infiltration by lymphocytes in early-stage primary biliary cirrhosis. Note lobular lymphoplasmacytic infiltrate with mild interface inflammation.
imparts a pale appearance to periportal or periseptal hepatocytes (Figure 3.1.6). Canalicular bile plugs are scarce to nonexistent. Although interface inflammation may be identified, it is not as prominent as in AIH, and lobular inflammation is minimal. Bile ductular reaction is common in intermediate stages of PBC but disappears in late-stage disease, leaving empty-appearing fibrous septa. Periportal and periseptal hepatocytes accumulate copper in chronic cholestasis but not in AIH, and this copper storage can be demonstrated with a variety of special stains. Orcein or aldehyde fuchsin stains, which are generally used to demonstrate accumulation of hepatitis B surface antigen, will also highlight increased copper-binding protein, which, like hepatitis B surface antigen, contains a large number of sulfhydryl groups. Positive
28
CHAPTER
3:
AU T O I M M U N E
H E PAT I T I S / O V E R L A P
SYNDROMES
FIGURE 3. 1. 6 Feathery degeneration in late stage primary biliary cirrhosis is a result of chronic cholestasis and imparts a pale appearance to periportal or periseptal hepatocytes.
F I G U R E 3 . 1 . 7 Zone 3 hepatocyte necrosis and lymphoplasmacytic
staining is seen as granular deposition in periportal hepatocytes. Any of the special stains for copper itself, such as rhodanine or rubeanic acid stain, may also be used. This copper accumulation is not specific, but when found in a precirrhotic liver biopsy, it is highly suggestive of chronic cholestasis, if rare entities such as Wilson disease and Indian childhood cirrhosis have been excluded. Mallory’s hyaline may also be found in periportal hepatocytes in chronic cholestasis and is morphologically indistinguishable from that found in alcoholic liver disease but is not a feature of AIH.
whereas the fact that in PBC an appreciable number of the plasma cells express IgM has prompted investigation as to whether this differential expression can be used to distinguish among the autoimmune liver diseases. However, the utility of this approach has not been validated (9). IgG4-bearing plasma cells have been described in liver disease associated with autoimmune pancreatitis, but are not present in significant numbers in typical AIH (10).
Morphologic Overlap Between PBC and AIH
The liver biopsy findings in AIH vary with the stage of the disease but always reflect a hepatitis pattern of injury. Although distinction of AIH with interface and lobular hepatitis from the classic form of PBC with florid duct lesions and ductopenia is not difficult, some cases of PBC contain appreciable numbers of plasma cells in the portal inflammatory infiltrate and exhibit interface inflammation. However, AIH lacks florid duct lesions and extensive bile duct injury and loss. Prominent lobular necroinflammatory activity is not a feature of PBC, and its presence favors AIH. Centrilobular injury with prominent hepatocellular necrosis and mononuclear inflammation in some cases of AIH, probably representing an early stage of the disease (4,5) (Figure 3.1.7), is also not seen in PBC. Bile duct destruction is generally not prominent in AIH, but up to 12% of AIH biopsies may show duct destruction, and an additional 12% show lymphocytic infiltration of bile duct epithelium without ductopenia (6,7). Taken in isolation the bile duct injury may be indistinguishable from early-stage PBC (8), but consideration of the overall histologic pattern of injury and correlation with serologic, laboratory, and clinical findings will help establish the diagnosis of AIH. It has been observed that plasma cells in AIH are predominantly IgG positive, with few IgM-expressing cells,
infiltration is prominent in some cases of autoimmune hepatitis and may represent an early phase of the disease.
CLINIC A L SIGNIFIC A NCE O F DIST INGUISH ING A IH FRO M P BC
Accurate distinction of AIH from PBC is critically important for selection of appropriate medical therapy. The majority of cases of AIH respond to immunosuppressive therapy, which can be lifesaving in severely active cases and can prevent progression to cirrhosis. In contrast, PBC is treated with the choleretic agent UDCA; in patients with a biochemical response, defined as a greater than 40% decrease in alkaline phosphatase levels within 1 year, survival is improved (11) and transplantation may be avoided. Immunosuppression has been shown in randomized clinical trials to be ineffective in PBC. In addition to being ineffective, steroids can exacerbate osteoporosis that commonly affects elderly women with PBC.
References 1. Batts KP, Ludwig J. Chronic hepatitis: an update on terminology and reporting. Am J Surg Pathol. 1995;19(12):1409–1417. 2. Scheuer PJ. Classification of chronic viral hepatitis: a need for reassessment. J Hepatology. 1991;13:372–374. 3. Czaja AJ, Carpenter HA. Sensitivity, specificity, and predictability of biopsy interpretations in chronic hepatitis. Gastroenterology. 1993;105:1824–1832. 4. Singh R, Nair S, Farr G, Mason A, Perrillo R. Acute autoimmune hepatitis presenting with centrizonal liver disease: case report and review of the literature. Am J Gastroenterol. 2002;97:2670–2673.
CASE
3.1:
AU T O I M M U N E
5. Pratt DS, Frwaz KA, Rabson A, Dellelis R, Kaplan MM. A novel histological lesion in glucocorticoid-responsive chronic hepatitis. Gastroenterology. 1997;113:664–668. 6. Czaja AJ, Carpenter HA. Autoimmune hepatitis with incidental features of bile duct injury. Hepatology. 2001;34:659–665. 7. Bach N, Thung SN, Schaffner F. The histological features of chronic hepatitis C and autoimmune chronic hepatitis: a comparative analysis. Hepatology. 1992;15:572–577. 8. Zen Y, Harada K, Sasaki M, et al. Are bile duct lesions of primary biliary cirrhosis distinguishable from those of autoimmune hepatitis? Interobserver histological agreement on trimmed bile ducts. J Gastroenterol. 2005;40:164–170.
H E PAT I T I S
29
9. Daniels JA, Torbenson M, Anders RA, Boitnott JK. Immunostaining of plasma cells in primary biliary cirrhosis. Am J Clin Pathol. 2009;131(2):243–249. 10. Umemura T, Zen Y, Hamano H, Kawa S, Nakanuma Y, Kiyosawa K. Immunoglobin G4-hepatopathy: association of immunoglobin G4bearing plasma cells in liver with autoimmune pancreatitis, Hepatology. 2007;46(2):463–471. 11. Kuiper EM, Hansen BE, De Vries RA, et al. Improved prognosis of patients with primary biliary cirrhosis that have a biochemical response to ursodeoxycholic acid. Gastroenterology. 2009;136(4):1281–1287.
Case 3.2
Autoimmune Hepatitis–Primary Biliary Cirrhosis Overlap Syndrome KAY WASHINGTON
with numerous collections of mononuclear cells within the lobule and scattered acidophilic bodies (Figure 3.2.3), without bridging necrosis. No hepatocyte rosettes are identified.
C L I N I C AL I N F OR M AT I ON
A 68-year-old woman presented with fatigue. She had elevated liver tests (ALT and AST 4 times normal, alkaline phosphatase 3 times normal, total bilirubin twice normal) and high titer autoantibodies, with ANA, SMA, and AMA positive at 1:1280. Liver biopsy was performed to clarify the nature of the liver disease and to assess for fibrosis.
DIAGNO SIS
Autoimmune hepatitis/primary biliary cirrhosis overlap syndrome .
R E A S ON F OR R E F E R R A L
The liver biopsy shows a combination of bile duct loss and prominent interface and lobular hepatitis with plasma cells. Do the findings represent AIH, PBC, or AIH/PBC overlap syndrome? PAT H OL OG I C F E AT U R E S
The biopsy shows a hepatitis pattern of injury with significant bile duct loss. Most portal tracts contain a moderately dense lymphoplasmacytic infiltrate, with circumferential interface hepatitis and minimal bile ductular reaction (Figure 3.2.1). Focal emperipolesis is seen. The majority of portal tracts in the biopsy lack identifiable bile ducts, although granulomatous florid duct lesions are not seen (Figure 3.2.2). Periportal hepatocytes show mild swelling and cytoplasmic rarefaction, without Mallory’s hyaline accumulation. Plasma cells are present in clusters in the portal inflammatory infiltrate and at the limiting plate. There is moderate to severe lobular hepatitis
F I G U R E 3 . 2 . 2 Portal tract with bile duct loss, without significant
ductular reaction.
FIGURE 3. 2. 1 Portal mononuclear cell infiltrate with plasma cells, interface inflammation, and with no identifiable interlobular bile duct.
F I G U R E 3 . 2 . 3 Prominent lobular necroinflammatory activity similar to autoimmune hepatitis.
30
CASE
3.2:
AU T O I M M U N E
H E PAT I T I S – P R I M A RY
D I S C U S S I ON
The diagnosis of AIH/PBC overlap syndrome in this case is based upon an aggregate score of “probable AIH” using the modified scoring system, the presence of high-titer AMA, and the combination of bile duct loss with a hepatitic pattern of injury. The existence of such overlap cases has been controversial, but there is agreement that the term should not be overused, and cases must demonstrate histologic, clinical, and serologic overlap to be classified as such. Up to 10% to 14% of AIH or PBC patients in some series may belong in the AIH/PBC overlap category (1,2), although criteria have varied among studies. Aggregate scores of definite or probable AIH using the modified scoring system of the International Autoimmune Hepatitis Group (3) and cholestatic biochemical studies combined with seropositivity for AMA are helpful. Alternatively, 2 out of 3 criteria for diagnosis of AIH and PBC have been required (Table 3.2.1) (1,4) for diagnosis. Liver biopsies in AIH/PBC overlap cases show features of both PBC (granulomatous inflammation and bile duct lesions) and autoimmune hepatitis (lobular hepatitis with spotty hepatocyte necrosis and acidophil bodies and moderate to severe interface hepatitis) (5). Granulomatous bile duct destruction and well-developed florid duct lesions are not seen in typical AIH. Because interface hepatitis and portal inflammatory infiltrates are common in typical PBC cases, these features alone cannot be used as discriminating features for AIH/PBC overlap. Lobular hepatitis with acidophilic bodies has been reported to be a more useful feature for discriminating between AIH and a spurious AIH/PBC overlap syndrome and may be more reliable than identification of mild forms of bile duct injury (6). Because some patients with clinical and histologic features of autoimmune hepatitis will have serum antimitochondrial antibodies, the presence of AMA in a patient with otherwise typical AIH is not considered sufficient for diagnosis of overlap syndrome. In some cases this apparent AMA positivity is caused by misreading of IF-type tests (confusing anti-LKM antibodies with AMA). In other patients, however, the AMA is truly positive but usually in low titer. Conversely, patients with PBC may have a positive ANA and are not considered to have AIH/PBC overlap unless there is clear evidence of a hepatitic component of liver injury. TABLE 3.2.1 Diagnostic criteria for AIH/PBC overlap syndrome AIH features (2 out of 3 criteria required) • ALT levels 5 times upper limit of normal • Serum IgG levels 2 times upper limit of normal or positive SMA • Liver biopsy with moderate to severe interface hepatitis or lobular acidophilic bodies PBC features (2 out of 3 criteria required) • Alkaline phosphatase levels 2 or GGT levels 5 times upper limit of normal • Positive AMA • Liver biopsy showing florid duct lesions Abbreviations: AIH, autoimmune hepatitis; ALT, alanine aminotransferase; AMA, antimitochondrial antibodies; GGT, gamma glutamyl transpeptidase; IgG, immunoglobulin G; PBC, primary biliary cirrhosis. From Refs. 1, 4.
BILIARY
CIRRHOSIS
31
The combination of seropositivity for AMA and for antibodies directed against double-stranded DNA has been described in almost 50% of AIH/PBC overlap syndrome patients compared with 2% of AIH or PBC patients and may represent a useful serologic profile for diagnosis of a subset of overlap syndrome cases (7), although validation studies are needed. In addition to the more commonly encountered cases with “overlapping” simultaneous features of both diseases, rare patients who switch from one disease to another over time have been reported (8). Implications for Therapy and Outcome
Combination therapy with ursodeoxycholic acid (UDCA) and immunosuppression has been advocated, largely on an empiric basis, for treatment of AIH/PBC overlap syndrome, although in some retrospective cohorts in which AIH/PBC overlap patients were identified from treatment trials of PBC patients, response to UDCA without immunosuppression was similar for both groups (4). However, some small nonrandomized studies do offer support for combined UDCA and immunosuppression, and some clinicians advocate dual therapy especially in cases that meet classic criteria for diagnosis of AIH (9) or when treatment with UDCA alone does not result in rapid biochemical response. Higher rates of liver transplantation and complications of portal hypertension in AIH/PBC overlap compared with PBC have been reported, suggesting that the higher risk of adverse outcome associated with overlap with AIH may justify the addition of immunosuppression to the therapeutic regimen in selected patients (10).
References 1. Chazouilleres O, Wendum D, Serfaty L, Montembault S, Rosmorduc O, Poupon R. Primary biliary cirrhosis-autoimmune hepatitis overlap syndrome: clinical features and response to therapy. Hepatology. 1998;28:296–301. 2. Heurgue A, Vitry F, Diebold MD, et al. Overlap syndrome of primary biliary cirrhosis and autoimmune hepatitis: a retrospective study of 115 cases of autoimmune liver disease. Gastroenterol Clin Biol. 2007;31(1):17–25. 3. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31(5):929–938. 4. Joshi S, Cauch-Dudek K, Wanless IR, et al. Primary biliary cirrhosis with additional features of autoimmune hepatitis: response to therapy with ursodeoxycholic acid. Hepatology. 2002;35(2):409–413. 5. Terracciano LM, Patzina RA, Lehmann FS, et al. A spectrum of histopathologic findings in autoimmune liver disease. Am J Clin Pathol. 2000;114(5):705–711. 6. Suzuki S, Arase Y, Ikeda K, et al. Clinical and pathologic characteristics of the autoimmune hepatitis and primary biliary cirrhosis overlap syndrome. J Gastroenterol Hepatol. 2004;19:699–706. 7. Muratori P, Granito A, Pappas G, et al. The serological profile of the autoimmune hepatitis/primary biliary cirrhosis overlap syndrome. Am J Gastroenterol. 2009;104(6):1420–1425. 8. Angulo P, El-Amin O, Carpenter HA, Lindor KD. Development of autoimmune hepatitis in the setting of long-standing primary biliary cirrhosis. Am J Gastroenterol. 2001;3021–3027. 9. Poupon R. Autoimmune overlapping syndromes. Clin Liver Dis. 2003;7:865–878. 10. Silveira MG, Talwalkar JA, Angulo P, Lindor KD. Overlap of autoimmune hepatitis and primary biliary cirrhosis: long-term outcomes. Am J Gastroenterol. 2007;102(6):1244–1250.
Case 3.3
Autoimmune Hepatitis–Primary Sclerosing Cholangitis Overlap Syndrome KAY WASHINGTON
tract contains an interlobular bile duct, but many ducts are distorted and show reactive epithelial change; concentric periductal fibrosis is not noted. There is a moderate degree of lobular inflammation (Figure 3.3.3). Canalicular bile plugs are not present.
C L I N I C AL I N F OR M AT I ON
A 16-year-old girl presented with chronic diarrhea and elevated liver tests (serum ALT and AST 4 times normal, alkaline phosphatase 2 times normal), with normal total bilirubin. The ceruloplasmin level was slightly higher than normal. ANA and anti-SMA were positive, both at 1:80. Endoscopic retrograde cholangiopancreatography (ERCP) showed minimal narrowing of large bile ducts just proximal to the common bile duct, with normal intrahepatic ducts. R E A S ON F OR R E F E R R A L
The liver biopsy shows features of chronic hepatitis and subtle evidence of bile duct loss and injury. Do the findings represent AIH or PSC? PAT H OL OG I C F E AT U R E S
The biopsy shows a hepatitic pattern of injury with a moderate portal chronic inflammatory infiltrate consisting primarily of lymphocytes and plasma cells, and a moderate degree of interface hepatitis (Figure 3.3.1). Portal tracts are enlarged by periportal fibrosis and bile ductular reaction, with focal bridging fibrosis (Figure 3.3.2). Each portal
F I G U R E 3 . 3 . 2 Prominent portal plasma cells with degenerative
changes in interlobular bile duct. Note lack of concentric periductal fibrosis.
FIGURE 3. 3. 1 Lymphoplasmacytic portal inflammatory infiltrate
and mild interface hepatitis. Note degenerative changes in interlobular bile duct, without bile duct loss.
F I G U R E 3 . 3 . 3 Lobular hepatitis and sinusoidal lymphocytosis.
32
CASE
3.3:
AU T O I M M U N E
H E PAT I T I S – P R I M A RY
D I AG N OS I S
Autoimmune hepatitis/primary sclerosing cholangitis overlap syndrome.
D I S C U S S I ON
AIH/PSC overlap has been reported in up to 50% of pediatric patients initially identified as PSC or AIH (1,2). In adults, this overlap syndrome is rarer and is reported in roughly 6% of patients initially diagnosed with AIH (3) and 8% of PSC patients (4). Many pediatric cases are initially diagnosed as AIH; in one 16-year prospective study in which 55 children with AIH were followed, 27 (49%) developed cholangiographic findings typical of PSC (1). Such overlap cases tend to have a worse outcome, with reduction in survival (5) compared with classic AIH and are more likely to require liver transplantation (1). The term “autoimmune sclerosing cholangitis” has been proposed by Gregorio and colleagues for AIH/PSC overlap syndrome (1) in children. Criteria for Diagnosis
AIH/PSC overlap is defined as definite or probable AIH (see Table 3.3 in Chapter 3) with cholangiographic evidence of PSC (6). Use of magnetic resonance cholangiopancreatography (MRCP) imaging techniques is becoming more common in the diagnostic evaluation of autoimmune liver disease and may facilitate early identification of AIH/PSC overlap cases. The liver biopsy in AIH/PSC overlap syndrome (Figures 3.3.1–3.3.3) shows portal inflammation, interface hepatitis, and lobular necroinflammatory activity typical of AIH combined with varying degrees of bile duct injury. As in typical PSC, the liver biopsy may show subtle bile duct injury or loss, and the presence of biliary lesions must be established by cholangiographic studies. The classic lesion for PSC, well-established periductal fibrosis, is rarely seen in biopsies from children with PSC or with AIH/PSC overlap. The finding of ductopenia in a liver biopsy from a young patient suspected of having AIH should prompt consideration of AIH/PSC overlap syndrome. Simultaneous or Sequential Presentation
Sequential syndromes of AIH evolving into PSC are rare in adults but have also been reported (7); such patients may be identified by biliary imaging techniques when their disease
SCLEROSING
CHOLANGITIS
33
becomes refractory to immunosuppression. Early studies interpreted as evolution of AIH to PSC must be interpreted in light of new information regarding prevalence of bile duct changes at diagnosis of AIH (8). In studies performed without ERCP or MRCP at diagnosis, concurrence of AIH and PSC may not have been adequately excluded. Implications for Therapy and Outcome
In most series, these AIH/PSC overlap patients are treated empirically with a combination of immunosuppression and ursodeoxycholic acid (UDCA) (9,10). Although UDCA may lead to improvement in liver tests in PSC, it has not been shown to improve overall survival. As in classic PSC, there is an increased risk of cholangiocarcinoma.
References 1. Gregorio GV, Portmann B, Karani J, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology. 2001;33(3):544–553. 2. el-Shabrawi M, Wilkinson ML, Portmann B, et al. Primary sclerosing cholangitis in childhood. Gastroenterology. 1987;92(5 pt 1):1226–1235. 3. Czaja AJ. The variant forms of autoimmune hepatitis. Ann Intern Med. 1996;125:588–598. 4. van Buuren HR, van Hoogstraten HJE, Terkivatan T, Schalm SW, Vleggaar FP. High prevalence of autoimmune hepatitis among patients with primary sclerosing cholangitis. J Hepatol. 2000; 33:543–548. 5. Al-Chalabi T, Portmann BC, Bernal W, McFarlane IG, Heneghan MA. Autoimmune hepatitis overlap syndromes: an evaluation of treatment response, long-term outcome and survival. Aliment Pharmacol Ther. 2008;28(2):209–220. 6. Kaya M, Angulo P, Lindor KD. Overlap of autoimmune hepatitis and primary sclerosing cholangitis: an evaluation of a modified scoring system. J Hepatol. 2000;33(4):537–542. 7. Abdo AA, Bain VG, Kichian K, Lee SS. Evolution of autoimmune hepatitis to primary sclerosing cholangitis: a sequential syndrome. Hepatol. 2002;36(6):1393–1399. 8. Abdalian R, Dhar P, Jhaveri K, Haider M, Guindi M, Heathcote EJ. Prevalence of sclerosing cholangitis in adults with autoimmune hepatitis: evaluating the role of routine magnetic resonance imaging. Hepatology. 2008;47(3):949–957. 9. Floreani A, Rizzotto ER, Ferrara F, et al. Clinical course and outcome of autoimmune hepatitis/primary sclerosing cholangitis overlap syndrome. Am J Gastroenterol. 2005;100:1516–1522. 10. Miloh T, Arnon R, Shneider B, Suchy F, Kerkar N. A retrospective single-center review of primary sclerosing cholangitis in children. Clin Gastroenterol Hepatol. 2009;7(2):239–245.
Case 3.4
Chronic Hepatitis C With Autoantibodies Versus Autoimmune Hepatitis KAY WASHINGTON
C L I N I C AL I N F OR M AT I ON
A 30-year-old woman with hypothyroidism presented with fatigue and had elevated transaminase levels (AST and ALT 3 times normal, alkaline phosphatase normal, total bilirubin normal). Hepatitis virus serologies were negative for hepatitis A and B; hepatitis C antibody test was positive. Serum ANA was positive at 1:640 and SMA at 1:320. R E A S ON F OR R E F E R R A L
The biopsy shows chronic hepatitis with prominent plasma cells. Do the findings represent AIH or chronic hepatitis C? PAT H OL OG I C F E AT U R E S
The liver biopsy shows a portal lymphoplasmacytic infiltrate (Figure 3.4.1) without nodular lymphoid aggregates and germinal centers. Circumferential interface hepatitis is present, with minimal bile ductular reaction. Interlobular bile ducts are unremarkable, and there is no cholestasis. Numerous aggregates of small lymphocytes associated with hepatocyte loss and scattered acidophilic bodies are noted in the lobule (Figure 3.4.2). Periportal fibrosis is present, without bridging. No steatosis is seen. Further investigation reveals that the patient is negative for hepatitis C RNA by PCR-based testing methods and is positive for p-ANCA.
F I G U R E 3 . 4 . 2 Spotty hepatocyte necrosis and sinusoidal lymphocytosis.
DIAGNO SIS
Chronic hepatitis with moderate activity and periportal fibrosis, favor autoimmune hepatitis.
DISCUSSIO N
This patient was considered to have autoimmune hepatitis with false-positive HCV antibodies based on negative PCR for HCV-RNA and clinical features, such as concurrent thyroid disease and positive p-ANCA, favoring an autoimmune etiology. Such cases can present diagnostic and management challenges, as antiviral therapy with interferon may exacerbate autoimmune conditions, whereas immunosuppression, the mainstay of therapy for AIH, can enhance viral replication. Of the patients with chronic HBV or HCV 20% to 40% are persistently positive for various autoantibodies, usually at low titers (~1:20 or 1:40) (1). The use of the AIH Scoring System can be helpful in excluding AIH in patients known to have HCV (2). Conversely, patients with AIH can have a false-positive test for anti-HCV antibodies but will have undetectable HCV-RNA. Three categories of patients with potential concurrent AIH/hepatitis C may be identified: • Patients with true AIH and false-positive anti-HCV antibodies (undetectable HCV-RNA)
FIGURE 3. 4. 1 Prominent plasma cells in portal tracts, with interface
hepatitis.
34
CASE
3.4:
CHRONIC
H E PAT I T I S
C
35
• Patients with true HCV and autoantibodies at low titers but no other signs of AIH • Patients with true HCV and features of AIH including young age, female gender, high autoantibody titers (1:320), hypergammaglobulinemia, and history of extrahepatic autoimmune disorders There has been considerable controversy about the diagnosis and management of chronic hepatitis patients with autoimmune markers, particularly regarding the presence of anti-LKM antibodies in HCV patients in southern Europe. It has recently been shown that anti-LKM1 antibodies directed against cytochrome P450 2D6 (CYP2D6) can cross-react with HCV proteins, perhaps as a result of molecular mimicry at the B cell level (3). The current general consensus is that interferon therapy is usually safe in HCV patients with anti-LKM1 F I G U R E 3 . 4 . 4 Minimal lobular inflammation in chronic hepatitis C.
autoantibodies. Patients with chronic viral hepatitis should be screened for autoantibodies before starting interferon therapy and monitored carefully. p-ANCA testing may be useful, as this autoantibody is rare in chronic hepatitis C, but relatively common in AIH. Liver biopsy in hepatitis C often shows less interface hepatitis (Figure 3.4.3) and lobular activity (Figure 3.4.4) than in AIH, although the histologic findings can be indistinguishable.
References
FIGURE 3. 4. 3 Portal lymphocytic inflammatory infiltrate in chronic
hepatitis C, with reactive changes in bile ducts. Note minimal interface hepatitis.
1. McFarlane IG. Autoimmune hepatitis: diagnostic criteria, subclassifications, and clinical features. Clin Liver Dis. 2002;6(3):605–621. 2. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31(5):929–938. 3. Marceau G, Lapierre P, Beland K, Soudeyns H, Alvarez F. LKM1 autoantibodies in chronic hepatitis C infection: a case of molecular mimicry? Hepatology. 2005;42:675–682.
Case 3.5
Syncytial Giant Cell Hepatitis KAY WASHINGTON
C L I N I C AL I N F OR M AT I ON
A 20-year-old man developed progressive fatigue over a 2-month span and sought medical attention when he noticed scleral icterus. There was no history of alcohol or drug use, and he was taking no medications. Serum ALT and AST were 10 times normal. Serum total bilirubin was 3 times normal. Serologic studies for hepatitis A and B, as well as molecular studies for hepatitis C were negative. ANA test was positive at 1:160; anti-SMA was negative. R E A S ON F OR R E F E R R A L
The liver biopsy shows active hepatitis with numerous syncytial giant cells. What is the etiology of the giant cell hepatitis in this adult patient? PAT H OL OG I C F E AT U R E S
F I G U R E 3 . 5 . 2 Multinucleated giant cells in zone 3.
The biopsy shows active hepatitis, with numerous apoptotic hepatocytes and spotty hepatocyte necrosis. The portal tracts contain a mononuclear inflammatory infiltrate with scattered plasma cells, and circumferential interface hepatitis is present (Figure 3.5.1). Areas of bridging necrosis are present, and some portal tracts show early fibrosis and bile ductular reaction. The most striking feature is the presence of numerous multinucleated hepatocytes, more prominent in this case in zone 3 of the lobule. The hepatocyte giant cells have a fused or syncytial appearance, and contain from 5 to more than 20 nuclei (Figure 3.5.2). Mild hepatocellular cholestasis is noted.
DIAGNO SIS
Syncytial giant cell hepatitis, probably autoimmune in etiology, with bridging necrosis.
DISCUSSIO N
Giant cell transformation of hepatocytes is a nonspecific tissue reaction in neonates that is rarely seen outside of infancy. In children it is particularly striking in cholestatic disorders and is prominent in neonatal hepatitis, but syncytial hepatocyte giant cells are also seen to a lesser extent in extrahepatic biliary atresia, paucity of intrahepatic bile ducts, alpha-1-antitrypsin deficiency, and many other disorders. In older patients, the presence of numerous hepatocyte giant cells is termed adult or postinfantile giant cell hepatitis or syncytial giant cell hepatitis. This pattern of injury is seen in autoimmune liver disease, drug reactions, and is occasionally reported in human immunodeficiency virus (HIV), hepatitis C (1), and hepatitis B infections. Some cases of adult giant cell hepatitis do not have a clearly established etiology, however, and are regarded as idiopathic. Adult giant cell hepatitis has the same range of histologic features seen in viral or autoimmune hepatitis (AIH), with the additional finding of multinucleated hepatocyte giant cells. The number of multinucleated giant cells, defined as hepatocytes with more than 4 nuclei per cell, is variable, but these cells should be more than a rare isolated occurrence.
FIGURE 3. 5. 1 Interface hepatitis with ballooning degeneration of
hepatocytes.
36
CASE
3.5:
SYNCYTIAL
GIANT
The giant cells may be located anywhere within the lobule and sometimes contain Mallory’s hyaline. A 1991 report of paramyxovirus-like particles in biopsies from 8 of 10 patients with adult syncytial giant cell hepatitis stimulated intense interest in an infectious case among idiopathic cases. However, most investigators have been unable to confirm this finding, and the nature of the ultrastructural findings has been questioned. In one series, 40% of patients (2) had evidence of autoimmune disease such as positive ANA or SMA. About 25% of reported patients with postinfantile giant cell hepatitis remain stable or have gradual resolution of the disease. Corticosteroid treatment results in improvement in some but not all patients, and in the setting of autoimmune hepatitis the clinical course may be severe, with most patients progressing to cirrhosis (3). One-third of patients die of liver
CELL
H E PAT I T I S
37
disease, and the remainder progress to cirrhosis or liver failure. Disease recurrence after liver transplantation has been reported and may require retransplantation (4).
References 1. Moreno A, Moreno A, Perez-Elias MJ, et al. Syncytial giant cell hepatitis in human immunodeficiency virus-infected patients with chronic hepatitis C: 2 cases and review of the literature. Hum Pathol. 2006;37(10):1344–1349. 2. Devaney K, Goodman ZD, Ishak KG. Post-infantile giant-cell transformation in hepatitis. Hepatol. 1992;16:327–333. 3. Estradas J, Pascual-Ramos V, Martinez B, Uribe M, Torre A. Autoimmune hepatitis with giant-cell transformation. Ann Hepatol. 2009;8(1):68–70. 4. Pappo O, Yunis E, Jordan JA, et al. Recurrent and de novo giant cell hepatitis after orthotopic liver transplantation. Am J Surg Pathol. 1994;18(8):804–813.
This page intentionally left blank
4 Fatty Liver Disease MATTHEW M. YEH AND ELIZABETH M. BRUNT
I N T ROD U C T I ON
Nonalcoholic steatohepatitis (NASH) is nearly universally associated with metabolic syndrome, which includes central obesity, hypertension, type 2 diabetes or glucose intolerance (insulin resistance), dyslipidemia (hypertriglyceridemia, low high-density triglycerides, and elevated low-density triglycerides), and systemic hypertension. Metabolic syndrome has become an emerging epidemic in the 21st century, not only in the Western countries, but also across the world. NASH may progress to advanced fibrosis, that is, cirrhosis, and may develop hepatocellular carcinoma; therefore, pathologists often encounter liver biopsies in their daily practice to determine the diagnosis of NASH, its differential diagnosis, and activity (grade) and fibrosis (stage) of the disease (1,2). In adults the injury pattern of NASH is zone 3 based and includes the constellation of macrovesicular steatosis, lobular inflammation, liver cell injury, particularly ballooned hepatocytes (Figure 4.1), with or without centrizonal “chicken-wire” fibrosis (Figure 4.2), as cardinal histologic features (3–6). The injury pattern in children may differ from adults. This is discussed in more detail in Chapter 4.10. Clinical evaluation, serologic and laboratory tests, and current imaging modalities, can strongly suggest the presence of hepatic steatosis, but none can distinguish steatohepatitis from uncomplicated steatosis; likewise, these evaluations can generally detect advanced liver disease (ie, portal hypertension), but none can truly assess the degree of liver necroinflammatory injury, lesser stages of fibrosis,
F I G U R E 4 . 2 Masson’s Trichrome stain demonstrates chicken-
wire fibrosis with collagen fibers within the sinusoidal spaces and surrounding the hepatocytes.
and architectural remodeling. Liver biopsy evaluation, therefore, remains the “gold standard” to unequivocally diagnose steatohepatitis and to document the severity of hepatic injury and fibrosis (3,4,7). Of importance, biopsy assessment has also shown that not all individuals with unexplained liver test elevations have fatty liver (8), not all phenotypic “NASH” patients have steatosis (9), and that the full spectrum of fatty liver disease may be present when laboratory values are normal (10). In addition, examination of liver biopsy is useful for detecting possible concomitant or alternative processes (11–13). In fact, studies have reported the observation of histologic features of NASH in biopsies from patients with other serologically and/or clinically defined chronic liver diseases (14,15). Finally, it needs to be emphasized that pathologists can make the diagnosis of steatohepatitis but cannot make the diagnosis of nonalcoholic steatohepatitis as that diagnosis requires the clinical exclusion of alcohol use.
References 1. Bugianesi E, Leone N, Vanni E, et al. Expanding the natural history of nonalcoholic steatohepatitis: from cryptogenic cirrhosis to hepatocellular carcinoma. Gastroenterology. 2002;134–140. 2. Ratziu V, Bonyhay L, Di Martino V, et al. Survival, liver failure, and hepatocellular carcinoma in obesity-related cryptogenic cirrhosis. Hepatology. 2002;35:1485–1493. 3. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999;94:2467–2474.
FIGURE 4. 1 Features of NASH including macrovesicular steatosis,
lobular inflammation, and ballooned hepatocytes.
39
40
CHAPTER
4:
FAT T Y
4. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41:1313–1321. 5. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc. 1980;55:434–438. 6. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology. 1999;116:1413–1419. 7. Farrell GC, Larter CZ. Nonalcoholic fatty liver disease: from steatosis to cirrhosis. Hepatology. 2006;43:S99–S112. 8. Skelly MM, James PD, Ryder SD. Findings on liver biopsy to investigate abnormal liver function tests in the absence of diagnostic serology. J Hepatol. 2001;35:195–199. 9. Merriman RB, Ferrell LD, Patti MG, et al. Correlation of paired liver biopsies in morbidly obese patients with suspected nonalcoholic fatty liver disease. Hepatology. 2006;44:874–880.
LIVER
DISEASE
10. Mofrad P, Contos MJ, Haque M, et al. Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values. Hepatology. 2003;37:1286–1292. 11. Brunt EM. Grading and staging the histopathological lesions of chronic hepatitis: the Knodell histology activity index and beyond. Hepatology. 2000;31:241–246. 12. Brunt EM. Liver biopsy interpretation for the gastroenterologist. Curr Gastroenterol Rep. 2000;2:27–32. 13. Ishak KG. Chronic hepatitis: morphology and nomenclature. Mod Pathol. 1994;7:690–713. 14. Brunt EM, Ramrakhiani S, Cordes BG, et al. Concurrence of histologic features of steatohepatitis with other forms of chronic liver disease. Mod Pathol. 2003;16:49–56. 15. Sanyal AJ, Contos MJ, Sterling RK, et al. Nonalcoholic fatty liver disease in patients with hepatitis C is associated with features of the metabolic syndrome. Am J Gastroenterol. 2003;98:2064–2071.
Case 4.1
Steatosis With Inflammation Versus Steatohepatitis MATTHEW M. YEH AND ELIZABETH M. BRUNT
C L I N IC AL I N F OR M AT I ON
A 40-year-old man underwent gastric bypass because of morbid obesity. He had not had routine checkup for his liver tests. As the liver was mildly enlarged at surgery, a liver biopsy was performed by the surgeon during the gastric bypass surgery. R E A SON F OR R E F E R R AL
There is steatosis and inflammation in the liver biopsy. The referring pathologist’s specific question is whether a diagnosis of steatohepatitis can be established. PAT H OL OG I C F E AT U R E S
The liver biopsy shows a moderate amount of macrovesicular steatosis in zone 3 (Figure 4.1.1). There are also scattered inflammatory foci in the lobules, composed predominantly of lymphocytes and Kupffer cells (Figure 4.1.2). Ballooned hepatocytes are not identified, and a Masson’s Trichrome stain shows no significant fibrosis.
F I G U R E 4 . 1 . 2 The biopsy also shows inflammatory foci in the hepatic lobules, composed predominantly of lymphocytes, Kupffer cells, and pigmented macrophages. Ballooned hepatocytes are not identified.
D I AG N OS I S DISCUSSIO N
Moderate steatosis and lobular inflammation, no evidence of steatohepatitis, history of morbid obesity.
The diagnosis of steatohepatitis may not be as straightforward as the diagnosis of chronic viral hepatitis such as hepatitis B or C, as the latter depends on the combination of clinical, laboratory, and histologic findings. In particular, serology and virology play very critical roles in the clinical evaluation of these diseases; therefore, if virologic and serologic data are compatible, a diagnosis of chronic viral hepatitis is relatively easily reached, and biopsy is largely done for evaluation of activity and fibrosis (1). The diagnosis of steatohepatitis, on the other hand, is not analogous to that of viral hepatitis, as it has been generally conceptualized that steatosis is a necessary but not sufficient component to constitute a diagnosis of steatohepatitis, and inflammation and features of liver cell injury, in particular, a rather specific liver cell injury form, that is, ballooning degeneration (Figure 4.1.3), not typically seen in other liver diseases, are also required features for a confident diagnosis of steatohepatitis to be established. In fact, it has become clearer that steatosis, ballooning, and lobular inflammation are considered the common constellation of minimal criteria for the diagnosis of NASH in adults (2,3). Although it is well known that significant liver disease may exist with liver enzymes in the normal range among nonalcoholic fatty liver disease (NAFLD) patients, it is generally recognized that patients with NASH typically present with mild elevation of serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) (4).
FIGURE 4.1.1 Moderate macrovesicular steatosis in zone 3
distribution pattern.
41
42
CHAPTER
4:
FAT T Y
LIVER
DISEASE
References 1. Brunt EM. Grading and staging the histopathological lesions of chronic hepatitis: the Knodell histology activity index and beyond. Hepatology. 2000;31:241–246. 2. Brunt EM. Nonalcoholic steatohepatitis. Semin Liver Dis. 2004;24:3–20. 3. Yeh MM, Brunt EM. Pathology of fatty liver: differential diagnosis of nonalcoholic fatty liver disease. Diagn Path. 2008;14:586–597. 4. Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis: summary of an AASLD single topic conference. Hepatology. 2003;37:1202–1219.
FIGURE 4. 1. 3 Several ballooned and swollen hepatocytes with rarefied and clumped cytoplasm.
Case 4.2
Steatohepatitis With Minimal Ballooning MATTHEW M. YEH AND ELIZABETH M. BRUNT
C L I N IC AL I N F OR M AT I ON
DISCUSSIO N
A 55-year-old woman was evaluated for persistent mildly elevated liver tests. She has a medical history of type 2 diabetes. All viral and autoimmune serologic tests were negative. A liver biopsy was performed for further evaluation of her abnormal liver tests.
Hepatocytic ballooning is characterized by cytoplasmic alterations, swelling, and enlargement of hepatocytes, which result in loss of normal hexagonal shape (Figure 4.2.2). The cytoplasm is rarefied or finely reticulated, suggesting accumulation of intracellular fluid; at the ultrastructural level, minute vacuoles of steatosis, and/or other acute or chronic toxic cell injuries have been reported (1,2). Microvesicular steatosis may resemble ballooning degeneration as the hepatocytes may also be enlarged; nonetheless, the nuclei remain centrally located and may appear to be indented by the small fat droplets in microvesicular steatosis (Figure 4.2.3). The presence of larger and hyperchromatic nuclei in the ballooned hepatocytes may also be a subtle difference in distinguishing those with microvesicular steatosis. Predominant microvesicular steatosis is a histologic manifestation of severe liver disease caused by massive mitochondrial dysfunction and should not be seen as the primary type of steatosis in individuals with NAFLD (1,2,3). The enlarged and swollen, glycogen-rich hepatocytes of glycogenosis may occur in glycogen storage diseases or in glycogenic hepatopathy (Figure 4.2.4), which may also mimic ballooned hepatocytes. Glycogenic hepatopathy is a well-characterized entity commonly associated with type 1 diabetics with poor glycemic control (4). Typically, biopsies from patients with glycogenic hepatopathy show diffuse involvement of all hepatocytes, lack zonal and significant steatosis, lobular or portal inflammation, lobular spotty necrosis, and the characteristic perisinusoidal fibrosis of NASH. Megamitochondria can be seen in affected hepatocytes. The clinical setting is also very different (see Chapter 17).
R E A SON F OR R E F E R R AL
There is a moderate amount of macrovesicular steatosis as well as foci of lobular inflammation in the liver biopsy. Only 1 to 2 hepatocytes with possible ballooning are identified. The question from the referring pathologist is whether the findings in the liver biopsy are sufficient for the diagnosis of steatohepatitis. PAT H OL OG I C F E AT U R E S
The liver biopsy shows liver parenchyma with moderate macrovesicular steatosis, in a zone 3 distribution. There are scattered foci of inflammatory infiltrates in the hepatic lobules, composed predominantly of lymphocytes and Kupffer cells. Rare ballooned hepatocytes are identified, located in zone 3 (Figure 4.2.1). Masson’s Trichrome stain shows focal perivenular/pericellular fibrosis.
D I AG N OS I S
Steatohepatitis, with type 2 diabetes.
FIGURE 4. 2. 1 The liver biopsy shows moderate macrovesicular
F I G U R E 4 . 2 . 2 Ballooned hepatocytes are characterized by swelling and enlargement of hepatocytes, loss of the normal hexagonal shape, and rarefied or finely reticulated cytoplasm.
steatosis, foci of mononuclear inflammatory infiltrates in the lobule, and adjacent two ballooned hepatocytes.
43
44
CHAPTER
4:
FAT T Y
FIGURE 4. 2. 3 The hepatocytes in microvesicular steatosis are enlarged, with centrally located nuclei.
LIVER
DISEASE
Hepatocellular cholestasis may result in swollen and enlarged hepatocytes with flocculent cytoplasm, so-called feathery degeneration in the perivenular region in processes that result in acute cholestasis; in many ways, these may also mimic ballooned hepatocytes. However, the clinical setting of biliary obstruction or drug/toxin injury, the absence of the typical perisinusoidal/pericellular fibrosis, and the presence of canalicular cholestasis differentiates this from the ballooning degeneration of NASH. Cholate stasis, the morphologic alteration in periportal hepatocytes in chronic cholestasis, is also manifested by cellular swelling and cytoplasmic material clumped around the nucleus and large lucent peripheral areas of the cell. Sometimes Mallory-Denk bodies (Mallory’s hyaline) may also be present (5). These changes are noted, however, in the periportal or periseptal regions. Although these findings may be discerned from ballooning degeneration in NASH by the clinical setting and zonal location in the noncirrhotic liver, it is difficult to differentiate these, in advanced fibrosis, as the remodeled parenchyma usually no longer has a distinctly zonal architecture. A copper stain may be able to highlight cholate stasis in this setting. A recent study using immunohistochemistry shows that loss of keratin 8/18 immunostaining can serve as an objective marker of ballooning degeneration (6). Although not widely used, it may be useful in equivocal cases.
References
FIGURE 4. 2. 4 Enlarged and swollen, glycogen-rich hepatocytes in glycogenic hepatopathy. Note that there is no significant steatosis or inflammation.
1. Burt AD, Mutton A, Day CP. Diagnosis and interpretation of steatosis and steatohepatitis. Semin Diagn Pathol. 1998;15:246–258. 2. Fujii H, Ikura Y, Arimoto J, et al. Expression of perilipin and adipophilin in nonalcoholic fatty liver disease; relevance to oxidative injury and hepatocyte ballooning. J Atheroscler Thromb. 2009;16:893–901. 3. Pessayre D, Berson A, Fromenty B, Mansouri A. Mitochondria in steatohepatitis. Semin Liver Dis. 2001;21:57–69. 4. Torbenson M, Chen YY, Brunt E, et al. Glycogenic hepatopathy: an underrecognized hepatic complication of diabetes mellitus. Am J Surg Pathol. 2006;30:508–513. 5. Li MK, Crawford JM. The pathology of cholestasis. Semin Liver Dis. 2004;24:21–42. 6. Lackner C, Gogg-Kamerer M, Zatloukal K, Stumptner C, Brunt EM, Denk H. Ballooned hepatocytes in steatohepatitis: the value of keratin immunohistochemistry for diagnosis. J Hepatol. 2008;48:821–828.
Case 4.3
Steatohepatitis Without Activity MATTHEW M. YEH AND ELIZABETH M. BRUNT
C L I N IC AL I N F OR M AT I ON DIAGNO SIS
A 65-year-old diabetic and overweight man had evidence of portal hypertension, but no known underlying etiology on clinical testing. The patient denied past or current alcohol use; medications included antidiabetic, antihyperlipidemic, and antihypertensive agents.
Cirrhosis, etiology not known, clinical history of overweight, and diabetes.
DISCUSSIO N R E A SON F OR R E F E R R AL
Cryptogenic cirrhosis is the clinical term applied to cases such as this. The findings represent “burned-out” liver disease. This process may occur with a variety of liver diseases in which not only the serologic findings, but also the histopathologic features of the active disease are no longer evident. The common diseases that may result in this process include alcoholic and NASH, autoimmune hepatitis (AIH), and, less commonly, drug-induced liver disease and viral hepatitides. The biliary and metabolic diseases (alpha-1-antitrypsin deficiency and hemochromatosis) would not be in the differential, as histopathologic features of these entities can be documented with careful evaluation. Wilson disease is diagnosed or excluded by tissue copper quantitation. There is a large volume of literature that supports a linkage of biopsy-proven NASH with subsequent “cryptogenic” cirrhosis (1–4). A semantic problem arises with the concept of “cryptogenic”: if the underlying cause of burned-out cirrhosis is known, then, by definition, the cirrhosis is no longer “cryptogenic.” If active features remain, the potential etiology of NAFLD/NASH may be suggested (Figures 4.3.2 and 4.3.3). This, however, is not a small concern; many articles refer to “cryptogenic cirrhosis” as an identified outcome of,
The liver biopsy shows cirrhosis; the referring physician wonders whether the cause can be determined. PAT H OL OG I C F E AT U R E S
Low power evaluation shows that the liver is cirrhotic, and mild to moderate septal inflammatory mixed mononuclear infiltrates admixed with ductular reaction are noted. Occasional lipogranulomas are noted in the septa. On higher power evaluation, no steatosis, ballooning, or Mallory-Denk bodies are seen (Figure 4.3.1). Features of acute cholestasis and chronic cholestasis (cholate stasis) are not appreciated. Lobular inflammation is absent or only focal. Trichrome stain highlights the cirrhotic remodeling; focal perisinusoidal fibrosis is present. Periodic acid–Schiff diastase (PASd) is negative for globules, and iron stain shows varying amounts of periseptal hepatocellular granular reactivity among the nodules.
FIGURE 4. 3. 1 There is obvious cirrhosis with almost no inflammation or hepatocellular alterations to suggest an underlying etiology. The differential diagnosis of “cryptogenic cirrhosis” is broad and includes, but is not limited to, alcoholic and nonalcoholic fatty liver diseases as well as “burned-out” autoimmune liver disease.
F I G U R E 4 . 3 . 2 Same patient as in Figure 4.3.1, but 12 years prior. There is active liver disease noted on this low power evaluation.
45
46
CHAPTER
4:
FAT T Y
LIVER
DISEASE
cirrhosis due to ALD. A distinguishing feature of cirrhosis due to ALD is micronodular cirrhosis; however, ALD may also result in mixed micro- and macronodular cirrhosis, and thus, if not present uniformly, micronodularity should not be relied upon for characterization. In summary, cirrhosis secondary to either alcoholic or NAFLD may result in features of inactive liver disease; focal perisinusoidal fibrosis, when present, may strongly support either of these as an underlying etiology. Accurate and astute clinical history may be the only avenue of arriving at the correct diagnosis. Cryptogenic cirrhosis is a term that should only be applied with truly unknown etiology.
References FIGURE 4. 3. 3 On higher power evaluation of the cirrhotic biopsy, there is evidence of hepatocellular ballooning and Mallory-Denk bodies. These features would suggest the possibility of “steatohepatitis” as the initiating cause of cirrhosis. However, they cannot distinguish alcoholic from nonalcoholic fatty liver.
or synonymously with, NASH, which completely ignores the possibility that cirrhosis of “unknown etiology” could be due to the various other processes discussed. In fact, some investigators have utilized the presence of features of NAFLD in an allograft liver as evidence for NASH as the underlying cause of the original disease in the native liver (2,5–8). For pathologists, it is important to keep the differential diagnostic possibilities in mind and to maintain the usage of the term “cryptogenic” for liver disease of truly unknown etiology. Cirrhosis due to burned-out steatohepatitis, whether of alcoholic or nonalcoholic etiology, may, or may not, have residual foci of perisinusoidal fibrosis by histologic evaluation. Steatosis is most commonly absent, as it is recognized that with progression, the active lesions of steatohepatitis may abate (9–12). Hepatocellular ballooning may be present rarely; affected hepatocytes may contain Mallory-Denk bodies. Lipogranulomas in the septa are described as a feature of prior alcoholic steatohepatitis (13), but they may also occur in NAFLD. Either process may result in cholestasis if there is decompensated liver disease and/or sepsis, but features of chronic cholestasis such as copper deposition, have only been described in alcoholic liver disease (ALD) (14). Iron deposition may also occur in cirrhosis and likely is a result of the diminution of hepcidin production due to decreased liver cell mass. Alcohol is also a direct suppressant of hepcidin production by hepatocytes and, thus, iron deposition is common in
1. Browning JD, Kumar KS, Saboorian MH, Thiele DL. Ethnic differences in the prevalence of cryptogenic cirrhosis. Am J Gastroenterol. 2004;99:292–298. 2. Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH, Driscoll CJ. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. Hepatology. 1999;29:664–669. 3. Poonawala A, Nair SP, Thuluvath PJ. Prevalence of obesity and diabetes in patients with cryptogenic cirrhosis: a case-control study. Hepatology. 2000;32:689–692. 4. Struben VM, Hespenheide EE, Caldwell SH. Nonalcoholic steatohepatitis and cryptogenic cirrhosis within kindreds. Am J Med. 2000;108:9–13. 5. Charlton MR, Kondo M, Roberts SK, Steers JL, Krom RA, Wiesner RH. Liver transplantation for cryptogenic cirrhosis. Liver Transpl Surg. 1997;3:359–364. 6. Contos MJ, Cales W, Sterling RK, et al. Development of nonalcoholic fatty liver disease after orthotopic liver transplantation for cryptogenic cirrhosis. Liver Transpl. 2001;7:363–373. 7. Maor-Kendler Y, Batts KP, Burgart LJ, et al. Comparative allograft histology after liver transplantation for cryptogenic cirrhosis, alcohol, hepatitis C, and cholestatic liver diseases. Transplantation. 2000;70:292–297. 8. Ong J, Younossi ZM, Reddy V, et al. Cryptogenic cirrhosis and posttransplantation nonalcoholic fatty liver disease. Liver Transpl. 2001; 7: 797–801. 9. Abdelmalek M, Ludwig J, Lindor KD. Two cases from the spectrum of nonalcoholic steatohepatitis. J Clin Gastroenterol. 1995;20:127–130. 10. Brunt EM. Nonalcoholic steatohepatitis: definition and pathology. Semin Liver Dis. 2001;21:3–16. 11. Brunt EM. Nonalcoholic steatohepatitis: pathologic features and differential diagnosis. Semin Diagn Pathol. 2005;22:330–338. 12. Brunt EM, Tiniakos DG. Alcoholic and nonalcoholic fatty liver disease. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract and Pancreas. 2nd ed. Philadelphia, PA: Elsevier; 2009:1007–1014. 13. Yip WW, Burt AD. Alcoholic liver disease. Semin Diagn Pathol. 2007;23:149–160. 14. Hall PDLM. Alcoholic liver disease. In: MacSween RNM, Burt AD, Portmann BC, Ishak KG, Scheuer PJ, Anthony PP, eds. Pathology of the Liver. 4th ed. London, UK: Churchill Livingstone; 2002:273–311.
Case 4.4
Nonalcoholic Steatohepatitis With Moderate/Marked Portal Inflammation MATTHEW M. YEH AND ELIZABETH M. BRUNT
C L I N IC AL I N F OR M AT I ON
A 45-year-old woman with phenotypic metabolic syndrome (increased waist to hip ratio, hypertension, elevated fasting glucose, and high high-density lipoprotein (HDL) levels) and elevated ALT had been biopsied for confirmation of clinical diagnosis of fatty liver and evaluation of severity of injury (grade and stage). The clinical workup was negative for viral hepatitis markers; ceruloplasmin and alpha-1-antitrypsin levels were normal. Antinuclear antibody (ANA) was positive but in low titer. Globulin levels were normal. Platelets were slightly depressed, but other hematologic values were normal. R E A SON F OR R E F E R R AL
The liver biopsy, otherwise suggestive of steatohepatitis, showed bridging fibrosis with prominent portal inflammation. The clinical concern was for AIH or AIH-overlap with steatohepatitis.
F I G U R E 4 . 4 . 2 Higher power evaluation of the same biopsy as in
Figure 4.4.1 shows ballooned hepatocytes, pigmented Kupffer cells, and a suggestion of perisinusoidal fibrosis. Mononuclear lobular inflammation is also present.
PAT H OL OG I C F E AT U R E S
The liver architecture is altered by bridging fibrosis and nodularity (Figure 4.4.1). There is evidence of hepatitis and hepatocellular injury including mixed small- and large droplet macrovesicular steatosis in an azonal distribution, hepatocyte ballooning, scattered inflammatory foci of mononuclear cells, single droplet lipogranulomas surrounded by Kupffer cells and an occasional eosinophil, and microgranulomas (Figure 4.4.2). Masson’s Trichrome stain highlights enlarged
F I G U R E 4 . 4 . 3 Trichrome stain highlights the dense portal chronic inflammation of a residual portal tract and zone 3 perisinusoidal fibrosis.
portal areas, septal fibrosis, and residual foci of perisinusoidal fibrosis (Figures 4.4.3 and 4.4.4). On higher magnification, expansion of septa by chronic inflammation with mixed mononuclear cells and focal “interface hepatitis” are noted (Figure 4.4.5). Plasma cells were not dominant in either portal or lobular infiltrates. Hepatitic rosettes were not present. No evidence of confluent or bridging necroses was seen.
FIGURE 4. 4. 1 On low power evaluation, there is evidence of fatty liver, bridging fibrosis, and nodularity. Portal/septal chronic inflammation is brisk, and interface activity can be seen.
47
48
CHAPTER
4:
FAT T Y
FIGURE 4. 4. 4 On trichrome stain the perisinusoidal fibrosis is evi-
dent and “dense.”
FIGURE 4. 4. 5 Interface hepatitis can be appreciated on this high
power view.
D I AG N OS I S
Nonalcoholic steatohepatitis with moderate activity and bridging fibrosis.
LIVER
DISEASE
with notable polymorphonuclear leukocytes is most likely indicative of alcoholic rather than nonalcoholic origin. The finding is common in association with ductular reaction and may represent cholangiolitis due to pancreatitis. In this setting, the portal areas take on a somewhat “holly-leaf” configuration; the proliferated ductular structures are intimately admixed with the spikes. The most common infiltrates in steatohepatitis, however, are “chronic” mixed mononuclear cells and occasionally, eosinophils. An important consideration in any liver biopsy with steatosis or steatohepatitis with or without notable portal chronic inflammation is Wilson disease (see Chapter 21). This disease has unknown prevalence, but studies have shown that Wilson disease is not restricted to the pediatric age group and may be diagnosed in individuals in middle age or older age groups (1). The key to the diagnosis is a high index of suspicion; this may first come from the pathologist. The primary clinical finding is often otherwise unexplained elevated liver tests. Histopathologic features can be as variable as the clinical features; none are absolutely diagnostic or required, and they may be present in varying combinations (1,2). Macrovesicular steatosis and portal chronic inflammation, swollen hepatocytes with or without Mallory-Denk bodies, “atypical lipofuscin” characterized by either non–zone 3 restriction or by large chunky granules, canalicular bile stasis, and megamitochondria are described. There may be portal expansion and portalbased fibrosis, or cirrhosis may be present. It is not uncommon for clinical evidence of either fulminant or submassive necrosis to show histologic features of cirrhotic remodeling with marked lobular activity. Copper staining itself is not the diagnostic standard for this disease. Although very high hepatic copper is present, histochemical stains for copper or copperbinding protein may not be able to detect it. Thus, the standard for diagnosis is copper quantitation from liver tissue; this can be obtained from the paraffin block (1). Explant livers from treated patients commonly show cirrhosis without activity and copper deposition extensively throughout the nodules. Portal accentuation (inflammation, fibrosis) is a common finding in pediatric NAFLD (3) (Figure 4.4.6). Portal chronic inflammation has also been noted as a feature of resolution of NASH in treatment trials (3,4) (Figure 4.4.7). Portal inflammation may also be a component of the injury in NASH (5) and has recently been shown to correlate with both increased fibrosis as well as more severe clinical features of NASH in adults and children (5,6). Finally, portal inflammation that is “disproportionate” to the lobular injury may be an indication of concurrent liver disease such as chronic viral hepatitis, AIH, and other forms of chronic liver disease (7,8). NASH and Other Concurrent Disease(s)
D I SC U SSI ON Significance of Portal Inflammation in NASH
Portal infiltrates can be characterized as either “chronic inflammation” dominant or polymorphonuclear leukocyte dominant. In the setting of steatohepatitis, portal expansion
The concept that more than one disease process could occur and could be detectable by liver biopsy evaluation was proposed and initially published by 2 groups (9–11) and subsequently confirmed by others. Two strong arguments allowed this concept to be accepted: (1) NAFLD was increasingly recognized as a widespread entity in the United States, and (2)
CASE
4.4:
NONALCOHOLIC
S T E AT O H E PAT I T I S
49
Diagnostic Criteria of Concurrent Disease
FIGUR E 4. 4. 6 Liver biopsy from a pediatric NAFLD patient high-
lights portal accentuation and portal chronic inflammation.
Investigators’ criteria for diagnosing both steatohepatitis and another liver disease differ. The authors’ criteria are more “strict” than for otherwise uncomplicated steatohepatitis so that this condition is not overdiagnosed (15). Thus, in addition to steatosis, ballooning may or may not be seen, Mallory-Denk bodies may or may not be present; but in noncirrhotic livers, zone 3 perisinusoidal fibrosis is required (see Figure 4.4.4). Lobular inflammation is a nondiscriminatory finding of most forms of chronic hepatitis and thus is not a required feature. Zone 3 accentuation of the necroinflammatory lesions may be lost with progression of fibrosis and architectural remodeling. Other groups have utilized less stringent criteria and discuss steatosis, lobular inflammation, and ballooning as features to characterize concurrent steatohepatitis (13,16). Suggesting that another form of chronic liver disease is present in a clinical and histologic setting of steatohepatitis can be characterized by finding the diagnostic lesions of the entity, such as periportal globules to suggest alpha-1antitrypsin deficiency, or markedly increased hepatocellular iron to suggest possible iron overload. Other, more common processes are discussed below. Hepatitis C Virus With Steatosis: Concurrent or Synonymous?
Steatosis alone has been recognized as a cofactor of progression in most liver diseases, including viral hepatitides, AIH, primary biliary cirrhosis, hereditary hemochromatosis (15). Steatosis can be found in the majority of hepatitis C virus (HCV) cases; the incidence of steatohepatitis, however, is significantly less (16) (Figures 4.4.8 and 4.4.9). Steatosis is thought to be related to “host” effects of weight, insulin resistance (IR), and diabetes mellitus (DM) in genotypes non-3 HCV, whereas genotype 3 HCV commonly results in hepatic steatosis due to viral
FIGUR E 4. 4. 7 Relative increase of portal inflammation compared
with the decrease in lobular inflammation and other features of active steatohepatitis has been reported in responders in treatment trials. This was demonstrated in the case of a patient who had clinical and histologic response to rosiglitazone.
chronic hepatitis C patients who lost weight showed improved liver tests and histologic findings in spite of lack of viral load alterations (10). Although exact figures are virtually impossible to discern, published work from referral centers has shown that around 5% of subjects with other diagnosed forms of chronic liver disease also have NAFLD or NASH (11-13). It is now currently accepted that even alcoholic liver disease can be concurrent with NAFLD/NASH, as proposed in the 1990s (14), although distinguishing the 2 typically relies on clinical findings.
F I G U R E 4 . 4 . 8 This biopsy is from a patient with HCV; it was done for grading and staging of HCV. The biopsy clearly shows the lesions of active steatohepatitis: zone 3 steatosis, ballooning, and perisinusoidal fibrosis. This is, therefore, a case of concurrent HCV and steatohepatitis.
50
CHAPTER
4:
FAT T Y
LIVER
DISEASE
F I G U R E 4 . 4 . 1 0 The trichrome stain demonstrates the diagnosis of
steatohepatitis in this case; zone 3 perisinusoidal fibrosis. The patient, however, is ANA positive. FIGURE 4. 4. 9 The trichrome stain of the same case highlights the
zone 3 perisinusoidal fibrosis.
promotion of hyperlipidemia but not the metabolic effects of IR. Whether steatosis promotes fibrosis, lack of response to treatment, and hepatocellular carcinoma in HCV-infected subjects are ongoing discussions recently reviewed (17). In a prospective study of biopsies and demographics in 278 hepatitis C patients (16), steatosis was noted in 34% and concurrent steatohepatitis in 9%. Two retrospective series documented steatohepatitis and HCV in 5% of cases (11,12). Bedossa et al showed that nongenotype 3 biopsies with steatohepatitis had greater fibrosis as well as elevated liver tests and serum triglycerides compared with controls. The study used ballooning and perisinusoidal fibrosis as features of concurrent steatohepatitis. The impact of concurrent underlying IR and/or DM on response to treatment for HCV has been reviewed (18,19). In brief, treatment of either has a potentially positive effect on both; whereas the presence of IR and/or DM negatively impacts response to HCV therapy and is related to progression of fibrosis. ANA/ASMA/AMA Pos NAFLD/NASH: What Does This Mean to the Pathologist?
Up to 5% of biopsies with features of other forms of chronic liver disease may have histopathologic features of concurrent steatohepatitis (11). This number may actually have increased since the original studies were published, as the prevalence of obesity has not abated. However, studies have also shown a range of non–organ-specific autoantibody positive cases of NAFLD. Typically, the autoantibody is ANA and/or anti–smooth muscle antibody (ASMA), but less commonly, antimitochondrial antibodies (AMA) or protoplasmic antineutrophil cytoplasmic antibodies (p-ANCA) have been noted (Figure 4.4.10). The prevalence has varied but may be as high as 30% overall (20,21). For a pathologist, the presence of a positive serologic test should initiate care in evaluation of liver biopsies in order to either document or exclude a concurrent disease. In a case
F I G U R E 4 . 4 . 1 1 This is an example of AMA positive primary biliary
cirrhosis (PBC) and steatohepatitis concurrent disease. The photomicrograph illustrates a florid duct lesion and shows some of the steatosis present elsewhere in the biopsy. The biopsy also had zone 3 perisinusoidal fibrosis.
with positive AMA, florid duct lesion and other features of chronic cholestasis (such as the presence of periportal copper in a noncirrhotic liver) that suggest primary biliary cirrhosis should be noted (Figure 4.4.11). A positive ANA with or without positive ASMA, however, is more common than AMA, and potentially raises considerable difficulties. Just as the presence of ANA cannot be considered diagnostic of AIH, simply the presence of plasma cells in portal inflammation cannot be considered sufficient criteria; in overlap of AIH and NASH, portal inflammation is marked and plasma cells are abundant, are present at the interface, and are present in the lobules. In addition, hepatitic rosettes and/or confluent necroses may be noted. A final finding for consideration of overlap with AIH is the presence of elevated globulins, particularly immunoglobulin G (IgG), in the serum. Without this, the diagnosis may be suggested but not unequivocally made. Alternatively, patients
CASE
4.4:
NONALCOHOLIC
with AIH may be receiving steroid therapy and may have steatosis (but not steatohepatitis with zone 3 perisinusoidal fibrosis) due to the medications. At the current time, the meaning of ANA-positive steatohepatitis is not clear; some authors have found increased severity of steatohepatitis (22–24), but this is not necessarily the case (25).
References 1. Ferenci P, Caca K, Loudianos G, et al. Diagnosis and phenotypic classification of Wilson disease. Liver Int. 2003;23:139–142. 2. Ishak KG. Inherited metabolic diseases of the liver. Clin Liver Dis. 2002;6:455–479, viii. 3. Schwimmer JB, Behling C, Newbury R, et al. Histopathology of pediatric nonalcoholic fatty liver disease. Hepatology. 2005;42:641–649. 4. Brunt EM. Pathology of nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol. 2010;7:195–203. 5. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999;94:2467–2474. 6. Brunt EM, Kleiner DE, Wilson LA, et al. Portal chronic inflammation in nonalcoholic fatty liver disease (NAFLD): a histologic marker of advanced NAFLD-clinicopathologic correlations from the nonalcoholic steatohepatitis clinical research network. Hepatology. 2009;49:809–820. 7. Brunt EM. Nonalcoholic steatohepatitis. Semin Liver Dis. 2004;24:3–20. 8. Brunt EM, Clouston AD. Histologic features of fatty liver disease. In: Bataller R, Caballeria J, eds. Nonalcoholic Steatohepatitis (NASH). Barcelona: Permanyer; 2007:95–110. 9. Clouston AD, Powell EE. Interaction of non-alcoholic fatty liver disease with other liver diseases. Best Pract Res Clin Gastroenterol. 2002;16: 767–781. 10. Hickman IJ, Clouston AD, Macdonald GA, et al. Effect of weight reduction on liver histology and biochemistry in patients with chronic hepatitis C. Gut. 2002;51:89–94. 11. Brunt EM, Ramrakhiani S, Cordes BG, et al. Concurrence of histologic features of steatohepatitis with other forms of chronic liver disease. Mod Pathol. 2003;16:49–56.
S T E AT O H E PAT I T I S
51
12. Ramesh S, Sanyal AJ. Hepatitis C and nonalcoholic fatty liver disease. Semin Liver Dis. 2004;24:399–413. 13. Ong JP, Younossi ZM, Speer C, Olano A, Gramlich T, Boparai N. Chronic hepatitis C and superimposed nonalcoholic fatty liver disease. Liver. 2001;21:266–271. 14. Naveau S, Giraud V, Borotto E, Aubert A, Capron F, Chaput JC. Excess weight risk factor for alcoholic liver disease. Hepatology. 1997;25: 108–111. 15. Yeh MM, Brunt EM. Pathology of nonalcoholic fatty liver disease. Am J Clin Pathol. 2007;128:837–847. 16. Bedossa P, Moucari R, Chelbi E, et al. Evidence for a role of nonalcoholic steatohepatitis in hepatitis C: a prospective study. Hepatology. 2007;46:380–387. 17. Powell EE, Jonsson JR, Clouston AD. Steatosis: co-factor in other liver diseases. Hepatology. 2005;42:5–13. 18. Cross TS, Rashid MM, Berry PA, Harrison PM. The importance of steatosis in chronic hepatitis C infection and its management: a review. Hepatology Research. 2010;40:237–247. 19. Blonsky JJ, Harrison SA. Review article: nonalcoholic fatty liver disease and hepatitis C virus—partners in crime. Aliment Pharmacol Ther. 2008;27:855–865. 20. Loria P, Lonardo A, Leonardi F, et al. Non-organ-specific autoantibodies in nonalcoholic fatty liver disease: prevalence and correlates. Dig Dis Sci. 2003;48:2173–2181. 21. Loria P, Carulli L, Lonardo A. The prevalence of autoantibodies and autoimmune hepatitis in patients with nonalcoholic fatty liver disease. Am J Gastroenterol. 2005;100:1200–1201. 22. Adams LA, Lindor KD, Angulo P. The prevalence of autoantibodies and autoimmune hepatitis in patients with nonalcoholic fatty liver disease. Am J Gastroenterol. 2004;99:1316–1320. 23. Niwa H, Sasaki M, Haratake J, et al. Clinicopathological significance of antinuclear antibodies in non-alcoholic steatohepatitis. Hepatol Res. 2007;37:923–931. 24. Bacon BR, Farahvash MJ, Janney CG, Neuschwander-Tetri BA. Nonalcoholic steatohepatitis: an expanded clinical entity. Gastroenterology. 1994;107:1103–1109. 25. Brunt EM. Pathology of hepatic iron overload. Semin Liver Dis. 2005;25:392–401.
Case 4.5
Steatohepatitis With Elevated Serum Iron Indices and Siderosis MATTHEW M.YEH AND ELIZABETH M. BRUNT
C L I N I C AL I N F OR M AT I ON DIAGNO SIS
A 48-year-old man presented with abnormal liver tests. He had metabolic syndrome including type 2 diabetes, obesity, and hyperlipidemia. Laboratory data also showed elevated serum iron indices, including increased transferrin saturation and ferritin. There was no family history of liver diseases. Physical examination did not reveal abnormal skin pigmentation and he had no joint pain. A liver biopsy was performed while results for genetic testing were awaited.
Steatohepatitis with mild iron deposition in the hepatocytes and sinusoidal lining cells.
DISCUSSIO N
Similar to alcoholic liver disease, iron deposition in NAFLD is not uncommon, especially in end-stage liver cirrhosis. Previous studies examining liver explants found increased iron in a significant number of nonbiliary cirrhosis cases, including NAFLD, with hepatic iron indices or 1.9 (the value commonly associated with hereditary hemochromatosis), indicating that nonbiliary cirrhosis may result in abnormal iron accumulation (1,2). The current explanation is that endstage liver disease leads to decreased expression of hepcidin, a negative regulator of iron homeostasis. Loss of hepcidin results in promoting excess release of iron from the reticuloendothelial cells, enhancing duodenal iron absorption and increasing iron uptake by the hepatocytes. Therefore overinterpretation of iron overload in biopsies from NAFLD-induced cirrhosis needs to be cautioned against. Iron deposition in NAFLD may be hepatocellular (Figure 4.5.1), reticuloendothelial (Figure 4.5.2), or both.
R E A S ON F OR R E F E R R A L
The liver biopsy shows features of steatohepatitis. In addition, there is significant hepatocyte iron deposition on using Perls’ iron stain. The specific question from the referring pathologist and hepatologist is whether the patient has hemochromatosis. PAT H OL OG I C F E AT U R E S
Besides the features of steatohepatitis, the liver biopsy also shows a mild degree of (grade 2 of 4) iron granules deposition within the hepatocytes and sinusoidal lining cells on using Perls’ iron stain (Figure 4.5.1).
FIGURE 4. 5. 1 Perls’ iron stain shows mild iron deposition in the hepatocytes and Kupffer cells.
F I G U R E 4 . 5 . 2 Perls’ iron stain demonstrates deposition in Kupffer
cells in an NAFLD case.
52
CASE
4.5:
NASH
WITH
E L E VAT E D
The pattern of iron deposition in the hepatocytes may inhomogenously involve zone 1, or may be similar to hereditary hemochromatosis, with a decreasing gradient from zone 1 to zone 3. Iron deposition in the reticuloendothelial cells, is typically punctuate and may be panacinar. NAFLD has been reported to be associated with increased serum iron indices including transferrin saturation and ferritin independent of genetic hemochromatosis. However, other studies showed the C282Y HFE gene mutation was enriched in NAFLD patients (3,4) and there was a significant correlation among the mutation, hepatic iron concentration or tissue iron stain, and degree of fibrosis (3). Thus liver biopsy in NAFLD patients may also provide prognostic implication in assessing hepatic iron stores, in addition to confirming the diagnosis. Although elevated serum ferritin may be a manifestation of aberrant HFE genetics (4), it has to be noted that elevated serum ferritin levels do not necessarily reflect increased hepatic iron stores, as ferritin is a known acute phase reactant protein and may be nonspecifically elevated in NAFLD alone or in insulin resistance (5,6). Whether
SERUM
IRON
INDICES
53
aberrant genetics, iron deposition, and localization of increased iron in NAFLD relate to progression of disease are areas of active investigation.
References 1. Ludwig J, Hashimoto E, Porayko MK, Moyer TP, Baldus WP. Hemosiderosis in cirrhosis: a study of 447 native livers. Gastroenterology. 1997;112:882–888. 2. Deugnier Y, Turlin B, le Quilleuc D, et al. A reappraisal of hepatic siderosis in patients with end-stage cirrhosis: practical implications for the diagnosis of hemochromatosis. Am J Surg Pathol. 1997;21:669–675. 3. George DK, Goldwurm S, MacDonald GA, et al. Increased hepatic iron concentration in nonalcoholic steatohepatitis is associated with increased fibrosis. Gastroenterology. 1998;114:311–318. 4. Bonkovsky HL, Jawaid Q, Tortorelli K, et al. Non-alcoholic steatohepatitis and iron: increased prevalence of mutations of the HFE gene in nonalcoholic steatohepatitis. J Hepatol. 1999;31:421–429. 5. Bugianesi E, Manzini P, D’Antico S, et al. Relative contribution of iron burden, HFE mutations, and insulin resistance to fibrosis in nonalcoholic fatty liver. Hepatology. 2004;39:179–187. 6. Zelber-Sagi S, Nitzan-Kaluski D, Halpern Z, Oren R. NAFLD and hyperinsulinemia are major determinants of serum ferritin levels. J Hepatol. 2007;46:700–707.
Case 4.6
Alcoholic Steatohepatitis MATTHEW M.YEH AND ELIZABETH M. BRUNT
As can be recalled, the term “nonalcoholic steatohepatitis,” credited to Ludwig et al from the 1980 publication (1) derived from the initial consideration of liver biopsy findings identical to ALD in patients who did not consume alcohol. These findings included macrosteatosis, ballooning, the presence of Mallory’s hyaline, now referred to as Mallory-Denk bodies (2), lipogranulomas, and varying degrees of fibrosis, including perisinusoidal, “chicken-wire” fibrosis. The clinical features included female gender, overweight, diabetes, and hyperlipidemia. These same clinical and histologic findings had been previously or concurrently reported in the similar setting of overweight and/or diabetes by others in the United States (3,4), Europe (5), and Japan (6) and were subsequently confirmed from studies around the world. Unfortunately, both the negative nomenclature (“non” alcoholic) and the concept of inseparability from “alcoholic histologic features” are embedded in the literature. Currently, there is discussion in the literature as to the amount of alcohol intake required for consideration of ALD (7). However, as pointed out, more than simply an amount may be causative; in fact, types of alcohol consumed and patterns are also important considerations (8). Most NAFLD literature lists consumption of 20g/d for men and 10g/d for women as the upper limit for “exclusion of ALD” as a contributor to liver disease.
F I G U R E 4 . 6 . 1 Active steatohepatitis with perisinusoidal fibrosis. Whether the lesions are caused by the patient’s diabetes or “mild alcohol” use, or both, cannot be discerned by histology.
DIAGNO SIS
Steatohepatitis, history of obesity, diabetes, and alcohol use.
DISCUSSIO N
The overlapping and often inseparable features of ALD and NAFLD are likely a reflection of the overlapping underlying pathophysiology of these 2 processes. The features in common include steatosis, that is, macrovesicular steatosis with or without lobular inflammation and with or without foci of microvesicular steatosis, with or without megamitochondria; and steatohepatitis, that is, steatosis with hepatocyte ballooning, with or without Mallory-Denk bodies, lobular inflammation, varying degrees of portal inflammation, and with or without varying amounts of perisinusoidal fibrosis. Cirrhosis, with or without the features above, may also occur in both processes. It is for these reasons, as illustrated in the case, pathologists cannot always assign an etiology, and it is recommended that the report simply state what is present, that is, steatohepatitis, and what is known clinically, that is obesity, diabetes, alcohol use. However, there are features of ALD that, to date, have not been reported in NAFLD. First, alcoholic hepatitis or cirrhosis due to alcohol may have a veno-occlusive lesion of subendothelial fibrosis or obliteration (9). Secondly, alcoholic hepatitis may or may not have steatosis but is characterized by numerous hepatocytes with Mallory-Denk bodies and satellitosis (polymorphonuclear leukocytes surrounding the affected cells). The affected hepatocytes may be ballooned or undergoing apoptosis and are eosinophilic and shrunken.
C L I N I C AL I N F OR M AT I ON
The patient is a 58-year-old obese, diabetic man with elevated liver tests: alanine aminotransferase (ALT) 59 IU/L, aspartate transaminase (AST) 76 IU/L, and hemoglobin A1C (HbA1C) 7.7. The patient’s history includes “mild” alcohol use. R E A S ON F OR R E F E R R A L
The question in this liver biopsy is one of attribution of cause of elevated liver tests: specifically, are they due to alcoholic steatohepatitis or nonalcoholic steatohepatitis? PAT H OL OG I C F E AT U R E S
The histologic features of the biopsy are those of markedly active steatohepatitis with grade 1 (5%–33%) macrovesicular steatosis in a nonzonal distribution, marked ballooning and evidence of Mallory-Denk bodies, moderate lobular and portal chronic inflammation. Perisinusoidal and periportal fibrosis can be discerned even on the hematoxylin and eosin (HE)–stained slides (Figure 4.6.1).
54
CASE
4.6:
ALCOHOLIC
Thirdly, canalicular cholestasis has yet to be reported in NAFLD. Finally, alcoholic foamy degeneration has also not been reported in NAFLD. This is an uncommon form of ALD; the hepatocytes appear uniformly foamy and there is little to no portal and/or lobular inflammation (10).
References 1. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 1980 Jul;55(7);434–438. 2. Zatloukal K, French SW, Denk H, et al. From Mallory to MalloryDenk inclusion bodies: what, how and why? Exp Cell Res. 2007;313: 2033–2049. 3. Miller DJ, Ishimaru H, Klatskin GGA. Nonalcoholic liver disease mimicking alcoholic hepatitis and cirrhosis. Gastroenterology. 1979;77:27A.
S T E AT O H E PAT I T I S
55
4. Schaffner F, Thaler H. Nonalcoholic fatty liver disease. Progr Liver Dis. 1986;8:283–298. 5. Thaler H. Die fettleber und ihre pathogenetische Beziehung zur Leberzirrhose. Virchows Arch. 1962;335:180–188. 6. Itoh S, Tsukada N, Motomura Y, Ichinoe A. Five patients with nonalcoholic diabetic cirrhosis. Acta Hepato-Gastroenterol. 1979;26:90–97. 7. Falck-Ytter Y, Younossi ZM, Marchesini G, McCullough AJ. Clinical features and natural history of nonalcoholic steatosis syndromes. Semin Liver Dis. 2001;21:17–26. 8. Day CP. Who gets alcoholic liver disease: nature or nurture? J R Coll Physicians Lond. 2000;34:557–562. 9. Goodman ZD, Ishak KG. Occlusive venous lesions in alcoholic liver disease. A study of 200 cases. Gastroenterology. 1982;83:786–796. 10. Hall P. Alcoholic liver disease. In: MacSween RNM BA, Burt AD, Portmann BC, et al. eds. Pathology of the Liver. 4th ed. London, UK: Churchill Livingstone; 2002; 273–311.
Case 4.7
Drugs and NAFLD MATTHEW M.YEH AND ELIZABETH M. BRUNT
The associations of medications with NAFLD continues to expand. As noted by authorities in the field, it is important to consider that the patient may have NAFLD and may be on a particular medication, but the disease may be related to the patient’s underlying metabolic condition rather than to the medication (1). Two drugs deserve specific comment: methotrexate and amiodarone. The injury of both may be exacerbated by the presence of underlying steatohepatitis due to insulin resistance; however, there are clues to distinguish the injuries. Methotrexate use may result in hepatocellular anisonucleosis and focal necrosis, nonzonally distributed macrovesicular steatosis that is commonly ”mild” (30) were noted as an incidental finding. Repeat imaging confirmed multiple 1- to 4-cm liver cysts with no solid component or contrast enhancement. The cysts were not connected to the biliary system. The kidneys were normal. Laboratory findings included normal creatinine and liver function tests. Persistent abdominal pain and an increase in the size of one cyst led to partial hepatectomy.
Multiple liver cysts of unclear nature. PAT H O LO GIC FEAT UR ES
The gross examination revealed multiple cysts with scant normal hepatic parenchyma. The cysts showed a denuded or partly preserved lining consisting of low columnar to cuboidal bland-appearing epithelium (Figure 8.3.1). Several small von
FIGURE 8. 3. 1 (A) Partial hepatectomy from a patient with adult polycystic liver disease showing numerous cysts. (B) Cut surface shows
multiple smooth-walled cysts. (C) Noncystic areas of the liver show von Meyenburg complexes. (D) Liver cysts are similar to the cysts in autosomal dominant polycystic kidney disease.
147
148
CHAPTER
8:
CYSTIC
Meyenburg complexes were present in the adjacent liver. The cysts were not associated with bile ducts, but one appeared to arise from a dilated ductule at the edge of a von Meyenburg complex. The hepatic parenchyma at the periphery of the cyst showed nonspecific fibrosis without typical features of congenital hepatic fibrosis.
D I AG N OS I S
Autosomal dominant polycystic liver disease.
D I SC U SSI ON
The differential diagnosis in this case included autosomal recessive polycystic kidney disease (ARPKD), autosomal dominant polycystic kidney disease (ADPKD), autosomal dominant polycystic liver disease (ADPLD), and Caroli syndrome, all of which can present with liver cysts that can be indistinguishable on imaging studies and histology. Detailed family history revealed cystic liver disease without renal impairment in two family members. Presence of isolated cystic liver disease without any renal involvement after careful imaging analysis makes ADPLD the most likely diagnosis. Recent confirmation of unique mutations in the PRKSCH and SEC63 genes has confirmed that ADPLD is distinct from its kidney-related counterparts. Genetic sequencing in a research lab within the institution was undertaken, and the patient was found to have
DISEASE
OF
THE
LIVER
a germline mutation in the SEC63 gene on chromosome 6p, consistent with ADPLD. Many ADPLD patients are asymptomatic. In some cases, abdominal pain or fullness prompts clinical work-up. The finding of at least 5 liver cysts on imaging is enough to warrant further investigation for ADPLD. Due to autosomal dominant inheritance, other family members may have cystic liver disease as well, but may be asymptomatic. Of crucial importance is the lack of renal cystic disease, both in the patient and in family members. Typically, liver synthetic function is well preserved, and liver function tests do not show significant abnormalities. Biopsy of the cysts is generally not indicated, and the pathologist is involved either when cysts are resected/unroofed, or if the patient undergoes transplantation. Following partial excision of the cysts or unroofing procedure, often only a very limited amount of adjacent liver parenchyma is present, precluding proper assessment. The histology of the cysts is similar to those seen in ADPKD. In particular, von-Meyenburg complexes may be prominent, and cysts may be seen to emanate from them. Cysts are completely separate from bile ducts, as opposed to ARPKD, although this may be difficult to assess on histology. A combination of family history, detailed imaging findings, exclusion of associated renal disease, and genetic tests help in establishing a final diagnosis in such cases. Often a diagnosis is established without tissue confirmation in the appropriate clinical setting. Genetic testing for ADPLD remains in its infancy but may become commercially available in the near future.
Case 8.4
Choledochal Cyst BARTON KENNEY AND DHANPAT JAIN
C L I N IC AL I N F OR M AT I ON
R EA SO N FO R R EFER R A L
A 12-week-old female infant presented with jaundice. The child was not breast-fed and had no history of neonatal jaundice. Physical examination revealed fullness in the right upper quadrant. Laboratory findings included elevated total (2.4 mg/dL) and direct (1.8 mg/dL) bilirubin. Alkaline phosphatase was markedly elevated (2418 U/L); AST and ALT were normal. Ultrasound revealed a 4.5-cm extrahepatic cyst involving the common bile duct (CBD). The gallbladder and liver appeared normal, and no stones were visualized. The patient underwent resection of the cyst and hepaticojejunostomy.
Classification of extrahepatic cyst involving the CBD. The differential diagnosis included foregut cyst, abscess, and choledochal cyst. PAT H O LO GIC FEAT UR ES
The resection specimen revealed a fusiform cystic dilation involving most of the CBD (Figure 8.4.1). On sectioning, the lumen contained thick dark-green biliary sludge. The wall was smooth and without masses or excrescences. On histologic examination, the wall was thickened and fibrotic, with an
FIGURE 8. 4. 1 (A) Choledochocystectomy specimen with saccular dilation of the common bile duct suggestive of a choledochal cyst. (B) Cholangiogram showing the same finding. (C) Cyst wall consisting of fibrosis and a single layer of biliary type lining epithelium. (D) Higher magnification showing benign columnar biliary epithelium overlying fibrous tissue.
149
150
CHAPTER
8:
CYSTIC
associated lymphocytic infiltrate. The lining epithelium was mostly denuded, but what remained was cuboidal to columnar biliary epithelium with reactive atypia. There was no evidence of malignancy.
D I AG N OS I S
Choledochal cyst.
D I SC U SSI ON
The differential diagnosis of solitary cysts near the hepatic hilum in children includes enterogenous type foregut cysts/duplication cysts, choledochal cysts, and abscess. The gross pathology of choledochal cysts is characteristic, and microscopic examination typically reveals a fibrotic wall lined in places by biliary epithelium. Enterogenous foregut or duplication cysts are lined by squamous, ciliated columnar or gastric type mucosa and may be separate from the biliary tree. The diagnosis of choledocal cyst is often easily established clinically due to the location of the cyst and its biliary connection. These cysts should be resected followed by subsequent hepaticojejunostomy or biliary reconstruction, both due to obstructive symptoms and to a potential risk of malignancy. The incidence of choledochal cysts (biliary cysts) in the West varies from 1:100 000 to 150 000 live births. The frequency is much higher in Asia, where some locations show incidence rates as high as 1:1000 live births (1). There is a female predominance. Most cases present in infancy or childhood, and the majority of the cases are diagnosed in the first
DISEASE
OF
THE
LIVER
decade of life (2). Young children may present with the classic triad of abdominal pain, jaundice, and a palpable mass in approximately 20% of cases, but at least 2 of these 3 findings are present in the majority of patients (2). Older children, and rarely adults, may present with recurrent pancreatitis, fever, abdominal pain, and jaundice. Once suspected, the cysts are usually readily characterized on abdominal imaging as saccular or fusiform dilations of the extrahepatic bile duct(s). Several types of choledochal cysts are recognized (see Table 8.1.3, Case 8.1) (1,3). Type I cysts are by far the most common, representing 80% to 90% of all choledochal cysts. These involve dilation of a segment of the CBD or of the entire CBD. Types IV and V include intrahepatic duct dilation and overlap with, and can be indistinguishable from Caroli disease. In addition to cholangitis and pancreatitis, these cysts are associated with a significant risk of cholangiocarcinoma ranging from 9% to 28% (4,5).
References 1. Singham J, Yoshida EM, Scudamore CH. Choledochal cysts: part 1 of 3: classification and pathogenesis. Can J Surg. 2009;52(5):434–440. 2. Singham J, Yoshida EM, Scudamore CH. Choledochal cysts: part 2 of 3: Diagnosis. Can J Surg. 2009;52(6):506–511. 3. Todani T, Watanabe Y, Narusue M, Tabuchi K, Okajima K. Congenital bile duct cysts: classification, operative procedures, and review of thirtyseven cases including cancer arising from choledochal cyst. Am J Surg. 1977;134(2):263–269. 4. Chaturvedi A, Singh J, Rastogi V. Case report: cholangiocarcinoma in a choledochal cyst. Indian J Radiol Imaging. 2008;18(3):236–238. 5. Franko J, Nussbaum ML, Morris JB. Choledochal cyst cholangiocarcinoma arising from adenoma: case report and a review of the literature. Curr Surg. 2006;63(4):281–284.
Case 8.5
Solitary Hepatic Cyst Versus Hydatid Cyst BARTON KENNEY AND DHANPAT JAIN
C L I N IC AL I N F OR M AT I ON
A 32-year-old male presented with abdominal fullness. Physical examination showed mild right upper quadrant tenderness. Laboratory findings showed leukocytosis with 15% eosinophils. Liver transaminases were normal, whereas total bilirubin and alkaline phosphatase were mildly elevated. CT scan showed an 8-cm irregular cyst with water attenuation. There was focal pericystic calcification, with focal cystwithin-cyst pattern.
R E A SON F OR R E F E R R AL
Hepatic cyst of presumed infectious etiology. The differential diagnosis includes simple hepatic cyst, pyogenic abscess, amoebic abscess, or hydatid cyst. After consultation with the infectious disease team, the patient’s serum was sent for Echinococcus serological testing, which was negative.
PAT H OL OG I C F E AT U R E S
At resection, the right hepatic lobe was sectioned to reveal a single large cyst containing serous fluid and a thick fibrous wall, but without an identifiable lining (Figure 8.5.1). A scant amount of adjacent preserved liver parenchyma was present that appeared unremarkable.
DIAGNO SIS
Simple hepatic cyst.
DISCUSSIO N
The lack of any specific histologic features and the presence of normal adjacent liver parenchyma suggests the diagnosis of simple hepatic cyst. The characteristic laminated wall of hydatid cyst is not seen. Typical features of hepatobiliary cystadenoma/carcinoma such as columnar or mucinous lining epithelium and ovarian-type stroma are not present. Simple cysts are typically solitary, and there is usually no familial history of liver cysts. The cysts tend to grow at a slow rate and can remain asymptomatic for many years. Larger cysts come to clinical attention due to mass effect or incidental detection on imaging studies. The cyst fluid may re-accumulate soon after needle aspiration, which is best avoided as a definitive therapeutic intervention. Due to the risk of anaphylaxis, needle aspiration is contraindicated if hydatid cyst is a consideration. Imaging studies are often nonspecific. Simple hepatic cysts are uniloculated, and the remaining liver generally appears normal. Pathologic examination is important for excluding an infectious cyst or cystic neoplasm. Classification into a hepatic or biliary-type cyst is often not possible and is not clinically relevant.
FIGURE 8. 5. 1 (A) Simple hepatic cyst opened to reveal luminal hemorrhage. (B) Simple hepatic cyst with a thin fibrous cyst wall and smooth inner surface. (C) Histology of the cyst wall showing denuded epithelium, fibrous wall, and mural hemorrhage.
151
This page intentionally left blank
9 Hereditary Hyperbilirubinemias SARANGARAJAN RANGANATHAN
I N T ROD U C T I ON
Elevation of bilirubin is a physiological phenomenon in the first week of life. This is controlled by the complex system of bilirubin conjugation and excretion and is the result of interplay of several enzymes and factors responsible for the uptake of bilirubin by the hepatocyte, its conjugation, release into bile, reabsorption in the enterohepatic circulation, and its ultimate excretion into placenta or urine. Neonatal hyperbilirubinemias are a group of genetic disorders that are characterized by disorder and defects at various steps in this process resulting in pathological elevations of unconjugated or conjugated bilirubins in the neonatal period (1). They need to be differentiated as a group from physiologic hyperbilirubinemia of the newborn that is the result of increased bilirubin production and delayed maturation resulting in decreased excretion by the neonatal liver.
due to a defect in the 39 end of the gene results in the severe form (Type I CJS), and a defect in the variable region results in partial defect in the isoforms causing the less severe Type II CJS (3,4). CJS Type I
This is the more severe form of the disease manifesting in infancy with serum bilirubin levels ranging from 20 to 25 mg dL and may exceed 50 mg/dL. Almost all the bilirubin
U N C O N J U G AT E D H Y P E R B I L I RU B I N E M I A Gilbert Syndrome
Gilbert syndrome (GS) is a familial, autosomal-dominant or recessive disease that is characterized by a mild, chronic unconjugated hyperbilirubinemia. There is a strong male predominance and most patients are diagnosed as young adults. The serum bilirubin level is usually less than 3 mg/dL but can fluctuate and rise during intercurrent illnesses. The remaining liver function tests and liver biopsy are normal with only some lipofuscin accumulation. Association with other genetic defects such as glucose-6-phosphate dehydrogenase (G6PD) deficiency, b-thalassemia, or hereditary spherocytosis can result in neonatal hyperbilirubinemia. It is now known to be caused by mutations or polymorphisms in the bilirubin UDPglucuronosyltransferase (UGT1A1) gene (2,3). Insertion of TA dinucleotide in the TATA box of the UGT1A1 promoter is the underlying abnormality in most Caucasians, with most affected individuals being homozygous. Heterozygosity can also result in higher bilirubin levels compared with normal. GS is not associated with any significant morbidity or mortality.
A
C R I G L E R -N AJ J A R S Y N D ROM E
The Crigler-Najjar syndrome (CJS) was first described by Crigler and Najjar in 1952 as a form of severe congenital unconjugated hyperbilirubinemia that presents in the first few days of life and is fatal due to kernicterus developing as a complication. There are 2 types: CJS Type I and CJS Type II. Both types are associated with the same UGT1A1 gene wherein a complete mutation of all isoforms of bilirubin-UGT cDNA
B F I G U R E 9 . 1 (A) Explanted liver in CJS type I showing a glistening smooth surface with minimal discoloration and no evidence of cirrhosis. (B) Canalicular bile plugs with preserved liver parenchyma and no hepatocellular pigment (H&E 400).
153
154
CHAPTER
9:
H E R E D I TA RY
is unconjugated due to the inability to conjugate bilirubin with uridine diphosphoglucuronic acid. Unlike GS, there is no response of bilirubin to phenobarbital administration. Phototherapy is the treatment of choice together with plasmapheresis and liver transplantation in older children (5). Liver biopsy is normal except for bile plugs in canaliculi. The explanted liver is normal in color and architecture with
HYPERBILIRUBINEMIAS
no evidence of cirrhosis (Figure 9.1). There is canalicular cholestasis with normal hepatocytes and no significant fibrosis (Figure 9.2). CJS Type II
Also known as Arias syndrome, it commonly affects patients in later life. It is associated with marked unconjugated hyperbilirubinemia with a dominance of bilirubin monoglucuronide in the bile, as opposed to CJS Type I that has diglucoronide as the dominant component. Although it is less severe than Type I disease, it can cause kernicterus and lead to death. CO NJUGAT ED H Y P ER BILIRUBINEMIA S Rotor Syndrome
FIGURE 9. 2 DJS with the characteristic pigment accumulation in the
cytoplasm of hepatocytes in the centrilobular region (H&E 400).
This is a rare, familial hyperbilirubinemia that is characterized by a chronic conjugated hyperbilirubinemia without hemolysis. It was originally thought to be related to Dubin-Johnson syndrome (DJS), but its genetic defect is unknown (6). Laboratory tests show a delay in excretion of all forms of organic anions including bromosulphthalein (BSP) and indocyanine green (ICG). The increase in BSP excretion is usually at 45 minutes, but the secondary increase at 90 minutes noted in DJS is not seen in Rotor syndrome due to the absence of reflux of conjugated BSP. Urinary coproporphyrin levels are also markedly increased in Rotor syndrome as compared with other hereditary hyperbilirubinemias and may be 25% to 50% higher
TA B LE 9. 1 Salient laboratory and genetic features of hereditary hyperbilirubinemias Gilbert Syndrome (GS)
Crigler-Najjar Syndrome (CJS)
Rotor Syndrome (RS)
Dubin-Johnson Syndrome (DJS)
Inheritance
Autosomal dominant
Autosomal recessive or dominant
Autosomal recessive
Autosomal recessive
Prevalence
3%
Rare
Rare
1:1300 Iranian Jews
Serum bilirubin
3–10 Unconjugated
20–50 unconjugated
Up to 20 conjugated
Up to 20 conjugated
Serum bilirubin decrease with phenobarbitol
70%
0%—CJS I 77%—CJS II
Not seen
Not seen
BUGT activity
5–53% controls
Severe ↓—CJS I 2–23%—CJS II
Normal
Normal
BSP clearance
Normal
Normal
Delayed
Normal early; delayed 90–120 min
ICG clearance test
Normal
Normal
Delayed
Normal
Urine coproporphyrin
Normal
Normal
Up to 5 × ↑ Isomer 1 < 80%
Normal Isomer I > 80%
Cholecystogram
Visualized biliary tree
Visualized biliary tree
Gallbladder visualized
Gallbladder not visualized
Other LFT
Normal
Normal
Normal
Normal
Genetics
UGT1A1 gene
UGT1A1 gene
Unknown
cMOAT/MRP2/ABCC2 gene
Prognosis
Benign
Mortality due to kernicterus
Benign
Benign
Abbreviations: BSP, bromosulphthalein; BUGT, bilirubin UDP glucoronyltransferase; ICG, indocyanine green; cMOAT, canalicular multispecific organic anion transporter; MRP2, multidrug resistant protein, LFT, liver function tests; UGT1A1, UDP glucuronosyltransferase; ABCC2, ATP-binding cassette, subfamily C, member 2.
CHAPTER
9:
H E R E D I TA RY
HYPERBILIRUBINEMIAS
155
than in DJS. The coproporphyrin isomer I is, however, around 65% of the total as compared with DJS. The increase in coproporphyrin is due to decreased biliary excretion with concomitant increase in filtration and excretion by the kidneys. The disease course is benign and liver histology is normal. There is no pigmentation as seen with DJS. Dubin- Johns on Synd rome
A
B
C FIGURE 9.3 (A) Preserved liver architecture with hepatocyte bal-
looning and minimal intracellular pigment (H&E 400). (B) Normal liver with cMOAT expression in canaliculi. Stain courtesy of Dr. AS Knisely, UK ( 400). (C) Absence of staining cMOAT staining in DJS ( 400).
This is an uncommon disease except in certain populations such as the Persian Jews that have a high prevalence of 1:1300. Originally described by Dubin and Johnson (7), it is characterized by conjugated hyperbilirubinemia that can present at any age with constitutional symptoms. The salient laboratory features are listed in Table 9.1. The significant laboratory abnormality in DJS is in the transport of the organic anion BSP. Following intravenous injection the initial plasma BSP concentration is normal, and at 45 minutes it is normal to elevated. The diagnostic test is at 90 minutes when BSP levels are higher than at 45 minutes in more than 90% of DJS patients. This secondary increase is due to the reflux of conjugated BSP into the plasma. The same is not true for other anions including ICG. Since the original description of elevated urinary coproporphyrin in DJS, it has been proven that although the total urinary coproporphyrin level may be normal, in DJS it is the isomer I form that constitutes over 80% of the urinary coproporphyrin rather than the normal excretion of coproporphyrin isomer III in all other patients and normal individuals. DJS has been the only hyperbilirubinemia that is characterized by coproporphyrin isomer I elevation, and hence this test is diagnostic of DJS. Recent molecular studies have demonstrated a defect in the transport of non–bile salt organic anions at the apical canalicular membranes by the ATP-binding cassette (ABC) transport system that is regulated by a single gene on chromosome 10q24. This gene has been variably called as cMOAT (canalicular multispecific organic anion transporter) or MRP2 (multidrug resistant protein 2), or ABCC2 (8–10). The classic pathologic finding is a black liver that is characterized by accumulation of brown-black pigment in the hepatocytes in a pericentral location (Figure 9.3). This pigment stains variably with Oil red O, PAS with diastase, and Fontana-Masson stain. The exact nature of the pigment is still unknown and debate continues on whether it is related to melanin, metabolites of epinephrine, or other compounds that are not excreted from the hepatocytes due to the defect (7,11). Unusual presentations include neonatal presentation, absence of pigment especially in younger children (12), and simultaneous occurrence of GS and DJS (13).
References 1. Gourley GR. Neonatal jaundice and disorders of bilirubin metabolism. In: Suchy FJ, Sokoi RJ, Balisteri WF, eds. Liver Disease in Children. 3rd ed. New York, NY: Cambridge University Press;2007:270–309. 2. Bosma PJ, Chowdhury JR, Bakker C, et al. The genetic basis of the reduced expression of bilirubin UDP-glucoronyltransferase 1 in Gilbert’s syndrome. N Engl J Med. 1995;333:1171–1175.
156
CHAPTER
9:
H E R E D I TA RY
3. Kadakol A, Ghosh SS, Sappal BS, et al. Genetic lesions of bilirubin uridine-diphosphoglucuronate glucuronosyltransferase (UGT1A1) causing Crigler-Najjar and Gilbert syndromes: correlation of genotype to phenotype. Hum Mutat. 2000;16(4):297–306. 4. Sneitz N, Bakker CT, de Knegt RJ, Halley DJ, Finel M, Bosma PJ. Crigler-Najjar syndrome in The Netherlands: identification of four novel UGT1A1 alleles, genotype-phenotype correlation, and functional analysis of 10 missense mutants. Hum Mutat. 2009;30:1–8. 5. Kaufman SS, Wood RP, Shaw BW Jr, et al. Orthotopic liver transplantation for Type I Crigler-Najjar syndrome. Hepatology. 1986;6:1259–1262. 6. Hrebicek M, Jirasek T, Hartmannova H, et al. Rotor-type hyperbilirubinemia has no defect in the canalicular bilirubin export pump. Liver In. 2007;27:485–491. 7. Dubin IN, Johnson FB. Chronic idiopathic jaundice with unidentified pigment in liver cells; a new clinicopathologic entity with a report of 12 cases. Medicine (Baltimore). 1954;33:155–198. 8. Paulusma CC, Kool M, Bosma PJ, et al. A mutation in the human canalicular multispecific organic anion transporter gene causes the Dubin-Johnson syndrome. Hepatology. 1997;25:1539–1542.
HYPERBILIRUBINEMIAS
9. Keppler D, Konig J. Hepatic canalicular membrane 5: Expression and localization of the conjugate export pump encoded by MRP2 (cMRP/ cMOAT) gene in liver. FASEB J. 1997;11:509–516. 10. Kanda D, Takagi H, Kawahara Y, et al. Novel large-scale deletion (whole exon 7) in the ABCC2 gene in a patient with the Dubin-Johnson syndrome. Drug Metab Pharmacokinet. 2009;24:464–468. 11. Rastogi A, Krishnani N, Pandey R. Dubin-Johnson syndrome— a clinicopathologic study of 20 cases. Indian J Pathol Microbiol. 2006;49:500–504. 12. Lee JH, Chen HL, Chen HL, Ni YH, Hsu HY, Chang MH. Neonatal Dubin-Johnson syndrome: long term follow-up and MRP2 mutations study. Pediatr Res. 2006,59:584–589. 13. Cebecauerova D, Jirasek T, Budisova L, et al. Dual hereditary jaundice: simultaneous occurrence of mutations causing Gilbert’s and Dubin-Johnson syndromes. Gastroenterology. 2005,129:315–320.
Case 9.1
Dubin-Johnson Syndrome SARANGARAJAN RANGANATHAN
C L I N IC AL I N F OR M AT I ON
A 6-week-old child presented with hyperbilirubinemia since birth with a total bilirubin of 6.1 mg/dL and a direct component of 3.1 mg/dL. She did not have any constitutional symptoms, but there was elevation of gamma glutamyl transpeptidase (GGT) that was decreasing on its own but had not normalized. There was some elevation of urinary coproporphyrin. A liver biopsy was done. R E A SON F OR R E F E R R AL
To evaluate for causes of hyperbilirubinemia and to exclude other causes of neonatal cholestasis M I C ROSC OP I C F E AT U R E S
The liver biopsy shows a core of parenchyma with preserved architecture. There is no portal fibrosis or ductular proliferation to suggest biliary obstruction. There is ballooning of hepatocytes without significant steatosis or dropout. There is scant pigment within hepatocytes representing iron and minimal steatosis. No dark-brown pigment of DJS is noted, and a Fontana-Masson stain was negative with only staining of the iron pigment. An immunohistochemical stain was performed for cMOAT and showed complete loss of expression of this antigen in the apical canalicular membrane (see Chapter 9, Figure 9.3). The control showed appropriate staining. A diagnosis of DJS was made. D I S C U S S I ON
Although the classic picture is to identify the black pigment of DJS, neonates may not show this pigment, and hence immunohistochemistry and molecular studies may be needed to make this diagnosis. Other causes of conjugated hyperbilirubinemia such as obstruction, alpha-1-antitrypsin deficiency, total parenteral nutrition (TPN), and familial cholestasis may need to be excluded in some cases. The classic features of obstruction are portal expansion with bile ductular proliferation and ductular and canalicular cholestasis: features that are usually not present in DJS other than the canalicular cholestasis. Fibrosis is not an integral part of the histology of DJS. Familial cholestasis may also present with canalicular cholestasis, but the Byler type (ATP8B1 disease) and PFIC-2 (ABCB11 disease) are both characterized by a low GGT in
view of the persistent cholestasis. Byler disease is characterized by canalicular bile plugs with pale staining bile. Electron microscopy shows the characteristic granular, “coarse” bile in the canaliculus. Immunohistochemical staining frequently shows aberrant staining for cytokeratin 7 (CK7), a bile duct marker within hepatocytes within the lobules, either singly or in clusters. There is preserved immunoreactivity for GGT in the apical portions of cholangiocytes along canaliculi, usually at the periphery of the lobule. Some bile may also be seen in the hepatocytes and Kupffer cells, but again no bile plugs are noted. Progressive familial intrahepatic cholestasis-2 (PFIC2), on the other hand, shows a striking inflammatory background with the appearance of a giant cell hepatitis with prominent giant cell change in hepatocytes, foci of necrosis with inflammation, and portal fibrosis besides the canalicular cholestasis. Hepatocytes may express CK7, while GGT staining is preserved in the canaliculus. There is, however, loss of bile salt export pump (BSEP) protein expression in many cases. The third important subtype of PFIC that is most likely to mimic DJS is the ABCB4 disease that is usually a high-GGT conjugated hyperbilirubinemia of infancy and is the result of failure of activity of the ABCB4/MDR3 protein. The children present with neonatal hepatitis that rapidly progresses to cirrhosis unlike the clinical picture of DJS. Histology may show giant cell change, disarray of hepatocytes, portal fibrosis, inflammation, and biliary proliferation with subsequent bridging fibrosis and cirrhosis (1). TPN-induced liver damage shows abundant inflammation with portal expansion, cholestasis, both ductular and canalicular and prominent ballooning, and fatty change in hepatocytes. There is ductular reaction with pericholangitis with evolving bridging fibrosis and centrilobular fibrosis. In conclusion, liver biopsy is not warranted for diagnosis of DJS and is done only when there is clinical overlap with other conditions or if coproporphyrin excretion studies are not available (2). A biopsy diagnostic of DJS with or without the pigment is now possible by performing cMOAT staining. Genetic testing is now the preferred modality for diagnosis.
References
157
1. Knisely AS. Hepatocellular and familial cholestasis. In: Russo P, Ruchelli E, Piccoli DA, eds. Pathology of pediatric gastrointestinal and liver disease. 1st ed. New York, NY: Springer-Verlag;2004:237–250. 2. Frank M, Doss MO. Relevance of urinary coproporphyrin isomers in hereditary hyperbilirubinemias. Clin Biochem. 1989;22:221–222.
This page intentionally left blank
10 Neonatal Cholestatic Liver Disease GRACE E. KIM AND LINDA D. FERRELL
I N T ROD U C T I ON
TA BL E 1 0 . 2 Neonatal hepatitis—associated etiologies
The differential diagnosis of neonatal jaundice is lengthy. Many normal-term newborns have transient neonatal jaundice with elevated unconjugated bilirubin. The unconjugated hyperbilirubinemia in some neonates results from an immature hepatic enzyme glucuronosyl transferase activity or can be associated with breast-feeding. Other etiologies include hemolysis, sepsis, hypothyroidism, or pyloric stenosis, and inherited disorders such as Crigler-Najjar and Gilbert syndromes (see Chapter 9). In contrast, conjugated hyperbilirubinemia nearly always reflects hepatic dysfunction. The current practice is to investigate jaundice in any infant who is more than 14 days old to determine whether the hyperbilirubinemia is unconjugated or conjugated. Unfortunately, jaundiced infants often escape clinical attention until the first well-baby examination at 6 to 8 weeks of age. For neonatal jaundice that represents hepatic dysfunction, one of the most frequent entities that presents within the first 2 months of life with conjugated hyperbilirubinemia is biliary atresia (BA), which typically causes an obstructive pattern of injury with a prominent portal-based fibrosis and ductular reaction in addition to bile stasis. This pattern is not specific for BA, and many of the other lesions with this pattern are listed in Table 10.1. The other most common pattern of injury in this age group is neonatal hepatitis (NH), which in contrast has a more lobular prominence of injury, with proportionately less portal-based injury and ductular reaction. As with the obstructive pattern, the neonatal hepatitic pattern also can be caused by a long list of entities (Table 10.2). The obstructive and hepatitic patterns can have overlapping features, and liver biopsy is often performed to establish the TA B LE 10. 1 Obstructive-type pattern of ductular reaction—
associated etiologies Biliary atresia Total parenteral nutrition effect Cholestasis-associated sepsis Inspissated bile syndrome, as in cystic fibrosis, prematurity, congenital heart disease, starvation, dehydration, diuretic therapy, congenital anomalies of the biliary tract, gut dysfunction or ileal resection, ABO incompatibility/hemolytic anemia, idiopathic (3,4) Infections, including CMV Alpha-1-antitrypsin deficiency Paucity of bile ducts, some nonsyndromatic types, and rarely in Alagille syndrome Abbreviation: CMV, cytomegalovirus.
Idiopathic forms (most common) Hypopituitarism, including septo-optic dysplasia Infectious Hepatotropic viruses: A, B, C Herpes viruses: CMV, HSV, varicella Other viruses: rubella, reovirus-3, ECHO, Coxsackie, adenovirus, parvovirus B19, HIV Bacteria: Listeria monocytogene, Mycobacterium tuberculosis,Treponema pallidum Protozoa: Toxoplasma gondii Metabolic Progressive familial intrahepatic cholestasis Alpha-1-antitrypsin deficiency Niemann-Pick disease type-C Cystic fibrosis Bile acid synthesis defects Zellweger syndrome Tyrosinemia Immunologic Autoimmune hepatitis Severe combined immunodeficiency Abbreviations: CMV, cytomegalovirus; ECHO, enteric cytopathic human orphan; HSV, herpes simplex virus; HIV, human immunodeficiency virus.
diagnosis. This distinction is important since early surgical intervention is required in BA (1). Neonatal hemochromatosis, one of the most common causes of liver failure at birth, is discussed with iron overload in Case 20.6. BILIA RY AT R ESIA
BA is an idiopathic, necroinflammatory process of the bile ducts that leads to periductal fibrosis and obliteration and secondary biliary cirrhosis. This entity typically presents within the first 2 months of life, but earlier and later presentations can occur. Numerous etiologic factors, including congenital, infectious, immunologic, vascular, and toxic have been proposed; however, the cause of BA remains unknown, and multiple factors resulting in the same outcome may be involved. About 20% of the cases are associated with other congenital anomalies, such as polysplenia, situs inversus, genetic trisomies, vascular anomalies, and perhaps maternal diabetes (2). In addition, numerous other entities can mimic the obstructive pattern of injury seen in BA (see Table 10.1) (3,4). Thus, correlating the biopsy findings with clinical setting, laboratory studies, intraoperative cholangiogram, and other biliary tree imaging results is needed. The reported accuracy of the liver 159
160
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
biopsy to determine an obstructive form of neonatal cholestasis approaches 95% when 5 to 7 portal tracts are present. Histology
Typically, the liver shows morphologic signs of biliary obstruction with cholestasis, ductular reaction, and portal edema or fibrosis. Greater than 90% of the cases involve obliteration of the hilar ducts at the porta hepatis. The remaining cases show patency of the large ducts to the levels of the common hepatic duct or common bile duct (2). BA is a dynamic process, and the histology transforms with the time course of the disease. During the neonate’s first 4 to 6 weeks of life, nonspecific features of cholestasis including swollen hepatocytes are present. Ductular reaction, the most reliable criterion for diagnosing biliary obstruction, develops around 6 to 8 weeks of age (Figure 10.1). During this time, the portal tracts may show a variable density of lymphocytes and neutrophils, damaged interlobular duct epithelium, similar to that observed in the extrahepatic ducts, and portal edema or early portalbased fibrosis. At about 8 weeks of age, the periportal fibrosis begins or may have progressed into portal-to-portal bridging fibrosis. The amount of portal inflammation may decrease and damaged interlobular ducts can have concentric fibrosis. Cholestasis persists in the parenchyma with pigment-laden Kupffer cells. Eventually, secondary biliary cirrhosis, the socalled jigsaw pattern of cirrhosis, forms at 1 to 6 months of age with a subset developing loss of interlobular ducts. The suggested timeline can be quite variable, and at least some cases described as perinatal, or neonatal sclerosing cholangitis may represent late onset BA. At the time of hepatoportoenterostomy, a segment of the porta hepatis and gallbladder, if present, is removed. Previously, a bile duct diameter of more than 150 μm within the ductal remnant within the portal hepatis was critical in
LIVER
DISEASE
determining postoperative bile flow. Subsequent studies have refuted this criterion, and, hence, surgeons no longer request frozen sections on bile duct remnant. The ductular structures in the portal hepatitis typically demonstrate a destructive fibro-obliterative process, often with intraductal inflammatory or cholestatic changes. The periductal fibrotic reaction typically has increased fibroblastic changes. NEO NATA L H EPAT IT IS
The other major pattern of liver injury in the newborn period is NH, which is an intrahepatic cholestatic disease, typically diagnosed within 2 months after birth. Most NH cases are idiopathic, but some cases have been linked to hypopituitarism (5), including the entity of septo-optic dysplasia, a syndrome with congenital hypoplasia of the optic nerve, absent septum pellucidum, and hypopituitarism. Other associated lesions are noted in Table 10.2 (6). Histologic features are not thought to distinguish the various possible etiologies of NH but can often differentiate NH from BA (Table 10.3). Rarely, BA and Alagille syndrome can also present with this pattern of injury. However, typical features like portal-based ductular reaction and ductopenia respectively predominate as histologic findings later in the clinical course. The clinical outcome in NH is often dependant on the etiology. Most idiopathic or infectious forms resolve without significant liver injury with appropriate therapy and supportive care. Histology
NH is also known as neonatal giant cell hepatitis because of the frequent finding of giant cell transformation in the lobular hepatocytes (Figures 10.1 and 10.2). These “giant cells” usually have greater than 4 nuclei that are often centrally clustered. Another almost universal feature is the presence of extramedullary hematopoiesis (EMH) in portal and lobular areas. Lobular hepatocyte necrosis, evidence of previous TA BL E 1 0 . 3 Comparison of histologic features of biliary atresia
and neonatal hepatitis
FIGURE 10. 1 Biliary atresia pattern of portal ductular reaction with fibrosis. A few giant multinucleate hepatocytes are present, but these are not a prominent pattern in most cases.
Histologic Features Biliary Atresia
Neonatal Hepatitis
Giant cell change
Focal
Diffuse
Lobular inflammation
Absent
Variable, can be absent
Hepatocyte loss
Rare
Present as focal necrosis, dropout, or intraparenchymal pericellular fibrosis
Fibrosis pattern
Portal-based, early feature
Intraparenchymal pericellular, variable periportal fibrosis
Ductular reaction
Present, becomes more prominent over time
Rare
Extramedullary hematopoiesis
Present in variable degrees
Typically present and often prominent
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
LIVER
DISEASE
161
TA BL E 1 0 . 4 Paucity of bile ducts in the neonatal period—
associated disorders Inheritable disorders Alagille syndrome Alpha-1-antitrypsin deficiency Zellweger syndrome (lack of peroxisomes) Progressive familial intrahepatic cholestatic defects type 1 and/or 2 Chromosomal defects including Turner syndrome, Trisomy 17-18, 21 Infections Congenital rubella Congenital CMV Abbreviation: CMV, cytomegalovirus.
FIGURE 10.2 Neonatal hepatitis with giant cell change, mild lobular
inflammation, and no significant ductular reaction. Focal cholestasis is present. Some hepatocytes are smaller than normal, and plates are wider than 2 cells thick, consistent with regenerative change.
FIGURE 10. 3 Neonatal hepatitis—follow-up biopsy 12 weeks later at age 5 months (after resolution of symptoms and enzymatic changes) demonstrates a few residual multinucleate hepatocytes.
hepatocyte dropout, or presence of pericellular fibrosis as a sign of lobular injury is also typically present. The degree of inflammatory infiltrates other than EMH can be variable and is scant in many cases. Hepatocanalicular bile stasis is common. Portal tracts have bile ducts, but these may be smaller than expected (hypoplastic) and difficult to identify on routine staining. Cytokeratin (CK)7 or CK19 immunostaining may be helpful to identify the ducts in order to differentiate NH from diseases related to paucity of bile ducts. Mild ductular reaction can be present. The giant multinucleate hepatocytes can persist for some months after symptoms have resolved (Figure 10.3). PAU C I T Y OF B I L E D U C T S
In the pediatric neonatal setting, paucity of bile ducts can be typically divided into 2 types: the syndromatic type (Alagille
syndrome, also known as arteriohepatic dyplasia), or nonsyndromatic types, associated with a long list of entities, some of which are noted in Table 10.4 (7,8). Paucity of ducts is defined by evaluation of the ratio of bile ducts per portal tract, with normal range from 0.9 to 1.8 per portal tract, and paucity as less than 0.5 ducts per portal tract with at least 10 portal tracts for evaluation. Values between 0.5 and 0.8 are considered borderline or nondiagnostic. Alagille syndrome is a variably penetrant, autosomal dominant disorder that typically has 5 major components. These consist of the paucity of ducts as the liver manifestation, associated with other extrahepatic abnormalities including peripheral pulmonary artery stenosis as the cardiac component, typical facial appearances, butterfly-shaped vertebral arch defects, and ocular defects; the latter 2 entities having no functional significance (9). This disorder is caused by a mutation in JAG 1 (JAGGED 1), resulting in overexpression of human growth factor (HGF), which in turn may have an effect on hepatic stem cells, pushing them toward hepatocytic rather than biliary differentiation (10). In addition to paucity of bile ducts, there may be hypoplasia of the gallbladder, common bile duct, and portal vein. Destructive lesions of the bile ducts, bile stasis, and giant cell transformation are often seen in the early stages. Ductular reaction and periportal fibrosis are not typical features but can rarely occur. In addition, a hypoplastic common bile duct might mimic the large duct lesions of BA on imaging. Of the nonsyndromatic forms, Zellweger syndrome (11), an autosomal recessive defect characterized by lack of peroxisomes and defect in bile acid synthesis, alpha-1-antitrypsin deficiency (AATD) (see below), and cystic fibrosis are other rare genetic forms of ductopenia in early infancy (6,7). A LP H A - 1- A NT IT RY P SIN DEFICIENCY
AATD is one of the more common inheritable defects that can present in the neonatal period but may be difficult to diagnose by routine histology as the typical (periodic acid–Schiff diastase) PASd+ cytoplasmic globules are almost always not yet present in a diagnostic pattern at this young age. Fortunately, the deposits of AAT in distended endoplasmic reticulum can be usually noted on electron microscopic examination, and immunohistochemistry for AAT may demonstrate granular
162
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
cytoplasmic positivity in hepatocytes (6). However, in this age group, the diagnosis is often more one of clinical suspicion and testing for serum AAT and phenotyping (protease inhibitor, or PI typing) for the abnormal alleles. Typically, only the patients with PiZZ phenotype would have a severe enough defect, if fully penetrant, to present in the neonatal period, as the other SZ and MZ phenotypes tend to present later in life if they are associated with high penetrance and fibrotic changes. The spectrum of changes in AATD are more fully discussed in Chapter 16, but for the purpose of this chapter on neonatal cholestasis, AAT can present with either an NH-like pattern of injury, or an obstructive pattern more like that of BA, and an element of ductopenia can occur with either pattern of injury. Progressive Familial Intrahepatic Cholestasis
Progressive familial intrahepatic cholestasis (PFIC) was initially categorized into 3 clinical diseases as PFIC1, PFIC2, and PFIC3. Recent molecular advances have added to our knowledge of these diseases. The involved inherited defects for some of the forms of PFIC are known, and this is reflected in the nomenclature. PFIC1 is due to ATP8B1 mutation encoding familial intrahepatic cholestasis 1 (FIC1), PFIC2 has mutations on ABCB11 that encodes bile salt export pump (BSEP), and PFIC3 is associated with a defect in the multidrugresistance-3 (MDR3) pathway on the encoding gene ABCB4. Hence PFIC1, PFIC2, and PFIC 3 are also termed FIC1 deficiency, BSEP deficiency, MDR3 deficiency, respectively. Based on the specific gene defect, the spectrum of these diseases can vary from mild to severe. PFIC1 and PFIC2 are characterized by normal serum gamma-glutamyl transferase (GGT) activity, which can be differentiated from the higher levels typically found in PFIC3. PFIC2 patients have been found to have a higher risk for neonatal jaundice than PFIC1 and may thus present with an NH-like pattern of injury with high serum alanine aminotransferase and elevated serum alpha-fetoprotein levels (12). A recent study identified that at presentation, serum aminotransferase and bile salt levels were higher in PFIC2, whereas serum alkaline phosphatase levels were higher and serum albumin levels were lower in PFIC1 (13). Morphologically, bland canalicular cholestasis between compact hepatocytes and small multinucleated hepatocytes with increasing portal fibrosis are characteristic features of PFIC1. Although PFIC2 can also demonstrate fibrosis, this typically begins in the pericentral regions. An NH pattern of injury with giant cell transformation can be observed. Some patients with PFIC1 or PFIC2 have paucity of ducts (14). PFIC3 displays a biliary pattern of injury depicted as expanded portal tracts with ductular reaction and mixed inflammatory infiltrate. Bile plugs can be present in ductules and eventual biliary cirrhosis can develop. Bile Acid Synthesis Defects
Bile acid synthesis defects can be due to a variety of enzymatic defects. Table 10.5 contains a short list of some of
LIVER
DISEASE
TA BL E 1 0 . 5 Bile acid defects: common histologic patterns Bile Acid Deficiency
Patterns of Histology
Oxysterol 7␣-hydroxylase
NH as well as ductular reaction, early progression to cirrhosis
3-OH steroid dehydrogenase
NH, progresses to cirrhosis
5-Reductase
NH, with progression to liver failure; neonatal hemochromatosis-like pattern
Mitochondrial sterol 27-hydroxylase (cerebrotendinous xanthomatosis)
NH as rare presentation, but neurologic dysfunction and xanthomas typically present later in life
C-27 peroxisomal side-chain oxidation (Zellweger syndrome)
NH, also with paucity of ducts and fibrosis EM: absence of perioxisomes
Acyl-CoA racemase (AMACR)
Similarities to Zellweger syndrome, with decreased size and number of perioxisomes
Abbreviations: EM, erythema multiforme; NH, neonatal hepatitis. From Refs. 11, 15.
the enzymatic defects and common histologic patterns that are typically present (11,15,16). A specific diagnosis for individual defects is important, as the efficacy of therapy with oral bile acid therapy is dependant on type of defect present (11,16). This chapter will not attempt to illustrate these entities, but excellent reviews by Bove et al (11), Heubi et al (15), and Sundaram et al (16) provide more detailed descriptions. Most of these lesions present in the neonatal period, with neonatal giant cell hepatitis being the most common pattern of injury. Some cases may demonstrate ductular reaction, and/or progression to cirrhosis at variable rates. Zellweger syndrome, which is characterized by absence of peroxisomes on electron microscopic examination, is one of the bile acid defects that can present in this period and progress to cirrhosis. This entity is also one of the few bile acid synthesis defects that can demonstrate duct paucity.
References 1. Mieli-Vergani G, Howard ER, Portman B, Mowat AP, et al. Late referral for biliary atresia: missed opportunities for effective surgery. Lancet. 1989;1:412–423. 2. Hartley JL, Kelly DA, Davenport M. Biliary atresia. Lancet. 2009;374:1704–1713. 3. Miloh T, Rosenberg HK, Kochin I, Kerkar N. Inspissated bile syndrome in a neonate treated with cefotaxime, sonographic aid to diagnosis, management, and follow-up. J Ultrasound Med. 2009;28:541–544. 4. Brown DM. Bile plug syndrome: successful management with a mucolytic agent. J Ped Surg. 1990;25:351–352. 5. Spray CH, McKiernan P, Waldron KE, Shaw N, Kirk J, Kelly DA. Investigation and outcome of neonatal hepatitis in infants with hypopituitarism. Acta Paediatr. 2000;89:951–954. 6. Jevon GP, Dimmick JE. Histopathologic approach to metabolic liver disease: Part 1. Pediatr Dev Pathol. 1998;1:179–199. 7. Kahn E, Daum F, Markowitz J, Teichberg S, Duffy L, Harper R, Aiges H. Nonsyndromatic paucity of interlobular bile ducts: light and electron
CHAPTER
8. 9. 10.
11.
12.
10:
N E O N ATA L
C H O L E S TAT I C
microscopic evaluation of sequential liver biopsies in early childhood. Hepatology. 1986;6:890–901. Dimmick JE. Intrahepatic bile duct paucity and cytomegalovirus infection. Pediatr Pathol. 1993;13(6):847–852. Hadchouel, M. Alagille syndrome. Indian J Pediatr. 2002;69:815-818. Yuan A, Kobayashi N, Kohsaka T. Human Jagged 1 mutants cause liver defect in Alagille syndrome by overexpression of hepatocyte growth factor. J Molecular Biol. 2006;356:559–568. Bove KE, Heubi JE, Balistreri WF, Setchell KD. Bile acid synthetic defects and liver disease: a comprehensive review. Pediatr Dev Pathol. 2004;7:315–334. Davit-Spraul A, Fabre M, Branchereau S, et al. ATP8B1 and ABCB11 Analysis in 62 children with normal gamma-glutamyl transferase progressive familial intrahepatic cholestasis (PFIC): phenotypic differences
13.
14.
15. 16.
LIVER
DISEASE
163
between PFIC1 and PFIC2 and natural history. Hepatology. 2010;51(5):1645–1655. Pawlikowska L, Strautnieks S, Jankowska I, et al. Differences in presentation and progression between severe FIC1 and BSEP deficiencies. J Hepatol. 2010;13. [Epub ahead of print]) Naveh Y, Bassan L, Rosenthal E, et al. Progressive familial intrahepatic cholestasis among the Arab population in Israel. J Pediatr Gastroenterol Nutr. 1997;24:548–554. Heubi JE, Setchell KD, Bove KE. Inborn errors of bile acid metabolism. Semin Liver Dis. 2007;3(27):282–294. Sundaram SS, Bove KE, Lovell MA, Sokol RJ. Mechanisms of disease: inborn error of bile acid synthesis. Nat Clin Pract Gastroenterol Hepatol. 2008;5:456–468.
Case 10.1
Biliary Atresia GRACE E. KIM AND LINDA D. FERRELL
C L I N I C AL I N F OR M AT I ON
A 5-week-old baby boy presented with pale stool, dark urine, delayed growth, and jaundice. He was a full-term infant with normal delivery. Laboratory data showed elevated alkaline phosphatase at 649 U/L (normal 110-302), gamma-glutamyl transferase of 616 U/L (normal 7-71), total bilirubin of 10.9 mg/dL with conjugated bilirubinemia of 6.2 mg/dL, and elevated transaminases at 3 to 4 times normal. Serum albumin, prothrombin time, alpha-1-antitrypsin levels, thyroidstimulating hormone (TSH), and free T4 were normal. Family history was noncontributory. Ultrasound showed a normalsized liver and spleen, with no biliary tract dilation or other anomalies, but the gallbladder was not visualized. HIDA scan (hepatobiliary [scintigraphy] imino-diacetic acid) showed no excretion of radiotracer (as a marker for bile excretion) into the duodenum. A liver biopsy was done. F I G U R E 1 0 . 1 . 1 Biliary atresia at age 6 weeks. Fibrosis around residual
R E A S ON F OR R E F E R R A L
To determine the cause of neonatal jaundice, the primary concern being biliary atresia. PAT H OL OG I C F E AT U R E S
The liver biopsy at 6 weeks of age demonstrated periportal and porto-portal bridging fibrosis. Ductular reaction was prominent in the fibrous bands, with many of the ductules closely approximating the edge of the fibrotic zone near the residual hepatic parenchyma; these ductules often contained bile (Figure 10.1) (Figures 10.1.1 and 10.1.2). The ductules also had a somewhat anastomosing appearance (Figure 10.1.2). The liver parenchyma was nodular; bile plugs were also noted. A few scattered giant, multinucleated hepatocytes were present, but this was not a prominent feature. Hepatocytic acinar change was focally noted (as a feature of chronic cholestasis), but there was no significant hepatocellular necrosis (Figure 10.1.2). Lobular and portal inflammation was scant other than some focal periductular mixed inflammatory reaction (Figure 10.1.1). A few foci of extramedullary hematopoiesis were also present. Based on these findings, the patient had an intraoperative cholangiogram, followed by a Kasai procedure. The intraoperative cholangiogram confirmed complete blockage of the ductal system and only minimal luminal patency of an atretic gallbladder. During Kasai portoenterostomy, the portal plate appeared very fibrotic with no visible bile duct. No drainage was noted into the bowel after surgery, and living-related donor liver transplantation was successfully completed at age 5 months. The porta hepatis was excised during Kasai portoenterostomy demonstrated complete obliteration of the large bile duct, confirming the findings of the intraoperative cholangiogram.
portal areas is prominent and contains a prominent ductular reaction. Note bile stasis in some of the ductules. The parenchyma shows a nodular appearance. Hepatocytes are crowded, some are smaller than normal and arranged in an acinar pattern, all of which are regenerative changes related to cholestasis. Inflammatory infiltrates are scant.
F I G U R E 1 0 . 1 . 2 Biliary atresia at age 6 weeks with fibrosis and ductu-
lar reaction (higher magnification of Figure 10.1). The parenchyma demonstrates a few multinucleate hepatocytes.
Histologically, the ducts were small, with minimal residual lumina, surrounded by a fibroblastic reaction and mild inflammatory changes. The lumina contained a few residual epithelial cells (Figure 10.1.3). At age 5 months (21 weeks), the liver explant, obtained during a living-related donor
164
CASE
10.1:
BILIARY
FIGURE 10. 1. 3 Portal hepatitis at age 7 weeks. The ducts are essentially obliterated by fibroblastic and mild inflammatory reaction.
FIGURE 10. 1. 4 Explanted liver at age 5 months demonstrates fully developed cirrhosis and very prominent ductular reaction as highlighted on cytokeratin 7 (CK7) immunoperoxidase staining.
transplantation, demonstrated cirrhosis with biliary pattern of injury (Figure 10.1.4).
D I AG N OS I S
Biliary obstructive pattern, consistent with biliary atresia.
D I S C U S S I ON
In this case, the histology was consistent with BA, demonstrating the typical portal-based fibrosis with ductular reaction. These numerous ductules with anastomosing, chain-like pattern may have a ductal plate malformation–like appearance. This feature has been proposed to be a marker of congenital
AT R E S I A
165
or fetal form of biliary atresia that is frequently associated with congenital anomalies, most commonly biliary atresia splenic malformation syndrome. This has led to the suggestion that the etiology of biliary atresia dates to fetal life (1) and is an indicator of poor prognosis (2). A recent study (3) noted the challenges on what features actually constituted this ductal plate malformation–like appearance and did not conclude these features identified the fetal form of biliary atresia. This finding is not specific for BA. Thus, correlation with other clinical features, particularly the conjugated bilirubinemia, preoperative radiographic imaging to exclude other biliary tree anomalies, duct dilation, and intraoperative cholangiogram to confirm the obliterative changes of the hilar ducts, and absence of gallbladder are very important. Untreated BA is typically fatal within 2 years, typically as a result of biliary cirrhosis and hepatic failure. The first step in the treatment remains hepatoportoenterostomy (Kasai procedure) early in the course of the condition, usually before 100 days of birth, and before prolonged complete obstruction leading to extensive liver fibrosis and/or cirrhosis (4). The next step is liver transplantation; BA represents more than 50% of pediatric liver transplants. Response to the Kasai procedure (as defined by normal bilirubin levels by 6 months after surgery) can result in prolonged survival of the native liver, which can postpone liver transplantation (5–7). Approximately 25% to 35% of patients with successful hepatoportoenterostomy survive more than 10 years without liver transplantation. One-third of the patients drain bile but develop complications of cirrhosis and require transplantation before age 10. In the remaining one third of the patients, bile flow is inadequate following heptoportoenterostomy leading to progressive fibrosis. These children usually die by age 2, if liver transplantation is not performed. A common complication after the Kasai procedure is recurrent cholangitis resulting from direct biliary-enteric connection (7,8). Essentially all patients develop progressive fibrosis and cirrhosis, leading to complications of portal hypertension including esophageal varices and ascites, and increased risk for hepatocellular carcinoma, and, rarely, hepatoblastoma or cholangiocarcinoma (9,10). Hepatopulmonary syndrome, a state of hypoxia probably due to unmetabolized vasoactive substances causing the shunting of blood away from the pulmonary vasculature, can also occur (4). Another feature often seen in BA and present in this case is extramedullary hematopoeisis. This is a common finding in many forms of neonatal liver diseases (particularly in neonatal hepatitis), and is thought to be a nonspecific finding related to stress in the intrauterine and neonatal period.
References 1. Desmet VJ. Ludwig symposium on biliary disorders—Part 1. Pathogenesis of ductal plate abnormalities. Mayo Clin Proc. 1998;73:80–89. 2. Low Y, Vijayan V, Tan CE. The prognostic value of ductal plate malformation and other histologic parameters in biliary atresia: an immunohistochemical study. J Pediatr. 2001;139(2):320–322.
166
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
3. Pacheco MC, Campbell KM, Bove KE. Ductal plate malformation-like arrays in early explants after a Kasai procedure are independent of splenic malformation complex (heterotaxy). Pediatr Dev Pathol. 2009;12(5): 355–360. 4. Hartley JL, Kelly DA, Davenport M. Biliary atresia. Lancet. 2009;374:1704–1713. 5. Lykavieris P, Chardot C, Sokhn M, Gauthier F, Valayer J, Bernard O. Outcome in adulthood of biliary atresia: a study of 63 patients who survived for over 20 years with their native liver. Hepatology. 2005;41: 366–371. 6. Shinkai M, Ohhama Y, Take H, et al. Long-term outcome of children with biliary atresia who were not transplanted after the Kasai operation: >20-year experience at a children’s hospital. J Pediatr Gastroenterol Nutr. 2009;48:443–450.
LIVER
DISEASE
7. Houben C, Phelan S, Davenport M. Late-presenting cholangitis and Roux loop obstruction after Kasai portoenterostomy for biliary atresia. J Pediatr Surg. 2006;41:1159–1164. 8. Wu ET, Chen HL, Ni YH, et al. Bacterial cholangitis in patients with biliary atresia: impact on short-term outcome. Pediatr Surg Int. 2001;17:390–395. 9. Tatekawa Y, Asonuma K, Uemoto S, Inomata Y, Tanaka K. Liver transplantation for biliary atresia associated with malignant hepatic tumors. J Pediatr Surg. 2001;36:436–439. 10. Kulkarni PB, Beatty E Jr. Cholangiocarcinoma associated with biliary cirrhosis due to congenital biliary atresia. Am J Dis Child. 1977;131: 442–444.
Case 10.2
Neonatal Hepatitis With Hypopituitarism GRACE E. KIM AND LINDA D. FERRELL
C L I N IC AL I N F OR M AT I ON
PAT H O LO GIC FEAT UR ES
A full-term neonate born to a 16-year-old mother had an uncomplicated prenatal course and normal vaginal delivery, other than a hypoglycemic episode that was successfully treated with dextrose infusion. The infant was readmitted at age 8 weeks with jaundice. Ultrasound demonstrated a small gallbladder, and a liver biopsy was done (first liver biopsy, see section on pathologic features), followed by Kasai hepatoportoenterostomy for probable BA. The patient was transferred to the referral center for further testing. On admission, the patient appeared to have some hypotonia and hyperreflexia. Laboratory values included anemia, thrombocytopenia, aspartate aminotransferase (AST) and alanine aminotransferase (ALT) of 3 and 2 times normal respectively, total bilirubin 5.7 mg/dL with direct bilirubin 4.1 mg/dL. Other findings included negative markers for viral hepatitis A, B, C, as well as autoimmune hepatitis, and normal serum alpha-1-antitrypsin Pi typing. Ultrasound of the liver demonstrated a normal-sized liver with normal intrahepatic bile ducts, portal and hepatic arteries, and moderate ascites. In addition, an enlarged spleen was noted, but no gallbladder was visualized. Because of the abnormal neurologic examination, an ultrasound of the head was done and demonstrated normal ventricles with no evidence of hemorrhage, but no definitive septum pellucidum was visualized and the corpus callosum was thin. Ophthalmologic consultation revealed evidence of bilateral optic nerve hypoplasia. Magnetic resonance imaging (MRI) confirmed hypoplastic anterior pituitary, with ectopic posterior pituitary in the tuber cinereum. The optic chiasm and optic nerve appeared small. The patient had hypoglycemia, and other pituitary function hormones such as adrenocorticotropic hormone (ACTH), growth hormone, and thyroid stimulating hormone (TSH) were low. The patient was treated for hypopituitarism. Initially, the patient did well but then became increasingly jaundiced, with elevations of total and direct bilirubins. At age 7 months, ultrasound of the liver revealed dilated intrahepatic bile ducts with distal obstruction at the hepatoportoenterostomy site, and the patient was taken to surgery for revision of the anastomosis. A liver wedge biopsy was taken at that time (second biopsy, see section on Pathologic Features.). After that revision, the jaundice resolved and the patient was in a stable condition 1 year later.
Two liver biopsies were obtained, one at age 9 weeks and the other at 8 months of age. First biopsy, age 9 weeks: This liver biopsy demonstrated features of mild hepatitis with expanded portal zones, focal swollen hepatocytes as well as scattered necrotic hepatocytes in the lobules, canalicular bile stasis, and a mild increase in sinusoidal cells (Figures 10.2.1–10.2.5). The cellular infiltrates in the portal areas and lobules consisted predominantly of EMH, with a few portal areas containing prominent eosinophilic or megakaryocytic components (Figures 10.2.4 and 10.2.5). No periportal fibrosis or ductular reaction was present and ducts were intact and present (Figures 10.2.6 and 10.2.7). No prominent giant, multinucleate cell transformation of hepatocytes was present (Figure 10.2.3). No porta hepatic sample was available for review. Second biopsy, age 8 months: This liver biopsy was a wedge biopsy performed at the time of revision of the Kasai procedure deemed necessary due to poor drainage of the hepatoportoenterostomy. The liver now had extensive portalbased fibrosis with prominent ductular reaction, a pattern typical of bile duct obstruction, as could be seen in BA or other forms of mechanical obstruction (Figures 10.2.8–10.2.10). No hepatitic changes were noted, no multinucleate giant hepatocytes of any significant number were seen, and the interlobular bile ducts were intact, without evidence of ductopenia on CK7 staining (Figure 10.2.10). Cholestasis as evidenced by bile plugs was easily identified (Figure 10.2.9).
R E A SON F OR R E F E R R AL
This patient was referred with the diagnosis of neonatal hepatitis–like syndrome, status-post Kasai procedure (hepatoportoenterostomy) for possible BA.
F I G U R E 1 0 . 2 . 1 First liver needle biopsy. Two portal tracts contain a cellular infiltrate, with a duct visible in the larger portal zone. The parenchyma contains a mild cellular infiltrate and shows mild hepatocellular disarray.
167
168
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
FIGURE 10.2.2 First liver biopsy. A small canalicular bile plug is noted directly below the small central vein in this figure. A couple of clusters of EMH are present in sinusoids above the vein. The hepatocytes do not show giant cell change, but have focal pale cytoplasm and swelling (just right of EMH) as well as appear to be somewhat smaller than normal, the latter a possible result of regenerative change.
FIGURE 10. 2. 3 First liver biopsy. A necrotic hepatocyte (apoptotic
body) is noted (center), and acinar change is present (lower center), the latter a feature commonly seen in cholestatic disease. A mild cellular infiltrate is present in sinusoids.
D I AG N OSI S
Liver, needle biopsy: neonatal hepatitis, mild, consistent with hypopituitarism and septo-optic dysplasia. Liver, wedge biopsy (8 months): Biliary pattern of cirrhosis, likely due to obstruction. D I SC U SSI ON
This case demonstrates a neonatal hepatitis pattern of injury but does not have the typical feature of multinucleate hepatocytes, reinforcing the idea that this feature is not necessary for
LIVER
DISEASE
F I G U R E 1 0 . 2 . 4 First liver biopsy. This portal zone is mildly expanded
by a dense cellular infiltrate, most of which is EMH, with prominent eosinophilia. An artery is present at the lower part of the portal zone, with the duct in upper right, but somewhat obscured by the infiltrate.
F I G U R E 1 0 . 2 . 5 First liver biopsy. This portal zone contains prominent megakaryocytes, and EMH is also seen in adjacent periportal sinusoids. An apoptotic body (necrotic hepatocyte) is also present (directly below megakaryocytes in lower center).
the diagnosis, and is a nonspecific manifestation of hepatocellular injury. The diagnostic features include lobular hepatocyte necrosis and swelling, canalicular cholestasis. EMH is a common finding in this setting and is prominent in this case. All 3 hematopoietic cell lines can be seen, but 1 cell line might predominate. EMH should not be confused with lymphocytes, the blasts of a leukemic infiltrate, or other malignancies. Hypopituitarism is one of the more common associated etiologies for hepatitis presenting in the neonate (1), in particular, septo-optic dysplasia spectrum of changes including optic nerve hypoplasia and absence of the septum pellucidum, but more commonly, hypoglycemia and evidence of endocrine failure, including low TSH and free T4, serum cortisol, and human growth hormone. The hepatitis is thought to be due to low
CASE
10.2:
N E O N ATA L
H E PAT I T I S
WITH
H Y P O P I T U I TA R I S M
169
F I G U R E 1 0 . 2 . 9 Second liver biopsy. Higher magnification demonFIGURE 10. 2. 6 First liver biopsy, higher magnification of Figure
strates cholestatic changes.
10.2.8. The portal zone contains mostly EMH, and the duct is intact. The hepatocytes show focal hepatocyte swelling and pale cytoplasm.
F I G U R E 1 0 . 2 . 1 0 Second liver biopsy. Cytokeratin 7 immunostain confirms a prominent ductular reaction and presence of an interlobular bile duct. FIGURE 10. 2. 7 First liver biopsy, trichrome stain. No periportal
fibrosis has occurred, and the bile duct is present.
cortisol and/or growth hormone levels, and appropriate hormone replacement therapy will almost always lead to resolution of the cholestasis and hepatitis if the lesion is diagnosed in early infancy. Thus, it is important to follow any hypoglycemic episodes in childhood associated with jaundice with examination for pituitary function, which is often done by monitoring serum TSH and free T4 and/or early morning serum cortisol. If the diagnosis is made at a later stage, liver fibrosis may be present. Other noninfectious causes of a mild hepatitic or a neonatal hepatitic pattern with giant cells (see Table 10.2) include alpha-1-antitrypsin deficiency (Case 10.4) and rare entities such as progressive familial intrahepatic cholestatic disorders (types 1–3) and bile salt defects, described in the introduction and in Table 10.2.
Reference FIGURE 10. 2. 8 Second liver biopsy, wedge taken at time of revision of hepatoportoenterostomy. Prominent portal-based fibrosis with ductular reaction is present.
1. Spray CH, McKiernan P, Waldron KE, Shaw N, Kirk J, Kelly DA. Investigation and outcome of neonatal hepatitis in infants with hypopituitarism. Acta Paediatr. 2000;89:951–954.
Case 10.3
Paucity of Intrahepatic Bile Ducts GRACE E. KIM AND LINDA D. FERRELL
C L I N I C AL I N F OR M AT I ON
This neonate presented at 12 days of age with direct hyperbilirubinemia. Ultrasound of the abdomen revealed normal-sized liver and spleen with no bile duct dilation or other anomalies, and a normal gallbladder. HIDA scan showed no biliary excretion of radiotracer into the bowel and was interpreted as consistent with either BA or severe neonatal hepatitis. Laboratory examination showed total bilirubin of 7.1 mg/dL (direct 4.4 mg/dL), elevated alkaline phosphatase of 537 U/L, gammaglutamyl transferase of more than 1000 U/L, and mild elevation of transaminases. A liver biopsy was performed at age 8 weeks. During this time period, echocardiography revealed bilateral branch pulmonary vascular stenosis, and a bone survey showed butterfly vertebral bodies at the level of T6-8. R E A S ON F OR R E F E R R A L
The patient was referred for evaluation for persisting neonatal jaundice and concerns for biliary atresia.
F I G U R E 1 0 . 3 . 2 This smaller portal zone contains a small hepatic arteriole and portal venule, as well as mild cellular infiltrate. Acinar structures and some crowding of hepatocyte nuclei are present in adjacent liver parenchyma.
PAT H OL OG I C F E AT U R E S
The liver biopsy shows 12 portal zones, 8 of which lacked an interlobular bile duct, confirming ductopenia. In addition, there is lobular canalicular cholestasis with acinar transformation of hepatocytes (Figures 10.3.1–10.3.4). No significant amount of giant cell transformation was present. The CK7 stain highlighted periportal hepatocyte as well as minimal ductular reaction (Figure 10.3.4) The portal zones did not show significant inflammation or fibrosis (Figures 10.3.1 and 10.3.2). EMH was present but was not prominent (Figure 10.3.3).
F I G U R E 1 0 . 3 . 3 A small central vein is present on the left. Note the
prominent small acinar formation by hepatocytes, and many hepatocytes are pale, or have a somewhat biphasic pink and pale cytoplasm. The sinusoidal infiltrate is mostly that of EMH with a cluster of erythroid precursors on the right.
DIAGNO SIS
Paucity of bile ducts, consistent with Alagille syndrome. FIGURE 10. 3. 1 The portal zone contains a mild cellular infiltrate, and the lobule also shows very mild sinusoidal cellular infiltrate. The portal zone also contains an artery but no obvious duct of corresponding caliber is noted.
DISCUSSIO N
Paucity of ducts and the identification of other features of Alagille syndrome are important to differentiate from BA in
170
CASE
10.3:
PAU C I T Y
OF
I N T R A H E PAT I C
FIGURE 10. 3. 4 Cytokeratin 7 immunostain: A small portal tract demonstrates some CK7 immunostaining around the edges of the portal zone but no interlobular bile duct is present, consistent with ductopenia. Some of the CK7 staining of single cells may represent periportal stem cells, and some of the lobular staining likely represents cholestatic hepatocytes. A few forms likely represent a minimal periportal ductular reaction.
the early postnatal period in order to avoid hepatoportoenterostomy (Kasai procedure). The primary histologic differentiating features in the liver of Alagille syndrome are lack of interlobular bile ducts in a significant number of portal zones (usually more than 50%) and lack of a significant portal-based fibrosis or ductular reaction at age 6 to 8 weeks. Alagille syndrome can also usually be distinguished from the neonatal hepatitic pattern of injury by the relative lack of giant multinucleate hepatocytes and/or portal/lobular mononuclear infiltrate. The presentation of jaundice and liver injury in Alagille syndrome usually occurs before 6 months of age, and intermittent jaundice is typical after the initial presentation. The other important syndromatic features that are helpful in the diagnosis of Alagille syndrome are cardiovascular,
BILE
DUCTS
171
vertebral, and ocular abnormalities. The most common cardiovascular defect is peripheral pulmonic stenosis, but aortic coarctation, stenosis of other main arteries, or more complex cardiac defects can also occur. The more complex cardiac defects may cause a significant number of deaths. The ocular and vertebral abnormalities are not significant clinically, but split lamp examination and dorsal spine radiographic examination (spinal survey) are part of the workup to exclude BA. Other associations include abnormal facies and voice, renal tubular acidosis, other renal malformations, retinal pigmentation, inner ear abnormalities, and unexplained intracranial hemorrhage. Abnormal facies are difficult to recognize in the neonate, and are not reliable for diagnosis. The prognosis for Alagille syndrome is much better than for BA, but there is no specific treatment other than fatsoluble vitamins and nutritional support. About 75% of patients reach age 20, and about 25% require liver transplant due to failure to thrive, bone fractures, end-stage liver disease, disfiguring xanthomas, or refractory pruritus. Only 25% to 30% of deaths are due to liver disease; the severity of the cardiac lesion has a major effect on prognosis. Hepatocellular carcinoma can occur in those surviving into adulthood. If the associated lesions are not seen, but the biopsy has paucity of ducts, nonsyndromatic causes for ductopenia must be considered (Table 10.4). Electron microscopic examination for absence of peroxisomes, sweat test or genetic testing for cystic fibrosis, phenotyping for alpha-1-antitrypsin deficiency (AAT), testing for progressive familial intrahepatic cholestatic disorders, chromosomal abnormalities (trisomies including chromosome 6) (1), or examination for congenital infections such as rubella or cytomegalovirus (CMV) may be helpful to determine etiology of nonsyndromatic cases.
Reference 1. Kenny AP, Crimmins NA, Mackay DJG, Hopkin RJ, Bove KE, Leonis MA. Concurrent course of transient neonatal diabetes with cholestasis and paucity of interlobular bile ducts: a case report. Pediatr Dev Pathol. 2009;12:417–420.
Case 10.4
Neonatal Hepatitis Due to Alpha-1-Antitrypsin Deficiency GRACE E. KIM AND LINDA D. FERRELL
C L I N I C AL I N F OR M AT I ON
A 6-week-old boy developed upper respiratory symptoms and jaundice. On examination, the liver was palpable 1 cm below the right costal margin. Laboratory studies included total bilirubin of 6.3 mg/dL, direct bilirubin 3.7 mg/dL, gamma glutamyl transpeptidase (GGT) 719, AST 104, and ALT 84. A HIDA scan did not show excretion of bile into the duodenum. The patient was then referred for further evaluation at age 8 weeks. Physical examination at this time revealed a jaundiced infant. Further laboratory workup also revealed a low serum alpha-1-antitrypsin (AAT) level and Pi typing was then pursued, demonstrating a ZZ phenotype. A liver biopsy was done. The patient was followed for more than 2 years, gradually developing more severe jaundice and suffering from various complications of liver disease and cirrhosis including portal hypertension, ascites, and spontaneous bacterial peritonitis. Living donor liver transplantation was performed, and the patient has had a relatively unremarkable course for 2 years posttransplant.
F I G U R E 1 0 . 4 . 1 Liver biopsy at 8 weeks showed a mild lobular and
portal infiltrate, but no prominent pattern of multinucleate hepatocyte formation, consistent with a mild hepatitis.
R E A S ON F OR R E F E R R A L
To determine the etiology of the cholestasis with concern for biliary atresia. PAT H OL OG I C F E AT U R E S
Liver biopsy at age 8 weeks showed a mild hepatitis-like injury with a few giant/multinucleate hepatocytes. A mild degree of mononuclear infiltrate was present in lobules and portal zones (Figure 10.4.1). Focal apoptosis and mild fatty change was also present, and periportal hepatocytes contained iron (Figures 10.4.2 and 10.4.3 including iron stain). Canalicular cholestasis was present (Figure 10.4.3). Ductular reaction was minimal (Figure 10.4.2). Trichrome stain did not show significant fibrosis. Immunohistochemistry for CK7 confirmed the presence of 5 complete portal zones, with most containing an interlobular bile duct, but a rare small portal zone lacked a bile duct. Periodic acid–Schiff diastase (PASd) stain demonstrated rare positivity in small granules in periportal hepatocytes (Figure 10.4.4). Immunohistochemical stain for AAT demonstrated granular cytoplasmic positivity in about majority of the hepatocytes, with the most intense staining in the periportal region (zone 1) (Figure 10.4.5) and a decreasing gradient into zone 2 of the acinus. Electron microscopic examination confirmed the presence of finely granular homogeneous material in the endoplasmic reticulum, supporting the diagnosis of AAT deficiency (Figure 10.4.6).
1 0 . 4 . 2 Higher magnification of the same area as Figure 10.4.1 shows a mild, mostly mononuclear infiltrate of portal zone, mild fatty change, and periportal pigment. Only minimal ductular reaction is present.
FIGURE
The patient underwent transplant approximately 2.5 years later, and the explanted liver demonstrated wellestablished cirrhosis (Figure 10.4.7), with ductopenia.
172
C A S E
1 0 . 4 :
N E O N ATA L H E PAT I T I S D U E T O A L P H A - 1 - A N T I T RY P S I N D E F I C I E N C Y
173
FIGURE 10. 4. 3 Perls’ iron stain demonstrates periportal iron deposits in hepatocytes as well as acinar change of hepatocytes in the lobule, with bile stasis present in the lumina of the acini.
F I G U R E 1 0 . 4 . 5 AAT immunostain showed focal granular zone 1 staining for AAT with most prominent staining in the periportal hepatocytes, supporting the diagnosis of AAT deficiency.
FIGURE 10. 4. 4 PASd stain demonstrated small PASd+ granules in
F I G U R E 1 0 . 4 . 6 Electron microscopic examination confirmed the presence of finely granular, amorphous, and homogenous material in the endoplasmic reticulum typical of AAT deposits.
periportal hepatocytes, but with the prominent iron staining present as seen in Figure 10.4.3, this staining might represent lysosomes with iron. In addition, the size of the granules was not diagnostic for AAT globules.
DISCUSSIO N D I AG N OS I S
Alpha-1-antitrypsin deficiency.
Alpha-1-antitrypsin deficiency (AATD) is one of the most common etiologies of neonatal jaundice (1), or neonatal hepatitis–like syndrome, especially in societies with large populations of Northern European descent (2). The pattern of
174
CHAPTER
10:
N E O N ATA L
C H O L E S TAT I C
FIGURE 10. 4. 7 The explant at age 2.5 years demonstrated cirrhosis
and ductopenia.
LIVER
DISEASE
who present earlier with jaundice (within first 10 weeks of life) are the most likely to develop cirrhosis. Those that do not develop cirrhosis typically have liver biopsies with significantly less fibrosis at presentation. Variation severity of liver disease and pulmonary symptoms is common and may in part be due to genetic modifiers/variants of the defect (2). The diagnosis can often be histologically confirmed by immunohistochemical staining for AAT globules in older patients, but this is not reliable in infants as the globules are often not present until after 3 months of age. AAT can also be positive in unaffected infants, typically as mild cytoplasmic staining without significant granularity or globules. Likewise, PASd+ globules may not be present in neonates, and care must be taken as PASd may also highlight bile and lysosomes in histiocytes. The latter may contain iron or bile in granular forms. Electron microscopic examination will be positive for deposits in the smooth endoplasmic reticulum at this age, but examination for serum phenotype is the confirmatory test of choice.
References presentation of AATD in infancy can vary and include neonatal hepatitis (with or without giant cells), biliary obstruction– like changes (mimicking biliary atresia), paucity of ducts, or a combination of these presentations either simultaneously or sequentially. Children can also present later in life with chronic hepatitis or even cirrhosis. Patients who present in the first 6 months of life with a neonatal hepatitis–like syndrome typically have fully penetrant disease with ZZ phenotype (3,4) and more than 50% progress to cirrhosis in childhood (5). Some progressive cases can be associated with SZ and MZ phenotypes (3,4). Those
1. Jevon GP, Dimmick JE. Histopathologic approach to metabolic liver disease: Part 1. Pediatr Dev Pathol. 1998;1:179–199. 2. Fregonese L. Stolk J. Hereditary alpha-1-antitrypsin deficiency and its clinical consequences. Ophanet J Rare Dis. 2008;3:16. 3. Pittschieler K, Massi G. Liver involvement in infants with PiSZ phenotype of AATD. J Pediatr Gastroenterol Nutr. 1992;15:315–318. 4. Asarian J, Archibald RW, Lieberman J. Childhood cirrhosis associated with AATD. A genetic, biochemical, and morphologic study. J Pediatr. 1975;86;844–850. 5. Nebbia G, Hadchouel M, Odievre M, Alagille D. Early assessment of evolution of liver disease associated with AATD in childhood. J Pediatr. 1983;102:661–665.
11 Sinusoidal Dilatation and Congestion SANJAY KAKAR
Dilatation and congestion of hepatic sinusoids is most commonly observed as a manifestation of vascular congestion due to obstruction of venous outflow from the liver (1,2). However, in around one-third of biopsies with sinusoidal dilatation and congestion, there is no evidence of venous outflow obstruction. The most common clinical scenarios associated with this pattern are discussed below (Table 11.1). 1. Venous outflow obstruction: Obstruction to venous outflow from the liver leads to passive venous congestion manifested as sinusoidal dilatation and congestion that is most pronounced in the zone 3 of the acinus, but can involve zones 1 and 2 in more severe cases (3–7). The increased sinusoidal pressure leads to compression and atrophy of the hepatic plates, a useful diagnostic feature. The increased pressure also leads to extravasation of red blood cell in the space of Disse, hepatocyte atrophy, and necrosis. The congestion and necrosis are most marked in zone 3 of the acinus but can involve zones 1 and 2 in more severe cases. As the disease process becomes chronic, perivenular and sinusoidal fibrosis develops, which can eventually progress to bridging fibrosis and cirrhosis. Venous outflow obstruction can occur at the 3 levels: a. Heart due to cardiac disease (right heart failure, constrictive pericarditis, tricuspid valve disease). b. Large hepatic veins or inferior vena cava (Budd-Chiari syndrome).
2.
3.
TA B LE 11. 1 Sinusoidal dilatation and congestion: differential
diagnosis
4.
Venous Outflow Obstruction Budd-Chiari syndrome
Systemic Inflammatory Diseases Crohn disease
Membranous occlusion of inferior vena cava
Castleman disease
Cardiac disease (right heart failure, tricuspid insufficiency, constrictive pericarditis) Sinusoidal obstruction syndrome (chemotherapeutic drugs, infiltrative disorders like leukemia/lymphoma, sickle cell anemia, extramedullary hematopoiesis, malaria) Other Vascular Etiologies Portal vein thrombosis
Rheumatoid arthritis/Still disease Polymyalgia rheumatica Granulomatous disorders (saroidosis, tuberculosis)
5.
Extrahepatic Neoplasms RCC
6.
Hodgkin lymphoma Carcinoma (stomach, uterus, colon)
Nodular regenerative hyperplasia Abbreviation: RCC, renal cell carcinoma.
175
c. Small hepatic veins (sinusoidal obstruction syndrome, formerly veno-occlusive disease). This most commonly results from chemotherapeutic drugs used in colorectal cancer (oxaliplatin), leukemia/lymphoma (actinomycin D, cytosine arabinoside, anti-CD33 antibody), and immunosuppressive agents like azathioprine, due to injury to sinusoidal endothelial cells and hepatic stellate cells (8,9). This manifests histologically as sinusoidal dilatation and hepatocellular necrosis. It can lead to centrilobular fibrosis, resulting in occlusion of the central veins. Infiltrative disorders that block the sinusoids like sickle cell anemia, leukemia, malaria, and extramedullary hematopoiesis can also lead to dilatation and congestion of sinusoids. Other vascular causes like portal vein thrombosis and nodular regenerative hyperplasia can also lead to sinusoidal dilatation and congestion in the liver. Obstruction of portal vein blood flow to the liver can lead to hepatocyte atrophy and an appearance of dilated sinusoids because of excess sinusoidal volume (6,7,10). Nodular regenerative hyperplasia can lead to sinusoidal dilatation by a combination of compression of hepatic microvasculature and areas of hepatocyte atrophy (11,12). Systemic inflammatory conditions like Crohn disease, rheumatoid arthritis, Still disease, and polymyalgia rheumatica (2,11) as well as granulomatous conditions like sarcoidosis without direct liver involvement (1,11). There are several reports in the literature that describe marked sinusoidal dilatation in liver biopsies in patients with Castleman disease (Figure 11.1) (2,13). Extrahepatic neoplasms without liver involvement: Sinusoidal dilatation has been most commonly associated with renal cell carcinoma (RCC) and Hodgkin lymphoma in the absence of liver involvement (11,14). RCC accompanied by liver dysfunction has been termed Stauffer syndrome (14,15). Sinusoidal dilatation can be seen in around 10% of patients with RCC in the absence of metastatic disease (14). Other tumors that have been associated with sinusoidal dilatation include carcinoma of the stomach, uterus, and colon (11). Drugs: Hormonal agents like oral contraceptives can cause sinusoidal dilatation, especially in zone 1 of the liver. Other: a. Mechanical artifact: Artifactual sinusoidal dilatation can result from mechanical reasons like rough handling or stretching of the biopsy (Figure 11.2). These changes are often more pronounced toward the edges. There is no hepatic plate atrophy or extravasation of RBC into the space of Disse.
176
CHAPTER
11:
SINUSOIDAL
D I L ATAT I O N
FIGURE 11. 1 Marked sinusoidal dilatation with plate atrophy in
Castleman disease.
b. Intraoperative biopsies: Sinusoidal dilatation can be observed in biopsies obtained during abdominal surgeries (2). The mechanism of sinusoidal dilatation and congestion (SDC) in these situations is unclear, but may be the result of alterations in portal blood flow during abdominal surgery that leads to transient SDC. c. Transplant biopsies: Sinusoidal dilatation is commonly seen in transplant biopsies in the absence of venous outflow obstruction. The reason is not clear, but it may be related to hemodynamic changes related to the vascular anastomosis. d. Cirrhosis and nonneoplastic liver adjacent to tumors: Localized venous outflow obstruction can result from regenerative nodules in cirrhosis or adjacent to mass lesions.
CONGESTION
F I G U R E 1 1 . 2 Artifactual sinusoidal dilatation not accompanied by hepatic plate atrophy.
5.
6.
7.
8.
9.
10.
Based on morphological features, it can be difficult to distinguish venous outflow obstruction from other causes of sinusoidal dilatation and congestion. The presence of prominent congestion and centrizonal fibrosis favor venous outflow obstruction but are not totally reliable for this distinction (2).
11.
References
13.
1. Poulsen H, Winkler K, Christoffersen P. The significance of centrilobular sinusoidal changes in liver biopsies. Scand J Gastroenterol Suppl. 1970;7:103–109. 2. Kakar S, Kamath PS, Burgart LJ. Sinusoidal dilatation and congestion in liver biopsy: is it always due to venous outflow impairment? Arch Pathol Lab Med. 2004;128:901–904. 3. Dilawari JB, Bambery P, Chawla Y, et al. Hepatic outflow obstruction (Budd-Chiari syndrome). Experience with 177 patients and a review of the literature. Medicine (Baltimore). 1994;73:21–36. 4. Tanaka M, Wanless IR. Pathology of the liver in Budd-Chiari syndrome: portal vein thrombosis and the histogenesis of veno-centric cirrhosis,
AND
12.
14.
15.
veno-portal cirrhosis, and large regenerative nodules. Hepatology. 1998; 27:488–496. Iwai M, Kitagawa Y, Nakajima T, et al. Clinical features, image analysis, and laparoscopic and histological liver findings in Budd-Chiari syndrome. Hepatogastroenterology. 1998;45:2359–2368. Wanless IR. Vascular disorders. In: MacSween RNM, Burt AD, Portman BC, Ishak KG, Scheuer PJ, Anthony PP, eds. Pathology of the Liver. 4th ed. New York, NY: Churchill Livingstone, Inc.; 2002:539–574. Kakar S, Batts KP, Poterucha JJ, Burgart LJ. Histologic changes mimicking biliary disease in liver biopsies with venous outflow impairment. Mod Pathol. 2004;17:874–878. Kandutsch S, Klinger M, Hacker S, Wrba F, Gruenberger B, Gruenberger T. Patterns of hepatotoxicity after chemotherapy for colorectal cancer liver metastases. Eur J Surg Oncol. 2008;34:1231–1236. Kumar S, DeLeve LD, Kamath PS, Tefferi A. Hepatic veno-occlusive disease (sinusoidal obstruction syndrome) after hematopoietic stem cell transplantation. Mayo Clin Proc. 2003;78:589–598. Shimamatsu K, Wanless IR. Role of ischemia in causing apoptosis, atrophy, and nodular hyperplasia in human liver. Hepatology. 1997;26: 343–350. Bruguera M, Aranguibel F, Ros E, Rodes J. Incidence and clinical significance of sinusoidal dilatation in liver biopsies. Gastroenterology. 1978;75:474–478. Reshamwala PA, Kleiner DE, Heller T. Nodular regenerative hyperplasia: not all nodules are created equal. Hepatology. 2006;44:7–14. Curciarello J, Castelletto R, Barbero R, et al. Hepatic sinusoidal dilatation associated to giant lymph node hyperplasia (Castleman’s): a new case in a patient with periorbital xanthelasmas and history of celiac disease. J Clin Gastroenterol. 1998;27:76–78. Aoyagi T, Mori I, Ueyama Y, Tamaoki N. Sinusoidal dilatation of the liver as a paraneoplastic manifestation of renal cell carcinoma. Hum Pathol. 1989;20:1193–1197. Delpre G, Ilie B, Papo J, Streifler C, Gefel A. Hypernephroma with nonmetastatic liver dysfunction (Stauffer’s syndrome) and hypercalcemia. Am J Gastroenterol. 1979;72:239–247.
Case 11.1
Budd-Chiari Syndrome Versus Biliary Disease SANJAY KAKAR
C L I N IC AL I N F OR M AT I ON
A 35-year-old woman presented with fatigue and abdominal pain. She has been on oral contraceptives for 5 years; there is no other history of medications. Physical exam revealed mild ascites, jaundice, mildly enlarged liver, and splenomegaly. Liver enzyme tests revealed elevation of alkaline phosphatase (5 times normal) and transaminases were slightly elevated (twice normal). Serological tests for hepatitis viruses and autoantibodies were negative. Ultrasound showed an enlarged liver, but the large bile ducts were normal and there was no evidence of bile duct obstruction.
lymphocytic infiltration of the bile duct is seen (lymphocytic cholangitis), but overt bile duct damage or duct loss was not seen (Figure 11.1.6).
DIAGNO SIS
Budd-Chiari syndrome (BCS) associated with portal changes mimicking chronic biliary disease.
R E A SON F OR R E F E R R AL
The biopsy shows features suggestive of bile duct obstruction, but there is no clinical or radiological evidence of biliary disease.
PAT H OL OG I C F E AT U R E S
The liver biopsy shows zone 3 sinusoidal dilatation and congestion with hepatic plate atrophy (Figure 11.1.1) and hepatocellular necrosis (Figure 11.1.2). These changes are accompanied by extravasation of red blood cells into the space of Disse (Figure 11.1.3). Fibrosis is present in the pericentral region. In addition, most of the portal tracts show mild lymphocytic infiltrate and mild bile ductular reaction (Figure 11.1.4) with florid reaction in 1 portal tract (Figure 11.1.5). Focal
FIGURE 11. 1. 1 Sinusoidal dilatation and congestion predominantly
in zone 3 accompanied by hepatic plate atrophy.
F I G U R E 1 1 . 1 . 2 Hepatocellular necrosis around the central vein.
F I G U R E 1 1 . 1 . 3 The red blood cells have been pushed from the sinusoids into the potential space between the sinusoidal endothelium and hepatocytes (space of Disse).
177
178
CHAPTER
11:
SINUSOIDAL
D I L ATAT I O N
FIGURE 11. 1. 4 Mild bile ductular reaction with intact bile duct.
AND
CONGESTION
F I G U R E 1 1 . 1 . 6 Mild portal lymphocytic inflammation with focal
bile duct involvement. TA BL E 1 1 . 1 . 1 Budd-chiari syndrome: etiology
FIGURE 11. 1. 5 Mild portal inflammation and prominent ductular
reaction in 1 portal tract.
D I SC U SSI ON
The pathologic features strongly raise the possibility of venous outflow obstruction. Based on the biopsy features, a CT scan was performed and demonstrated hepatic vein thrombosis, confirming the diagnosis of BCS. No abnormalities were seen in the biliary tree. Based on the history, the BCS is likely to be due to hypercoaguable state as a result of oral contraceptive use. Most cases of BCS present in a subacute fashion with painful hepatomegaly and ascites. In some instances, the diagnosis is made during the workup for chronic liver disease or cirrhosis. Rare cases can present with an acute fulminant picture; this presentation carries a very poor prognosis with 80% mortality (1,2). The important etiologies of BCS are listed in Table 11.1.1. The diagnosis of BCS is made by imaging techniques and biopsy. Ultrasound with Doppler flow studies is the
Thrombotic Causes
Mechanical Causes
(a) Hypercoaguable states Myeloproliferative disorders Pregnancy and oral contraceptives Antiphospholipid antibodies (SLE) Paroxysmal nocturnal hemoglobinuria Factor V Leiden
Membranous obstruction of inferior vena cava
(b) Factor deficiencies Protein C deficiency Protein S deficiency Antithrombin deficiency
Posttransplant kinking of hepatic venous outflow
Obstruction by tumor Extrinsic compression by tumor or mass
initial tool of choice; hepatic scintigraphy, CT, and MRI can also contribute to the diagnosis. Hepatic venography was considered the gold standard, but is now restricted to diagnostically challenging cases. The inflammation and ductular reaction in the portal tracts in this case raise the possibility of biliary disease. Since venous outflow obstruction can be associated with cholestasis and elevated alkaline phosphatase, the clinical profile can further reinforce the suspicion of a biliary etiology. The literature and textbooks have emphasized centrizonal changes in venous outflow obstruction, whereas portal changes are often overlooked. Portal tract changes that mimic biliary disease are frequently present in venous outflow obstruction. In a study of 34 patients with a confirmed diagnosis of venous outflow obstruction, pathologic changes in the portal tracts were present in more than half of the cases (3). The liver biopsies in these cases showed portal expansion with bile ductular reaction with mild lymphoplasmacytic infiltrate and portal or periportal fibrosis in nearly half of the cases. The ductular reaction is generally mild but can occasionally be florid. In few cases, there can be bile duct damage in the form of lymphocytic cholangitis. In some instances, periportal fibrosis can be seen without ductular reaction.
CASE
11.1
:
BUDD-CHIARI
SYNDROME
Although imaging studies can be done to confirm the absence of bile duct abnormalities depending on the clinical context, awareness of these morphologic changes can prevent overemphasizing the risk of biliary disease in this setting. The mechanism of bile ductular reaction in venous outflow obstruction is not known. It may be the result of low level of ischemic damage to the biliary tree as a result of increased venous pressure or compromised arterial flow due to abnormal shunts. Fibrosis in BCS typically starts as sinusoidal fibrosis around the central vein, but can progress to cirrhosis. Fibrosis is often present in cases with acute presentation, indicating that longstanding subclinical obstruction has been present (3). Based on the biopsy findings, other causes of venous outflow like right heart failure, tricuspid valve disease, and
VERSUS
BILIARY
DISEASE
179
constrictive pericarditis can be considered. However, there is no evidence of cardiac or pericardial disease based on clinical information. Similarly, there is no clinical basis to consider other causes of sinusoidal dilatation and congestion (Table 11.1.1).
References 1. Powell-Jackson PR, Ede RJ, Williams R. Budd-Chiari syndrome presenting as fulminant hepatic failure. Gut. 1986;27:1101–1105. 2. Sandle GI, Layton M, Record CO, Cowan WK. Fulminant hepatic failure due to Budd-Chiari syndrome. Lancet. 1980;1:1199. 3. Kakar S, Batts KP, Poterucha JJ, Burgart LJ. Histologic changes mimicking biliary disease in liver biopsies with venous outflow impairment. Mod Pathol. 2004;17:874–878.
Case 11.2
Sinusoidal Obstruction SANJAY KAKAR
C L I N I C A L I N F OR M AT I ON
A 27-year-old woman presented with acute onset of right upper quadrant pain and nausea. There is no significant medication history. There were 2 similar episodes in the past that had resolved spontaneously. On examination, there was jaundice and tender hepatomegaly. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were 350 IU/L; alkaline phosphatase was slightly elevated. Serum bilirubin was 10 mg/dL, with 70% being conjugated. Serological studies for hepatitis A, B, and C were negative. There were no autoantibodies and serum ceruloplasmin is normal. Gallbladder and large bile ducts were normal on ultrasound. R E A S ON F OR R E F E R R A L
The history and liver enzymes suggest acute hepatocellular injury, but there is minimal inflammation or hepatocellular injury on the biopsy.
F I G U R E 1 1 . 2 . 2 Mild dilatation and prominent congestion of sinu-
soids observed throughout the biopsy.
PAT H OL OG I C F E AT U R E S
The liver biopsy shows normal portal tracts lacking significant inflammation and intact bile ducts. The hepatic parenchyma shows minimal hepatocellular damage, scattered prominent Kupffer cells and no significant inflammation (Figure 11.2.1). Sinusoidal dilatation is present throughout the biopsy but is not marked (Figure 11.2.2). Many of the dilated sinusoids are congested and are clogged with red blood cells (Figure 11.2.3). There is no fibrosis.
F I G U R E 1 1 . 2 . 3 Marked congestion is uniformly present in the
sinusoids.
DIAGNO SIS
Acute hepatic sickle cell crisis.
DISCUSSIO N FIGURE 11. 2. 1 Normal portal tracts and no significant inflamma-
tion or hepatocellular damage.
The clinical information indicates an acute process with hepatocellular injury evidenced by rise in transaminases. However, there is no significant inflammation and hepatocellular damage
180
CASE
11.2
:
SINUSOIDAL
is minimal. The sinusoidal dilatation and congestion raises the possibility of venous outflow obstruction. Based on the clinical information, there is no history of cardiac disease and the imaging studies were negative for Budd-Chiari syndrome. The clinical scenario does not support other etiologies of sinusoidal dilatation and congestion like systemic autoimmune disorders, granulomatous disorders, Castleman disease, and neoplasms. A closer examination of the congested hepatic sinusoids reveals that many are packed with sickle-shaped red blood cells (Figure 11.2.4). Further clinical information revealed that the patient had sickle cell anemia.
OBSTRUCTION
181
Hyperbilirubinemia in sickle cell anemia can result from hemolysis, biliary obstruction due to stones, and sickle cell hepatopathy (1–5). There is no evidence of active hemolysis or biliary obstruction in this patient. The clinical and pathologic features establish the diagnosis of acute hepatic sickle cell crisis. Sickle cell crisis selectively involves the liver in 10% of patients with sickle cell anemia, and less commonly with sickle cell trait (3). In most instances, it is a self-limited disease that results from ischemia due to vascular occlusion by sickled RBCs. The liver biopsy shows sinusoidal dilatation and congestion, sickled RBCs, focal necrosis, variable cholestasis, and prominent Kupffer cells (2,3). In rare instances, a severe variant of hepatic sickle cell crisis can result in widespread sickling and ischemic hepatocellular damage (3,4). The liver shows swollen hepatocytes, prominent necrosis, and marked cholestasis. This complication can be fatal, and the liver disease can be complicated by renal failure, hemorrhage, and encephalopathy. In severe cases, liver biopsy can be hazardous due to coagulopathy (4).
References
FIGURE 11. 2. 4 The sinusoids are packed with sickled red blood cells (sickle cell thrombi).
1. Mills LR, Mwakyusa D, Milner PF. Histopathologic features of liver biopsy specimens in sickle cell disease. Arch Pathol Lab Med. 1988;112:290–294. 2. Charlotte F, Bachir D, Nénert M, et al. Vascular lesions of the liver in sickle cell disease. A clinicopathological study in 26 living patients. Arch Pathol Lab Med. 1995;119:46–52. 3. Banerjee S, Owen C, Chopra S. Sickle cell hepatopathy. Hepatology. 2001;33:1021–1028. 4. Ahn H, Li CS, Wang W. Sickle cell hepatopathy: clinical presentation, treatment, and outcome in pediatric and adult patients. Pediatr Blood Cancer. 2005;45:184–190. 5. Bandyopadhyay R, Bandyopadhyay SK, Dutta A. Sickle cell hepatopathy. Indian J Pathol Microbiol. 2008;51:284–285.
Case 11.3
Veno-Occlusive Disease (Sinusoidal Obstruction Syndrome) SANJAY KAKAR
C L I N C A L I N F OR M AT I ON
A 42-year-old male with acute myeloid leukemia underwent stem cell transplantation. Two weeks later, he presented with weight gain, right upper quadrant pain, and jaundice. Physical examination revealed tender hepatomegaly and ascites. The liver enzymes showed elevated ALT and AST (both 500 IU/L) and alkaline phosphatase was twice normal. Doppler ultrasound revealed blood flow away from the liver in the portal vein. R E A S ON F OR R E F E R R A L
To establish the etiology of centrizonal necrosis in the setting of stem cell transplantation. PAT H OL OG I C F E AT U R E S
The liver biopsy showed marked dilatation and congestion of sinusoids around the central vein (Figure 11.3.1). The central veins show marked edema in the subendothelial region and perivenular necrosis (Figure 11.3.2). Similar changes are seen in the small hepatic vein (Figure 11.3.3). There is no significant inflammation and the portal tracts are normal. Trichrome stain showed mild sinusoidal fibrosis around the central vein.
F I G U R E 1 1 . 3 . 2 Subendothelial edema in the central vein with surrounding hepatocellular swelling and necrosis.
D I AG N OS I S
Sinusoidal obstruction syndrome (SOS) related to stem cell transplantation.
F I G U R E 1 1 . 3 . 3 Subendothelial edema in small hepatic venule with
perivenular necrosis.
DISCUSSIO N
FIGURE 11. 3. 1 Marked sinusoidal dilatation, congestion, and hepatic plate atrophy.
The striking sinusoidal changes and centrizonal injury implicate venous outflow obstruction. The clinical information and imaging studies do not support cardiac disease, Budd–Chiari syndrome, or other causes of sinusoidal dilatation and congestion. The history of stem cell transplantation and the biopsy findings confirm the diagnosis of veno-occlusive disease or SOS. SOS is thought to result from injury to the sinusoidal and venular endothelial cells (1–3). Recent studies have indicated
182
CASE
11.3
:
VENO-OCCLUSIVE
that the primary event is sinusoidal injury and involvement of hepatic venules may not be always present. Hence, the term SOS has been proposed to replace veno-occlusive disease (1). The endothelial injury leads to edematous thickening of the subintimal zone of small hepatic veins, which in turn causes venous outflow obstruction with sinusoidal dilatation, congestion, and hepatocellular necrosis. Fibrin thrombi can be seen, but are uncommon. With time, venular and sinusoidal fibrosis can develop. In addition to stem cell or bone marrow transplant (1,4), SOS can also be observed due to endothelial injury as a result of chemotherapy (Table 11.3.1). A wide variety of drugs have been implicated. Oxaliplatin is used to reduce the size of liver metastasis from colorectal cancer before surgery and has been associated with SOS (5–7). Several drugs used for the treatment of leukemia/lymphoma and long-term use of immunosuppressive agent azathioprine have also been associated with SOS (4). Historically, veno-occlusive disease has been described as a result of toxicity of pyrrolizidine alkaloids in herbal medications as well as in herbal tea in southern Africa (8,9). Rare cases of veno-occlusive disease have been reported with liver transplantation and have a strong association with azathioprine use and acute rejection (10,11). The clinical diagnosis of SOS can be challenging. In the setting of stem cell transplantation, it typically occurs in the first 3 weeks. Clinical and laboratory features like weight gain due to salt and water retention by the kidneys, hyperbilirubuinemia, and reverse flow in portal vein on Doppler ultrasound raise the suspicion for SOS. Plasma levels of plasminogen activator inhibitor-1 are often elevated. Liver biopsy TA B LE 11. 3. 1 Etiologies of hepatic veno-occlusive disease (sinusoi-
dal obstruction syndrome) Chemotherapeutic Agents Colorectal cancer: oxaliplatin Leukemia/lymphoma: actinomycin D, mithramycin, dacarbazine, cytosine arabinoside, 6-thioguanine, anti-CD33 monoclonal antibody Immunosuppressive: azathioprine Radiation Therapy Abdominal radiation for Wilms tumor Transplant Related Stem cell transplantation Liver transplantation Toxins Pyrrolizidine alkaloids in African bush tea Herbal medications
DISEASE
183
can establish the diagnosis, but changes can be patchy in early disease leading to false-negative results (4). Overall, 50% to 80% of patients with SOS recover. The outcome of severe SOS is poor with high mortality (4). Other causes of hepatic dysfunction after stem cell transplant include acute graft versus host disease (GVHD), drug-induced hepatitis, infections, and relapse of leukemia with liver involvement. Typical features of GVHD like bile duct damage and apoptosis are not seen in SOS, whereas centrizonal hepatocellular damage is not characteristic of GVHD. Immunosuppressive agents like cyclosporine, a wide variety of antibiotics, and antifungal drugs can lead to drug-induced hepatitis. The presence of prominent inflammation, cholestasis, and noncentrizonal pattern of hepatocellular injury favors drug-induced hepatitis. The liver biopsy should always be evaluated for fungal organisms and involvement by leukemia/ lymphoma.
References 1. Shulman HM, Fisher LB, Schoch HG, Henne KW, McDonald GB. Venoocclusive disease of the liver after marrow transplantation: histological correlates of clinical signs and symptoms. Hepatology. 1994;19:1171– 1181. 2. DeLeve LD, Shulman HM, McDonald GB. Toxic injury to hepatic sinusoids: sinusoidal obstruction syndrome (veno-occlusive disease). Semin Liver Dis. 2002;22:27–42. 3. DeLeve LD. Hepatic microvasculature in liver injury. Semin Liver Dis. 2007;27(4):390–400. 4. Kumar S, DeLeve LD, Kamath PS, Tefferi A. Hepatic veno-occlusive disease (sinusoidal obstruction syndrome) after hematopoietic stem cell transplantation. Mayo Clin Proc. 2003;78:589–598. 5. Kandutsch S, Klinger M, Hacker S, Wrba F, Gruenberger B, Gruenberger T. Patterns of hepatotoxicity after chemotherapy for colorectal cancer liver metastases. Er J Surg Oncol. 2008;34:1231–1236. 6. Karoui M, Penna C, Amin-Hashem M, et al. Influence of preoperative chemotherapy on the risk of major hepatectomy for colorectal liver metastases. Ann Surg. 2006;243:1–7. 7. Nordlinger B, Benoist S. Benefits and risks of neoadjuvant therapy for liver metastases. J Clin Oncol. 2006;24:4954–4955. 8. Datta DV, Khuroo MS, Mattocks AR, Aikat BK, Chhuttani PN. Herbal medicines and veno-occlusive disease in India. Postgrad Med J. 1978;54:511–515. 9. Ridker PM. Hepatic veno-occlusive disease and herbal teas. J Pediatr. 1989;115:167. 10. Dhillon AP, Burroughs AK, Hudson M, Shah N, Rolles K, Scheuer PJ. Hepatic venular stenosis after orthotopic liver transplantation. Hepatology. 1994;19:106–111. 11. Sebagh M, Debette M, Samuel D, et al. “Silent” presentation of veno-occlusive disease after liver transplantation as part of the process of cellular rejection with endothelial predilection. Hepatology. 1999;30: 1144–1150.
Case 11.4
Amyloidosis SANJAY KAKAR
C L I N I C AL I N F OR M AT I ON
A 55-year-old woman with compensated biventricular heart failure had elevated liver enzymes (ALT 135, AST 125, ALP 425 U/L). These abnormalities were thought to be due to passive venous congestion. The ultrasound showed a nodular appearance raising concern for cirrhosis. Occasional larger nodules (2–3 cm) were also noted. A wedge liver biopsy was performed to assess the degree of liver damage and evaluate the liver nodules.
R E A S ON F OR R E F E R R A L
The liver showed a nodular appearance, but definite fibrosis or venous congestion was not seen. F I G U R E 1 1 . 4 . 2 The hepatocyte cell plates at the periphery of the
PAT H OL OG I C F E AT U R E S
nodules are compressed and atrophic.
The liver showed a nodular architecture (Figure 11.4.1) with compression of cell plates at the periphery of the nodules (Figure 11.4.2). Reticulin stain highlighted the nodular architecture. Trichrome stain showed lack of fibrous septa between the nodules (Figure 11.4.3). The portal tracts lacked significant inflammation, and bile ducts were intact. Globular eosinophilic deposits were seen in the sinusoids in some areas (Figure 11.4.4). The deposits showed homogenous light staining on trichrome (Figure 11.4.5). Congo red stain was positive in the deposits and showed apple-green birefringence under polarized light.
F I G U R E 1 1 . 4 . 3 Trichrome stain highlights the nodules and shows
lack of fibrous septa.
DIAGNO SIS
Amyloidosis with nodular regenerative hyperplasia.
DISCUSSIO N FIGURE 11. 4. 1 Nodular architecture of the liver without fibrous
septa.
Amyloidosis is a heterogeneous group of diseases characterized by deposition of glycoprotein fibrils in the extracellular matrix and blood vessel walls. The most common presenting
184
CASE
11.4
:
FIGURE 11. 4. 4 The sinusoids show globular eosinophilic deposits in
some areas of the biopsy.
AMYLOIDOSIS
185
has not been confirmed in a large Mayo Clinic series (33). Hepatic involvement portends a poor prognosis, as it is a reflection of advanced disease. Median survival is only 9 months and the 5-year survival 17%. Elevated bilirubin and congestive heart failure are adverse prognostic factors. Depending on the biochemical composition of the fibrils, amyloidosis can be divided into several subtypes. In primary or AL amyloidosis, the deposits are composed of monoclonal immunoglobulin light chain fragments. Some cases occur in association with systemic hematologic diseases like multiple myeloma and Waldenstrom macroglobulinemia. In secondary or AA amyloidosis, the deposits are composed of fragments of serum amyloid A, an acute phase reactant. The liver can be involved in dialysis-related amyloidosis, in which the deposits are composed of beta-2-microglobulin and is a consequence of decrease in renal excretory function. Liver involvement can be seen in up to 70% of cases and occurs in both AL (primary) and AA (secondary) amyloidosis. Amyloid deposition can occur in the sinusoids or blood vessel walls; both locations are affected in 20% of cases. The deposits are more often seen in blood vessels in AA amyloidosis (Figures 11.4.6 and 11.4.7), whereas the sinusoids are more commonly affected in AL amyloidosis (Figure 11.4.8). However, there is considerable overlap in these distribution patterns, and these are not reliable for definite distinction between AL and AA amyloidosis (35,36). The amyloid fibrils stain with Congo red and show apple-green birefringence under polarized light. The color varies from yellow-green to blue-green and can change with the rotation of the polarizer or analyzer. Different colors may be seen in different areas of the biopsy. Demonstration of birefringence is facilitated by using thick sections (10 microns), turning light to the maximum and pulling out color filters. Congophilia can be reduced after prolonged fixation. Examination of amyloid fibrils on Congo red-stained sections by a fluorescence microscope using fluorescein
FIGURE 11. 4. 5 Pale homogenous appearance of the sinusoidal
deposits on trichrome stain.
symptoms include weight loss, fatigue, abdominal pain, anorexia, early satiety, nausea, and dysgeusia. Hepatomegaly, ascites, edema, purpura, and splenomegaly can be present (32–34). In majority of the cases, extrahepatic manifestations such as nephrotic syndrome, heart failure, peripheral neuropathy, orthostatic hypotension, and/or carpal tunnel syndrome are also present. The most common biochemical abnormality is elevated alkaline phosphatase; mild elevation of liver transaminases occurs in one-third of cases (32–34). Due to the nonspecific symptoms and laboratory abnormalities, the diagnosis can easily be overlooked clinically. The presence of involuntary weight loss, hepatomegaly, and unexplained elevation of ALP, proteinuria, or hyposplenism should raise the clinical suspicion for hepatic amyloidosis. There has been concern about the risk of hepatic rupture following liver biopsy for amyloidosis, but
F I G U R E 1 1 . 4 . 6 Amyloid deposits in the blood vessels walls in the
portal tract.
186
CHAPTER
11:
SINUSOIDAL
D I L ATAT I O N
FIGURE 11. 4. 7 Congo red stain highlights the amyloid deposits in
the vessel walls.
AND
CONGESTION
of nonimmunological origin. Electron microscopy is seldom required for diagnosis but can demonstrate central electronlucent core and nonbranching fibrils of indefinite length with a mean diameter of 10 nm. Deposition of monoclonal protein (M-protein) in tissue can occur in light-chain deposition disease, heavy-chain deposition disease, and light- and heavy-chain deposition disease (37,38). Deposition of amyloid-like noncongophilic occurs in the kidney in fibrillary and immunotactoid glumerulopathies. In rare instances, these are associated with extrarenal sites including liver (39,40). The fibrils in these conditions have larger diameter compared to amyloid fibrils, and electron microscopy is required for the distinction. In most cases, sinusoidal deposition of amyloid is obvious on the biopsy (Figure 11.4.8). In some instances, the deposits can be subtle, and the diagnosis requires a high index of suspicion. Nodular sinusoidal deposition of amyloid as in this case can be easily overlooked. The pale homogenous staining of amyloid on trichrome stain can be very helpful in raising the suspicion of amyloidosis. In this case, the amyloidosis was accompanied by nodular regenerative hyperplasia, a nonspecific response to vascular injury as a result of diverse etiologies (see Chapter 13). Further investigations revealed that the heart failure was a result of cardiac amyloidosis.
References
FIGURE 11. 4. 8 Congo red stain shows extensive sinusoidal amyloid
deposits with atrophy of hepatic plates.
isothiocyanate filter shows yellow fluorescence but is not specific for amyloid. Glycoprotein P component is present in all amyloid deposits, which can be demonstrated by immunohistochemistry. Immunohistochemistry for specific proteins like immunoglobulin light chains (AL amyloid) and SAA (AA amyloid) can also be done for further classification. Background staining can make interpretation difficult. Immunoglobulin deposits can occasionally be seen in amyloid
1. Gertz MA, Kyle RA. Hepatic amyloidosis (primary [AL], immunoglobulin light chain): the natural history in 80 patients. Am J Med. 1988;85:73–80. 2. Park MA, Mueller PS, Kyle RA, Larson DR, Plevak MF, Gertz MA. Primary (AL) hepatic amyloidosis: clinical features and natural history in 98 patients. Medicine (Baltimore). 2003;82:291–298. 3. Malnick S, Melzer E, Sokolowski N, Basevitz A. The involvement of the liver in systemic diseases. J Clin Gastroenterol. 2008;42:69–80. 4. Chopra S, Rubinow A, Koff RS, et al. Hepatic amyloidosis. A histopathologic analysis of primary (AL) and secondary (AA) forms. Am J Pathol. 1984;115:186–193. 5. Buck FS, Koss MN. Hepatic amyloidosis: morphologic differences between systemic AL and AA types. Hum Pathol. 1991;22:904–907. 6. Faa G, Van Eyken P, De Vos R, Fevery J, Van Damme B, De Groote J, Desmet VJ. Light chain deposition disease of the liver associated with AL-type amyloidosis and severe cholestasis. J Hepatol. 1991;12:75–82. 7. Buxbaum J, Gallo G. Nonamyloidotic monoclonal immunoglobulin deposition disease. Light-chain, heavy-chain, and light- and heavy-chain deposition diseases. Hematol Oncol Clin North Am. 1999;13:1235–1248. 8. Strøm EH, Hurwitz N, Mayr AC, Krause PH, Mihatsch MJ. Immunotactoid-like glomerulopathy with massive fibrillary deposits in liver and bone marrow in monoclonal gammopathy. Am J Nephrol. 1996; 16:523–528. 9. Hvala A, Ferluga D, Vizjak A, Koselj-Kajtna M. Fibrillary noncongophilic renal and extrarenal deposits: a report on 10 cases. Ultrastruct Pathol. 2003;27:341–347.
12 Peliosis Hepatis SANDRA FISCHER AND MAHA GUINDI
I N T ROD U C T I ON
TA BL E 1 2 . 1 Causes of peliosis hepatis
Peliosis hepatis is a condition with many underlying etiologies. It is defined as cystic blood-filled spaces in the liver and other organs (eg, spleen, lymph nodes). Peliosis hepatis has been recognized of clinical significance only since the middle of the 20th century. The term peliosis was originally applied to macroscopic lesions that look dusky or purple in color. Microscopic lesions may be confused with severe sinusoidal dilatation or with “evacuation of the liver cell plates,” a lesion seen after zonal hepatocellular dropout but without loss of the normal reticulin fibers (1) (Figure 12.1). The pathogenesis of peliosis hepatis is unknown. The lesion has been produced in experimental animals by administration of phalloidin presumably as a result of damage to the sinusoidal wall. Focal apoptosis of hepatocytes or sinusoidal endothelial cells and disruption of liver extracellular matrix seem to play a role in the pathogenesis (2). According to the mechanism of injury to the sinusoidal wall, the lesions can be classified as phlebectatic type, if associated with an intrinsic weakness of the wall, or parenchymal type, when secondary to hepatocyte necrosis (1). The blood-filled cystic spaces may have no endothelial lining initially, but subsequent re-endothelialization probably occurs. The lesions are randomly distributed without zonal preference. Anabolic steroids, diethylstilboestrol contraceptive steroids, tamoxifen, azathioprine, vitamin A, and thorotrast have been implicated as causative agents in humans (see Table 12.1). Microscopic lesions occur in patients receiving thiopurines for renal transplantation, liver transplantation, or various malignancies. A variety of chronic diseases including malnutrition,
Drugs
Androgens, arsenic compounds, azathioprine, busulfan, chemotherapeutic agents, contraceptive steroids, corticosteroids, danazol, diethylstilboestrol, estrone sulfate, fluoxymesterone, glucocorticoids, hydroxyprogesterone, hydroxyurea, medroxyprogesterone, mercaptopurine, methandrostenolone, methotrexate, methyltestosterone, synthetic estrogens, tamoxifen, testosterone, thioguanine, thorotrast, vinyl chloride, and vitamin A.
Chronic conditions
Malnutrition, tuberculosis, leprosy, vasculitis, bartonellosis, AIDS, and other immunosuppressed states.
Neoplasms
Hepatocellular adenoma/carcinoma, angiosarcoma, hairy cell leukemia, angioimmunoblastic lymphadenopathy, myeloproliferative disorders, Waldenström’s macroglobulinemia with light chain deposition, and Hodgkin disease.
leukemia, tuberculosis, leprosy, vasculitis, and AIDS have also been reported with macroscopic peliotic lesions. Peliotic lesions found in AIDS and other immunosuppressed patients are caused by rickettsial organisms (Bartonella species). The lesions are visible as red spots beneath the capsule of the liver, which may be greatly enlarged. Histologically, they appear as 1 to 4 mm cysts within the parenchyma, containing blood. These lesions have a myxoid stroma that has a bluish haze on routine haematoxylin and eosin staining and contain clumps of organisms that can be identified with Warthin-Starry staining. Patients often have peliosis of the spleen and lymph nodes and cutaneous angiomatous lesions. Increased alkaline phosphatase is usually present. The lesions do respond to antibiotics and should be distinguished from Kaposi sarcoma (1). Peliosis of the liver and spleen may be the result of sinusoidal wall injury induced by tumor cells. Peliosis hepatis also occurs within hepatocellular neoplasms such as adenoma and carcinoma, and angiosarcoma. Peliosis hepatis has rarely been associated with hairy cell leukemia, angioimmunoblastic lymphadenopathy, myeloproliferative disorders, and Waldenström’s macroglobulinemia. Peliosis hepatis and sinusoidal dilatation involving the perivenular and midzones have been reported in patients with Hodgkin disease. Although peliosis hepatis is usually of no clinical significance, macroscopic peliosis of liver or spleen may rupture spontaneously or after trauma.
References
FIGURE 12. 1 Peliosis hepatis. Low power view of blood-filled space
within liver parenchyma.
187
1. Wanless IR. Vascular disorders. In: Burt AD, Portmann BC, Ferrell LD, eds. MacSween’s Pathology of the Liver. 5th ed. Philadelphia, PA: Churchill Livingstone/Elsevier; 2007:628–630. 2. Crawford JM, Liu C. Liver and biliary tract. In: Robbins, Cotran, eds. Pathologic Basis of Disease. 8th ed. Philadelphia, PA: Saunders/Elsevier; 2010:872.
Case 12.1
Alcoholic Lipopeliosis SANDRA FISCHER AND MAHA GUINDI
C L I N I C AL I N F OR M AT I ON
A 53-year-old man presented with decompensated end-stage liver disease due to alcoholic cirrhosis. The patient had been abstinent from alcohol for approximately 8 months prior to receiving a liver transplant. He was admitted 2 weeks prior to transplantation, with hepatorenal syndrome, encephalopathy, and refractory ascites. On the day preceding transplantation he developed hematemesis and hypotension and underwent upper endoscopy but no upper gastrointestinal source of bleeding was identified and no varices were seen. R E A S ON F OR R E F E R R A L
Decompensated cirrhosis likely due to alcoholic cirrhosis. PAT H OL OG I C F E AT U R E S
FIGURE 12.1.1 Alcoholic lipopeliosis. High power view of regenerative
The explanted liver showed severe cirrhosis, mild steatosis, but no active steatohepatitis. Many of the parenchymal nodules showed hepatocyte dropout of varying degrees, including dropout of the entire nodule, consistent with the ischemic complications of end-stage cirrhosis. There was a recanalized thrombus in the main portal vein, a common finding in end-stage cirrhosis and a feature that might lead to the necrosis of nodules. The sinusoids were dilated in areas and filled with large fat globules (Figure 12.1.1). The amount of lipopeliosis was small and focal, likely mirroring the small amount of steatosis.
nodule from explanted liver with alcohol-related cirrhosis showing fat globules released in distended sinusoids (elastic trichrome stain).
188
DIAGNO SIS
Alcoholic lipopeliosis arising in a setting of alcoholic cirrhosis.
Case 12.2
Lipopeliosis in Transplanted Donor Livers SANDRA FISCHER AND MAHA GUINDI
C L I N IC AL I N F OR M AT I ON
This is a deceased donor liver that had been retrieved at another hospital. Cold ischemic time was prolonged and the left lobe appeared ischemic. This was resected and the right lobe was engrafted. R E A SON F OR R E F E R R AL
Liver transplantation. PAT H OL OG I C F E AT U R E S
Histologic examination of the resected left lobe revealed zone 3 necrosis, which is in keeping with ischemic injury. There were perivenular (zone 3) large fat globules present within the sinusoids but not accompanied by an inflammatory reaction (Figure 12.2.1). A posttransplant liver biopsy of the right lobe showed typical lipopeliosis and cholestasis (Figure 12.2.2).
D I AG N OS I S
Lipopeliosis related to preservation injury.
D I S C U S S I ON
The term “lipopeliosis” had been introduced in 1992 to characterize an unusual liver lesion in which the sinusoids appear to become engorged by large fat globules (1). The mechanism
FIGURE 12. 2. 1 Early posttransplant cholestasis with lipopeliosis.
Low power view of biopsy showing steatosis and fat globules released in distended sinusoids.
F I G U R E 1 2 . 2 . 2 Alcoholic lipopeliosis. Macrophages around released fat (CD68 immunostain).
by which the lesion develops is that hepatocyte necrosis occurs in a steatotic graft after transplantation due to ischemia or preservation injury. The fat globules are then released from the injured hepatocytes and are sequestered in the sinusoids and/or the space of Disse (2). Lipopeliosis documented 1 week after transplantation can be cleared within 25 days, probably by macrophages. The clinical outcome can vary greatly and most probably depends on the extent of hepatocellular necrosis. Lipopeliosis seems to be fairly common; when it occurs it is not the primary cause of graft dysfunction (3). The “lipopeliosis” lesion has also been reported in native livers with alcoholic and nonalcoholic steatohepatitis and chronic hepatitis C with steatosis (4). In the nontransplant setting, a superimposed insult (eg., portal vein thrombosis, shock, hypovolemia associated with severe ascites and diuresis, or toxic/drug-related injury) may contribute to liver cell injury and necrosis, with the release of fat globules into the space of Disse. The lesion is easily detectable when florid but can be very subtle when mild, or when the biopsy is done later on in the course of the lesion, it may by then have started to resolve. A trichrome stain may help to make the extruded fat droplets stand out in contrast against the darker-staining surrounding hepatocytes (5). CD68 immunoperoxidase stain demonstrates the cytoplasm of macrophages surrounding the “empty spaces” that represent the extruded fat droplets indicating that the fat is no longer within the hepatocytic cytoplasm (Figure 12.2.2). Factor VIII–related antigen and type IV collagen immunoperoxidase stains help to delineate the contours of dilated sinusoids, or may show that fat droplets are present just outside the sinusoid, in the space of Disse and are compressing the sinusoids (Figure 12.2.3) (2).
189
190
CHAPTER
12:
PELIOSIS
H E PAT I S
References 1. Ferrell L, Bass N, Roberts J, Ascher N. Lipopeliosis: fat induced sinusoidal dilatation in transplanted liver mimicking peliosis hepatis. J Clin Pathol. 1992;45:1109–1110. 2. Cha I, Bass N, Ferrell LD. Lipopeliosis. An immunohistochemical and clinicopathologic study of five cases. Am J Surg Pathol. 1994;18: 789–795. 3. Bioulac-Sage P, Balabaud C, Ferrell L. Lipopeliosis revisited: should we keep the term? Am J Surg Pathol. 2002;26:134–135. 4. Guindi M. The “lipopeliosis” lesion can occur in native livers too, not in transplanted ones only [abstract 1286]. Mod Pathol. 2007;20(suppl 2); 280A. 5. Adeyi O, Fischer SE, Guindi M. Liver allograft pathology: approach to interpretation of needle biopsies with clinicopathological correlation. J Clin Pathol. 2010;63:47–74.
FIGURE 12. 2. 3 Alcoholic lipopeliosis. Type IV collagen immunostain. Immunoreactivity highlighting the sinusoids (arrows). Released fat droplets (asterisks) outside sinusoids in space of Disse, compressing sinusoids.
13 Portal Hypertension Without Cirrhosis
Case 13.1
Hepatoportal Sclerosis SANDRA FISCHER AND MAHA GUINDI
C L I N IC AL I N F OR M AT I ON
A 27-year-old man was admitted with massive variceal bleed. Physical examination revealed palm erythema, clubbing and leukonychia, and splenomegaly only. His investigations showed low hemoglobin at 108, platelets 96 with a mean corpuscular volume (MCV) of 75, and a low ferritin; international normalized ratio (INR) was 1.32 and albumin normal. Viral markers and other etiological tests to reveal the cause of any underlying liver disease were negative. Computed tomography (CT) showed lobar redistribution in the liver, with patent paraumbilical vein, and major vessels grossly unremarkable. R E A SON F OR R E F E R R AL
The history and subsequent investigations suggest portal hypertension, but the biopsy does not show unequivocal features of cirrhosis.
F I G U R E 1 3 . 1 . 1 Hepatoportal sclerosis. Liver needle biopsy with no evidence of cirrhosis (Masson’s Trichrome stain).
PAT H OL OG I C F E AT U R E S
Sections from liver tissue showed mild architectural changes with only few fibrous septae. Features of cirrhosis were not appreciated (Figure 13.1.1). There was marked portal fibrosis and obliteration of intrahepatic terminal portal veins (Figure 13.1.2). Iron and periodic acid–Schiff diastase (PASd) stains were negative. Biliary-type lesions were not identified, and the bile ducts were preserved in numbers. Also, there was no significant necroinflammation to suggest hepatitis, and no steatosis (Figure 13.1.1).
D I AG N OS I S
Loss of portal veins with secondary arterialization of the liver, consistent with hepatoportal sclerosis.
D I S C U S S I ON
The pathologic features raise the possibility of noncirrhotic portal hypertension. Obliteration of small portal veins (obliterative portal venopathy) may develop secondary to local
inflammation, thrombosis, congestive portal venopathy, and toxic injury (eg, chronic exposure to arsenic). Hepatoportal sclerosis is essentially a diagnosis of exclusion, and other causes of portal vein loss should be excluded before a diagnosis of hepatoportal sclerosis is made. Highly regressed cirrhosis, extrahepatic portal vein thrombosis, hepatic vein thrombosis, intrabiliary parasites, and schistosomiasis represent important differential diagnoses (1). Other conditions associated with noncirrhotic portal hypertension include myeloproliferative syndromes, human immunodeficiency virus (HIV) infection, and the rare Adams-Oliver syndrome. In the 1960s, investigators in India introduced the term noncirrhotic portal hypertension to describe an entity that is characterized histologically by “obliterative portovenopathy” (2). Mikkelsen and coworkers introduced the term hepatoportal sclerosis in 1965 to describe similar histopathological findings in noncirrhotic liver (3). Other diagnostic names given to this condition include Banti disease, noncirrhotic portal fibrosis, noncirrhotic intrahepatic portal hypertension, benign intrahepatic portal hypertension, and idiopathic presinusoidal portal hypertension.
191
192
CHAPTER
13:
P O RTA L
HYPERTENSION
WITHOUT
CIRRHOSIS
identifiable alteration in hepatic features and the portal vein. In a patient with portal hypertension, it is crucial to confirm the presence or absence of cirrhosis, including the possibility of regressed (remodeled) cirrhosis. Size of the biopsy specimen size is very important, to accurately assess fibrosis. Regressed cirrhosis in particular can be easily missed when biopsy specimens are shorter than 2 cm. Regressed cirrhosis should be suspected when there is a reduction of both portal and hepatic veins, delicate remnants of fibrous septa, and an irregular arrangement of portal structures and hepatic veins, typically when hepatic veins are located in close approximation to portal tracts (1). If small portal veins are obliterated, hepatic veins are patent, and no architectural distortion is present, the disease likely involves only the portal veins or portal tracts.
References FIGURE 13.1.2 Hepatoportal sclerosis. Portal tract with fibrous expansion and obliteration of portal vein, hematoxylin and eosin (HE) stain.
The diagnosis of hepatoportal sclerosis requires consideration of clinical and imaging information in addition to biopsy findings. Idiopathic or noncirrhotic portal hypertension can be diagnosed on the basis of imaging studies that show no
1. Wanless IR. Vascular disorders in the liver. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract, and Pancreas. 2nd ed. Philadelphia, PA: Saunders/Elsevier;2009:1169–1229. 2. Dhiman RK, Chawla Y, Vasishta RK, et al. Non-cirrhotic portal fibrosis (idiopathic portal hypertension): experience with 151 patients and a review of the literature. Gastroenterol Hepatol. 2002;17(1):6–16. 3. Sarin SK, Kapoor D. Non-cirrhotic portal fibrosis: current concepts and management. J Gastroenterol Hepatol. 2002;17(5):526–534.
Case 13.2
Portal Vein Thrombosis SANDRA FISCHER AND MAHA GUINDI
C L I N I C A L I N F OR M AT I ON
This 55-year-old male, who had a diagnosis of Crohn disease 15 years ago, complicated by bowel perforation for which he underwent a terminal ileum resection, developed abdominal pain and loss of weight. Imaging studies showed that he had disease in the terminal ileum. He was started on antibiotics, steroids, and Imuran. A complete blood count (CBC) showed that his hemoglobin, white blood cell count, and platelets were low. A CT scan of the abdomen at that time showed that he had proximal superior mesenteric vein thrombosis, and he was started on Coumadin. At the same time, an upper gastrointestinal (GI) endoscopy showed that he had grade 2 to 3 esophageal varices, and he was started on Nadolol. A needle biopsy was performed to confirm/exclude cirrhosis. He later underwent liver transplantation. F I G U R E 1 3 . 2 . 2 Portal vein thrombosis. Portal tract containing ob-
R E A SON F OR R E F E R R AL
The history and clinical investigations suggest cirrhosis with portal hypertension, but a liver biopsy did not show underlying chronic liver disease. PAT H OL OG I C F E AT U R E S
A needle biopsy showed the majority of portal areas to have a small or absent portal vein (Figure 13.2.1). Slight ductular reaction was seen but no dilated forms to suggest congenital hepatic fibrosis. No evidence of cirrhosis was present. Furthermore, the explanted liver showed focal intimal thickening involving the left and right portal veins and
literated portal vein and several large arteries. The surrounding liver parenchyma has sinusoidal dilatation secondary to increased arterial supply (Masson’s Trichrome stain).
some smaller portal veins consistent with healed thrombus. The main portal vein was widely patent. CD34 immunostaining showed increased angiogenesis and some arterialization of the liver (Figure 13.2.2).
DIAGNO SIS
Portal vein thrombosis (PVT).
DISCUSSIO N
FIGURE 13. 2. 1 Portal vein thrombosis: large portal area with obliterated portal vein showing smooth muscle hypertrophy (Masson’s Trichrome stain).
Both local (hepatobiliary) and systemic (thrombophilic) risk factors are associated with thrombosis of the portal vein (1). In children, infectious causes of PVT, such as sepsis or omphalitis, are frequently present. Particularly in neonates, catheterization of the umbilical vein is an important risk factor for development of PVT. In the adult population, liver cirrhosis and hepatobiliary malignancies are the most common local precipitating factors that together account for a large proportion of cases of PVT. Based on clinical presentation and results of imaging, acute and chronic PVT can be identified. An acute obstruction of the main portal vein usually manifests itself as a sudden onset of abdominal pain, which may be very severe. On physical examination the majority of patients will exhibit splenomegaly, but ascites is usually absent. Doppler ultrasound, computerized tomography (CT), or magnetic resonance imaging (MRI) can all be applied to demonstrate either the absence
193
194
CHAPTER
13:
P O RTA L
HYPERTENSION
of flow or the presence of a thrombus in the portal vein. The diagnosis of chronic PVT is often incidental or made later on when complications of chronic PVT occur. In response to PVT, portoportal and portosystemic collateral veins will develop to compensate for the decreased portal blood flow. The amount, size, and localization of collaterals differ strongly between patients. On imaging, the presence of a network of collateral vessels around the portal vein, a so-called portal cavernoma, is a typical feature of chronic PVT (1). Obliteration of small portal veins as seen in needle biopsies is a nonspecific finding. A peripheral liver biopsy is not indicated to establish a diagnosis of PVT as large vessels are not sampled with this procedure. Recanalization of large portal vein thrombi make these lesions difficult to recognize from
WITHOUT
CIRRHOSIS
a clinical and radiological point of view. Prior thrombosis is suspected when there is prominent intimal fibrosis of portal veins, especially those larger than 200 μm in diameter (2). Sometimes, a compensatory increase in the arterial supply to the liver following PVT can generate features that mimic nodular regenerative hyperplasia with parenchymal atrophy in zone 3 and periportal hyperplasia.
References 1. Hoekstra J, Janssen HL. Vascular liver disorders (II): portal vein thrombosis. Neth J Med. 2009;67(2):46–53. 2. Wanles IR. Vascular disorders in the liver. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract, and Pancreas. 2nd ed. Philadelphia, PA: Saunders/Elsevier; 2009:1169–1229.
Case 13.3
Budd-Chiari Syndrome SANDRA FISCHER AND MAHA GUINDI
C L I N IC AL I N F OR M AT I ON
A 47-year-old woman developed increased abdominal girth and ascites. Abdominal ultrasound showed irregular liver contour and hypertrophy of the caudate lobe. The large hepatic veins were not well visualized. Angiography revealed narrowing of the right and middle and left hepatic veins, and the right hepatic vein appeared obstructed. A diagnosis of Budd-Chiari syndrome (BCS) was made. The patient was referred to a hematologist, but no hypercoagulable state was found. Repeat ultrasound showed a hypervascular nodule in the right lobe approximately 1.8 cm in diameter. Its exact nature was not determined, but hepatocellular carcinoma was in the differential diagnosis. The patient subsequently developed liver failure and underwent cadaveric orthotopic liver transplantation. R E A SON F OR R E F E R R AL
Characterization of pattern of parenchymal injury and classification of right lobe nodule. PAT H OL OG I C F E AT U R E S
The features are of congestive liver disease in keeping with hepatic venous outflow obstruction evidenced by the fibrous intimal thickening and luminal obstruction by organized thrombus involving hepatic veins of all sizes (Figure 13.3.1). There is associated sinusoidal dilatation and congestion (Figure 13.3.2) with areas of hepatocyte necrosis and dropout. The pattern of parenchymal injury is variable. In areas where there is mainly hepatic vein obstruction without secondary
FIGURE 13. 3. 1 Hepatic vein thrombosis. Medium- and small-sized obliterated hepatic veins; perivenular congestive necrosis (elastic trichrome stain).
portal vein disease, there is a venocentric pattern of hepatocyte loss and/or septation (Figure 13.3.2). Hepatocyte necrosis near the obstructed hepatic veins results in a central-veinto-central-vein pattern of necrosis and, subsequently, fibrosis (Figure 13.3.2). In areas where there is secondary portal vein obstruction, venoportal septa are present. The right lobe nodule is a large parenchymal regenerative nodule (Figure 13.3.3). DISCUSSIO N
Acute obstruction of all 3 hepatic veins causes massive hepatomegaly with prominent engorgement of the liver parenchyma. With time, the affected lobes become smaller, whereas unaffected regions undergo compensatory hypertrophy, a phenomenon that most frequently affects the caudate lobe (1). Biopsies or explants show severe congestion and dilatation of sinusoids especially on zone 3. There may be hemorrhage within the liver cell plates, and recent thrombi within hepatic veins of any size and/or intimal fibrous thickening of hepatic veins larger than 100 μm in diameter. There is intimal thickening by fibrosis of the large- and medium-sized hepatic veins. Most veins undergo some recanalization with the formation of multiple luminal channels or delicate webs (1). Two patterns of parenchymal injury may be seen: venocentric and venoportal (1,2). Cases of pure hepatic vein obstruction typically lead to a characteristic venocentric pattern of septation also known as reverse nodularity. Hepatocyte necrosis occurs mainly near the obstructed hepatic veins, which results in a central-vein-to-central-vein pattern of necrosis and, subsequently, fibrosis. The portal tracts are seldom incorporated into the fibrous septa. When secondary portal vein thrombosis occurs, venoportal septa and, eventually, venoportal cirrhosis may develop (a pattern similar to that in posthepatitic cirrhosis) (1,2). Regions of liver with only mild parenchymal disease may develop a pattern of atrophy and compensatory hyperplasia that can be mistaken for nodular regenerative hyperplasia (1). Liver biopsies are helpful in confirming the diagnosis of hepatic venous outflow obstruction and exclude other entities that may be associated with very high transaminase elevations, but are immune-mediated, such as viral or autoimmune hepatitis. Questions regarding the severity of the disease including the presence of necrosis and fibrosis may be posed by the clinician at the time of biopsy. Regional variation in pathologic features may lead to spurious conclusions. Therefore, biopsy specimens taken for this purpose should be obtained from two sites to minimize sampling error (1). Further confirmation can be achieved by correlation with serum aminotransferase levels, liver function tests, and imaging studies of the large hepatic veins.
195
196
CHAPTER
13:
P O RTA L
HYPERTENSION
WITHOUT
CIRRHOSIS
FIGURE 13.3.2 (A) Hepatic vein (HV) with surrounding marked sinusoidal congestion in zone 3 (asterisk) (Masson’s Trichrome stain). (B) Chronic congestion with sinusoidal fibrosis and red blood cells present in the space of Disse (asterisk) (Masson’s Trichrome stain). (C) venocentric pattern of necrosis. Hepatocyte necrosis mainly near obstructed hepatic veins, causing a central-vein-to-central-vein pattern of necrosis and, subsequently, fibrosis. The portal tracts (PT) (arrow) are seen at the center of the periportal sleeve of residual hepatocytes (arrowheads) (Masson’s Trichrome stain). (D) Venoportal necrosis. Hepatocyte necrosis extending between portal tract (arrow pointing to bile duct in portal area) and hepatic vein, and, eventually, venoportal septation and venoportal cirrhosis when secondary portal vein thrombosis occurs (Masson’s Trichrome stain).
In severe cirrhosis, of any cause, there is often intimal thickening of medium-sized and large hepatic veins due to congestive venopathy, or thrombosis (3). As a result, accurate histologic differentiation of hepatic vein thrombosis from other types of cirrhosis often requires large enough tissue samples to allow for an evaluation of the large hepatic veins (1). Whether obstruction to blood flow is at the level of the small or large hepatic veins, histologic features are similar. Thus the differential diagnosis of BCS in biopsies includes congestive heart failure, constrictive pericarditis, and chronic veno-occlusive disease; and interpretation of the biopsy needs to be in the context of clinical history. In BCS, venous outflow obstruction may be high grade. Associated disturbed portal flow without extrahepatic portal vein thrombosis leads to decreased portal venous blood flow and increased arterial flow (4). A marked increase in hepatic
arterial perfusion was observed mainly in patients with longstanding BCS where it is preceded by severely impaired portal perfusion (4). Enlargement of the hepatic artery in this setting is regarded as a progressive and reactive compensatory adaptation. In order to maintain hepatic blood inflow in the face of a decrease in portal perfusion, arterial enlargement occurs analogous to the effect of “hepatic arterial buffer response” described by Lautt et al (5). Focal nodular hyperplasia (FNH)-like large regenerative nodules develop independent of cirrhosis in patients with chronic BCS where there is an enlarged hepatic artery (2,4). FNH-like nodules, that is nodular reactive hepatocytic hyperplastic lesions, occur in areas of arterial hyperemia which are highly perfused and shown to possess arborization in a nodular arrangement by arteriography (4). Thus, they appear similar in pathogenesis to the sporadic cases of FNH reported by
CASE
13.3:
BUDD-CHIARI
SYNDROME
197
FIGURE 13.3.3 (A) Large regenerative nodule with some focal nodular hyperplasia (FNH)-like features (Masson’s Trichrome stain). (B) Large
regenerative nodule. Artery associated with ductules (arrow head) and cholestatic hepatocytes (arrow) (Masson’s Trichrome stain). (C) Large regenerative nodule. Unpaired artery amidst benign hepatocytes (Masson’s Trichrome stain).
Wanless et al (6). These FNH-like lesions or large regenerative nodules are composed of benign hepatocytes arranged in double cell plates (2). The lesions, similar to sporadic FNH, do not contain portal veins. Cholestatic features may be present with rosettes, canalicular bile plugs, and collections of foamy macrophages. Cholestatic features tend to be more prominent in the larger nodules. Ductular reaction is seen in approximately 50% of nodules. The nodules appear to be supplied by an arterial tree with radial branches extending from the center of the nodule (2). These vessels are accompanied by variable amounts of fibrous stroma. The distinction of large adenoma-like or FNH-like regenerative nodules from true neoplastic lesions such as hepatic adenoma or well-differentiated hepatocellular carcinoma can be difficult. The difficulty stems from the fact that large adenoma-like or FNH-like regenerative nodules share a number of histologic features with well-differentiated hepatocellular neoplasms: lack of significant nuclear and architectural atypia and an arterial supply as shown by angiographic
and histological findings (7). In one report, large nodules were indistinguishable from adenomas (7). In another study, Wanless et al found large regenerative nodules differ from hepatic adenoma in that they have a single branching arterial supply that is often accompanied by a duct and/or ductular reaction. In routine forms of FNH, glutamine synthetase (GS) immunostaining marks hepatocytes in large anastomosing areas in a “map-like” pattern, often surrounding hepatic veins, whereas GS was not expressed in hepatocytes close to fibrotic bands containing arteries and ductules (8). In comparison, in hepatocellular adenomas or well-differentiated hepatocellular carcinoma, GS was shown to be positive but with a different perivenular or diffuse pattern rather than “map-like.” Thus GS represents a useful marker to distinguish true FNH from other hepatocellular nodules (8), but this same map-like pattern is not a reliable marker for FNH-like lesions, which can show variable staining patterns, as compared with true FNH. As noted above, distinction from well-differentiated hepatocellular carcinoma can also be difficult. Lack of any
198
CHAPTER
13:
P O RTA L
HYPERTENSION
cytologic atypia and absence of reticulin deficiency is helpful, and absence of immunoreactivity with Glypican-3 immunostain can be helpful. Glypican-3 has high sensitivity for the diagnosis of hepatocellular carcinoma, but one should note that this stain is much less sensitive in extremely well-differentiated hepatocellular carcinomas (9). In one series, Glypican-3 expression was seen in only 50% of well-differentiated hepatocellular carcinomas. All hepatic adenomas and large regenerative nodules were negative, but in the same series 43% of high-grade dysplastic nodules were positive for Glypican-3 (9). Thus Glypican-3 is often not helpful in this setting. Clinical context and correlation with newer imaging techniques such as contrast ultrasound is also needed in these situations.
References 1. Wanless IR. Vascular disorders in the liver. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI Tract, Liver, Biliary Tract, and Pancreas. 2nd ed. Philadelphia, PA: Saunders/Elsevier;2009:1169–1229. 2. Tanaka M, Wanless IR. Pathology of the liver in Budd-Chiari syndrome: portal vein thrombosis and the histogenesis of veno-centric cirrhosis,
3.
4.
5.
6. 7.
8.
9.
WITHOUT
CIRRHOSIS
veno-portal cirrhosis, and large regenerative nodules. Hepatology. 1998;27:488–496. Wanless IR, Wong F, Blendis LM, et al. Hepatic and portal vein thrombosis in cirrhosis: possible role in development of parenchymal extinction and portal hypertension. Hepatology. 1995;21:1238–1247. Cazals-Hatem D, Vilgrain V, Genin P, et al. Arterial and portal circulation and parenchymal changes in Budd-Chiari syndrome: a study in 17 explanted livers. Hepatology. 2003;37:510–519. Lautt WW, Legare DJ, d’Almeida MS. Adenosine as putative regulator of hepatic arterial flow (the buffer response). Am J Physiol. 1985;248: 331–338. Wanless IR, Mawdsley C, Adams R. On the pathogenesis of focal nodular hyperplasia of the liver. Hepatology. 1985;5:1194–1200. de Sousa JM, Portmann B, Williams R. Nodular regenerative hyperplasia of the liver and the Budd-Chiari syndrome. Case report, review of the literature and reappraisal of pathogenesis. J Hepatol. 1991;12:28–35. Bioulac-Sage P, Laumonier H, Rullier A, et al. Over-expression of glutamine synthetase in focal nodular hyperplasia: a novel easy diagnostic tool in surgical pathology. Liver Int. 2009;29(3):459–465. Shafizadeh N, Ferrell LD, Kakar S. Utility and limitations of glypican-3 expression for the diagnosis of hepatocellular carcinoma at both ends of the differentiation spectrum. Mod Pathol. 2008;21:1011–1018.
Case 13.4
Regressed Cirrhosis Case SANDRA FISCHER AND MAHA GUINDI
C L I N IC AL I N F OR M AT I ON
This 54-year-old male, born in Taiwan, immigrated to Canada 1 year ago, when he was found to have mildly elevated liver enzymes by his family physician: alanine transaminase (ALT) and aspartate transaminase (AST) in the 40 to 50 U/L range. Workup revealed positive hepatitis B (HBV) serology including positive HBV anticore antibodies (HBcAb) and HBV e-antigen (HBeAg). Abdominal ultrasound showed no mass lesions. There was hypertrophy of the left lobe of the liver and very mild nodularity of the liver contour raising the possibility of cirrhosis. An upper endoscopy showed grade esophageal varices. Platelet count was at the low end of normal. R E A SON F OR R E F E R R AL
Clinical history and investigations suggest cirrhosis with portal hypertension but a liver biopsy only showed a small amount of fibrosis and no obvious parenchymal nodules.
F I G U R E 1 3 . 4 . 2 Regressed cirrhosis. Obliterated terminal hepatic vein and perivenular sinusoidal dilatation (Masson’s Trichrome stain).
PAT H OL OG I C F E AT U R E S
The biopsy is of adequate size. The portal tracts appear remodeled with reduced collagen and many lack the portal vein (portal remnants) (Figure 13.4.1). Small hepatic veins are reduced in number and appear obliterated (Figure 13.4.2). There is loss of the normal alternation of portal tracts with hepatic veins. There is a small amount of visible fibrosis in the form of delicate fibrous spikes on portal areas and delicate septation
F I G U R E 1 3 . 4 . 3 Regressed cirrhosis. Expanded portal areas showing
long and delicate bridging fibrous septa (Masson’s Trichrome stain).
(Figure 13.4.3). Some septa appear incomplete or interrupted/ perforated (Figure 13.4.4). A few collagen fibrils are scattered in the parenchyma (Figure 13.4.5). There are a few occasions of ectopically placed hepatic veins (approximated to or abutting the adjacent portal tract (Figure 13.4.6).
FIGURE 13. 4. 1 Regressed cirrhosis. Arrow pointing to portal rem-
nant (portal tract lacking the portal vein); small bile duct apparent in portal remnant (arrowhead). Sinusoidal dilatation likely reflects shunting in cirrhosis (Masson’s Trichrome stain).
199
DIAGNO SIS
Fibrosis with features of regressed cirrhosis.
200
CHAPTER
13:
P O RTA L
HYPERTENSION
WITHOUT
CIRRHOSIS
FIGURE 13.4.4 Regressed cirrhosis. Interrupted portal-to-portal sep-
F I G U R E 1 3 . 4 . 6 Regressed cirrhosis. Portal-to-central approxima-
tum, arrow pointing to gap in septum. Notice each of the 2 portal areas joined by this septum is a portal remnant (Masson’s Trichrome stain).
tions (adhesions) (Masson’s Trichrome stain). HV, hepatic vein; PT, portal tract.
FIGURE 13. 4. 5 Regressed cirrhosis. Collagen fibrils (arrows) interspersed in the parenchyma (Masson’s Trichrome stain).
amount was less than found in normal portal tracts, so that hepatocyte plates closely approximated the portal tract elements (1). Wanless et al suggest that livers with micronodular cirrhosis, macronodular cirrhosis, and incomplete septal cirrhosis (ISC) demonstrate a histologic continuum and that a continuum of regressive changes can be present within individual livers as well (1). Incomplete septal cirrhosis (ISC) is characterized histologically by delicate and incomplete fibrous septa (3,4). Livers with ISC demonstrated all of the features of the hepatic repair complex, including venoportal adhesions and aberrant veins that would not be expected in early-stage disease (1). These results supported the notion that ISC may arise from complete cirrhosis as regression of fibrosis occurs. It appears cirrhosis must become inactive with little or no ongoing parenchymal injury to allow for repair such that fibrous scars have the opportunity to become delicate and highly regressed. The differential diagnosis of regressed cirrhosis includes causes of noncirrhotic portal hypertension. Of special note are hepatoportal sclerosis, portal vein obstruction by any cause, for example portal vein thrombosis and nodular regenerative hyperplasia. Hepatoportal sclerosis (HPS) is discussed in detail in case 13.1. The Asian Pacific Association for the Study of the Liver (APASL) set up a Working Party on Portal Hypertension in 2002 with a mandate to develop a consensus on the various clinical aspects of portal hypertension, including consensus guidelines on the definition of HPS. The consensus was published in 2007 (5). The APASL consensus stipulated that HPS/noncirrhotic portal fibrosis (NCPF)/idiopathic portal hypertension (IPH) comprises a group of diseases that are characterized by an increase in portal pressure in the absence of cirrhosis of the liver and that the entity is of uncertain etiology. Subcapsular septation can be noted, whereas deeper parenchyma shows normal
D I SC U SSI ON
The features of regressed cirrhosis have been described by Wanless in several publications (1,2). Fibrosis gradually disappears leaving behind a remodeled parenchyma as a clue that overt cirrhosis was once present. The regression parameters (hepatic repair complex) are delicate perforated septa, isolated thick collagen fibers, delicate periportal fibrous spikes, portal tract remnants, hepatic vein remnants with prolapsed hepatocytes, hepatocytes within portal tracts or splitting septa, minute regenerative nodules, and aberrant parenchymal veins. Aberrant parenchymal veins were veins close to portal tracts within 5 hepatocyte diameters. Portal tract remnants had several appearances, including artery/duct pairs, unaccompanied arteries, or unaccompanied ducts, usually with absent portal vein. Portal tract collagen
CASE
13.4:
REGRESSED
architecture. The liver functions and structure primarily remain normal, and usually there are no signs of chronic liver disease (5). Histological features noted in autopsies include occasional portal to portal or portal to central bridging septa may be present (5,6). Thus HPS and regressed cirrhosis share minimal visible fibrosis, loss of small portal veins, and absence of overt parenchymal nodules surrounded by fibrosis. Dense portal tract fibrosis may also be the residuum of regressed cirrhosis (2). The distinction between HPS and regressed cirrhosis could be especially difficult in limited tissue such as a liver needle biopsy, especially if small in size. The presence of liver parenchymal architectural distortion (features listed above) in addition to the portal tract changes would favor regressed cirrhosis. Previous liver biopsies should be sought in a given patient in whom a current biopsy raises the possibility of regressed cirrhosis. Previous biopsy(ies), may show overt features of severe fibrosis or cirrhosis, thus providing a major clue to the diagnosis. This was indeed the case in the patient reported by Wanless et al, with chronic hepatitis B virus (1). The patient’s first biopsy showed cirrhosis with sinusoidal fibrosis and severe chronic hepatitis. A second biopsy 1 year later after lamivudine therapy showed no sinusoidal fibrosis, and apparent enlargement of cirrhotic nodules compared to the first biopsy. A third biopsy after another 2.5 years of therapy showed no overt nodularity or cirrhosis, only 1 incomplete fibrous septum, and a few small regions of atrophy. The APASL consensus defined HPS as being of uncertain etiology, thus the presence of known causes of chronic liver disease, for example positive viral serology for hepatitis B or hepatitis C, would make HPS an unlikely contender for portal hypertension (5). Since the APASL consensus regarded HPS as a “group of diseases,” it is possible that some cases of regressed cirrhosis were inadvertently included in reported cases of HPS, especially those reported prior to the description of regression in human cirrhosis. This notion is strengthened by some of the reported liver capsular appearances in HPS as showing “some
CIRRHOSIS
CASE
201
nodularity resembling cirrhosis” (a feature of cirrhosis) upon gross examination of autopsy livers (5). Cirrhotic livers may show similar hypertrophic/atrophic variation within nodules and might be mistaken for nodular regenerative hyperplasia. In cirrhosis, especially after PVT, the centers of ischemic cirrhotic nodules become atrophic with some secondary hyperplasia usually at the periphery of these nodules. This pattern is similar to that in NRH, except that the underlying stroma is scarred in cirrhosis (7). It is not deemed suitable to use the term “nodular regenerative hyperplasia” in cirrhotic livers as that term has been applied to livers having little or no fibrous septation in the original description of NRH (8,9).
References 1. Wanless IR, Nakashima E, Sherman M. Regression of human cirrhosis. Morphologic features and the genesis of incomplete septal cirrhosis. Arch Pathol Lab Med. 2000;124:1599–1607. 2. Wanless IR. Cirrhosis. In: Odze RD, Goldblum JR, eds. Surgical Pathology of the GI tract, Liver, Biliary Tract, and Pancreas. 2nd ed. Philadelphia, PA: Saunders/Elsevier; 2009:1115–1145. 3. Popper H. What are the major types of hepatic cirrhosis? In: Ingelfinger F, Relman A, Finland M, eds. Controversy in Internal Medicine. Philadelphia, PA: Saunders; 1966:233–243. 4. Nevens F, Staessen D, Sciot R, et al. Clinical aspects of incomplete septal cirrhosis in comparison with macronodular cirrhosis. Gastroenterology. 1994;106:459–463. 5. Sarin SK, Kumar A, Chawla YK, et al. Noncirrhotic portal fibrosis/ idiopathic portal hypertension: APASL recommendations for diagnosis and treatment. Hepatol Int. 2007;1:398–413. 6. Schiano TD, Kotler DP, Ferran E, Fiel MI. Hepatoportal sclerosis as a cause of noncirrhotic portal hypertension in patients with HIV. Am J Gastroenterol. 2007;102:1–5. 7. Shimamatsu K, Wanless IR. Role of ischemia in causing apoptosis, atrophy, and nodular hyperplasia in human liver. Hepatology. 1997;26:343–350. 8. Wanless IR. Micronodular transformation (nodular regenerative hyperplasia) of the liver: a report of 64 cases among 2,500 autopsies and a new classification of benign hepatocellular nodules. Hepatology. 1990;11: 787–797. 9. Terminology of nodular hepatocellular lesions. International Working Party. Hepatology. 1995;22:983–993.
This page intentionally left blank
14 Clinical and Morphological Spectrum of Liver Diseases in Pregnancy ANDREW KENNETH BURROUGHS AND AMAR PAUL DHILLON
I N T ROD U C T I ON
Liver disease in pregnancy can be considered in 2 categories. The first group consists of all of the acute and chronic diseases of the liver that can affect women of child-bearing age, including viral hepatitis, gallstone disease, and drug reactions. The frequency, clinical aspects, and histopathology of these conditions are no different generally from their expression in the population from which the pregnant patient is drawn (1). In regions such as Asia, Africa, and Central America where hepatitis E virus (HEV) is endemic, fulminant HEV hepatitis can occur with a high mortality. Herpes simplex virus is uncommon, but because these patients may present with severe fulminant hepatitis and effective antiviral treatment is available, a low threshold of suspicion for HSV infection must be maintained. The second group of diseases are particular to pregnancy. These comprise hyperemesis gravidarum, benign recurrent cholestasis (recurrent intrahepatic cholestasis of pregnancy, RCP), acute fatty liver of pregnancy (AFLP), and toxaemia/HELLP syndrome. The tendency of pregnancy to provoke itching and cholestasis can lead to the discovery of previously unsuspected chronic liver disease such as primary biliary cirrhosis, primary sclerosing cholangitis, or HCV infection. All pregnant women should be tested for HBV infection at the first antenatal visit, but this is a council of perfection, and HBV disease may be unsuspected until a later stage of the pregnancy when it can cause diagnostic confusion (1). The fact that pregnancy-independent diseases share morphological features with pregnancy-associated liver diseases means that before concluding a diagnosis of the latter, the possibility of hepatic drug reaction, viral, alcoholic, obesityrelated, biliary, and autoimmune diseases must be considered carefully clinically. Pregnancy-associated liver diseases usually occur in late pregnancy, and so the pregnancy itself will usually be obvious clinically. However, in some cases, for example, cases of severe RCP or toxaemia in patients with pre-existing kidney disease and hypertension, the disease may occur sooner, and pregnancy may not be immediately apparent. Therefore, the possibility of pregnancy should be considered specifically, confirmed, and the relevant information must be transmitted to the hepatopathologist, together with the liver biopsy. Close clinicopathological communication is always advisable, and this is especially so when considering pregnancy-associated liver disease. With AFLP especially the diagnosis critically rests on the demonstration of microvesicular steatosis and unless this is considered at an early stage with retention of appropriate material for frozen section staining for
fat (or pre-embedment osmium impregnation), the diagnosis is bound to remain speculative. The role of liver biopsy in the diagnosis and management of pregnancy-related liver disease is quite limited. In most instances, the diagnosis has to be made with clinical and noninvasive investigations only. As obstetric indications for early delivery of the fetus take precedence, biopsy may be contraindicated in view of the coagulopathy that can exist in the more severe pregnancy-related liver diseases, although transjugular biopsy is safe. When a liver biopsy is obtained in cases of suspected liver disease in pregnant patients, one of the most important considerations is the use of the biopsy to identify or exclude additional liver diseases coincidental to the pregnancy. H Y P ER EMESIS GR AV IDA RUM
This is severe and intractable vomiting requiring intravenous hydration. It occurs early in about 0.3% of pregnancies (1). Liver biopsy is rarely indicated unless it is necessary to exclude other conditions such as drug reaction and viral hepatitis, which are usually coincidental to the pregnancy itself. In cases of hyperemesis when liver biopsy is performed and when there are no other superimposed conditions, the histopathological appearance is within normal limits, or there may be minor, nonspecific changes. R ECUR R ENT INT R A H EPAT IC CH O LESTA S IS O F P R EGNA NCY
Recurrent cholestasis of pregnancy occurs in late pregnancy and is probably a result of various hormonal (including pregnancy-related elevation of oestrogen) and genetic factors. Approximately 0.1% of pregnancies can be affected (1). Up to 15% of RCP cases may be associated with the MDR3 (ABCB4) gene mutation, and in these cases, the cholestasis may be severe with an earlier (mid-trimester) onset than otherwise (2). The defect can be associated with gallbladder and intrahepatic cholesterol stones (3). Other bile acid transporter defects may be relevant in other cases (4). A family history of RCP or familial cholestatic disease is relevant, and further consideration of MDR3 mutations is necessary particularly in patients with cholesterol gallstones below 40 years of age (3). Liver biopsy in typical cases of RCP will show canalicular parenchymal cholestasis with a centrilobular predominance. Little inflammation is seen.
203
204
CHAPTER
14:
LIVER
DISEASES
AC U T E FAT T Y L I V E R OF P R E G N A N C Y
AFLP is a mitochondrial cytopathy (5). It affects about 1:14 000 pregnancies (1), but is more frequent in patients with mitochondrial fatty acid oxidation defects, for example, mothers who are heterozygous for long chain 3 hydroxyl co-enzyme A dehydrogenase deficiency (LCHAD), carrying a fetus homozygous for LCHAD, as the mother cannot metabolise the excess free fatty acids. It is more common in pregnancies with males and twins. Patients present with worsening nausea, vomiting, and abdominal pain followed by jaundice. There is hypoglycaemia and coagulopathy, and there can be gastrointestinal bleeding and hyponatremia. Signs of pre-eclampsia may be present in 50% of patients. Lactic acidosis and metabolic acidosis are related to defective mitochondrial energy supply and defects in oxidative phosphorylation. The compromised maternal mitochondrial metabolic reserve leads to accumulation of hepatocytic trigycerides (6). This is evident as a predominantly centrilobular hepatic microvesicular steatosis, which characterises AFLP, and this feature is considered diagnostic. Hepatocytic mitochondria may be enlarged and show crystalline inclusions. The steatosis tends to obscure concomitant liver cell loss, which can be seen as areas of collapse in the reticulin preparation and reactive Kupffer cells in the periodic acid–Schiff diastase stain. Centrilobular hepatocytes can also show regenerative mitotic activity. In severe cases, imaging studies will show reduction of the size of the liver. However, microvesicular steatosis has also been found in liver tissue from patients with pre-eclampsia (7). Therefore, it could be argued that the microvesicular steatosis in AFLP is not necessarily pathognomonic of the “AFLP syndrome.” It could be that AFLP and eclampsia are either related conditions or that they can co-exist together because they are not mutually exclusive. As usual, the histopathology must be interpreted within the relevant clinical context. TOX E MI A / H E L L P S Y N D ROM E
The relationship between toxaemia/HELLP syndrome and their pathophysiology is unsettled. Toxemia has been considered an unsatisfactory term because no causative toxin has
IN
PREGNANCY
been identified. Toxemia/pre-eclampsia signifies the presence of edema, proteinuria, and hypertension (T140/90) in a previously normotensive patient or worsening hypertension in patients with pre-existing hypertension usually in late pregnancy. This occurs in 5% to 10% of pregnancies. Eclampsia is when the condition is severe and there is cerebral involvement, with convulsions. There may be particularly pronounced hepatic involvement with hemolysis (H), elevated liver “function” tests (EL), and low platelet count (LP) constituting the “HELLP” syndrome (1). However, many patients with HELLP syndrome (which overall occurs in 0.2% to 0.6% of pregnancies) do not exhibit preceding hypertension and proteinuria. HELLP syndrome is considered to be a hepatic microangiopathy with hemolysis and endothelial damage, microthrombi with fibrin deposition and platelet consumption causing areas of hemorrhage and necrosis possibly with hematomas, capsular tearing, and bleeding into the peritoneum. Mild pre-eclampsia can show sinusoidal fibrin deposition with focal Disse space hemorrhage, and the histopathological diagnostic distinction from severe eclampsia/HELLP syndrome is basically a matter of the extent and degree of the vascular changes (2).
References 1. Hay JE. Liver disease in pregnancy. Hepatology. 2008;47:1067–1076. 2. Hepburn IS, Schade RR. Pregnancy-associated liver disorders. Dig Dis Sci. 2008;53:2334–2358. 3. Wasmuth HE, Glantz A, Keppeler H, et al. Intrahepatic cholestasis of pregnancy: the severe form is associated with common variants of the hepatobiliary phospholipid transporter ABCB4 gene. Gut. 2007; 56:265–270. 4. Hardikar W, Kansal S, Elferink RPJO, Angus P. Intrahepatic cholestasis of pregnancy: when should you look further? World J Gastroenterol. 2009;15:1126–1129. 5. Sherlock S, Dooley J. Diseases of the Liver and Biliary System. p477 10th ed. Blackwell Science; 1997. 6. Lee NM, Brady CW. Liver disease in pregnancy. World J Gastroenterol. 2009;15:897–906. 7. Burt AD. p899 Chapter 17: Liver pathology associated with diseases of other organ systems. Burt AD, Portmann B, Ferrell L, eds. MacSween’s Pathology of the Liver. 5th ed. Elsevier; 2007.
Case 14.1
Recurrent Cholestasis of Pregnancy ANDREW KENNETH BURROUGHS AND AMAR PAUL DHILLON
C L I N IC AL I N F OR M AT I ON
A 23-year-old woman presented with jaundice and itching 1 week after delivery at term of a normal male baby. She had a history of recurrent jaundice including a postpartum episode 2 years ago. Serum bilirubin was 6 times normal, alkaline phosphatase was 2 times normal, aspartate aminotransferase (AST) was 2 times normal, alanine aminotransferase (ALT) was 2 times normal, and platelets were 400 109/L. There was a “large gallstone” in the gallbladder but no obvious biliary dilatation on ultrasound imaging. R E A SON F OR R E F E R R AL
Recurrent cholestasis of pregnancy versus gallstone disease. PAT H OL OG I C F E AT U R E S
The liver biopsy at low magnification shows disarray of the liver cell plates (Figure 14.1.1). Marked parenchymal canalicular cholestasis is present, with “cholestatic rosettes” (Figures 14.1.2 and 14.1.3). Compared to the degree of histological cholestasis, there is little inflammation. Occasional eosinophils are present (Figure 14.1.2). There is little portal inflammatory or ductular reaction, no ductular bile plugging, and no portal tract edema or fibrosis (Figure 14.1.4). No copperassociated protein deposition was seen with the orcein stain.
F I G U R E 1 4 . 1 . 2 Central venule with pericentral canalicular cholesta-
sis and a little inflammation including an eosinophil (right border of picture) ( 400).
D I AG N OS I S
Parenchymal cholestasis consistent with recurrent cholestasis of pregnancy.
F I G U R E 1 4 . 1 . 3 Parenchymal canalicular cholestasis and a choles-
tatic “rosette” ( 400). DISCUSSIO N
F I G U R E 1 4 . 1 . 1 Disorganization of liver cell plates, with portal tract at the bottom of the picture, and central venule at the top ( 25).
The appearance is of histologically “pure” parenchymal cholestasis (Table 14.1.1). There are no “biliary” features such as portal tract expansion, edema, fibrosis, and copper-associated protein deposition, and there is little portal inflammation and ductular reaction which argue against biliary obstruction related to the gallstone that was found on ultrasound imaging. On the other hand, the presence of a gallstone at the early age (23 years) of this patient is consistent with a bile acid transporter defect.
205
206
CHAPTER
14:
LIVER
DISEASES
FIGURE 14. 1. 4 Portal tract with a little inflammation and little duc-
tular reaction ( 100). TA B LE 14. 1. 1 Causes of histologically “pure” parenchymal cholestasis Drug reaction Sepsis Lymphoma (as a noninfiltrative paraneoplastic phenomenon, for example, Hodgkin’s) Biliary transporter defect (including BRIC and recurrent cholestasis of pregnancy) Abbreviations: BRIC, benign recurrent intrahepatic cholestasis.
IN
PREGNANCY
The clinical differential diagnosis includes biliary (gallstone, primary sclerosing cholangitis, primary bilary cirrhosis) tract disease, viral hepatitis, and drug reaction (1). Malignancy such as Hodgkin’s lymphoma may manifest as a noninfiltrative paraneoplastic phenomenon. Each of these possibilities must still be considered specifically and excluded clinically. No ductular bile plugging (which might have indicated sepsis) was present. The finding of scattered eosinophils is unhelpful: their presence does not confirm drug reaction and their absence does not exclude drug reaction. Bile acid transporter genetic testing and documentation of relevant family history should be performed. In specialist centers, where these investigations are available, immunostaining for different biliary transporters can help to characterise cases where there is a deficiency. Recurrent cholestasis of pregnancy cannot be diagnosed definitively by liver biopsy alone, and further confirmatory clinical and laboratory investigations are necessary.
Reference 1. Hepburn IS, Schade RR . Pregnancy-associated liver disorders. Dig Dis Sci. 2008;53:2334–2358.
Case 14.2
Acute Fatty Liver of Pregnancy ANDREW KENNETH BURROUGHS AND AMAR PAUL DHILLON
C L I N IC AL I N F OR M AT I ON
A 31-year-old woman presented at 37 weeks of gestation with fetal distress. An emergency Caesarean section was performed at which time she was noted to be jaundiced. Blood tests (taken prior to the Caesarean section) shows that serum bilirubin was 2 times normal, alkaline phosphatase was 2 times normal, AST was 7 times normal, ALT was 10 times normal, platelets were 101 3 109/L, and blood glucose was 3.9 mmol/L. Liver ultrasound showed no abnormality. R E A SON F OR R E F E R R AL
Acute fatty liver of pregnancy versus drug reaction versus toxemia/HELLP syndrome. PAT H O L OG I C A L F E AT U R E S
At low magnification (Figure 14.2.1), the biopsy shows centrilobular hepatocytic cytoplasmic microvesicular change with normal portal tracts, and at higher magnification (Figures 14.2.2 and 14.2.3), the microvesicular change is confirmed with centrally placed hepatocytic nuclei and expanded, foamy cytoplasm with small droplets of fat. There is centrilobular liver cell loss, collapse, and mild inflammation including pigmented macrophages (Figure 14.2.4). Scanty normoblasts are present (Figure 14.2.5).
F I G U R E 1 4 . 2 . 2 Centrilobular microvesicular change with loss of centrilobular hepatocytes and a mild inflammatory infiltrate ( 100).
D I AG N OS I S
Acute fatty liver of pregnancy.
F I G U R E 1 4 . 2 . 3 Microvesicular steatosis: the hepatocyte nucleus is centrally placed and the cytoplasm is expanded, foamy, and contains small droplets ( 400).
DISCUSSIO N
FIGURE 14. 2. 1 Microvesicular change of centrilobular hepatocytes, with little inflammation ( 50).
The clinical context and blood tests suggest the possibility of a “hepatitic” process, and indeed to some extent this is explained by the centrilobular liver cell loss, which can be obscured by the occupation of these areas by hepatocytes swollen with microvesicular fat in acute fatty liver of pregnancy (AFLP). Bearing in mind the other causes of microvesicular steatosis (Table 14.2.1), specific clinical exclusion of drug reaction remains necessary. However, the lack of a sharp cutoff between the areas with liver cell loss and microvesicular change and the periportal/midlobular hepatocytes, and the lack of coagulative
207
208
CHAPTER
14:
LIVER
DISEASES
IN
PREGNANCY
sized vacuoles, and centrally placed nuclei that may be indented by microvesicles. One should exclude a similar appearance due to liver cell ballooning as might be seen in alcoholic steatohepatitis or in association with cholestasis (a particular problem because the latter can also be a more prominent feature of AFLP than is evident in the present case). Staining for cytokeratin 8 can be helpful here: in microvesicular steatosis, the hepatocyte cytoskeleton is seen to be intact (Figure 14.2.6). Actually, this can also be appreciated in the haematoxylin and eosin preparation in well-fixed specimens (Figure 14.2.3). In ballooning degeneration, for example, in alcoholic steatohepatitis, apart from the clues rendered by the other features of alcoholic liver disease such as pericellular fibrosis and larger droplet fat, the ballooned hepatocytes of steatohepatitis show disruption of the cytoskeleton, which becomes clumped with other components of Mallory bodies (Figure 14.2.7). Ideally, of FIGURE 14. 2. 4 A periodic acid–Schiff diastase preparation highlights the centrilobular liver cell loss, collapse, and pigmented macrophages ( 200).
FIGURE 14. 2. 5 A few normoblasts are present in the centre of this
microscopic field ( 400).
F I G U R E 1 4 . 2 . 6 Cytokeratin 8 immunostaining demonstrates the
intact cytoskeleton of microvesicular steatotic hepatocytes ( 400).
TA B LE 14. 2. 1 Causes of microvesicular steatosis Acute fatty liver of pregnancy Drugs (valproate, acetaminophen, tetracycline, nucleoside analogues) Alcoholic foamy degeneration Urea cycle disorders Reye’s syndrome Total parenteral nutrition
necrosis are points in favour of AFLP, rather than, for example, acetaminophen toxicity. The occasional normoblasts seen in this case corresponds with the observation that circulating normoblasts are often found in AFLP (1). Without suitable material for the crucial histological demonstration of fat, close attention to morphological detail is required for the diagnosis of microvesicular steatosis. Microvesicular change is characterised by relatively uniformly sized hepatocytes, foamy cytoplasm with distinct uniformly
F I G U R E 1 4 . 2 . 7 Cytokeratin 8 immunostaining demonstrates the
disrupted cytoskeleton of ballooned hepatocytes in a case (not the present case) of alcoholic steatohepatitis ( 400).
CASE
14.2
:
ACUTE
FAT T Y
FIGURE 14. 2. 8 A case of microvesicular steatosis (not the present case) with a pre-embedment osmium stain (Marchi method) for fat ( 400).
course, if there is effective clinicopathological communication and if appropriate material can be organised, fat can be demonstrated directly and without equivocation (Figure 14.2.8).
LIVER
OF
PREGNANCY
209
Even when fatty microvesicular change has been established, this supports AFLP, but this feature alone is neither specific nor conclusive of “AFLP syndrome”; (fatty) microvesicular change can be found as a feature within the setting of other conditions such as toxemia/HELLP syndrome and drug reaction; conversely, relatively minor degrees of necrosis, hemorrhage, and fibrin deposition can be seen in AFLP; more hemorrhage, fibrin, thrombi, and necrosis than in the present case would be expected in toxemia/HELLP syndrome; a greater degree of inflammation would raise other questions such as viral hepatitis. Concluding that AFLP syndrome is the predominant diagnosis requires a balanced evaluation of all of the clinical and pathological features as a whole, and is a matter of the degree of each of the changes in any individual case that supports or argues against each of the competing diagnoses, which after all may co-exist and may not in fact be mutually exclusive.
Reference 1. Burroughs AK, Seong NH, Dojcinov DM, Scheuer PJ, Sherlock SV. Idiopathic acute fatty liver of pregnancy in 12 patients. Q J Med. 1982;51:481–497.
Case 14.3
Toxemia/HELLP Syndrome ANDREW KENNETH BURROUGHS AND AMAR PAUL DHILLON
C L I N I C AL I N F OR M AT I ON
A 32-year-old woman presented at term feeling unwell and nauseous. There was no documented preeclampsia. Blood tests were: serum bilirubin was 2 times normal, ALP was normal, AST was 89 times normal, ALT was 86 times normal, platelets were 71 109/L, international normalized ratio was 2.4, and blood glucose was 2.7 mmol/L. There was evidence of renal failure (creatinine 2 times upper limit of normal range). Microangiopathic hemolytic features were not evident on blood film examination. Clinically, an acute hepatitic injury was diagnosed and an emergency Caesarean section was performed.
7 immunostaining reveals bile ducts and ductules within the necrotic regions (Figure 14.3.4). Diagnostic features are absent of most types of drug- and virus-related massive hepatic necrosis (MHN), venous outflow obstruction (VOO), and chronic liver disease. Viral inclusions (eg, herpes simplex virus [HSV]) are not present.
DIAGNO SIS
Toxemia/HELLP syndrome of pregnancy.
R E A S ON F OR R E F E R R A L
Severe acute hepatitis versus toxaemia/HELLP syndrome. PAT H O L OG I C AL F E AT U R E S
The biopsy shows extensive, irregular areas of coagulative and hemorrhagic necrosis, and minimal inflammation (Figure 14.3.1). There is so much necrosis that the relationship of the lobular structures has been obscured. A small amount of parenchyma remains, and in places this is seen to have a perivenular location (Figure 14.3.2). Portal tracts are difficult to identify (Figure 14.3.3). However, cytokeratin
F I G U R E 1 4 . 3 . 2 A rim of hepatocytes remains around a central
venule ( 100).
FIGURE 14. 3. 1 Extensive coagulative and hemorrhagic necrosis with some small islands of residual hepatocytes ( 25).
F I G U R E 1 4 . 3 . 3 An unrecognizable structure lies buried in a necrotic
region ( 100).
210
CASE
14.3
:
TOXEMIA/HELLP
FIGUR E 14. 3. 4 Cytokeratin 7 immunostaining reveals that the struc-
ture is a portal tract including an abortive ductular reaction ( 100).
TA B LE 14. 3. 1 Causes of periportal hepatocellular damage Toxemia/HELLP syndrome of pregnancy Disseminated intravascular coagulation Toxins (phosphorus, ferrous sulphate, cocaine) Acute viral hepatitis (eg, HAV, HEV) Small for size liver grafts Abbreviations: HAV, hepatitis A virus; HEV, hepatitis E virus.
D I S C U S S I ON
The pathological features in this case support the interpretation of toxemia/HELLP syndrome rather than the alternative diagnoses. A key point in this case is the irregularity of the necrosis and hemorrhage (due to the superimposed vascular compromise of toxemia/HELLP syndrome) combined with a periportal propensity of the necrosis and preferential preservation of perivenular parenchyma. This combination favors toxemia/HELLP syndrome and effectively argues against some of the more important differential diagnostic possibilities such as necrosis due to drug toxicity and VOO that usually exhibit a more regular pattern of liver damage (see below). There may be little inflammation (as in this case) in severe hepatitis, but the rest of the histopathology in this case excludes the main clinical alternative diagnosis of MHN associated with fulminant hepatic failure due to drug reaction or viral infection (see below). Other causes of periportal hepatocellular damage have to be considered as well (Table 14.3.1). In the original histopathological assessment, there was difficulty in identifying the periportal predilection of the necrosis. Immunostaining for biliary cytokeratins can be helpful in orientation. Without specific attention to the combination of patchy hemorrhage and irregular areas of necrosis with a periportal predilection, or the recognition of residual
SYNDROME
211
centrilobular hepatocytes, there was difficulty in concluding any morphological diagnosis beyond “nonspecific hepatocellular necrosis.” In most cases of MHN caused by viral infection and drug reaction hepatitis, hepatic necrosis tends to be central/ midlobular. Consequently, even when most or all of the hepatocytes have been lost, with subtotal parenchymal reticulin collapse the portal tracts remain intact and are brought closer together (“approximated”) and are evenly spaced. They are made prominent with a florid ductular reaction and portal lymphoplasmacytic infiltrate. The necrosis of acetaminophen toxicity can be confused with toxemia, and exclusion of acetaminophen overdose should be addressed specifically by clinical enquiry and investigations. Microvesicular steatosis can be seen in both acetaminophen toxicity and toxemia. Attention to the morphology of acetaminophen-related centrilobular coagulative necrosis (in the early postacetaminophen time period) or the empty, collapsed macrophage-rich centrilobular regions (in subsequent postacetaminophen time periods) seen in acetaminophen-related injury is informative. In the present case, the retention of a rim of centrilobular hepatocytes (rather than the periportal rim of residual hepatocytes that would be expected in acetaminophen toxicity) is a strong pointer toward a diagnosis of toxemia/HELLP syndrome. In both toxemia/HELLP syndrome and VOO, liver hemorrhage can be seen in the space of Disse. In VOO, the hemorrhage is in a regular, repeated centrilobular location. The hemorrhage is accompanied by centrilobular hepatocytic atrophy and loss, centrilobular sinusoidal ectasia, and congestion, and there may be venous outflow stenosis and thrombosis. In toxemia/HELLP syndrome, the hemorrhage is irregular and unaccompanied by the other features characteristic of VOO. The patchy hemorrhage seen in the present case is also unusual in MHN caused by drug toxicity or viral infection, except HEV. Since HEV can cause severe liver disease in pregnant women with periportal injury as well, this possibility particularly should be excluded by the appropriate serological investigations (1). Appreciably more inflammation is usual in HEV, than was found in the present case. Another viral infection that is worthy of specific consideration and exclusion is HSV. Immunohistochemistry and serological investigation may be necessary (2). Even though in this case the biopsy appearance (detailed above) favors the morphological interpretation of toxemia/HELLP syndrome–related liver damage, it should be remembered that in cases of suspected toxemia/HELLP syndrome there is a large risk of biopsy sampling error because of the focal nature of the diagnostic features. Alternatively (as in this case), the tissue destruction can be so extreme that the relevant clues toward the correct diagnosis are obscured. Therefore, in either event, the absence of diagnostic features in a small biopsy sample cannot exclude, with confidence, a diagnosis of clinically suspected toxemia/HELLP syndrome. In cases with incomplete information and in atypical cases, biopsy is useful to confirm the clinical suspicion and to exclude additional and alternative diagnostic possibilities. For example, biopsy might form a useful part of the diagnostic
212
CHAPTER
14:
LIVER
DISEASES
jigsaw puzzle and may help to determine management in a patient presenting with fulminant hepatic failure, hypoglycemia, coagulopathy, and cerebral edema who could be suffering from acetaminophen toxicity (especially considering that toxicology several days after overdose when measurement of blood acetaminophen levels is unhelpful because serum levels have already fallen). Conversely, in a typical case of toxemia with a complete and consistent “noninvasive” dataset, liver biopsy is probably not necessary.
IN
PREGNANCY
References 1. Malcolm P, Dalton H, Hussaini HS, Mathew J. The histology of acute autochthonous hepatitis E virus infection. Histopathology. 2007;51:90–94. 2. Halliday J, Lokan J, Angus PW, Gow P. A case of fulminant hepatic failure in pregnancy. Hepatology. 2010;51:341–342.
15 Drug-Induced Liver Injury DAVID E. KLEINER
I N T ROD U C T I ON
Drug-induced liver injury (DILI) is the most common cause of acute liver failure (ALF) in the United States, with acetaminophen alone responsible for 46% of adult ALF cases between 1998 and 2007 (1). Other drugs accounted for a further 11%. Over a 24-year period, DILI was the most common cause for the FDA to withdraw or attach black box warnings to drugs (2). The incidence of DILI is difficult to estimate because of the difficulties in accurately capturing likely cases in national registries, but a careful prospective study in the Dijon region of France performed over a period of 3 years reported an incidence of 14 cases per 100 000 people (3). This rate is more than 10 times higher than previous estimates based on physicians reporting to national and international registries. Increased awareness of the significance of DILI in drug development and clinical care has led to the establishment of prospective registries in Spain, Sweden, and the United States (4–6). The U.S. network, known as the Drug-Induced Liver Injury Network (DILIN, www.dilin.org), is an NIDDK-funded consortium of academic centers spread across the country that not only collects new cases of DILI but also serves as a source of expertise in the diagnosis of DILI (7). Data from these national networks have demonstrated that most new (nonacetaminophen) cases of DILI are due to anti-infectives (27–46%), followed by central nervous system agents (13–17%) and anti-inflammatory and pain medications (5–17%) (4–6). In the Swedish and U.S. cohorts, there were more women than men (about 1.5–1), whereas in the Spanish study, the numbers were about equal. The death rate varied from 5% to 9.2%, whereas the rate of chronicity (defined as the persistence of biochemical injury more than 6 months after stopping the agent) was 11% to 13%. DILI can have medical, legal, and regulatory consequences, and it is critical that the diagnosis be made accurately. L I V E R B I OP S Y A N D D I L I
There are a number of reasons why a liver biopsy may be performed in a suspected case of DILI (8). Determination of the injury pattern can be important in implicating or excluding particular agents. Table 15.1 lists the major nonneoplastic injury patterns observed in DILI. DILI can mimic essentially all non-DILI patterns of injury, but any individual agent will have a more restricted injury profile. The first task of the pathologist is to classify the pattern(s) of injury present in the biopsy. The pattern of injury can then be compared to reported DILI patterns for the suspected agents and can also be used to focus the non-DILI differential diagnosis by eliminating diseases that do not cause that particular injury. The pattern of injury can also
have implications for the mechanism of injury. For example, acute or chronic hepatitis patterns may suggest an autoimmune component, granulomas and eosinophils may suggest hypersensitivity, zonal necrosis may suggest a metabolic toxin, and microvesicular steatosis may suggest mitochondrial injury. Of equal importance to the classification of pattern is the determination of the degree of severity of the injury. Discrepancy between the degree of serum biochemical abnormalities and the severity of histological injury is common. Some types of injury, such as the distinction between necrosis and hepatitis or the evaluation of duct integrity, can only be made on biopsy, and these features may have prognostic implications. Alternatively, if the injury is mild, a needed medication may be continued. This has been the role of liver biopsy in the case of methotrexate injury, but the principle can be applied in other situations. Liver biopsy may be performed when there are unexplained serum biochemistries. In these cases, the pathologist should be particularly alert to the possibility of DILI, especially caused by herbal or over-the-counter agents. Patients may not provide a history of taking herbal medications unless specifically asked, thinking that because they are “natural,” they are safe. Nevertheless, herbal agents are increasingly recognized as a cause of liver injury and accounted for 9% of cases in the DILIN experience (6). Some patterns of injury are more likely to be caused by drugs or other agents, including cholestatic hepatitis, granulomatous hepatitis, eosinophilia, microvesicular steatosis, and zonal necrosis (9). Vascular injury patterns in Table 15.1 should also prompt a detailed medication or occupational exposure history. Finally, whenever unusual mixed patterns are present, such as cholestasis with steatohepatitis, DILI should be suspected.
EVA LUAT IO N O F SUSP ECT ED DILI
Unlike other liver diseases where the diagnosis is established by clinical tests, like hepatitis C infection, or by identification of a characteristic pattern of injury, like steatohepatitis, DILI remains a diagnosis of exclusion. There is no single test or combination of tests that establishes the diagnosis of DILI, leaving it an area where thoughtful investigation and expert opinion are the gold standards (10). Because it is difficult for even experienced physicians to become experts in DILI, scoring systems have been created to fill the expertise gap. The 2 scales applicable to liver disease are the Roussel Uclaf Causality Assessment Method (RUCAM) (11) and the Clinical Diagnostic Scale (CDS) (12). Both systems take into account various factors, including the type of biochemical injury, timing of drug intake relative to the onset of injury, the
213
214
CHAPTER
15:
DRUG-INDUCED
LIVER
INJURY
TA B LE 15. 1 Non-neoplastic injury patterns in DILI Typical Biochemical Presentation1
Relative Frequency Within DILI
Examples of Non-DILI Etiologies
Zonal necrosis
H
Uncommon
Hypoxic injury
Acute hepatitis
H
Common
Acute viral hepatitis, AIH2
Chronic hepatitis
H
Uncommon
Chronic viral hepatitis, AIH2
Granulomatous
C
Rare
Sarcoidosis, Infections
Mononucleosis-like
M to H
Rare
EBV hepatitis
Cholestatic hepatitis
C to H
Common
Acute viral hepatitis, AIH2
Acute (intrahepatic) cholestasis
C to M
Uncommon
Sepsis, obstruction
Chronic cholestasis (including duct paucity)
C to M
Uncommon
PBC
Chronic cholestasis (duct sclerosis)
C to M
Rare
PSC
Microvesicular steatosis
M
Rare
Alcoholic foamy degeneration
Macrovesicular steatosis
M
Uncommon
NAFLD2
Steatohepatitis
M
Rare
NASH2
Rare
Artifact, mass lesions, outflow obstruction
Rare
Outflow obstruction
Budd–Chiari
Rare
Outflow obstruction
Hepatoportal sclerosis
Rare
Injury Pattern Necroinflammatory Patterns
Cholestatic Patterns
Steatotic Patterns
Vascular Injury Patterns Sinusoidal dilation/peliosis SOS/VOD2
Nodular regenerative hyperplasia
M to H
C to M
Rare
Collagen-vascular diseases, lymphoproliferative diseases
1 Biochemical presentation defined by ratio (R) of alanine aminotransferase (ALT) to alkaline phosphatase, normalized to the upper limit of normal. (H) Hepatocellular, R 5; (M) Mixed, 2 5 R 5; (C) Cholestatic, R 2. 2 Abbreviations: AIH, autoimmune hepatitis; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; SOS/VOD, sinusoidal obstruction syndrome/ veno-occlusive disease.
time to resolve the injury, age of the patient, exclusion of other etiologies and drugs, prior reports of toxicity for the suspected agent and positive rechallenge with the agent. These “one size fits all” systems cannot include the nuances of particular drug injuries and they tend to work best in simple clinical situations where the suspects are limited and the presentation falls into one of the more common hepatocellular or cholestatic patterns. Neither system uses information from the liver biopsy as it is not a required part of the clinical evaluation. In the 1970s, Dr. Irey, chair of the Department of Environmental Pathology at the Armed Forces Institute of Pathology, proposed a method of evaluation based on 6 general principles, outlined in Table 15.2 (13). The first principle is Temporal Eligibility, which considers the timing of drug exposure with the injury. Clearly, the suspect drug must have been taken prior to onset of injury, but it is not always clear when the onset was. The patient may have had symptoms for days or weeks prior to presentation. The duration of exposure is also important to consider, and it is somewhat drug-specific. As a general rule, a drug must be taken for a few weeks to 6 months before toxicity can develop.
This is a typical time period for metabolic injury to develop from the accumulation of toxic metabolite or for an immunologic response. There are many exceptions though. A toxic dose of acetaminophen can be taken all at once in a suicide attempt, and significant toxicity can develop in a short time period if maximum doses are taken by susceptible individuals (14). Toxicity to amoxicillin-clavulanate typically develops 1 to 3 weeks after a 10-day course of therapy is complete. A few drugs, like nitrofurantoin, may be taken for more than a year before the patient becomes symptomatic. Exclusion of Competing Causes is a necessary part of any DILI evaluation. The investigators of the DILIN network recently published criteria for documenting cases of DILI (15). Although the guidelines are meant to apply to publications, they serve as an outline of evaluation in cases of suspected DILI. It is important to consider the patient’s primary disease as well as comorbidities such as heart failure, hypotension, sepsis, diabetes, obesity, and alcohol use. Laboratory tests to exclude viral hepatitis and autoimmune liver disease are important, as are imaging tests to exclude abnormalities or obstruction of duct system. It may not be practical to perform
CHAPTER
15:
DRUG-INDUCED
TA B LE 15. 2 Elements of causality analysis Temporal eligibility
Was the suspected agent started prior to the onset of liver injury? Was the agent given for an appropriate time or at a dose consistent with causing injury?
Exclusion of competing causes
Have other non-DILI etiologies been excluded by appropriate history, laboratory testing, and imaging studies?
Known potential for injury
Has the agent been previously reported to cause DILI? How common is DILI from the agent?
Precedent for pathologic pattern
What biochemical and histological presentations have been reported for DILI caused by the agent?
Dechallenge/ rechallenge
Did stopping the agent result in recovery? Was the rate of recovery compatible with DILI? If the patient was given the agent again (either planned or unplanned), did the injury recur?
Toxicology
Were blood levels of the agent measured and were they in the known toxic range?
Conclusion
Based on the evaluation of presentation, exposure, pattern of injury and follow-up, estimate the probability of DILI due to the suspect agent (17)
Definite
95% chance—all competing causes excluded, typical injury pattern for agent, positive rechallenge (if attempted), characteristic clincopathologic pattern
Likely
75–95% chance—most other possibilities excluded, but clinicopathologic signature of DILI either not well described for agent or not perfectly matched
Probable
50–75% chance—competing causes unlikely but cannot be fully excluded
Possible
25–50% chance—other etiologies possible and cannot be excluded
Not DILI
25% chance—other etiology identified, pattern does not match agent
Abbreviation: DILI, drug-induced liver injury. From Ref. 13.
the full evaluation in all cases and the liver biopsy can used to help guide the evaluation. For example, obstruction of the biliary tree is unlikely to cause an acute hepatitis, and canalicular cholestasis would be an unlikely result of chronic viral hepatitis. On the other hand, it would be difficult to conclude that DILI was present in an obese, diabetic patient whose biopsy showed only steatohepatitis, regardless of how liver enzymes rose and fell in association with a particular drug. In consulting on a case of DILI, the pathologist should be clear about which diseases could result in the pattern of injury and therefore need to be rigorously excluded in order to conclude that a drug had caused injury. The drug’s Known Potential for Injury is an important piece of circumstantial evidence that may implicate one drug over another. Literature reports of DILI are very important in establishing a drug’s potential to cause serious harm. If only 1 or 2 case reports have been published despite widespread use
LIVER
INJURY
215
of a drug, then the likelihood is low that the drug is causing the patient’s liver injury. A patient taking both isoniazid and penicillin is much more likely to be harmed by the isoniazid, which has thousands of published examples of DILI, than the penicillin, for which only a couple of well-documented cases exist. There are many resources available to pathologists seeking information about DILI, including the online databases like Pubmed (www.ncbi.nlm.nih.gov/pubmed/) and other search engines, textbooks focused on DILI, and government agencies like the Food and Drug Administration (www.fda.gov). The literature search used to evaluate the drug’s potential for injury can be used to identify the biochemical and Pathologic Patterns associated with the suspect agent. The biochemical patterns are divided into hepatocellular, cholestatic, and mixed injury based on the ratio of ALT to alkaline phosphatase, after normalizing each laboratory value to its upper limit of normal (ULN) (16). The ratio, “R,” is defined at the point when the ALT or alkaline phosphatase rises to more than twice the ULN. When R is greater than 5, the injury is hepatocellular. Cholestatic injury is defined as an R value less than 2, whereas mixed injury has an R value between 2 and 5. The biochemical pattern is further refined by the presence of jaundice. Thus, if a drug typically causes hepatocellular jaundice, then this means that R is more than 5 and the bilirubin is elevated. Jaundice or elevated bilirubin levels are not required for an injury to be characterized as cholestatic. The biochemical pattern is poorly predictive of the histologic pattern, as can be seen in Table 15.1. A patient with hepatocellular jaundice might easily show zonal necrosis, acute hepatitis, or cholestatic hepatitis on biopsy. Cholestatic injury might turn out to be cholestatic hepatitis, granulomas, nodular regenerative hyperplasia, or even steatohepatitis on biopsy. The pathology pattern therefore provides much more information about the injury, which can also be compared to the literature reports of DILI by the suspected agent. Dechallenge is the act of observing what happens to the patient after the suspected drug is stopped. Most patients will show gradual recovery, with cholestatic injury typically taking longer to resolve than hepatocellular injury. About 10% of the time, the injury will take more than 6 months to resolve, despite therapy. In some patients with fulminant liver failure, it will not be possible to observe a full dechallenge because they undergo transplant or die. Similarly, a patient who develops drug-induced cirrhosis is unlikely to fully recover. In the absence of one of these special circumstances, failure of an injury to resolve is usually considered to be evidence against drug injury. Recurrence of an injury if a drug is restarted (rechallenge) is one of the strongest pieces of evidence for drug injury, but is unlikely to be used in practice except by accident or in situations where the drug is absolutely necessary. In immune-mediated injury, DILI may recur with as little as a single dose, whereas in injuries mediated by metabolic transformation, the drug may be needed to be given for weeks before recurrence of the injury. Toxicology is used rarely in DILI evaluation, and then only for specific agents like acetaminophen or for drugs with narrow therapeutic windows where blood-level monitoring is routinely performed. Most DILI is idiosyncratic, meaning that
216
CHAPTER
15:
DRUG-INDUCED
the dose is of little importance in deciding whether the patient will develop toxicity, and injury may occur while on therapeutic levels of medication. Toxicology may also serve to confirm that patients were taking specific medications when they are unable to provide history and other records are unavailable. After all of the evidence have been gathered, it may be possible to assign a likelihood of causality to the drug (Table 15.2). The terms presented here are those used by investigators in the DILIN network to assign levels of causality during expert review (17). Unfortunately, it may not be possible to complete the full analysis as outlined above during the time frame of liver biopsy evaluation. Depending on when the liver biopsy is performed during the course of the clinical workup, many of the tests excluding other causes of liver injury may not have been performed and DILI may only be suspected after reviewing the liver biopsy findings. In these situations, the pathologist’s diagnostic line should at least contain a description of the pattern of injury and its severity. The note should evaluate the findings in light of what is known at the time, including some evaluation of the potential of the suspect drugs to cause the observed histological changes. In the case examples that follow in this chapter, Irey’s method of evaluation will be used to show how the information in the history, on biopsy, and from the literature are combined to provide a full evaluation. AC K N OW L E D G M E N T
The author would like to acknowledge the DILIN network for the accrual and expert evaluation of 6 of the cases presented in this chapter (Cases 15.2, 15.3, 15.4, 15.5, 15.6, and 15.8).
References 1. Lee WM, Squires RH, Jr, Nyberg SL, Doo E, Hoofnagle JH. Acute liver failure: summary of a workshop. Hepatology. 2008;47:1401–1415.
LIVER
INJURY
2. Lasser KE, Allen PD, Woolhandler SJ, Himmelstein DU, Wolfe SM, Bor DH. Timing of new black box warnings and withdrawals for prescription medications. JAMA. 2002;287:2215–2220. 3. Sgro C, Clinard F, Ouazir K, et al. Incidence of drug-induced hepatic injuries: a French population-based study. Hepatology. 2002;36:451–455. 4. Andrade RJ, Lucena MI, Fernandez MC, et al. Drug-induced liver injury: an analysis of 461 incidences submitted to the Spanish registry over a 10-year period. Gastroenterology. 2005;129:512–521. 5. Bjornsson E, Olsson R. Outcome and prognostic markers in severe druginduced liver disease. Hepatology. 2005;42:481–489. 6. Chalasani N, Fontana RJ, Bonkovsky HL, et al. Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology. 2008;135:1924–1934, e1–e4. 7. Hoofnagle JH. Drug-induced liver injury network (DILIN). Hepatology. 2004;40:773. 8. Kleiner DE. The pathology of drug-induced liver injury. Semin Liver Dis. 2009;29:364–72. 9. Goodman ZD. Drug hepatotoxicity. Clin Liver Dis. 2002;6:381–397. 10. Rochon J, Protiva P, Seeff LB, et al. Reliability of the Roussel Uclaf causality assessment method for assessing causality in drug-induced liver injury. Hepatology. 2008;48:1175–1183. 11. Danan G, Benichou C. Causality assessment of adverse reactions to drugs—I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol. 1993;46:1323–1330. 12. Maria VA, Victorino RM. Development and validation of a clinical scale for the diagnosis of drug-induced hepatitis. Hepatology. 1997;26:664–669. 13. Irey NS. Teaching monograph. Tissue reactions to drugs. Am J Pathol. 1976;82:613–647. 14. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: a randomized controlled trial. JAMA. 2006;296:87–93. 15. Agarwal VK, McHutchison JG, Hoofnagle JH; Drug-Induced Liver Injury Network. Important elements for the diagnosis of drug-induced liver injury. Clin Gastroenterol Hepatol. 2010;8:463–470. 16. Benichou C. Criteria of drug-induced liver disorders. Report of an international consensus meeting. J Hepatol. 1990;11:272–276. 17. Fontana RJ, Watkins PB, Bonkovsky HL, et al. Drug-Induced Liver Injury Network (DILIN) prospective study: rationale, design and conduct. Drug Saf. 2009;32:55–68.
Case 15.1
Acetaminophen-Induced Fulminant Liver Failure DAVID E. KLEINER
C L I N IC AL I N F OR M AT I ON
A 6-year-old boy presented to the emergency room after being found poorly responsive and lethargic by his mother. The child had developed a febrile illness about 1 week prior to admission and was receiving regular doses of a pediatric formulation of acetaminophen. When this medication ran out, the mother had started giving him a regular adult strength pill but was not sure how much he had received. He also had gastrointestinal symptoms with nausea and vomiting and had had little to eat or drink in the past 3 days. Laboratory tests sent from the emergency room showed an ALT of 8200 IU/L, an aspartate aminotransferase (AST) of 9600 IU/L, a total bilirubin of 9 mg/dL, and an acetaminophen level of 79 μg/mL about 15 hours from his last dose. N-acetylcysteamine was administered, but over the next day the child developed grade 3 encephalopathy and hypotension. An emergency liver transplantation was performed.
F I G U R E 1 5 . 1 . 1 Acetaminophen toxicity with zonal necrosis of perivenular hepatocytes with periportal sparing.
R E A SON F OR R E F E R R AL
Determine etiology of liver failure. PAT H OL OG I C F E AT U R E S
Sections of the explanted liver showed preserved hepatic architecture. Confluent coagulative necrosis was present in zone 3 and focally extended to zone 2 (Figure 15.1.1). Focally, the necrosis appeared to bridge between central veins. A small amount of hemorrhage was present around some of the veins. Examination of the interface between necrotic and viable liver showed occasional apoptotic hepatocytes in the viable areas (Figure 15.1.2). The viable hepatocytes were swollen and ballooned, and some were clearly steatotic. Mitotic figures were easily identified among the viable hepatocytes (Figure 15.1.3). There was no visible bile in hepatocytes or canaliculi. The portal areas showed occasional inflammatory cells, but the portal structures were intact and unremarkable (Figure 15.1.4). The degree of necrosis varied across the section from 30% to 50% of the parenchyma. D I AG N OS I S
Zone 3 necrosis with steatosis due to acetaminophen injury. D I S C U S S I ON
Most drugs that cause liver injury are idiosyncratic hepatotoxins. The liver injury is uncommon to very rare, unpredictable, not dependant on dose and not reproducible in animal
F I G U R E 1 5 . 1 . 2 Interface between necrotic and viable hepatocytes.
models. Acetaminophen is one of the rare drugs that is an intrinsic hepatotoxin, causing reproducible, dose-dependent toxicity. Hepatic injury occurs when the capacity to detoxify acetaminophen by the normal mechanisms of glucuronidation and sulfation is exceeded leading to oxidation by the cytochrome P450 (CYP) system to a toxic metabolite, N-acetylp-benzoquinone (NAPQI) (1). This can occur through drug overdose, either accidentally or with suicidal intention or at normal doses where the balance of detoxification has shifted. This latter situation applies when the hepatocyte glutathione levels have been depleted from malnutrition or chronic alcoholism. Alcohol (and other drugs) increases the level of CYP
217
218
CHAPTER
15:
DRUG-INDUCED
FIGURE 15. 1. 3 Steatosis and ballooning of viable hepatocytes with
mitotic activity.
LIVER
INJURY
and jaundice may develop. During the next couple of days the patient may feel somewhat better, but serious signs of liver injury appear 3 to 4 days after a toxic dose. The transaminase levels can be very high, over 10 000 IU/L, and jaundice may appear or worsen. Gastrointestinal symptoms may recur along with hepatic encephalopathy. The prognosis is poor if there is marked prolongation of the prothrombin time, renal failure, lactic acidosis, cerebral edema, or grade 3 or 4 encephalopathy. If the patient survives without a need for liver transplant, there is gradual resolution of the injury over the next several months (1). Although biopsies are seldom performed during the acute episode, the histologic changes are characteristic and may be seen on explants or at autopsy. There is zonal necrosis that is centered in zone 3 but may involve other zones in severe cases. The necrosis is confluent and coagulative, although apoptotic bodies can be seen among the viable hepatocytes. Inflammation is usually minimal, but scattered neutrophils and macrophages may be seen at the edge of the necrotic zone. The remaining viable hepatocytes usually show macrovesicular steatosis. Zone 3 necrosis is an uncommon pattern in DILI but can be seen with exposure to organics and toxins such as carbon tetrachloride and tannic acid. Halothane is also reported to cause zonal necrosis. The main non-DILI cause of zone 3 necrosis is hypoxic-ischemic injury; so in patients with heart failure, hypotension, or shock, it is important to correlate changes in enzymes with other clinical findings. Acetaminophen levels can be tested, and there is a useful nomogram that relates blood levels and time from last dose to toxicity. Acetaminophencysteine adducts can also be measured and have proven useful in the evaluation of unexplained ALF (5). If the patient presents soon after taking an overdose of acetaminophen, N-acetylcysteine can be given to prevent hepatotoxicity. If, as in this case, toxicity has already developed, the chances of reversing or limiting the injury are less. The causality analysis is presented in Table 15.1.1. In this case, even though it was not known how much acetaminophen had
FIGURE 15. 1. 4 Representative portal area showing minimal inflam-
matory infiltrate.
2E1, allowing oxidation to occur more readily. Although serious toxicity occurs when the dose of acetaminophen overwhelms the detoxification pathways, liver enzyme elevations can occur at normal doses in healthy adults (2). Acetaminophen is the single most common cause of acute liver failure (ALF) in the United States according to data gathered by the Acute Liver Failure Study Group. Between 1998 and 2003, the proportion of acetaminophen cases in their cohort rose from 28% to 51% (3). The proportion of ALF due to acetaminophen is lower in children, being 18% in children of 3 to 18 years of age and only 3% of infants under 3 years of age (4). After an overdose of acetaminophen, several stages of toxicity are recognized. Within the first day there were nonspecific gastrointestinal symptoms with nausea, vomiting, and abdominal pain as well as lethargy and malaise. The transaminases start to rise during this period,
TA BL E 1 5 . 1 . 1 Causality analysis—acetaminophen Temporal eligibility
Acetaminophen given over the last week, dose unknown
Exclusion of competing causes
Hypotension developed after onset of symptoms, no other drugs given, other etiologies of liver failure not tested
Known potential for injury
Acetaminophen is a known intrinsic hepatotoxin
Precedent for pathologic pattern
Zone 3 necrosis with steatosis is the classic injury pattern of acetaminophen
Dechallenge/ rechallenge
Dechallenge unable to be performed due to transplant. Rechallenge not performed.
Toxicology
Blood levels of acetaminophen were 79 μg/mL at 15 hours, well within the toxic range
Conclusion
Zone 3 necrosis with steatosis due (95% chance) to acetaminophen
CASE
15.1:
A C E TA M I N O P H E N - I N D U C E D
been given, toxicology could be performed. This piece of information, combined with the characteristic histologic pattern allows the diagnosis of acetaminophen toxicity with near certainty. Even though hypotension developed just prior to transplant, the biochemical injury preceded that complication. As a final cautionary note, this case illustrates the unfortunate, unintentional overdosing of a child with acetaminophen, a situation that can have serious consequences (6).
References 1. Larson AM. Acetaminophen hepatotoxicity. Clin Liver Dis. 2007;11: 525–548 vi.
FULMINANT
LIVER
FAILURE
219
2. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: a randomized controlled trial. Jama. 2006;296:87–93. 3. Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: results of a United States multicenter, prospective study. Hepatology. 2005;42:1364–1372. 4. Lee WM, Squires RH Jr, Nyberg SL, et al. Acute liver failure: Summary of a workshop. Hepatology. 2008;47:1401–1415. 5. James LP, Alonso EM, Hynan LS, et al. Detection of acetaminophen protein adducts in children with acute liver failure of indeterminate cause. Pediatrics. 2006;118:e676–e681. 6. Heubi JE, Barbacci MB, Zimmerman HJ. Therapeutic misadventures with acetaminophen: hepatotoxicity after multiple doses in children. J Pediatr. 1998;132:22–27.
Case 15.2
Statin-Associated Acute Hepatotoxicity DAVID E. KLEINER
C L I N I C AL I N F OR M AT I ON
An 82-year-old man with a history of diabetes and gastroesophageal reflux presented with a several week history of epigastric pain and nausea. He was jaundiced on physical examination. One month prior his liver tests had been essentially normal, but now testing showed ALT of 1737 IU/L, AST of 1919 IU/L, alkaline phosphatase of 260 IU/L, and total bilirubin of 5.3 mg/dL. Serologic results were negative for hepatitis A, B, and C, and the ANA was negative. ASMA was weakly positive. An abdominal CT showed no evidence of gallstones or biliary dilation. A medication history revealed that he had started simvastatin about 4 months prior to presentation. Other current medications included metformin (long term), bupropion (long term), and escitalopram (4 days). All medications were stopped. F I G U R E 1 5 . 2 . 2 Plasmacellular infiltrate in a portal area with inter-
R E A S ON F OR R E F E R R A L
face hepatitis and duct injury.
To confirm suspected drug-induced liver injury. PAT H OL OG I C F E AT U R E S
Examination of the liver biopsy showed a diffuse hepatocellular injury characterized by numerous foci of lobular inflammation, scattered apoptotic hepatocytes, Kupffer cell hyperplasia, and lobular disarray (Figure 15.2.1). With the portal areas, the infiltrate varied from mild and periportal to dense and was associated with interface hepatitis and bile duct injury (Figure 15.2.2). Plasma cells were present in increased
F I G U R E 1 5 . 2 . 3 Steatosis and ballooning injury with lobular
inflammation.
FIGURE 15. 2. 1 Marked lobular inflammation and lobular disarray in
statin-induced liver injury.
numbers in both the portal areas and the parenchymal infiltrate. Small fat droplets were present in most hepatocytes. Occasional foamy cells with microvesicular steatosis were also present as well as rare ballooned hepatocytes (Figure 15.2.3). There was canalicular and hepatocellular cholestasis, which was pale and difficult to see on routine stains (Figure 15.2.4), but clearly visible on the copper stain (Figure 15.2.5). The Masson Trichrome stain demonstrated portal fibrotic expansion (Figure 15.2.6).
220
CASE
15.2:
S TAT I N - A S S O C I AT E D
ACUTE
H E PAT O T OX I C I T Y
221
DIAGNO SIS
Acute cholestatic hepatitis due to simvastatin. Probable underlying fatty liver disease.
DISCUSSIO N
FIGURE 15. 2. 4 Dilated canaliculi with pale-staining bile.
FIGURE 15. 2. 5 Pale green bile is visible within canaliculi and hepatocytes on rhodanine stain.
FIGURE 15. 2. 6 Portal fibrotic expansion on Masson Trichrome stain.
The statins are a class of lipid-lowering drugs that act by inhibiting hydroxymethylglutaryl coenzyme A reductase, a key enzyme in the cholesterol biosynthetic pathway. Their general effectiveness, safety profile, and expanding list of indications have made statins among the most commonly prescribed medications worldwide. Clinical trials of statins have reported a low (0.5–2%) incidence of ALT elevations greater than 3 times the ULN (1,2). Although these biochemical abnormalities have caused some concern, they are generally asymptomatic and without serious consequences (3). Studies have shown that statins can be safely used in the presence of underlying liver diseases such as nonalcoholic fatty liver disease (4). Of more concern are the relatively rare reports of serious hepatotoxicity. Drug-Induced Liver Injury Network investigators have recently reviewed literature evidence of serious statin hepatotoxicity (5). Statins have been reported to cause 2 major patterns of serious liver injury. The more common form is an acute cholestatic hepatitis, as demonstrated by this case. All statins currently in use have been reported to cause this injury. Patients have usually been taking the statin for several months prior to developing hepatotoxicity, but a few patients have been reported with injury only after 3 or 4 years of therapy. The biochemical pattern of injury is often hepatocellular with transaminase elevations 5 to 20 times the upper limit of normal and most cases are jaundiced as well. Biopsies show variable degrees of inflammation and intrahepatic cholestasis. The prognosis is generally favorable, with only a few patients dying or requiring liver transplant. Statins have also been implicated in causing an autoimmune-like hepatitis. It is thought that some of these cases may represent a triggering effect, since the hepatitis did not resolve when the drug was withdrawn. Patients have positive serology, mainly elevated titers of ANA, but other autoantibodies have also been reported including anti–smooth muscle and antihistone antibodies (6). Biopsies show necroinflammatory changes compatible with autoimmune hepatitis. Cholestasis on biopsy has not been reported in these cases, although the bilirubin can be elevated. The analysis of this case is shown in Table 15.2.1. The patient had been taking simvastatin for 4 months prior to developing liver injury, within the window of reported cases of statin hepatotoxicity. Since the pattern of injury on biopsy was acute cholestatic hepatitis, the differential diagnosis includes acute viral and autoimmune hepatitis, which were excluded by serological testing. It would be important in this case to repeat tests for hepatitis C after several months, because acute hepatitis C may not detectable in clinical tests during the episode of jaundice. Although duct obstruction is unlikely given
222
CHAPTER
15:
DRUG-INDUCED
TA B LE 15. 2. 1 Causality analysis—simvastatin Temporal eligibility
4 months of therapy is at the median for statin injury
Exclusion of competing causes
Viral and autoimmune hepatitis, obstruction excluded, other drugs less likely
Known potential for injury
Rare, but well documented reports of hepatocellular jaundice due to statins
Precedent for pathologic pattern
Biopsies of statin-induced injury have shown cholestatic hepatitis in multiple reports
Dechallenge/ rechallenge
Biochemical evidence of injury resolved over the next 6 weeks after stopping medication
Toxicology
Not performed
Conclusion
DILI very likely (75–95%) due to simvastatin
LIVER
INJURY
Based on this reasoned approach, it is very likely that simvastatin caused this patient’s liver injury. It is difficult to be absolutely certain because of the relative rarity of statininduced hepatotoxicity. It is important to note that not all of the pathology observed in this case may be attributed to the drug. The steatosis present in the background may have been due to underlying fatty liver disease since it has not been previously noted to be a part of the statin injury pattern. In other respects, this case matches the literature reports of statin injury, both clinically and on biopsy.
References
Abbreviaton: DILI, drug-induced liver injury.
the pattern of enzyme elevations and the pathology, gallstones were excluded by imaging studies. The patient was taking 3 other drugs which need to be considered. Metformin, bupropion, and escitalopram are very rare causes of cholestatic hepatitis, but the latency period in reported cases is about 1 to 3 months after starting the drugs (7–9). This makes the other drugs unlikely candidates for the patient’s injury. As discussed above, statins have been demonstrated to cause an acute cholestatic hepatitis, which matches the presentation of the current case. Finally, the patient’s symptoms and serum biochemistries resolved with discontinuation of the drug. No rechallenge was attempted and no toxicology was performed.
1. de Denus S, Spinler SA, Miller K, Peterson AM. Statins and liver toxicity: a meta-analysis. Pharmacotherapy. 2004;24:584–591. 2. Silva MA, Swanson AC, Gandhi PJ, Tataronis GR. Statin-related adverse events: a meta-analysis. Clin Ther. 2006;28:26–35. 3. Chalasani N, Aljadhey H, Kesterson J, et al. Patients with elevated liver enzymes are not at higher risk for statin hepatotoxicity. Gastroenterology. 2004;126:1287–1292. 4. Chalasani N. Statins and hepatotoxicity: focus on patients with fatty liver. Hepatology. 2005;41:690–695. 5. Russo MW, Scobey M, Bonkovsky HL. Drug-induced liver injury associated with statins. Semin Liver Dis. 2009;29:412–422. 6. Ahmad S. Lovastatin-induced lupus erythematosus. Arch Intern Med. 1991;151:1667–1668. 7. Alvaro D, Onetti-Muda A, Moscatelli R, Atili AF. Acute cholestatic hepatitis induced by bupropion prescribed as pharmacological support to stop smoking. A case report. Dig Liver Dis. 2001;33:703–706. 8. Del Val Antonana A, Ortiz Polo I, Rosello Sastre E, Moreno-Osset E. Hepatotoxicity related to escitalopram. Med Clin (Barc). 2008;131:798. 9. Desilets DJ, Shorr AF, Moran KA, Holtzmuller KC. Cholestatic jaundice associated with the use of metformin. Am J Gastroenterol. 2001;96: 2257–2258.
Case 15.3
Drug-Induced Autoimmune Hepatitis DAVID E. KLEINER
C L I N IC AL I N F OR M AT I ON
A 21-year-old man presented with nausea, vomiting, and jaundice. He had been placed on minocycline for acne, 100 mg/day about 11 weeks prior to presentation, but had stopped the medication 1 week earlier due to fatigue and unintended weight loss. His past medical history was only significant for mild asthma, treated with albuterol and fluticasone-salmeterol (both inhaled medications). Laboratory evaluation showed an ALT of 1343 IU/L, AST of 867 IU/L, alkaline phosphatase of 783 IU/L, and a total bilirubin of 5.7 mg/dL. Viral serologies for hepatitis A, B, and C, EBV and CMV were negative, but ANA was positive at 1:1280. ASMA and AMA were negative. Quantitative IgG was elevated at 1870 mg/dL. He was started on corticosteroids and responded rapidly, with normalization of serum biochemistries within 3 weeks. FIGURE 15.3.2
R E A SON F OR R E F E R R AL
Foci of lobular inflammation and apoptotic
hepatocytes.
Evaluate etiology of hepatitis. PAT H OL OG I C F E AT U R E S
On low power examination, the inflammation was mainly concentrated in the portal areas with scattered foci of lobular inflammation in the parenchyma (Figure 15.3.1). The hepatocytes were enlarged, with reactive nuclear changes, and acidophil bodies were easily identified (Figure 15.3.2). Hepatocyte rosetting was seen, but there was no parenchymal cholestasis or lobular disarray. The inflammatory infiltrate in
F I G U R E 1 5 . 3 . 3 Portal inflammation with increased numbers
of eosinophils.
FIGURE 15. 3. 1 Chronic hepatitis pattern with minocycline showing moderate portal inflammation and scattered foci of lobular inflammation.
the portal areas was a mix of lymphocytes, plasma cells, and histiocytes, with prominent eosinophils (Figure 15.3.3). Most of the edges of portal areas were involved by interface hepatitis (Figure 15.3.4). Bile duct injury was present in several portal areas (Figure 15.3.5). A Masson Trichrome stain showed mild portal fibrotic expansion but no bridging fibrosis. The overall histologic pattern was most similar to chronic hepatitis, because of the portal-dominant inflammation and lack of cholestatic features or lobular disarray.
223
224
CHAPTER
15:
DRUG-INDUCED
FIGURE 15. 3. 4 Near-circumferential interface hepatitis around a typical portal area.
FIGURE 15. 3. 5 Injured duct within a portal area showing marked
inflammation.
D I AG N OS I S
LIVER
that mimics idiopathic AIH. The drug injury can mimic all aspects of idiopathic disease, including clinical presentation, positive serologies for autoantibodies, elevated immunoglobulin levels, and histologic manifestations. The diagnostic criteria for AIH have been published, and there has been a recent proposal to simplify these criteria (3,4). Both systems assign points for various disease features with certain scores qualifying as probable or definite AIH. The simplified criteria consider autoantibodies, immunoglobulin levels, and absence of viral hepatitis in the score, but a diagnosis of definite AIH requires a liver biopsy. Given that one must exclude idiopathic AIH in order to make a diagnosis of DIAIH, it is reasonable to ask how one can ever make a diagnosis of DIAIH with certainty. First, certain drugs have characteristically been associated with AIHlike presentations (Table 15.3.1). These drugs, particularly nitrofurantoin and minocycline, have all been associated with drug injury that matches AIH. Some herbal agents, particularly germander, have also been associated with DIAIH (5). Although most studies of DIAIH are single case reports or small series, the Mayo Clinic has recently published their experience with cases of DIAIH and compared them with cases of AIH accrued during the same time period (6). They identified 24 cases of DIAIH and 237 cases of AIH over a 10-year period. Most of the DIAIH cases were due to either minocycline or nitrofurantoin. The patients with DIAIH did not differ from those with AIH in terms of gender, age, autoantibody positivity, immunoglobulin level, necroinflammatory activity, or fibrosis. None of their cases of DIAIH had cirrhosis on biopsy, although cirrhosis has been reported for nitrofurantoin (7,8). The one major difference was that the patients with DIAIH were more likely to be managed with steroids alone, and in all cases where immunosuppression was withdrawn, the hepatitis did not recur. The causality analysis for this case is presented in Table 15.3.2. The histologic considerations in the differential diagnosis are mainly other causes of chronic hepatitis such as the hepatitis viruses and AIH. Given his young age and the presence of duct injury, the overlap disease of primary sclerosing cholangitis (PSC) and AIH should be considered, TA BL E 1 5 . 3 . 1 Currently approved drugs associated with autoimmune
Minocycline-induced autoimmune hepatitis.
hepatitis Clometacine
D I SC U SSI ON
Immune-mediated drug injury may be divided into 2 basic forms, immunoallergic DILI (IA-DILI) and drug-induced autoimmune hepatitis (DIAIH) (1,2). Patients with IA-DILI have evidence of a hypersensitivity reaction with fever, skin rash, and peripheral eosinophilia. Biopsies may show a wide range of histologies, from pure cholestasis to acute hepatitis. Autoantibodies may be present but are not required, and the clinical presentation is usually distinct from that of idiopathic AIH. In contrast, DIAIH is the term applied to drug injury
INJURY
Diclofenac Fenofibrate Methyldopa Minocycline Nitrofurantoin Papaverine Propylthiouracil Statins
CASE
15.3:
DRUG-INDUCED
TA B LE 15. 3. 2 Causality analysis—minocycline Temporal eligibility
Minocycline may develop after weeks to years of therapy
Exclusion of competing causes
Viral hepatitis excluded, other medications were inhalants, which almost never cause liver injury
Known potential for injury
Minocycline is well documented to cause an AIH-like injury
Precedent for pathologic pattern
Both acute and chronic hepatitis patterns have been reported
Dechallenge/ rechallenge
The patient was treated with steroids and rapidly improved, a trial of steroid withdrawal was planned
Toxicology
Not done
Conclusion
DILI definitely (95% certainty) due to Minocycline
Abbreviations: AIH, autoimmune hepatitis; DILI, drug-induced liver injury.
but there were no features of chronic cholestasis in the biopsy. The patient was treated with steroids resulting in rapid symptomatic and biochemical resolution, which would also argue against an overlap syndrome. The hepatitis viruses were excluded by serological evaluation. EBV, which can sometimes cause a chronic hepatitis-like injury, was also excluded. In other respects, the clinical presentation had all of the features
AU T O I M M U N E
H E PAT I T I S
225
of AIH. Although the biopsy did not show a prominent plasmacellular infiltrate, the other changes were compatible with AIH. Because minocycline is well known to cause DIAIH and because there were no other apparent considerations for the etiology of the liver disease, this injury was considered to be due to minocycline.
References 1. Liu ZX, Kaplowitz N. Immune-mediated drug-induced liver disease. Clin Liver Dis. 2002;6:755–774. 2. Uetrecht J. Immunoallergic drug-induced liver injury in humans. Semin Liver Dis. 2009;29:383–392. 3. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31:929–938. 4. Hennes EM, Zeniya M, Czaja AJ, et al. Simplified criteria for the diagnosis of autoimmune hepatitis. Hepatology. 2008;48:169–176. 5. Ben Yahia M, Mavier P, Metreau JM, et al. Chronic active hepatitis and cirrhosis induced by wild germander. 3 cases. Gastroenterol Clin Biol. 1993;17:959–962. 6. Bjornsson E, Talwalkar J, Treeprasertsuk S, et al. Drug-induced autoimmune hepatitis: clinical characteristics and prognosis. Hepatology. 2010;51:2040–2048. 7. Iwarson S, Lindberg J, Lundin P. Nitrofurantoin-induced chronic liver disease. Clinical course and outcome of five cases. Scand J Gastroenterol. 1979;14:497–502. 8. Sharp JR, Ishak KG, Zimmerman HJ. Chronic active hepatitis and severe hepatic necrosis associated with nitrofurantoin. Ann Intern Med. 1980;92:14–19.
Case 15.4
Drug-Induced Cholestatic Hepatitis DAVID E. KLEINER
C L I N I C AL I N F OR M AT I ON
A 59-year-old African-American woman presented with fatigue, pruritus, dark urine, and jaundice. She had a history of hyperthyroidism and had been placed on methimazole, 20 mg/day, about 6 weeks earlier. She did not have fever or rash. Her past medical history was significant for sarcoidosis, asthma, and hypertension. Her laboratory data at presentation included an AST of 98 IU/L, ALT of 159 IU/L, alkaline phosphatase of 397 IU/L, and a total bilirubin of 5.6 mg/dL. The methimazole was discontinued. Serological studies for hepatitis viruses as well as ASMA and AMA were negative, but ANA was positive at 1:160. She was admitted to the hospital 2 days later where imaging studies showed a thickened gallbladder wall without evidence of cholelithiasis or intrahepatic duct dilation. She was discharged and followed in an outpatient clinic. Her bilirubin peaked at 18.4 mg/dL 2 weeks later. Resolution of jaundice was very gradual.
F I G U R E 1 5 . 4 . 2 Portal area with lymphohistiocytic inflammation
and duct injury. Several eosinophils are present. R E A S ON F OR R E F E R R A L
Exclusion of sarcoidosis or autoimmune hepatitis (AIH) as causes for the persistent jaundice. PAT H OL OG I C F E AT U R E S
On biopsy, there was a mild to moderate portal infiltrate with mild, focal interface hepatitis (Figure 15.4.1). The portal infiltrate was composed mainly of lymphocytes and histiocytes, but eosinophils were also present in most portal areas (Figure 15.4.2). The bile ducts were injured and showed
F I G U R E 1 5 . 4 . 3 Canalicular and hepatocellular cholestasis.
FIGURE 15. 4. 1 Mild portal inflammation with scattered spots of
lobular inflammation and mild steatosis in methimazole-induced cholestatic hepatitis.
reactive epithelial changes, but there was no evidence of duct paucity. The parenchyma revealed prominent canalicular and hepatocellular cholestasis (Figure 15.4.3). There was no evidence of chronic cholestasis, and the copper stain was negative. There were numerous foci of lobular inflammation, mostly in the form of microgranulomas, as well as rare acidophil bodies (Figures 15.4.4 and 15.4.5). Well-formed epithelioid granulomas were not present. Mild macrovesicular steatosis was present, but not in any zonal distribution. The trichrome stain showed mild concentric portal fibrotic expansion. The iron stain showed moderate hepatocellular iron in zone 1 distribution (Figure 15.4.6).
226
CASE
15.4:
DRUG-INDUCED
C H O L E S TAT I C
H E PAT I T I S
227
DIAGNO SIS
Cholestatic hepatitis, with mild hepatitis and marked cholestasis, likely due to methimazole. Moderate hepatocellular hemosiderosis and mild steatosis.
DISCUSSIO N
FIGURE 15. 4. 4 Acidophil body with hypertrophic Kupffer cells in
the sinuses.
FIGUR E 15. 4. 5 Spotty lobular inflammation with microgranulomas
and steatosis ( 400).
FIGUR E 15. 4. 6 Accumulation of iron in hepatocytes and portal
macrophages.
The biopsy was performed because of a clinical picture complicated by a history of sarcoidosis, the positive ANA, and imaging studies that showed some abnormalities of the gallbladder but no duct dilation or stones. The biopsy presents a variety of findings, including inflammation, cholestasis, steatosis, and iron overload. The steatosis was mild and azonal in distribution. There was evidence of steatohepatitis. The iron accumulation is concerning in that there is no obvious etiology in the history but the degree of hemosiderosis is not severe enough to warrant immediate investigation. Setting aside the steatosis and iron accumulation, what remains is cholestasis and hepatitis that are not easily explained by AIH, sarcoidosis, or intermittent biliary obstruction. Cholestatic hepatitis is a mixed injury pattern that combines the hepatocellular and canalicular bile accumulation of obstructive jaundice with the portal and lobular inflammation and hepatocyte injury of hepatitis. In his classic text on druginduced liver injury (DILI) (1), Dr. Hyman Zimmerman noted the use of several terms for this pattern, including “cholangiolitic cholestasis,” “hepatocanalicular jaundice,” and “hypersensitivity cholestasis,” but cholestatic hepatitis is simpler and preferred over other terms. The cholestatic component may consist of bile accumulation in hepatocytes, bile plugs in canaliculi or both, and is most prominent in zone 3. Ductal and cholangiolar cholestasis are not typically seen in cholestatic hepatitis. The hepatitic component may resemble acute hepatitis, with a predominance of lobular inflammation, lobular disarray, and confluent necrosis, or it may resemble chronic hepatitis, in which the portal inflammation stands out over the lobular inflammation, as it did in this case. As the inflammation becomes very mild, there is overlap with the pattern of pure intrahepatic or acute cholestasis. When the inflammation is severe, there is overlap with the pattern of acute hepatitis, particularly when the cholestatic changes are mild. It is important to exclude chronic cholestatic injury, particularly ductal sclerosis or paucity, because these changes lead to a different histologic differential diagnosis. It is important to recognize cholestatic hepatitis, particularly when the inflammation is distributed in a chronic hepatitis pattern, because there are very few non-DILI etiologies to consider. Large duct obstruction may lead to cholestasis with inflammation, but the inflammation is usually mild and there may be other evidence to point to obstruction, such as portal edema, ductal cholestasis, or acute cholangitis. Sepsis and postsurgical jaundice may both show zone 3 cholestasis on biopsy, but the cholestasis is usually bland and these
228
CHAPTER
15:
DRUG-INDUCED
etiologies are easily excluded. Chronic viral hepatitis, AIH, and the chronic cholestatic diseases of primary biliary cirrhosis and primary sclerosing cholangitis rarely have hepatocellular or canalicular cholestasis as a feature of early-stage disease. In contrast, cholestatic hepatitis is a commonly reported pattern of DILI. Some drug classes have been particularly associated with this pattern of injury, including the antibacterials, antifungals, statins, antithyroid agents (except propylthiouracil), and the major tranquilizers. The analysis of this case is summarized in Table 15.4.1. As noted above, there was clinical concern for competing causes of liver injury, notably sarcoidosis and AIH. The biopsy was effective in excluding these diseases. Sarcoidosis is unlikely to cause significant intrahepatic cholestasis without
TA B LE 15. 4. 1 Causality analysis—methimazole Temporal eligibility
Patient had 6 weeks of therapy, most cases of methimazole injury have presented after 2 to 12 weeks of therapy
Exclusion of competing causes
The pathology was not characteristic of sarcoidosis or AIH. Viral hepatitis excluded by serology. Imaging showed normal ducts
Known potential for injury
Methimazole has been implicated in multiple cases of DILI
Precedent for pathologic pattern
The pathology observed most often in methimazole DILI is cholestasis or cholestatic hepatitis
Dechallenge/ rechallenge
Gradual resolution of jaundice after cessation of therapy, no rechallenge attempted
Toxicology
Not done
Conclusion
DILI very likely (75–95%) due to methimazole
Abbreviations: AIH, autoimmune hepatitis; DILI, drug-induced liver injury.
LIVER
INJURY
extensive bile duct destruction by granulomas. The inflammatory infiltrate, degree of injury, and presence of cholestasis all make AIH unlikely. Extrahepatic obstruction should always be considered in a case of cholestatic hepatitis, but the imaging showed no duct dilation and the slow recovery from the jaundice would make stone-related obstruction unlike in retrospect. The biopsy did not show specific features of large duct obstruction (portal edema, ductal cholestasis). The degree and character of the inflammation as well as the presence of duct injury would argue against this possibility. Methimazole, however, is well reported to cause cholestasis and cholestatic jaundice. Some cases are associated with features of hypersensitivity, including fever and rash. Agranulocytosis (neutropenia) has also been reported (2,3). Most cases develop between 2 and 12 weeks after initiation of drug therapy. Cholestatic changes are the most commonly reported pathology present in 15 out of 23 cases summarized by Woeber (4). Some patients can develop severe hepatitis, and deaths have been reported (5). Recovery from jaundice can be very slow, taking 3 to 5 months in some cases (3). Based on the characteristic injury pattern and the exclusion of competing etiology of liver injury, the evidence points to methimazole hepatotoxicity.
References 1. Zimmerman HJ. Hepatotoxicity: the adverse effects of drugs and other chemicals on the liver. 2nd ed. Philadelphia, PA: Lippincott, Williams & Wilkins; 1999. 2. Specht NW, Boehme EJ. Death due to agranulocytosis induced by methimazole therapy. J Am Med Assoc. 1952;149:1010–1011. 3. Vitug AC, Goldman JM. Hepatotoxicity from antithyroid drugs. Horm Res. 1985;21:229–234. 4. Woeber KA. Methimazole-induced hepatotoxicity. Endocr Pract. 2002;8:222–224. 5. Baker B, Shapiro B, Fig LM, Woodbury D, Sisson JC, Beierwaltes WH. Unusual complications of antithyroid drug therapy: four case reports and review of literature. Thyroidology. 1989;1:17–26.
Case 15.5
Drug-Induced Ductopenia DAVID E. KLEINER
C L I N IC AL I N F OR M AT I ON
A 10-year-old girl developed a punctate rash over her chest, trunk, and neck 7 weeks after taking a 5-day course of azithromycin. During the next week, she developed cough, fever, and sore throat and went to a local emergency room. She was given a dose of cephtriaxone and another prescription for azithromycin. She took two doses of the azithromycin, but the rash continued to worsen and she was admitted to a hospital the following day. Work-up at this time revealed negative culture for Streptococcus as well as abnormal liver–associated enzymes: ALT of 411 IU/L, AST of 376 IU/L, alkaline phosphatase of 460 IU/L, and a total bilirubin of 3.4 mg/dL. The rash involved her oral and conjuctival mucosa, and a skin biopsy confirmed Stevens-Johnson syndrome. An abdominal ultrasound exam was normal. Viral and autoantibody serologies were negative. Her liver chemistry abnormalities persisted despite resolution of the rash, and 2 weeks after presentation, a liver biopsy was performed. Her bilirubin and alkaline phosphatase gradually returned to normal over the next 5 months, but her transaminases remained elevated more than a year after the initial event.
were enlarged, with pale staining, and slightly pigmented cytoplasm (Figure 15.5.2). There were enlarged vacuolated macrophages in the sinusoids. Occasional apoptotic hepatocytes were present. Although bile stasis was difficult to appreciate on the routine stains, the copper stain showed distinctive zones of hepatocytes with green-tinged cytoplasm (Figure 15.5.3). Some portal areas showed mild inflammation, but plasma cells and eosinophils were not evident (Figure 15.5.4). Close examination of the portal areas revealed that most lacked an identifiable duct and a formal count found only one duct in 18 portal tracts (Figure 15.5.5).
R E A SON F OR R E F E R R AL
Evaluation of possible etiologies for the persistently abnormal liver chemistries. PAT H OL OG I C F E AT U R E S
Liver biopsy at low magnification showed minimal necroinflammatory infiltrate (Figure 15.5.1). In zone 3, the hepatocytes
F I G U R E 1 5 . 5 . 2 Zone 3 injury with pale, swollen hepatocytes and sinusoidal macrophages.
FIGURE 15. 5. 1 Small portal area without inflammation and en-
F I G U R E 1 5 . 5 . 3 Green pigmentation of hepatocytes is apparent on
larged pale hepatocytes in zone 3.
the copper stain.
229
230
CHAPTER
15:
DRUG-INDUCED
FIGURE 15.5.4 Small portal area with mild lymphocytic inflammation.
LIVER
INJURY
infection of the biliary tree. Hodgkin’s lymphoma and histiocytosis X have also been associated with ductopenia, as have a variety of congenital and developmental diseases. If all known etiologies (including drugs and toxins) have been excluded, the term idiopathic adulthood ductopenia is applied. The list of drugs and other agents associated with VBDS is long and growing (Table 15.5.1). The initial presentation may be similar to cholestatic hepatitis. Many patients present with an acute onset of jaundice and elevated levels of alkaline phosphatase and gamma glutamyltransferase. The transaminases may be modestly elevated or normal. Unlike cholestatic hepatitis, the abnormalities in bilirubin or cholestatic enzymes persist for months and may worsen, leading to liver failure from biliary cirrhosis and the need for transplantation (Figure 15.5.6). The pathophysiology that leads to VBDS is poorly understood, although immune mechanisms are almost certainly involved in some cases. If biopsies are performed early TA BL E 1 5 . 5 . 1 Drugs associated with duct paucity
FIGURE 15. 5. 5 Small portal area without a visible duct.
Aceprometazine Ajmaline Amineptine Amitriptyline Amoxicillinclavulanate Ampicillin Arsenicals Azathioprine Azithromycin Barbiturates Candesartan Carbamazepine Carbutamide Chlorothiazide Chlorpromazine Cimetidine Clindamycin
Cromolyn Cyamemazine Cyclohexyl propionate Cyproheptadine Diazepam Erythromycin Estradiol Flucloxacillin Glibenclamide Glycyrrhizin Gold Haloperidol Ibuprofen Imipramine Itraconazole Methyltestosterone Moxifloxacin
Norandrostenolone Phenylbutazone Phenytoin Prochlorperazine Sulpiride Tenoxicam Terbinafine Tetracycline Thiabendazole Tiopronin Tolbutamide Trifluoperazine Trimethoprimsulfamethoxazole Troleandomycin Xenalamine
D I AG N OS I S
Cholestatic injury with marked bile duct paucity, probably related to azithromycin.
D I SC U SSI ON
This case is an example of ductopenia that was probably related to drug injury. Ductopenia is the hallmark of vanishing bile duct syndrome (VBDS), which is the term used for the collection of diseases that can give rise to ductopenia (1,2). The most common causes for VBDS are the chronic cholestatic diseases of primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC), where VBDS may develop with progression to cirrhosis. VBDS may also be associated with hepatic sarcoidosis, graft-versus-host disease and liver transplant rejection. A form of VBDS may be seen in HIV-infected patients where it is thought to be secondary to opportunistic
F I G U R E 1 5 . 5 . 6 Ductopenic portal area from an explant following
prolonged jaundice due to moxifloxacin.
CASE
15.5:
DRUG-INDUCED
in the course of the disease, there is often a pattern of cholestatic hepatitis with nonsuppurative duct injury. However, given the rarity of VBDS with respect to drug-induced cholestatic hepatitis, most patients with cholestatic hepatitis do not progress to ductopenia. Some cases have an immunoallergic phenotype, with fever, rash, and eosinophilia, and StevensJohnson syndrome is a complication of many drugs associated with ductopenia. In several reports, both Stevens-Johnson syndrome and ductopenia occur together (3). Drug-induced ductopenia should be suspected in any case of drug-induced cholestasis in which either the jaundice or the enzyme abnormalities persist beyond 6 months. Imaging is helpful to exclude a large duct problem, but a liver biopsy is the only way to assess for ductopenia, since the damage is almost always confined to the smallest branches of the duct tree. A diagnosis of ductopenia can be made when at least half of the portal areas lack a duct (1). Bile ducts may be differentiated from ductular reaction by location and size: the main duct is usually located near the artery and has approximately the same outer diameter. A cytokeratin 7 or 19 immunostain can also be helpful, particularly when inflammation may be obscuring duct remnants. A copper stain can be helpful in picking up early chronic cholestasis, but it may be negative, as it was in this case. The prognosis may depend on the extent of small duct injury, but there does not seem to be good correlation between the degree of ductopenia on biopsy and outcome. It should be noted that there is a separate type of chronic cholestatic injury that may be drug-induced and associated with ductopenia. Biliary sclerosis with fibro-obliterative PSClike lesions has been observed after hepatic artery floxuridine infusions (4) or after treatment with scolicides (5). The chemotherapeutic injury is thought to be related to microvascular injury of the peribiliary capillaries, whereas scolicides may directly injure the bile duct epithelium. The analysis of this case is summarized in Table 15.5.2. With respect to the case at hand, azithromycin causes cholestatic hepatitis that presents 1 to 3 weeks after a 5-day course of treatment (6–8). It had not been reported to cause ductopenic injury at the time the case was first seen and evaluated. However, there is now a case report of azithromycin-induced ductopenia in a 62-year-old man (9). In a 10-year-old child, the main differential diagnostic considerations would be exclusion of extrahepatic obstruction or a PSC autoimmune hepatitis syndrome. Both of these are adequately excluded by a combination of clinical testing and histologic evaluation. Consistent with the finding of duct paucity on biopsy, the patient had a very prolonged recovery, lasting more than a year after the initial injury.
DUCTOPENIA
231
TA BL E 1 5 . 5 . 2 Causality analysis—azithromycin Temporal eligibility
Latency of 2 months after taking azithromycin is longer than most reported cases, but within the 1- to 3-month window for immunoallergic reactions
Exclusion of competing causes
Viral and autoantibody studies were negative. Imaging did not identify an obstructive cause; other causes of ductopenia would be unusual at this age
Known potential for injury
Rare cause of cholestatic hepatitis
Precedent for pathologic pattern
One recent case report of ductopenic injury
Dechallenge/rechallenge
With short-course antibiotic therapy, the course is usually over by the time the symptoms begin. The patient was “rechallenged” with 2 doses of azithromycin after which her symptoms worsened
Toxicology
Not done
Conclusion
Ductopenia probably (50–75%) due to azithromycin. Case is less certain because of the long latency and the fact that ductopenia had not been reported at the time the patient presented.
References 1. Desmet VJ. Vanishing bile duct syndrome in drug-induced liver disease. J Hepatol. 1997;26(suppl 1):31–35. 2. Reau NS, Jensen DM. Vanishing bile duct syndrome. Clin Liver Dis. 2008;12:203–217. 3. Srivastava M, Perez-Atayde A, Jonas MM. Drug-associated acuteonset vanishing bile duct and Stevens-Johnson syndromes in a child. Gastroenterology. 1998;115:743–746. 4. Ludwig J, Kim CH, Wiesner RH, Krom RA. Floxuridine-induced sclerosing cholangitis: an ischemic cholangiopathy? Hepatology. 1989;9: 215–218. 5. Castellano G, Moreno-Sanchez D, Gutierrez J, Moreno-Gonzalez E, Colina F, Solis-Herruzo JA. Caustic sclerosing cholangitis. Report of four cases and a cumulative review of the literature. Hepatogastroenterology. 1994;41:458–470. 6. Chandrupatla S, Demetris AJ, Rabinovitz M. Azithromycin-induced intrahepatic cholestasis. Dig Dis Sci. 2002;47:2186–2188. 7. Longo G, Valenti C, Gandini G, Ferrara L, Bertesi M, Emilia G. Azithromycin-induced intrahepatic cholestasis. Am J Med. 1997;102:217–218. 8. Suriawinata A, Min AD. A 33-year-old woman with jaundice after azithromycin use. Semin Liver Dis. 2002;22:207–210. 9. Danica J, Irena H, Davor R, et al. Vanishing bile duct syndrome associated with azithromycin in a 62-year-old man. Basic Clin Pharmacol Toxicol. 2010;106:62–65.
Case 15.6
Methotrexate-Induced Chronic Liver Disease DAVID E. KLEINER
C L I N I C AL I N F OR M AT I ON
A 64-year-old woman was referred for evaluation of liver disease related to methotrexate therapy. She had psoriatic arthritis for 20 years, and 11 years prior to the current evaluation had been placed on methotrexate (20 mg/week). Her current total cumulative dose was about 2900 g of methotrexate. Her past medical history was also significant for obesity and hypothyroidism. She was not diabetic and did not drink alcohol. Serological tests for hepatitis viruses were negative as was anti-SMA. ANA was positive at 1:40. Three liver biopsies had already been performed and these were reported to show steatosis with minimal inflammation and possibly progressive periportal fibrosis. Her liver enzyme tests were mainly normal throughout her therapy, although recently she had mildly elevated tests, with ALT 52 IU/L, AST 51 IU/L, and alkaline phosphatase of 153 IU/L. F I G U R E 1 5 . 6 . 2 Lymphoplasmacytic infiltrate with mild interface
hepatitis.
R E A S ON F OR R E F E R R A L
Staging of liver disease and exclusion of obesity-related fatty liver disease. PAT H OL OG I C F E AT U R E S
The hepatic architecture was disrupted by wide bands of portal-based fibrosis that crossed the width of the biopsy (Figure 15.6.1). There was a mild to moderate portal inflammatory infiltrate associated with mild interface hepatitis (Figure 15.6.2). Increased numbers of plasma cells were seen in the portal infiltrate. The parenchyma
FIGURE 15.6.3 Mild macrovesicular steatosis and nuclear variability.
FIGURE 15. 6. 1 Moderate portal inflammation and mild steatosis.
showed mild macrovesicular steatosis in no particular zonal distribution (Figure 15.6.3). Mild nuclear variation was apparent. Ballooning injury was seen within the parenchyma (Figure 15.6.4) as well as near some portal tracts (Figure 15.6.5). No well-formed Mallory-Denk bodies were seen. Fibrosis was difficult to stage due to the narrow gauge of the core, but Masson stains did show probable bridging fibrosis based on the width of the fibrotic tracts (Figure 15.6.6). No zone 3 perisinusoidal fibrosis was present, although there were preserved central veins (Figure 15.6.7). 232
CASE
15.6:
M E T H O T R E X AT E - I N D U C E D
FIGURE 15. 6. 4 Ballooning injury with small foci of lobular
CHRONIC
LIVER
DISEASE
233
F I G U R E 1 5 . 6 . 7 No perisinusoidal fibrosis is present.
inflammation.
DIAGNO SIS
Chronic hepatic injury likely due to methotrexate, with mild inflammation, mild steatosis, and probable bridging fibrosis, Roenigk Stage IIIB.
DISCUSSIO N
FIGUR E 15. 6. 5 Periportal ballooning injury and mild portal
inflammation.
FIGURE 15. 6. 6 Wide fibrotic band with trapping of hepatocytes.
Methotrexate was introduced in the mid-1950s as a therapy for childhood leukemias but was soon noted to ameliorate the symptoms of arthritis in rheumatoid arthritis and psoriasis. Methotrexate inhibits intracellular metabolism and purine biosynthesis by inhibiting or interfering with dihydrofolate reductase and thymidylate synthase. Hepatotoxicity of high-dose methotrexate was recognized very soon after its introduction (1). Methotrexate has been associated with a variety of pathologic changes, but the most serious toxicity is fibrosis that leads to cirrhosis. Although many of the classical studies on the pathology of methotrexate were done without pretreatment biopsies and in an era prior to the discovery of hepatitis C and the appreciation of the prevalence of NAFLD and NASH, there is a consensus that long-term therapy with methotrexate increases the risk for cirrhosis. Several risk factors have been identified, including alcohol use, obesity, and diabetes (2). Interestingly, all of these risk factors can induce liver disease that is in the differential of the pattern of injury caused by methotrexate. Weekly, rather than daily, dosing of methotrexate also reduces the risk of serious liver injury. The histologic changes observed by light microscopy are not specific for methotrexate injury and include steatosis, inflammation, nuclear enlargement and variation (anisonucleosis), and fibrosis (3–5). The steatosis is typically macrovesicular and may be mild to marked. Ballooning hepatocellular injury may be seen. The inflammation is generally mild, lymphohistiocytic, and portal, with scattered foci of lobular
234
CHAPTER
15:
DRUG-INDUCED
inflammation as seen in other chronic liver diseases. There is often significant nuclear variability, with nuclear enlargement, multinucleation, and vacuolation. Methotrexate injury may show a stellate portal fibrosis early in the disease course that later progresses to portal-portal and portal-central bridging fibrosis and cirrhosis. The incidence of cirrhosis varies widely (8–46% in psoriasis), but most studies do not adequately control for underlying liver disease. Patients with psoriasis seem to be at somewhat greater risk for serious liver injury than those with rheumatoid arthritis. The liver biopsy has played an important role in monitoring liver disease in patients on long-term methotrexate therapy. Although recent guidelines have reduced the frequency of biopsy as well as tempering the need for pretreatment baseline biopsy, the liver biopsy remains the best method for gauging the severity of the liver disease (6). Newer, noninvasive techniques, such as ultrasound fibrosis measurement or serum fibrosis markers, are being explored but have not yet supplanted the liver biopsy. The mainstay of grading the hepatic injury in methotrexate therapy is the Roenigk scale, summarised in Table 15.6.1. Unlike more modern systems of grading and staging liver disease, the Roenigk classification does not explicitly separate fibrosis from other features. It is clear though that grades I and II deal with prefibrotic changes, whereas grades IIIA to IV only consider fibrosis in the stratification. Clinical decisions are based on the degree of fibrosis. Current guidelines recommend more frequent monitoring at grade IIIA and consideration for stopping therapy at grades IIIB and IV (6,7). Since fibrosis is the most important feature, it would not be unreasonable to substitute a more familiar fibrosis staging system for the Roenigk classification. The case under consideration here presents some difficulty in deciding the role of obesity-related NAFLD in this patient’s hepatic pathology, summarized in Table 15.6.2. Studies have reported NASH-like pathology in patients treated with methotrexate, usually in those with obesity or diabetes (8). The prior biopsies in this case were not available for review but were reported to show steatosis and inflammation with little or no fibrosis. A specific pattern of steatohepatitis was not noted. When the current biopsy was studied prior to reviewing the history, the overall pattern was thought to be more like a chronic hepatitis, because of the significant portal TA B LE 15. 6. 1 Roenigk classification of liver injury from methotrexate
Grade
Fibrosis
Changes of Steatosis, Portal Inflammation and Anisonucleosis
Grade I
None
Up to mild degrees of injury
Grade II
None
Moderate to marked changes
Grade IIIA
Mild (periportal, short septae)
Any degree
Grade IIIB
Moderate/severe (bridging fibrosis)
Any degree
Grade IV
Cirrhosis
Any degree
Adapted from Ref. 9.
LIVER
INJURY
TA BL E 1 5 . 6 . 2 Causality analysis—methotrexate Temporal eligibility
11 years of therapy with methotrexate, with a total dose of 2900 mg
Exclusion of competing causes
Viral studies were negative, ANA titer was low, no history of diabetes or alcohol; obesity-related NAFLD still possible, but the pathology did not show specific features of NASH
Known potential for injury
Methotrexate has been documented in numerous series to cause chronic liver disease
Precedent for pathologic pattern
The pattern of injury in this case is typical for what has been reported in methotrexate injury
Dechallenge/ rechallenge
Liver enzyme abnormalities resolved after stopping the drug. Rechallenge was not performed
Toxicology
Not done
Conclusion
Chronic liver disease with probable bridging fibrosis likely (75–95%) due to methotrexate
inflammation, interface hepatitis, and portal-based fibrosis, than steatohepatitis or fatty liver disease. Indeed, although there was ballooning injury, it was as much periportal as parenchymal and there was no perisinusoidal fibrosis. Still, portal inflammation may increase with increasing stage of fibrosis in NASH, and the characteristic zonal distribution of lesions may be lost. Nevertheless, because there were no specific features pointing toward steatohepatitis as the underlying disease, it was judged that the injury in this case was most likely due to methotrexate.
References 1. Colsky J, Greenspan EM, Warren TN. Hepatic fibrosis in children with acute leukemia after therapy with folic acid antagonists. AMA Arch Pathol. 1955;59:198–206. 2. Nyfors A, Poulse H. Liver biopsies from psoriatics related to methotrexate therapy. 1. Findings in 123 consecutive non-methotrexate treated patients. Acta Pathol Microbiol Scand [A]. 1976;84:253–261. 3. Aponte J, Petrelli M. Histopathologic findings in the liver of rheumatoid arthritis patients treated with long-term bolus methotrexate. Arthritis Rheum. 1988;31:1457–1464. 4. Kremer JM, Lee RG, Tolman KG. Liver histology in rheumatoid arthritis patients receiving long-term methotrexate therapy. A prospective study with baseline and sequential biopsy samples. Arthritis Rheum. 1989;32:121–127. 5. West SG. Methotrexate hepatotoxicity. Rheum Dis Clin North Am. 1997;23:883–915. 6. Reuben A. Methotrexate contraversies. In: Kaplowitz N, DeLeve LD, eds. Drug-Induced Liver Disease. New York, NY: Informa Healthcare; 2007:683–706. 7. Visser K, Katchamart W, Loza E, et al. Multinational evidence-based recommendations for the use of methotrexate in rheumatic disorders with a focus on rheumatoid arthritis: integrating systematic literature research and expert opinion of a broad international panel of rheumatologists in the 3E Initiative. Ann Rheum Dis. 2009;68:1086–1093. 8. Langman G, Hall PM, Todd G. Role of non-alcoholic steatohepatitis in methotrexate-induced liver injury. J Gastroenterol Hepatol. 2001;16:1395–1401. 9. Roenigk HH, Jr., Auerbach R, Maibach HI, et al. Methotrexate in psoriasis: revised guidelines. J Am Acad Dermatol. 1988;19:145–156.
Case 15.7
Liver Injury Due to Total Parenteral Nutrition DAVID E. KLEINER
C L I N C I A L I N F OR M AT I ON
A baby boy was born prematurely at 27 weeks and weighed 850 g at birth. The infant was admitted to a neonatal intensive care unit, where meconium ileus was diagnosed. A portion of the small intestine was removed and the infant was placed on total parenteral nutrition (TPN) for nutritional support until he could tolerate enteral feeding. Two weeks after starting TPN, he became jaundiced and was noted to have elevated alkaline phosphatase and gamma glutamyltransferase. The transaminases were initially normal. TPN feedings continued but he succumbed 6 weeks later to sepsis. At autopsy, the biliary tree was not dilated and there were no stones in the gallbladder or bile duct. R E A SON F OR R E F E R R AL
Etiology of jaundice: obstruction, sepsis, or TPN.
F I G U R E 1 5 . 7 . 2 There was no bile duct injury, cholangitis, or portal
edema. PAT H OL OG I C F E AT U R E S
Sections of liver taken at autopsy showed generally preserved hepatic architecture. There was fibrotic expansion of portal areas with early bridging fibrosis (Figure 15.7.1). There was mild lymphocytic inflammation and clusters of pigmented macrophages in the portal areas. There was no duct injury, duct paucity, portal edema, or bile stasis in the interlobular bile ducts (Figure 15.7.2). There was extensive rosette formation, and many rosettes contained a plug of inspissated bile (Figure 15.7.3). Many of the hepatocyte plates were 2 cells thick, consistent with regeneration. There was little or no lobular inflammation.
F I G U R E 1 5 . 7 . 3 Canalicular bile stasis, hepatocyte rosettes, and
widened liver cell plates.
DIAGNO SIS
Cholestatic hepatitis with mild inflammation, prominent cholestasis, and early bridging fibrosis due to TPN.
DISCUSSIO N
FIGURE 15. 7. 1 Portal expansion with inflammation and fibrosis.
TPN is an effective and generally safe technique to provide partial or complete nutritional support for patients with either acute or chronic intestinal failure (1,2). Because TPN is not 235
236
CHAPTER
15:
DRUG-INDUCED
a drug, per se, but a nutritional supplement, liver injury from TPN is not always considered together with drug-induced liver injury. The incidence of biochemical abnormalities is high, even during short courses of TPN. Between 42% and 93% of adults may have elevated transaminases with lower numbers showing abnormalities of bilirubin or alkaline phosphatase (3). There are 2 basic patterns of injury seen with TPN, steatosis, and cholestasis/cholestatic hepatitis. Steatosis is the most common complication, particularly in adults. The steatosis is macrovesicular and distributed in zone 1, allowing distinction from the zone 3 distribution of steatosis in NAFLD. The steatosis may appear after only a few weeks of TPN, accompanied by mild portal and lobular inflammation. There may be mild (1-3 ULN) elevations of transaminases (4). The bilirubin may rise slightly, but the development of jaundice suggests progression to cholestatic liver injury. In neonates treated for 1 to 2 weeks or adults on longterm TPN, cholestatic liver disease may develop. In neonates, TPN is generally started soon after birth and the liver disease progression has been carefully studied (5,6). Infants who die within a week of starting TPN generally show steatosis in a zone 1 distribution without cholestasis. After a couple of weeks of TPN, canalicular cholestasis appears, with bile plugs most prominent in zone 3. Infants treated for 1 to 3 months develop ductular reaction and periportal fibrosis, which can progress to bridging fibrosis and cirrhosis. Inflammation is usually mild or absent. Neutrophils can be seen around reactive ductules. Liver failure can occur with prolonged treatment. The etiology of TPN-induced liver injury is not entirely clear and probably multifactorial (1). Because the nutritional content reaches the liver via the hepatic artery rather than the portal vein, the flow of nutrients through the hepatic acinus is altered. This causes adaptive changes within the hepatic acinus. The lack of enteral nutrition causes changes in gastrointestinal hormone release, including loss of physiologic cholecystokinin release and decreased enterohepatic circulation (7,8). Reduced cholecystokinin levels also decrease gallbladder emptying and increase the risk for stones and sludge. Enterocyte atrophy from lack of stimulation may allow increased bacterial translocation across the gut (9). Bacterial overgrowth within the lumen of the unused gut increases the conversion of deoxycholate to lithocholate, one of the more toxic bile salts. Over the years, toxicity of various components of TPN has been recognized and formulations have been altered to make the mixture less injurious. The patient’s underlying disease may also increase the susceptibility to TPN-induced injury. Known risk factors include prolonged TPN therapy, prematurity, low birth weight, malnutrition (low serum albumin), sepsis, jejunostomy, and gastrointestinal surgery. Combining TPN with enteral feeding can reduce the potential for injury by countering some of the consequences of intestinal atrophy. The causality analysis is presented in Table 15.7.1. In this case, the infant received a total of 8 weeks of TPN prior to death from sepsis. The jaundice developed at 2 weeks, consistent with the pattern of TPN-induced cholestasis. At
LIVER
INJURY
TA BL E 1 5 . 7 . 1 Causality analysis—TPN Temporal eligibility
TPN was given for 8 weeks, well documented as sufficient time to develop cholestatic liver injury
Exclusion of competing causes
Obstruction excluded at autopsy, sepsis may have contributed to some of the cholestasis, but cannot be the primary cause based on the time course
Known potential for injury
TPN is well known to cause cholestatic injury in premature infants
Precedent for pathologic pattern
Cholestasis and cholestatic hepatitis with fibrosis progression are the typical pathology observed in neonates with TPN injury
Dechallenge/ rechallenge
Not applicable—infant received TPN until death
Toxicology
Not done
Conclusion
Cholestatic hepatitis due (95% chance) to TPN
Abbreviation: TPN, total parenteral nutrition.
autopsy, the liver showed a pattern of cholestatic hepatitis, with only mild portal inflammation but prominent cholestasis. Portal and early septal fibrosis was present, also consistent with the known progression of the cholestatic liver disease from TPN. Macroscopic and microscopic features of extrahepatic obstruction were not seen, so obstructive jaundice is easily dismissed from consideration. Sepsis may have contributed to bile stasis, but cholangiolar cholestasis was not seen and the jaundice developed weeks before the sepsis. This premature infant was clearly at risk to develop TPNinduced cholestasis, and after excluding the other possibilities, it is possible to state with confidence that the changes observed were due to TPN.
References 1. Guglielmi FW, Regano N, Mazzuoli S, et al. Cholestasis induced by total parenteral nutrition. Clin Liver Dis. 2008;12:97–110, viii. 2. Kwan V, Georg J. Liver disease due to parenteral and enteral nutrition. Clin Liver Dis. 2004;8:893–913. 3. Baker AL, Rosenber IH. Hepatic complications of total parenteral nutrition. Am J Med. 1987;82:489–497. 4. Bengoa JM, Hanauer SB, Sitrin MD, Baker AL, Rosenber IH. Pattern and prognosis of liver function test abnormalities during parenteral nutrition in inflammatory bowel disease. Hepatology. 1985;5:79–84. 5. Cohen C, Olse MM. Pediatric total parenteral nutrition. Liver histopathology. Arch Pathol Lab Med. 1981;105:152–156. 6. Mullick FG, Moran CA, Isha KG. Total parenteral nutrition: a histopathologic analysis of the liver changes in 20 children. Mod Pathol. 1994;7:190–194. 7. Hofman AF. Defective biliary secretion during total parenteral nutrition: probable mechanisms and possible solutions. J Pediatr Gastroenterol Nutr. 1995;20:376–390. 8. Forchielli ML, Walke WA. Nutritional factors contributing to the development of cholestasis during total parenteral nutrition. Adv Pediatr. 2003;50:245–267. 9. Alverdy J, Chi HS, Sheldo GF. The effect of parenteral nutrition on gastrointestinal immunity. The importance of enteral stimulation. Ann Surg. 1985;202:681–684.
Case 15.8
Amiodarone-Induced Phospholipidosis DAVID E. KLEINER
C L I N IC AL I N F OR M AT I ON
A 62-year-old man was admitted to the hospital for evaluation of jaundice. He had an extensive history of cardiac disease including ischemic cardiomyopathy, atrial fibrillation, coronary artery bypass grafts, and pacemaker placement. He had received amiodarone for 9 years. Three years prior to this admission, he had an episode of hypoxic hepatitis documented by biopsy. His current laboratory evaluation showed an ALT of 781 IU/L, AST 734 IU/L, alkaline phosphatase 119 IU/L, and total bilirubin 3 mg/dL. Serologies for hepatitis viruses were negative, but ANA was positive at 1:640. After admission, the transaminase levels gradually fell, but the bilirubin rose and a biopsy was performed. The patient died while still in the hospital from multiorgan failure.
macrovesicular steatosis, the hepatocytes showed microvesiculation (Figure 15.8.2). Pale pink Mallory-Denk bodies could also be seen in many hepatocytes, a finding accentuated by immunostaining for ubiquitin (Figure 15.8.3). Within the fibrous bands were many small collections of foamy macrophages containing coarsely granular, pale, pigmented material (Figure 15.8.4). The pigmented material stained purple on a diastase-digestion PAS stain (Figure 15.8.5).
R E A SON F OR R E F E R R AL
Determine etiology of cirrhosis. PAT H OL OG I C F E AT U R E S
The hepatic architecture was diffusely distorted by bands of bridging fibrosis that surrounded small regenerative nodules and isolated small groups of hepatocytes (Figure 15.8.1). Extensive sinusoidal fibrosis was present. The inflammatory infiltrate was generally mild and was mainly composed of lymphocytes. There was focal disruption of the limiting plate at many points by fibrosis and inflammation, associated with prominent ductular reaction. Although there was little
FIGURE 15. 8. 1 Micronodular cirrhosis with minimal steatosis and mild inflammation.
F I G U R E 1 5 . 8 . 2 Hepatocytes with microvesicular steatosis and pale eosinophilic Mallory-Denk bodies (arrow).
F I G U R E 1 5 . 8 . 3 Numerous Mallory-Denk bodies stained with
antiubiquitin.
237
238
CHAPTER
15:
DRUG-INDUCED
FIGURE 15. 8. 4 Foamy macrophages containing pale blue-gray
material.
FIGURE 15. 8. 5 PAS-positive, diastase-resistant, coarsely granular
material in the foamy macrophages.
LIVER
INJURY
presented with hepatosplenomegaly. When drug therapy is stopped, the drug gradually disappears from the cells and the phospholipidosis resolves. Besides amiodarone and Coralgil, a large number of drugs have been reported to cause phospholipidosis, including perhexiline maleate, chlorpheniramine, thioridazine, the aminoglycoside antibiotics, the selective serotonin reuptake inhibitors (SSRIs), tamoxifen, and others. Phospholipidosis is a dose-dependent process in which the drug gradually accumulates within lysosomes over time. There are several theories as to how the lamellar bodies form (2). Many of these amphiphilic drugs have been shown to interfere with lysosomal phospholipase either by direct inhibition or by binding to phospholipids and blocking digestion. These changes would result in the accumulation of phospholipids within the lysosome. The drugs may interfere with normal trafficking of membrane phospholipids between the plasma membrane and the Golgi so that increased amounts of phospholipids are sequestered in lysosomes. It is unclear if phospholipidosis, by itself, is toxic to cells. Hepatocyte culture studies with amiodarone showed that lamellar bodies developed prior to evidence of cell toxicity and at lower drug concentrations than were necessary to cause toxicity (3). Phospholipidosis has also been observed in patients in the absence of hepatoxicity (4). It is therefore possible that phospholipidosis may represent an adaptive response in which cells are able to sequester potentially harmful agents within lysosomes, thus tempering their toxicity. Histologically, the diagnostic feature is the lamellar body, which is a whorled, laminated collection of phospholipid membranes within a lysosome. Lamellar bodies are also seen in some genetic lipid storage diseases, such as Niemann-Pick disease (Figure 15.8.6). Lamellar bodies can be found in the liver within hepatocytes or macrophages as well as in other organs, such as lymph nodes and lung (5). The light microscopic appearance is not specific, but the presence of foamy cells with coarsely granular PAS-positive material is characteristic of phospholipidosis in amiodarone-induced liver injury. In addition to lamellar bodies, Mallory-Denk bodies are a typical finding. They are often more numerous than is typically seen in NASH or alcoholic liver disease and are more likely to
D I AG N OS I S
Cirrhosis with mild inflammatory activity, numerous Mallory-Denk bodies, and phospholipidosis due to amiodarone.
D I SC U SSI ON
This case highlights an unusual histological change observed in amiodarone injury—phospholipidosis. Drug-induced phospholipidosis is essentially a lipid-storage disease in which phospholipid membranes accumulate within lysosomes as inclusion bodies (1,2). Cationic amphiphilic drugs, such as amiodarone, are sequestered within these stacks of membranes. Phospholipidosis was first described in patients treated with diethylaminoethoxyhexestrol (trade name Coralgil) who
F I G U R E 1 5 . 8 . 6 Transmission electron micrograph of lamellar bodies in a case of Niemann-Pick disease.
CASE
15.8:
AMIODARONE-INDUCED
be found in zone 1. Amiodarone has been shown to increase cytokeratin 8 (CK8) monomers within treated cells and also activate the enzyme involved in cross-linking CK8 to form larger complexes which are the building materials for MalloryDenk bodies (6). Other histologic features include macro- and microvesicular steatosis, periportal and perisinusoidal fibrosis, lymphocytic inflammation, and cholestasis (7,8). Cirrhosis is uncommon, but well reported. An acute form of amiodaronerelated toxicity also exists, occurring within a day of intravenous loading (9). Patients develop a severe acute hepatocellular injury that usually resolves after stopping the infusion, but fulminant hepatic failure has occurred. It is possible that the drug vehicle, Polysorbate 80, is the true culprit. Although the case described here had a very complicated clinical presentation, with multiple co-morbidities, the biopsy showed characteristic changes of amiodarone toxicity. The causality analysis is presented in Table 15.8.1. The main
TA B LE 15. 8. 1 Causality analysis—amiodarone Temporal eligibility
Nine years of therapy with amiodarone
Exclusion of competing causes
Although not detailed, the other drugs he was taking were not reported to cause the pathologic pattern of cirrhosis with Mallory-Denk bodies; hepatitis virus screen was negative; ANA positive, but autoimmune hepatitis not consistent with pathology
Known potential for injury
Amiodarone is well described to cause DILI
Precedent for pathologic pattern
Cirrhosis with Mallory-Denk bodies and phospholipidosis is characteristic of amiodarone injury
Dechallenge/ rechallenge
Patient died soon after biopsy
Toxicology
Not done
Conclusion
Cirrhosis with Mallory-Denk bodies and phospholipidosis due (95% chance) to amiodarone
PHOSPHOLIPIDOSIS
239
diagnostic considerations from the pathology point of view would be other causes of cirrhosis, including viral and autoimmune hepatitis, as well as steatohepatitis. Viral hepatitis and alcohol were excluded by serologic tests and history, respectively. The pathology is helpful in excluding autoimmune hepatitis and steatohepatitis because both of these etiologies would be unlikely to result in cirrhosis with so many MalloryDenk bodies. A combined clinical-pathologic approach permits the confident elimination of all etiologies of liver injury except for amiodarone.
References 1. Anderson N, Borlak J. Drug-induced phospholipidosis. FEBS Lett. 2006;580:5533–5540. 2. Reasor MJ, Hastings KL, Ulrich RG. Drug-induced phospholipidosis: issues and future directions. Expert Opin Drug Saf. 2006;5:567–583. 3. Sun EL, Petrella DK, McCloud CM, Cramer CT, Reasor MJ, Ulrich RG. Amiodarone-induced cytoplasmic lamellar body formation in cultured primary rat and human hepatocytes: relationship to cell function and cytotoxicity. In Vitro Toxicol. 1997;10:459–470. 4. Guigui B, Perrot S, Berry JP, et al. Amiodarone-induced hepatic phospholipidosis: a morphological alteration independent of pseudoalcoholic liver disease. Hepatology. 1988;8:1063–1068. 5. Dake MD, Madison JM, Montgomery CK, et al. Electron microscopic demonstration of lysosomal inclusion bodies in lung, liver, lymph nodes, and blood leukocytes of patients with amiodarone pulmonary toxicity. Am J Med. 1985;78:506–512. 6. Robin MA, Descatoire V, Pessayre D, Berso A. Steatohepatitis-inducing drugs trigger cytokeratin cross-links in hepatocytes. Possible contribution to Mallory-Denk body formation. Toxicol In Vitro. 2008;22: 1511–1519. 7. Lewis JH, Mullick F, Ishak KG, et al. Histopathologic analysis of suspected amiodarone hepatotoxicity. Hum Pathol. 1990;21:59–67. 8. Lewis JH, Ranard RC, Caruso A, et al. Amiodarone hepatotoxicity: prevalence and clinicopathologic correlations among 104 patients. Hepatology. 1989;9:679–685. 9. Rizzioli E, Incasa E, Gamberini S, et al. Acute toxic hepatitis after amiodarone intravenous loading. Am J Emerg Med. 2007;25: 1082 e1–1082 e4.
Case 15.9
Drug-Induced Microvesicular Steatosis DAVID E. KLEINER
C L I N I C AL I N F OR M AT I ON
A 42-year-old man with chronic hepatitis B was presented to an emergency room with lethargy, weakness, myalgias, anorexia, and nausea. He had been enrolled in a clinical trial of an investigational new drug, fialuridine, for chronic hepatitis B, but had stopped taking the drug 3 weeks earlier due to symptoms of neurotoxicity after receiving a total of 11 weeks of therapy. Laboratory evaluation showed an ALT of 100 IU/L, total bilirubin 5.8 mg/dL, prothrombin time 15.4 s, albumin 2.8 g/dL, ammonia 77 umol/L, and lactate of 18.8 mmol/L. A diagnosis of hepatic failure was made, and he underwent transplantation about 1 week after initial presentation.
diffuse microvesicular steatosis (Figure 15.9.2). The microvesiculation was so fine in some cells that the cytoplasm had a granular appearance (Figure 15.9.3). Megamitochondria were seen in some hepatocytes, but no Mallory-Denk bodies were noted (Figure 15.9.4). There were no areas of confluent or zonal necrosis and only rare apoptotic hepatocytes. A marked ductular reaction extended from the portal areas deep into the parenchyma. Canalicular and hepatocellular cholestasis was present, best seen on the iron stain (Figure 15.9.5). Immunostains for hepatitis B surface and core stains were positive, but in fewer cells than in the pretreatment biopsy.
R E A S ON F OR R E F E R R A L
Progression of chronic hepatitis B versus drug-induced liver injury as the cause for the hepatic failure.
PAT H OL OG I C F E AT U R E S
A pretreatment biopsy was available for comparison with the explanted liver. The pretreatment biopsy showed mild portal inflammation with focal interface hepatitis (Figure 15.9.1). There were scattered foci of lobular inflammation. No steatosis was seen. Mild portal fibrotic expansion was present on the Masson Trichrome stain. Immunostains for hepatitis B surface and core antigens were positive. The explant revealed mild portal inflammation and interface hepatitis similar in degree to the pretreatment biopsy. The hepatocytes, however, showed
F I G U R E 1 5 . 9 . 2 Explant showing diffuse microvesicular steatosis and
persistent chronic hepatitis.
FIGURE 15. 9. 1 Pretreatment biopsy showing chronic hepatitis with
mild activity and no steatosis.
F I G U R E 1 5 . 9 . 3 Microvesicular steatosis.
240
CASE
15.9:
DRUG-INDUCED
MICROVESICULAR
S T E AT O S I S
241
TA BL E 1 5 . 9 . 1 Drugs and toxins associated with microvesicular
steatosis
FIGURE 15. 9. 4 Megamitochondria in a background of microvesicu-
lar steatosis.
FIGURE 15. 9. 5 Iron stain showing canalicular cholestasis.
D I AG N OS I S
Marked microvesicular steatosis and intrahepatic cholestasis due to fialuridine. Chronic hepatitis B with mild activity and portal fibrotic expansion.
D I S C U S S I ON
Microvesicular steatosis is a rare pattern of liver disease that is usually due to drug or toxic injury (Table 15.9.1). The associations outside of drugs and toxins are unusual and have rather specific clinical associations. These non–drug-induced liver injury (DILI) etiologies include fatty liver of pregnancy, alcoholic foamy degeneration (arguably a toxic injury), multiple hornet stings, and a variety of genetically transmitted metabolic diseases mainly related to lipid metabolism,
Acetylsalicylic acid (aspirin) Amineptine Amiodarone Calcium hopantenate Camphor Chlortetracycline Demeclocycline Desferoxamine Didanosine Fialuridine Hypoglycin (Jamaican vomiting sickness) Ibuprofen Indinivir Indomethacin Jin bu huan (herbal) Ketoprofen Margosa oil (herbal) Methyl salicylate Oxytetracycline Pennyroyal oil (herbal) Pentanoic acid (herbal) Piroxicam Pirprofen Riluzole Rolitetracycline Stavudine Syo-saiko-to (herbal) Tetracycline Tolmetin Valproic acid Vitamin A Warfarin
mitochondria, and the urea cycle. Essentially, all of these can be excluded by history, leaving only drug or toxic injury as a possibility. All these agents appear to cause mitochondrial injury (1,2) by inhibition of beta-oxidation (e.g., aspirin, amineptine, tetracycline, and valproate) or oxidative phosphorylation (e.g., amiodarone), or mitochondrial DNA replication (didanosine, stavudine, and fialuridine). Thus, identification of microvesicular steatosis as the pattern of injury in a new agent, as is the situation in this case example, can help narrow the possible mechanisms of injury (3). Microvesicular steatosis is frequently associated with lactic acidosis and hypoglycemia due to impaired mitochondrial function. If mitochondria are affected outside the liver, there may be evidence of other organ dysfunction, such as myopathy or pancreatitis. Patients may present with hepatomegaly and nonspecific symptoms of fatigue, nausea, vomiting, and weakness or they may present in fulminant failure, with hyperammonemia, jaundice, and hepatic encephalopathy. The transaminases may not be particularly elevated. On liver biopsy, microvesicular steatosis is recognized as a foamy change in the hepatocyte cytoplasm. The nucleus is not displaced as in macrovesicular steatosis but may be indented by the microvacuoles. The vacuoles may be smaller than the resolving power of a light microscope, in which case the hepatocyte cytoplasm may look finely granular. A fat stain such as
242
CHAPTER
15:
DRUG-INDUCED
oil red O or Sudan Black can be very helpful, and electron microscopy will also demonstrate fine vacuolation. The histologic differential diagnosis includes small droplet macrovesicular steatosis, in which there may be multiple small vacuoles instead of a single large one. The distinction is mainly based on size and quantity. If the cell contains uncountable numbers of small vacuoles, it is probably microvesicular steatosis. It is not unusual to see some macrovesicular steatosis in a case of microvesicular steatosis as is true for this case (Figure 15.9.2). Some diseases that are predominantly macrovesicular, such as NASH, will show some microvesicular steatosis, but unless the latter is diffuse, the case should not be diagnosed as microvesicular steatosis. Other histologic features of microvesicular steatosis include megamitochondria, which are seen as eosinophilic, PAS-negative, cytoplasmic inclusions about 1 to 3 microns in size and Mallory-Denk bodies. Hepatocellular and canalicular cholestasis can be seen, as in this case, but the overall pattern remains that of microvesicular steatosis. There is usually little inflammation, unless there is an underlying liver disease like viral hepatitis. Confluent necrosis is not a feature of microvesicular steatosis, although occasional acidophil bodies can be found. This case presented a challenge clinically (Table 15.9.2), because fialuridine was still an experiment agent and there TA B LE 15. 9. 2 Causality analysis—fialuridine Temporal eligibility
11 weeks of therapy with drug prior to the development of toxicity, but with a lag time of 3 weeks between stopping drug and DILI onset
Exclusion of competing causes
Not taking other drugs, exacerbation of hepatitis B excluded on pathologic exam
Known potential for injury
No prior reports of hepatotoxicity, but related drugs didanosine and zidovudine had both been implicated in syndromes of lactic acidosis and hepatic failure
Precedent for pathologic pattern
Didanosine had been reported to cause microvesicular steatosis
Dechallenge/rechallenge
Patient underwent liver transplantation, limiting observation of dechallenge; no rechallenge performed
Toxicology
Not done
Conclusion
Microvesicular steatosis due (95% chance) to fialuridine
Abbreviation: DILI, drug-induced liver injury.
LIVER
INJURY
had been no reports of hepatotoxicity in preclinical human or animal studies or in the phase I toxicity and pharmacokinetic studies (4,5). It was, unfortunately, the index case in a series of patients who developed hepatic failure from fialuridine. In the clinical trial that enrolled this patient, out of 15 patients enrolled, 5 died from liver failure and 2 survived with transplantation. Although drug toxicity was suspected clinically based on the development of lactic acidosis and hepatic failure, there were some doubts because the patient had stopped taking the drug and the serum half-life was thought to be short. Other nucleoside analogues, notably zidovudine and didanosine, have both been reported to cause lactic acidosis and liver failure (6–8). The finding of microvesicular steatosis on liver biopsy clinched the diagnosis of DILI. Didanosine has been implicated in causing microvesicular steatosis, providing a related precedent (9). In subsequent studies, it was shown that fialuridine was incorporated into mitochondrial DNA at a high level in addition to causing termination of DNA replication (10).
References 1. Fromenty B, Pessayre D. Inhibition of mitochondrial beta-oxidation as a mechanism of hepatotoxicity. Pharmacol Ther. 1995;67:101–154. 2. Pessayre D, Mansouri A, Haouzi D, et al. Hepatotoxicity due to mitochondrial dysfunction. Cell Biol Toxicol. 1999;15:367–373. 3. Tang W. Drug metabolite profiling and elucidation of drug-induced hepatotoxicity. Expert Opin Drug Metab Toxicol. 2007;3:407–420. 4. Kleiner DE, Gaffey MJ, Sallie R, et al. Histopathologic changes associated with fialuridine hepatotoxicity. Mod Pathol. 1997;10:192–199. 5. McKenzie R, Fried MW, Sallie R, et al. Hepatic failure and lactic acidosis due to fialuridine (FIAU), an investigational nucleoside analogue for chronic hepatitis B. N Engl J Med. 1995;333:1099–1105. 6. Chattha G, Arieff AI, Cummings C, et al. Lactic acidosis complicating the acquired immunodeficiency syndrome. Ann Intern Med. 1993;118:37–39. 7. Bissuel F, Bruneel F, Habersetzer F, et al. Fulminant hepatitis with severe lactate acidosis in HIV-infected patients on didanosine therapy. J Intern Med. 1994;235:367–371. 8. Olano JP, Borucki MJ, Wen JW, et al. Massive hepatic steatosis and lactic acidosis in a patient with AIDS who was receiving zidovudine. Clin Infect Dis. 1995;21:973–976. 9. Lai KK, Gang DL, Zawacki JK, et al. Fulminant hepatic failure associated with 2’,3’-dideoxyinosine (ddI). Ann Intern Med. 1991;115:283–284. 10. Lewis W, Griniuviene B, Tankersley KO, et al. Depletion of mitochondrial DNA, destruction of mitochondria, and accumulation of lipid droplets result from fialuridine treatment in woodchucks (Marmota monax). Lab Invest. 1997;76:77–87.
16 Cytoplasmic Globules ELAINE S. CHAN AND MATTHEW M.YEH
I N T ROD U C T I ON
A number of globules or inclusion bodies can be found within the cytoplasm of hepatocytes. These include alpha-1-antitrypsin globules, alpha-1-antichymotrypsin globules, Mallory-Denk bodies, megamitochondria, and pale bodies/fibrinogen storage inclusions, among others. Although cytoplasmic globules may be found in normal liver, when present, especially abundant, they are often associated with various pathologic states. Alpha-1-antitrypsin (AAT) is a serine protease inhibitor, encoded by SERPINA1, and is produced primarily in the hepatocytes. AAT deficiency is an autosomal recessive inherited metabolic disorder. The normal gene product is PiM; the most common deficiency alleles are PiS and PiZ, and PiZZ (where Pi stands for protease inhibitor and ZZ represents the banding pattern by isoelectric focusing) phenotype accounts for most cases of severe AAT deficiency (1). AAT deficiency is a codominant condition that predisposes to pulmonary emphysema and liver disease. As a serine protease inhibitor, AAT normally inhibits neutrophil elastase, thereby preventing the enzyme from disrupting elastin within the pulmonary alveolar connective tissue. In AAT-deficient patients, emphysema or chronic obstructive pulmonary disease is often noted. In addition, certain mutant genotypes are predisposed to hepatic disorders. For example, in individuals with the genotype PiZZ, there is pathologic polymerization of mutant AAT within the endoplasmic reticulum that cannot be transported to the Golgi apparatus, resulting in the intrahepatocytic accumulation of the mutant AAT proteins, which can be detected as cytoplasmic globules within the hepatocytes. This can in turn lead to the development of hepatic diseases, including cirrhosis and hepatocellular carcinoma (2,3). A1AT deficiency represents the most common genetic cause of neonatal liver disease. The disease usually manifests as persistent jaundice at 4 to 8 weeks of age. Morphologic changes include canalicular cholestasis, giant cell transformation of hepatocytes, and hepatocyte ballooning degeneration (see Chapter 10). Interlobular bile ducts may be reduced in number. Other histologic changes may also include ductular reaction and fibrosis, but the typical globules are usually not seen in the first few months after birth (4). In fetuses with PiZZ phenotype, granular deposits of A1AT can be detected by immunohistochemistry in liver by the second trimester (5). Similar to AAT, alpha-1-antichymotrypsin (ACT) is also a serine protease inhibitor. This plasma protein is encoded by SERPINA3 and is synthesized in the liver. Although its physiological function is not entirely clear, it can inhibit neutrophil cathepsin G and mast cell chymase, both of which are able to convert angiotensin-1 to the active angiotensin-2 (6). Defects in the SERPINA3 gene have been associated with chronic
obstructive pulmonary disease (7), as well as hepatic disorders. The most common type of ACT deficiency is caused by mutation in exon III leading to Pro to Ala substitution (8). Although the pathogenesis of ACT deficiency is not entirely clear, it has been thought that ACT deficiency leads to pulmonary and liver disease in a similar fashion as AAT deficiency, namely, through decreased suppression of biological activity of protease in the lung and through accumulation of abnormally folded protease inhibitor molecules in the hepatocytes, respectively. Aside from cytoplasmic globules caused by mutant AAT or ACT, a number of other cytoplasmic inclusions or globules can be observed in hepatocytes. These may mimic the mutant AAT or ACT globules and cause diagnostic pitfalls, such as Mallory-Denk bodies, intracellular hyaline bodies, megamitochondria, pale bodies in fibrinogen storage disease, and pseudoground glass hepatocytes and so on. The primary constituents of Mallory-Denk bodies are ubiquitinated keratins 8 and 18, small amounts of chaperones (heat shock proteins 25/60/70) and the polyubiquitin-binding protein p62 (9). Correspondingly, Mallory-Denk bodies can be labeled with antibodies to CK8, CK18, p62, and ubiquitin by immunohistochemistry. Mallory-Denk bodies are often seen in the ballooned hepatocytes in alcoholic hepatitis or nonalcoholic steatohepatitis (10–15) and even in the carcinoma cells of hepatocellular carcinoma. Although immunohistochemical staining represents a more sensitive method for Mallory-Denk body detection than conventional histologic staining (16), the use of immunohistochemistry is not as yet a routine practice in the clinical setting. Megamitochondria are observed in hepatocytes in both physiological and pathological conditions. It is not unusual to identify a few megamitochondria along with normal-sized mitochondria within a given cell. They are also detected more frequently in the livers of older normal subjects in comparison to younger individuals and are regarded as physiological changes related to aging (17). Megamitochondria formation has been described in numerous human pathological conditions, but they are most commonly seen in alcoholic liver injury. In fact, the presence of megamitochondria has been used as a diagnostic feature of alcoholic liver injury. The incidence of megamitochondria among chronic alcoholics vary from 25% to 93%, whereas the incidence of megamitochondria in nonalcoholics range from 0% to 37% (18–20). Bruguera et al (21) found that the presence of megamitochondria was correlated with the amount of alcohol consumed and to the shortness of abstinence prior to biopsy. They observed that megamitochondria may disappear within 1 month of abstinence, and therefore, the presence of megamitochondria should alert pathologists to the diagnosis
243
244
CHAPTER
16:
CYTOPLASMIC
of alcoholism. Besides, in alcoholic liver injury, megamitochondria have also been described in nonalcoholic steatohepatitis (22). Pale bodies are formed by intraceullar storage of secretory proteins, such as fibrinogen, albumin, C3, and C4 (23,24,25). These secretory proteins accumulate in large dilated cisternae of the rough endoplasmic reticulum, thought to be due to a defective intracellular transport or an excretion disturbance. Pale bodies have been identified in 5% to 6% (23,26) of resected hepatocellular carcinoma, especially of the fibrolamellar variant (27). They have also been observed in fibrinogen storage disease (28), which can be associated with hypofibrinoginemia and mutant variant of the fibrinogen molecule (29), but they can also be found transiently in the livers of patients who have acute infection without hypofibrinogenemia (30).
References 1. Fairbanks KD, Tavill AS. Liver disease in alpha-1-antitrypsin deficiency: a review. Am J Gastroenterol. 2008;103(8):2136–2141. 2. Lomas DA, Evans DL, Finch JT, Carrell RW. The mechanism of Z alpha-1-antitrypsin accumulation in the liver. Nature. 1992;357(6379): 605–607. 3. Sivasothy P, Dafforn TR, Gettins PG, Lomas DA. Pathogenic alpha-1antitrypsin polymers are formed by reactive loop-beta-sheet A linkage. J Biol Chem. 2000;275(43):33663–33668. 4. Mowat AP. Hepatitis and cholestasis in infancy: intrahepatic disorders. In: Mowat AP, ed. Liver disorders in children. London, UK: Butterworths & Co; 1982:50. 5. Malone M, Mieli-Vergani G, Mowat AP, Portmann B. The fetal liver in PiZZ alpha-1-antitrypsin deficiency: a report of five cases. Pediatr Pathol. 1989;9(6):623–631. 6. Rubin H, Wang ZM, Nickbarg EB, et al. Cloning, expression, purification, and biological activity of recombinant native and variant human alpha 1-antichymotrypsins. J Biol Chem. 1990;265(2):1199–1207. 7. Poller W, Faber JP, Weidinger S, et al. A leucine-to-proline substitution causes a defective alpha-1-antichymotrypsin allele associated with familial obstructive lung disease. Genomics. 1993;17(3):740–743. 8. Elzouki AN, Verbaan H, Lindgren S, Widell A, Carlson J, Eriksson S. Serine protease inhibitors in patients with chronic viral hepatitis. J Hepatol. 1997;27(1):42–48. 9. Strnad P, Zatloukal K, Stumptner C, Kulaksiz H, Denk H. Mallory-Denkbodies: lessons from keratin-containing hepatic inclusion bodies. Biochim Biophys Acta. 2008;1782(12):764–774. 10. Brunt EM. Nonalcoholic steatohepatitis. Semin Liver Dis. 2004;24(1): 3–20. 11. Zatloukal K, French SW, Stumptner C, et al. From Mallory to MalloryDenk bodies: what, how and why? Exp Cell Res. 2007;313(10): 2033–2049.
GLOBULES
12. Burt AD, Mutton A, Day CP. Diagnosis and interpretation of steatosis and steatohepatitis. Semin Diagn Pathol. 1998;15(4):246–258. 13. Pinto HC, Baptista A, Camilo ME, Valente A, Saragoça A, de Moura MC. Nonalcoholic steatohepatitis. Clinicopathological comparison with alcoholic hepatitis in ambulatory and hospitalized patients. Dig Dis Sci. 1996;41(1):172–179. 14. Yeh MM, Brunt EM. Pathology of fatty liver: differential diagnosis of nonalcoholic fatty liver disease. Diagnostic Pathology. 2008;14:586–597. 15. Li MK, Crawford JM. The pathology of cholestasis. Semin Liver Dis. 2004;24(1):21–42. 16. Ray MB. Distribution patterns of cytokeratin antigen determinants in alcoholic and nonalcoholic liver diseases. Hum Pathol. 1987;18(1):61–66. 17. Wilson PD, Franks LM. The effect of age on mitochondrial ultrastructure. Gerontologia. 1975;21(2):81–94. 18. Yokoo H, Singh SK, Hawasli AH. Giant mitochondria in alcoholic liver disease. Arch Pathol Lab Med. 1978;102(4):213–214. 19. Junge J, Horn T, Christoffersen P. Megamitochondria as a diagnostic marker for alcohol induced centrilobular and periportal fibrsis in the liver. Virchows Arch A Pathol Anat Histopathol. 1987;410(6):553–558. 20. Stewart RV, Dincsoy HP. The significance of giant mitochondria in liver biopsies as observed by light microscopy. Am J Clin Pathol. 1982;78(3):293–298. 21. Bruguera M, Bertran A, Bombi JA, Rodes J. Giant mitochondria in hepatocytes: a diagnostic hint for alcoholic liver disease. Gastroenterology. 1977;73(6):1383–1387. 22. Yeh MM, Brunt EM. Pathology of nonalcoholic fatty liver disease. Am J Clin Pathol. 2007;128(5):837–847. Review. 23. Moon WS, Yu HC, Chung MJ, Kang MJ, Lee DG. Pale bodies in hepatocellular carcinoma. J Korean Med Sci. 2000;15(5):516–520. 24. Callea F, de Vos R, Togni R, Tardanico R, Vanstapel MJ, Desmet VJ. Fibrinogen inclusions in liver cells: a new type of ground-glass hepatocyte. Immune light and electron microscopic characterization. Histopathology. 1986;10(1):65–73. 25. Ng IO, Ng M, Lai EC, Wu PC. Endoplasmic storage disease of liver: characterization of intracytoplasmic hyaline inclusions. Histopathology. 1989;15(5):473–481. 26. Nakashima O, Sugihara S, Eguchi A, Taguchi J, Watanabe J, Kojiro M. Pathomorphologic study of pale bodies in hepatocellular carcinoma. Acta Pathol Jpn. 1992;42(6):414–418. 27. Craig JR, Peters RL, Edmondson HA, Omata M. Fibrolamellar carcinoma of the liver: a tumor of adolescents and young adults with distinctive clinico-pathologic features. Cancer. 1980;46(2):372–379. 28. Callea F, Brisigotti M, Fabbretti G, Bonino F, Desmet VJ. Hepatic endoplasmic reticulum storage diseases. Liver. 1992;12(6):357–362. 29. Brennan SO, Wyatt J, Medicina D, Callea F, George PM. Fibrinogen brescia: hepatic endoplasmic reticulum storage and hypofibrinogenemia because of a gamma284 Gly-->Arg mutation. Am J Pathol. 2000;157(1):189–196. 30. Marucci G, Morandi L, Macchia S, et al. Fibrinogen storage disease without hypofibrinogenaemia associated with acute infection. Histopathology. 2003;42(1):22–25.
Case 16.1
Alpha-1-Antitrypsin Deficiency ELAINE S. CHAN AND MATTHEW M.YEH
C L I N IC AL I N F OR M AT I ON
A 52-year-old woman with no known history presents with elevated serum aminotransferase and liver cirrhosis of unknown etiology. She has no history of cigarette, alcohol, or illicit drug use. Family history is noncontributory. Serological markers are negative for viral or autoimmune hepatitis. There is no laboratory evidence of hemochromatosis, Wilson disease, or autoimmune diseases. A presumptive diagnosis of cirrhosis attributed to nonalcoholic steatohepatitis was made, and the patient was referred for a liver transplant consultation. A liver biopsy was performed. R E A S ON F OR R E F F E R AL
The liver biopsy shows cirrhosis with many intrahepatocytic globules that are periodic acid–Schiff (PAS) positive and diastase resistant. The referring pathologist’s specific question is whether the finding represents alpha-1-antitrypsin (AAT) deficiency. PAT H OL OG I C F E AT U R E S
Abnormal accumulation of mutant alpha-1antitrypsin as periodic acid-Schiff–positive, diastase-resistant globules in a patient with mutations in the SERPINA1 gene (PAS stain with diastase).
FIGURE 16.1.2
A liver biopsy shows distortion of the hepatic architecture and the formation of regenerative nodules consistent with cirrhosis. Intracytoplasmic globules can be seen on H&E–stained sections (Figure 16.1.1). The globules are round to oval, eosinophilic, and homogeneous with a waxy appearance. These cytoplasmic globules are strongly PAS positive and diastase resistant (Figure 16.1.2). Immunohistochemical stain for alpha1-antitrypsin is strongly positive (Figure 16.1.3).
F I G U R E 1 6 . 1 . 3 Alpha-1-antitrypsin (AAT) immunohistochemi-
cal stain highlighting the mutant AAT globules within the hepatocytes. Note the peripheral accentuation of the globules secondary to incomplete penetration of the immunohistochemical stain.
LA BO R ATO RY R ESULT S FIGURE 16. 1. 1 Eosinophilic cytoplasmic globules detected in the
hepatocytes of a patient with alpha-1-antitrypsin deficiency. The globules tend to have a periseptal and periportal distribution.
Serum electrophoresis reveals hypoalbuminemia with decrease in alpha-1-globulin fraction. Serum alpha-1-antitrypsin is lower than the normal reference range. Protease inhibitor phenotyping indicates Z homozygosity (PiZZ).
245
246
CHAPTER
16:
CYTOPLASMIC
D I AG N OS I S
Alpha-1-antitrypsin deficiency–associated cirrhosis.
D I SC U SSI ON
The patient has cirrhosis of unknown cause and was given a presumptive diagnosis of nonalcoholic steatohepatitis clinically. Microscopic examination reveals cirrhosis accompanied by the presence of intrahepatocytic globules. The cytoplasmic globules are accentuated by PAS stain with diastase. To elucidate the identity of the intracytoplasmic globules, immunohistochemical stain with alpha-1-antitrypsin is performed. The cytoplasmic globules show strong reactivity with AAT. Hence, the morphologic and immunohistochemical features are consistent with AAT deficiency–related cirrhosis. However, since the diagnosis of AAT deficiency is based on laboratory findings, additional laboratory testing is performed and the patient is found to have low plasma AAT and the homozygous PiZZ phenotype by isoelectric focusing, thus confirming the diagnosis of AAT deficiency. AAT deficiency was not initially suspected in the patient because it is a relatively uncommon cause of liver cirrhosis, especially in the absence of emphysema and a history of childhood liver disease. This case drives home the point the use of PASd stain in populations where this disease is prevalent and that the possibility of AAT deficiency should be considered when other more common causes of liver cirrhosis are excluded. AAT deficiency is persistently under-recognized, with AAT-deficient patients visiting an average of 3 different physicians and experiencing a mean delay of 7.2 years between initial symptoms and definitive diagnosis (1). Also, adult-onset cirrhosis secondary to AAT deficiency can occur without
GLOBULES
antecedent childhood liver disease. In cases of cirrhosis with other known etiology, the diagnosis of AAT deficiency should still be entertained, as it is not uncommon to have co-existing, previously undetected AAT deficiency in patients with other known liver disease (2). Therefore, if AAT-like globules are found in end-stage liver cirrhosis, they should be examined carefully, with the aid of special and immunohistochemical stains. Of note, AAT immunohistochemical stain typically stains small globules in their entirety and larger ones at the periphery (3). When a diagnosis of AAT deficiency is suspected based on positive PASd and immunohistochemical stains, the patient’s plasma AAT concentration should be tested. AAT is an acute phase reactant and may be elevated in conditions such as infection, hormonal stimulation, and cirrhosis. Plasma AAT can also be decreased in liver necrosis. Therefore, it is necessary to perform phenotyping by isoelectric focusing, currently considered to be the gold standard for the diagnosis of AAT deficiency as AAT-positive globules are not specific to AAT deficiency and can be found in acute and chronic hepatitis, cirrhosis, hepatocellular carcinoma, as well as other liver diseases (4,5). Typically, liver biopsy is not essential in the workup of AAT deficiency; however, histologic examination can be useful in assessing the extent of liver injury and fibrosis, and excluding other concomitant liver diseases. As mentioned at the beginning of the chapter, AAT globules should be distinguished from other cytoplasmic globules or inclusion bodies (Table 16.1.1) such as Mallory-Denk bodies, intracellular hyaline bodies, megamitochondria, pale bodies in fibrinogen storage disease, and pseudoground glass hepatocytes. Mallory-Denk bodies are present in alcoholic hepatitis and nonalcoholic steatohepatitis, chronic biliary diseases, Wilson disease, and amiodarone hepatotoxicity. Also,
TA B LE 16. 1. 1 Characteristics of different intracytoplasmic globules/inclusions Intracytoplasmic Globules
H&E
PASd
Alpha-1-antitrypsin deficiency
Round to oval, eosinophilic, and homogeneous globules
Positive
Alpha-1-antitrypsin
Alpha-1-antichymotrypsin deficiency
Granular
Positive
Alpha-1-antichymotrypsin
Mallory-Denk bodies
Eosinophilic, dense, branching, “twisted rope”
Negative
Blue with red center after chromotrope aniline blue labeling
CK8, CK18, p62, ubiquitin
Intracellular hyaline bodies
Globular, homogeneous, dense, eosinophilic cytoplasmic inclusions, often accompanied by a surrounding, clear halo
Negative
Congo red
p62
Megamitochondria
Round, oblong, or needle shaped, homogenous, regularly contoured
Negative
Pale bodies/fibrinogen storage bodies
Homogenous, palely eosinophilic, with a distinct border, and ground glass–like
Negative
Other Stains
Immunohistochemistry
Fibrinogen, occasionally weakly positive for albumin
CASE
16.1
:
ALPHA-1-ANTITRYPSIN
FIGURE 16. 1. 4 Mallory-Denk bodies with the characteristic twisted-rope appearance in alcoholic hepatitis or nonalcoholic steatohepatitis.
as Mallory-Denk bodies represent the result of chronic and accumulative injury to the liver (6), they can be found in advanced fibrosis, cirrhosis (7–10), or hepatocellular carcinoma (11–13). They are characterized as ropy, branched, and eosinophilic inclusions within the cytoplasm, typically in swollen or ballooned hepatocytes (Figure 16.1.4). In noncirrhotic liver, they are relatively easy to distinguish from AAT globules, as they are typically present with relevant histologic findings, such as steatosis and ballooned hepatocytes and in the appropriate clinical setting, such as metabolic syndrome or alcohol use. They are more challenging to distinguish when the liver is cirrhotic, as most of the above-mentioned features are no longer present. Intracellular hyaline bodies are globular, homogeneous, dense, and eosinophilic cytoplasmic inclusions that may mimic AAT globules, albeit they are typically accompanied by a surrounding clear halo and are most often identified in hepatocellular carcinoma and idiopathic copper toxicosis (14); therefore, the clinical scenario may differ from A1AT deficiency. Morphologically, hepatic megamitochondria are round, oblong, cigar or needle-shaped (15,16), homogenous, regularly contoured, and eosinophilic intracytoplasmic inclusions (Figure 16.1.5). Under light microscopy, the diameter of megamitochondria is roughly one-third that of the hepatocytic nucleus (17), but it may vary. Although morphologically resembling AAT globules, they are PAS-negative. Pale bodies (Figure 16.1.6) are homogenous, palely eosinophilic, with a distinct border, and ground glass–like, resembling the ground glass change characteristic of hepatitis B surface antigen (HbsAg)-containing hepatocytes. Similar to ground glass hepatocytes, pale bodies often occupy the entire cytoplasm, displacing the nuclei to the periphery and leaving a small rim of clear cytoplasm. Unlike AAT globules, they are negative for PAS stain, but are positive for fibrinogen, and are
DEFICIENCY
247
F I G U R E 1 6 . 1 . 5 Megamitochondria showing as round to oblong eosinophilic inclusions.
F I G U R E 1 6 . 1 . 6 Cytoplasmic, homogenous pale bodies seen in fibrolamellar hepatocellular carcinoma.
occasionally weakly positive for albumin. Ultrastructurally, pale bodies appear as amorphous, granular, or fibrillar material within dilated cisternae of the rough endoplasmic reticulum, which are moderately electron dense (18–20). Ground glass hepatocytes in chronic hepatitis B may mimic AAT globules, but typically they do not have the round and globular appearance of AAT globules and are PASnegative. Similarly, the recently described glycogen pseudoground glass change in hepatocytes may resemble AAT globules and is morphologically similar to the ground glass change seen in chronic hepatitis B infection, with distinct, circumscribed, gray-glassy inclusions surrounded by a rim of cytoplasm; however, the background livers showed only mild or no inflammation and mild or no fibrosis. Different from AAT globules and ground glass hepatocytes, the pseudoground glass change is PAS-positive and diastase-sensitive (21,22).
248
CHAPTER
16:
CYTOPLASMIC
AAT globules also need to be distinguished from the pale, round, or kidney-shaped inclusions in Lafora disease, which are weakly positive for PAS with diastase. They also stain with colloidal iron and silver stains.
References 1. Stoller JK, Fromer L, Brantly M, Stocks J, Strange C. Primary care diagnosis of alpha-1-antitrypsin deficiency: issues and opportunities. Cleve Clin J Med. 2007;74(12):869–874. 2. Iezzoni JC, Gaffey MJ, Stacy EK, Normansell DE. Hepatocytic globules in end-stage hepatic disease: relationship to alpha-1-antitrypsin phenotype. Am J Clin Pathol. 1997;107(6):692–627. 3. Ishak KG. Inherited metabolic diseases of the liver. Clin Liver Dis. 2002;6(2):455–479, viii. 4. Qizilbash A, Young-Pong O. Alpha-1-antitrypsin liver disease differential diagnosis of PAS-positive, diastase-resistant globules in liver cells. Am J Clin Pathol. 1983;79(6):697–702. 5. Hall P, Herrmann R, Brennan J, Mackinnon M. Detection of alpha1-antitrypsin in hepatocytes in acute and chronic hepatitis. Pathology. 1987;19(4):415–418. 6. Zatloukal K, French SW, Stumptner C, et al. From Mallory to MalloryDenk bodies: what, how and why? Exp Cell Res. 2007;313(10): 2033–2049. 7. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ. Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity. Gastroenterology. 1999;116(6):1413–1419. 8. Mendler MH, Kanel G, Govindarajan S. Proposal for a histological scoring and grading system for non-alcoholic fatty liver disease. Liver Int. 2005:294–304. 25(2):294–304. Erratum in: Liver Int. 2005;25(3): 682–683. 9. Cortez-Pinto H, Baptista A, Camilo ME, De Moura MC. Nonalcoholic steatohepatitis—a long-term follow-up study: comparison with alcoholic hepatitis in ambulatory and hospitalized patients. Dig Dis Sci. 2003;48(10):1909–1913.
GLOBULES
10. Gramlich T, Kleiner DE, McCullough AJ, Matteoni CA, Boparai N, Younossi ZM. Pathologic features associated with fibrosis in nonalcoholic fatty liver disease. Hum Pathol. 2004;35(2):196–199. 11. Jensen K, Gluud C. The Mallory body: morphological, clinical and experimental studies (Part 1 of a literature survey). Hepatology. 1994; 20(4 pt 1):1061–1077. 12. Lewis JH, Mullick F, Ishak KG, Ranard RC, Ragsdale B, Perse RM, Rusnock EJ, Wolke A, Benjamin SB, Seeff LB, et al. Histopathologic analysis of suspected amiodarone hepatotoxicity. Hum Pathol. 1990; 21(1):59–67. 13. Richer M, Robert S. Fatal hepatotoxicity following oral administration of amiodarone. Ann Pharmacother. 1995;29(6):582–586. 14. Denk H, Stumptner C, Fuchsbichler A, et al. Are the Mallory bodies and intracellular hyaline bodies in neoplastic and non-neoplastic hepatocytes related? J Pathol. 2006;208(5):653–661. 15. French SW, Nash J, Shitabata P, et al. Pathology of alcoholic liver disease. VA Cooperative Study Group 119. Semin Liver Dis. 1993;13(2):154–169. 16. Junge J, Horn T, Christoffersen P. Megamitochondria as a diagnostic marker for alcohol induced centrilobular and periportal fibrosis in the liver. Virchows Arch A Pathol Anat Histopathol. 1987;410(6):553–558. 17. Bruguera M, Bertran A, Bombi JA, Rodes J. Giant mitochondria in hepatocytes: a diagnostic hint for alcoholic liver disease. Gastroenterology. 1977;73(6):1383–1387. 18. Moon WS, Yu HC, Chung MJ, Kang MJ, Lee DG. Pale bodies in hepatocellular carcinoma. J Korean Med Sci. 2000;15(5):516–520. 19. Callea F, de Vos R, Togni R, et al. Fibrinogen inclusions in liver cells: a new type of ground-glass hepatocyte. Immune light and electron microscopic characterization. Histopathology. 1986;10(1):65–73. 20. Ng IO, Ng M, Lai EC, Wu PC. Endoplasmic storage disease of liver: characterization of intracytoplasmic hyaline inclusions. Histopathology. 1989;15(5):473–481. 21. Wisell J, Boitnott J, Haas M, et al. Glycogen pseudoground glass change in hepatocytes. Am J Surg Pathol. 2006;30(9):1085–1090. 22. Thomas RM, Schiano TD, Kueppers F, Black M. Alpha-1antichymotrypsin globules within hepatocytes in patients with chronic hepatitis C and cirrhosis. Hum Pathol. 2000;31(5):575–577.
Case 16.2
Alpha-1-Antichymotrypsin Deficiency ELAINE S. CHAN AND MATTHEW M.YEH
C L I N IC AL I N F OR M AT I ON
A 61-year-old man presents with malaise and weakness. Jaundice and mild hepatomegaly are noted. alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels are slightly elevated. Alkaline phosphatase, serum ferritin, serum copper, and serum ceruloplasmin are within normal reference ranges. Serological studies for hepatitis A, B, and C and autoimmune hepatitis are all negative. A liver biopsy was performed. R E A SON F OR R E F E R R AL
The liver biopsy shows PAS-positive, diastase-resistant globules. Immunohistochemical stain for alpha-1-antitrypsin (AAT) is negative. The referring pathologist is curious about the nature of these globules. PAT H OL OG I C F E AT U R E S
The liver biopsy shows patchy inflammatory mononuclear infiltrate with mild interface hepatitis. Intracytoplasmic globules are scattered within the hepatocytic cytoplasm, especially in the hepatocytes near the portal tracts. The globules are weakly periodic acid–Schiff (PAS) positive, diastase resistant. Immunohistochemistry shows no reactivity with AAT or fibringoen antibodies. There is, however, strong positivity with alpha-1-antichymotrypsin (ACT) antibodies in the intrahepatocytic globules.
D I AG N OS I S
Intrahepatocytic alpha-1-antichymotrypsin globules.
D I S C U S S I ON
The presence of PAS-positive, diastase-resistant, granules led initially to a differential diagnosis of AAT deficiency– associated chronic hepatitis. However, immunohistochemistry demonstrates a lack of AAT immunoreactivity in the cytoplasmic globules. Plasma level of AAT is within normal limits as well. Immunohistochemical staining with antibodies to ACT antigen is positive in these globules, with accentuation at the periphery (Figure 16.2.1). Multiple case reports in the literature have demonstrated that ACT deficiency is associated with chronic liver disease. However, since the prevalence of ACT deficiency in the population is very low, ACT deficiency is usually not in the differential diagnosis when chronic liver disease is
F I G U R E 1 6 . 2 . 1 Immunohistochemical staining for alpha-1-antichymotrypsin (ACT) demonstrates the presence of ACT globules within the cytoplasm of the hepatocytes. (Courtesy of Dr. Linda Ferrell, University of California, San Francisco.)
encountered and therefore rarely investigated in the clinical setting. Cytoplasmic globules have been demonstrated in hepatocytes of patients with cirrhosis associated with ACT deficiency. A case series shows that patients with hepatitis C virus (HCV) infection and low levels of serum ACT are more prone to developing cirrhosis to those with HCV only (1). In addition, there appears to be no established correlation between plasma levels of ACT and the abundance of globules in hepatocytes (2). It has been suggested that simultaneous heterozygous AAT and ACT mutations increase the risk of liver disease (2). Ultrastructurally, the inclusions have been described as fluffy material in the dilated cistern of the endoplasmic reticulum (3). This case and other published case reports (4) of hepatic diseases in patients with ACT deficiency illustrate the importance of including ACT deficiency in the differential diagnosis when intrahepatocytic globules/ inclusion bodies are encountered, especially when immunohistochemical stain for AAT is negative and the plasma level of AAT is normal. SUMMA RY
When intracytoplasmic globules/inclusion bodies are observed within the hepatocytes, it is crucial in distinguishing between physiologic and pathologic conditions, in arriving at a correct diagnosis, and in determining the severity of the hepatic involvement. Although immunohistochemistry with AAT and ACT is useful, it is by no means absolutely specific. Thus,
249
250
CHAPTER
16:
CYTOPLASMIC
immunohistochemical results should always be interpreted with caution. Also, it should be noted that AAT globules are not specific in individuals with AAT deficiency. They are not uncommonly found in liver cirrhosis due to various etiologies and in severely ill and older patients who have high concentration of AAT in the blood (5).
References 1. Thomas RM, Schiano TD, Kueppers F, Blac M. Alpha-1-antichymotrypsin globules within hepatocytes in patients with chronic hepatitis C and cirrhosis. Hum Pathol. 2000;31(5):575–577.
GLOBULES
2. Yoon D, Kueppers F, Genta RM, Klintmalm GB, Khaoustov VI, Yoff B. Role of alpha-1-antichymotrypsin deficiency in promoting cirrhosis in two siblings with heterozygous alpha-1-antitrypsin deficiency phenotype SZ. Gut. 2002;50(5):730–732. 3. Lindmark B, Millward-Sadler H, Callea F, Eriksso S. Hepatocyte incusions of alpha 1-antichymotrypsin in a patient with partial deficiency of alpha 1-antichymotrypsin and chronic liver disease. Histopathology. 1990;16(3):221–225. 4. Ortega L, Balboa F, González L. alpha(1)-Antichymotrypsin deficiency associated with liver cirrhosis. Pediatr Int. 2010;52(1):147–149. 5. Carlson J, Eriksson S, Hägerstran I. Intra- and extracellular alpha-1antitrypsin in liver disease with special reference to Pi phenotype. J Clin Pathol. 1981;34(9):1020–1025.
17 Glycogenic Abnormalities on Liver Biopsy MICHAEL TORBENSON
In most surgical pathology liver specimens, glycogen is inapparent on hematoxylin and eosin (HE) stains. However, the liver contains abundant glycogen that can be visualized with periodic acid–Schiff (PAS) stains in both healthy and diseased livers. As part of normal energy metabolism, the liver takes up excess glucose from the blood stream and converts it to glycogen by the process of glycogenesis. Glycogen is a polymeric chain of glucose molecules that serves as a storage form for glucose. Glycogen can be rapidly converted back to glucose if needed to meet the body’s energy demands. In some cases, glycogen accumulation in the liver reaches pathological levels and can cause disease. This occurs primarily in 2 conditions: (1) mutations that occur in various enzymes within glycogen metabolism pathways, leading to abnormal accumulation of glycogen; (2) glycogenic hepatopathy, a distinctive clinicopathological entity where the normal balance between glycogenesis and glycogenolysis is disrupted and excess glycogen accumulates within hepatocytes. Glycogenic hepatopathy is not associated with known mutations. Instead, glycogenic hepatopathy is caused in most cases by high and poorly controlled levels of serum glucose. G LY C O G E N S TOR AG E D I S E A S E S
There are many glycogen storage diseases that present as abnormal accumulations of glucose within the hepatocytes (1,2). These include primarily types Ia/b, IIIa/b, VI, IX, and XI. The hepatocytes in glycogen storage diseases are markedly swollen and filled with glycogen (Figure 17.1).
FIGURE 17. 1 Glycogen storage disease type IIIa. The hepatocytes are swollen and pale. There is no inflammation or fatty change in this case. Scattered glycogenated nuclei are seen.
A precise diagnosis of the subtype of glycogen storage disease cannot be reliably made on the basis of histology alone. Instead, the biopsy is useful to demonstrate the abnormal accumulation of glycogen. Additional biochemical assays are then needed to precisely classify the type of glycogen storage disease. Clinical features may also be helpful in some cases, but the degree of clinical overlap and the frequent presence of significant clinical ambiguities limit this approach and precise biochemical assays are typically needed. UR EA CY CLE DEFECT S
In addition to individuals with glycogen storage disease, marked glycogen accumulation is often seen in individuals with inherited defects in urea cycle enzymes. In one study, marked liver glycogen accumulation was reported in 8/11 children with urea cycle defects, including those with ornithine transcarbamylase deficiency, argininosuccinate lyase deficiency, and carbamoyl phosphate synthetase deficiency (3). The histological findings can be essentially identical to that of both glycogen storage disease as well as that of glycogenic hepatopathy (Figure 17.2), but the clinical settings are distinctively different. Classically, individuals with urea cycle defects present in infancy or childhood. However, increasing numbers of individuals with milder forms of disease are being reported in adults. The histological findings in adults have not been well described but presumably may resemble that of glycogenic hepatopathy.
Argininesuccinatelyase deficiency. The hepatocytes are swollen and pale, similar to that seen in glycogen storage diseases. FIGURE 17.2
251
252
CHAPTER
17:
G LY C O G E N I C
ABNORMALITIES
ON
LIVER
BIOPSY
G LY C O G E N I C H E PATOPAT H Y
Glycogenic hepatopathy closely resembles glycogen storage diseases on routine light microscopy (Figure 17.3). There may be subtle histological differences in some cases. For example, there tends to be more diffuse hepatocyte swelling and greater cytoplasmic clumping of glycogen in glycogen storage diseases, but these subtle findings are not sufficiently sensitive or specific to be of routine use and it is really the clinical parameters that differentiate these 2 entities. For example, glycogenic hepatopathy typically occurs in the setting of poorly controlled diabetes and is responsive to diabetic control. The presence of significant fibrosis, which can develop in some forms of glycogen storage disorders, would also argue against glycogenic hepatopathy (Figure 17.4). C L I N I C AL F I N D I N G S
The classic clinical setting for the development of glycogenic hepatopathy is the presence of elevated blood sugars and elevated insulin levels, usually in the setting of type I diabetes mellitus and poor glycemic control. Glycogenic hepatopathy has been referred to by other descriptive names over the years, all of which emphasize the key finding of excess glycogen accumulation (Table 17.1). Glycogenic hepatopathy was first documented in 1930 in a study that described the Mauriac syndrome (4). The Mauriac
F I G U R E 1 7 . 4 Cirrhosis associated with glycogen storage disease (same case as in Figure 17.1). The presence of advanced fibrosis or cirrhosis would be unusual for glycogenic hepatopathy and instead suggests a glycogen storage disease.
TA BL E 1 7 . 1 Other descriptive names for glycogenic hepatopathy Diabetes mellitus–associated glycogen storage Hepatomegaly (14) Hepatic glycogenosis (15) Liver glycogenosis (16) Liver glycogen storage (17,18)
FIGURE 17. 3 Glycogenic hepatopathy. Note that the histological changes are essentially identical to that of glycogen storage disease (Figure 17.1) in that the main finding is of swollen hepatocytes with clear cytoplasm. In glycogen hepatopathy, there can be a range of glycogen accumulation, from mild to marked, as shown in this image, where the findings will be essentially identical to that of inherited glycogen storage disease.
syndrome results from poorly controlled type I diabetes and includes features of growth retardation, delayed puberty, cushingoid features, and hypercholesterolemia. These clinical findings are accompanied by hepatomegaly, abnormal liver enzymes, and marked accumulation of glycogen within hepatocytes on liver biopsy. The Mauriac syndrome was first described soon after the introduction of insulin into clinical care, highlighting the strong association between this disease and insulin treatment of diabetics. The Mauriac syndrome is rarely seen today because of improved diagnosis and care of type I diabetes. However, less severe clinical manifestations of type I diabetes are still seen, including glycogenic hepatopathy (5). In fact, most current reports of glycogenic hepatopathy are in type I diabetic patients, typically in the setting of rapid reversal of ketosis or in periods of poor diabetic control. At clinical presentation, a wide variety of clinical signs and symptoms have been observed in patients with glycogenic hepatopathy (Table 17.2). Of note, elevated transaminase levels and hepatomegaly are almost universal findings. In some cases, the patient’s hepatic enzymes can be dramatically elevated and can exceed 10 times the upper limit of normal (6). Also of importance, the enzymes’ levels may fluctuate considerably over time (7). However, in all cases, the liver’s synthetic function is well preserved. For this reason,
CHAPTER
17:
G LY C O G E N I C
ABNORMALITIES
TA B LE 17. 2 Summary of clinical symptoms and signs. A single
representative reference is provided Findings at Presentation Clinical Symptoms Abdominal distension (9) Anorexia (9) Nausea (15)
ON
LIVER
BIOPSY
253
glycogen within the cytoplasm of hepatocytes. PAS staining will disappear after digestion with diastase. However, it should be remembered that even normal livers will have PAS positivity and the typical HE appearance is required to make the diagnosis of glycogenic hepatopathy. Electron microscopy is not needed to make the diagnosis but has been performed in previous studies, confirming the presence of glycogen accumulation at the ultrastructural level (15,5,6).
Right upper quadrant pain (6)
MECH A NISMS
Shortness of breath (6) Clinical Signs Ascites (6) Hepatomegaly (7,19) Poor growth (9) Vomiting (15) Laboratory Test Abnormalities Ketoacidosis (6) Elevated serum glucose (20) Increased liver enzymes (ALT/AST) (11) Liver enzyme flares (ALT/AST) (7) Increase serum lipids (13) Abbreviations: ALT, alanine aminotransferase; AST, aspartate transaminase.
glycogenic hepatopathy is not associated with low albumin levels or clotting abnormalities from decreased synthesis of clotting factors. Although rare, ascites is a dramatic presentation of glycogenic hepatopathy (6,8,15). The presence of ascites can be clinically misinterpreted as evidence for advanced liver disease. However, the ascites is not due to advanced fibrosis nor to low albumin levels but appears to result from compression of the sinusoids by the glycogen-laden hepatocytes, which restricts blood flow within the liver. The ascites of glycogenic hepatopathy typically resolves with adequate control of blood sugar (6,8,15). Also of note, both fatty liver disease as well as glycogenic hepatopathy can appear echogenic on ultrasound evaluation (9,15–17). Because of this, individuals with glycogenic hepatopathy may have an erroneous diagnosis of fatty liver disease at the time of liver biopsy. C LI N I C AL C OU R S E
The hepatomegaly and abnormal liver enzymes associated with glycogenic hepatopathy will improve with better glycemic control (15,16,9,10,11). In addition, histological resolution has also been documented in follow-up liver biopsies (6,12). H I S TOL OG I C F I N D I N G S
The distinctive histologic findings along with a history of diabetes are usually sufficient to make a diagnosis of glycogenic hepatopathy. A PAS stain can be used to highlight the
Glycogenic hepatopathy occurs in the setting of excess blood glucose that is also accompanied by high insulin levels. Glucose in the blood will passively diffuse into hepatocytes in a process that is driven largely by the glucose concentration gradient. Thus, high glucose levels lead to elevated levels of glucose diffusion into the hepatocytes. Once the glucose is in the hepatocyte cytoplasm, it is rapidly converted to glycogen and the glycogen is trapped within the cytoplasm. Over time, this leads to hepatocyte swelling. High levels of insulin can drive this process by increasing the conversion rate to glycogen. Interestingly, Manderson et al found that the hepatic levels of glycogen phosphorylase (PGYL), an enzyme that plays a key role in glycogenolysis, appeared lower than normal in 2 cases of glycogenic hepatopathy (13), but the precise role for genetic susceptibility remains undefined.
References 1. McAdams AJ, Hug G, Bove KE. Glycogen storage disease, types I to X: criteria for morphologic diagnosis. Hum Pathol. 1974;5:463–487. 2. Jevon GP, Finegold MJ. Reliability of histological criteria in glycogen storage disease of the liver. Pediatr Pathol. 1994;14:709–721. 3. Miles L, Heubi JE, Bove KE. Hepatocyte glycogen accumulation in patients undergoing dietary management of urea cycle defects mimics storage disease. J Pediatr Gastroenterol Nutr. 2005;40:471–476. 4. Mauriac P. Gros ventre, hepatomegalie, troubles de las croissance chez les enfants diabetiques traits depuis plusieurs annes par l’insuline. Gax Hebd Med Bordeaux. 1930;26:402–410. 5. Lorenz G, Barenwald G. Histologic and electron-microscopic liver changes in diabetic children. Acta Hepatogastroenterol (Stuttg). 1979;26:435–438. 6. Torbenson M, Chen YY, Brunt E, et al. Glycogenic hepatopathy: an underrecognized hepatic complication of diabetes mellitus. Am J Surg Pathol. 2006;30:508–513. 7. van den Brand M, Elving LD, Drenth JP, van Krieken JH. Glycogenic hepatopathy: a rare cause of elevated serum transaminases in diabetes mellitus. Neth J Med. 2009;67:394–396. 8. Bronstein HD, Kantrowitz PA, Schaffner F. Marked enlargement of the liver and transient ascites associated with the treatment of diabetic acidosis. N Engl J Med. 1959;261:1314–1318. 9. Munns CF, McCrossin RB, Thomsett MJ, Batch J. Hepatic glycogenosis: reversible hepatomegaly in type 1 diabetes. J Paediatr Child Health. 2000;36:449–452. 10. Tomihira M, Kawasaki E, Nakajima H, et al. Intermittent and recurrent hepatomegaly due to glycogen storage in a patient with type 1 diabetes: genetic analysis of the liver glycogen phosphorylase gene (PYGL). Diabetes Res Clin Pract. 2004;65:175–182. 11. Olsson R, Wesslau C, William-Olsson T, Zettergren L. Elevated aminotransferases and alkaline phosphatases in unstable diabetes mellitus without ketoacidosis or hypoglycemia. J Clin Gastroenterol. 1989;11: 541–545.
254
CHAPTER
17:
G LY C O G E N I C
ABNORMALITIES
12. Fridell JA, Saxena R, Chalasani NP, et al. Complete reversal of glycogen hepatopathy with pancreas transplantation: two cases. Transplantation. 2007;83:84–86. 13. Manderson WG, McKiddie MT, Manners DJ, Stark JR. Liver glycogen accumulation in unstable diabetes. Diabetes. 1968;17:13–16. 14. Nakamuta M, Ohashi M, Goto K, Tanabe Y, Hiroshige K, Nawata H. Diabetes mellitus-associated glycogen storage hepatomegaly: report of a case and review of the Japanese literature. Fukuoka Igaku Zasshi. 1993;84:354–358. 15. Chatila R, West AB. Hepatomegaly and abnormal liver tests due to glycogenosis in adults with diabetes. Medicine (Baltimore). 1996;75: 327–333.
ON
LIVER
BIOPSY
16. Carcione L, Lombardo F, Messina MF, Rosano M, De Luca F. Liver glycogenosis as early manifestation in type 1 diabetes mellitus. Diabetes Nutr Metab. 2003;16:182–184. 17. Torres M, Lopez D. Liver glycogen storage associated with uncontrolled type 1 diabetes mellitus. J Hepatol. 2001;35:538. 18. Evans RW, Littler TR, Pemberton HS. Glycogen storage in the liver in diabetes mellitus. J Clin Pathol. 1955;8:110–113. 19. Sweetser S, Kraichely RE. The bright liver of glycogenic hepatopathy. Hepatology. 2010;51:711–712. 20. Bassett JT, Veerappan GR, Lee DH. Glycogenic hepatopathy: a rare cause of increased aminotransferase levels in a diabetic patient. Clin Gastroenterol Hepatol. 2008;6:A26.
Case 17.1
Glycogenic Hepatopathy MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
A 16-year-old boy with a long history of type I diabetes presented with abdominal pain. He had been active on the swimming team and had been supplementing his diet because of the vigorous exercise. His blood glucose levels showed increased fluctuation during this time. Blood testing showed an ALT level of 450 and an AST level of 721 IU/mL. Ultrasound showed a strongly echogenic liver consistent with fatty liver disease. A biopsy was performed to investigate the marked AST and ALT elevations. R E A SON F OR R E F E R R AL
No fatty change was seen. There were scattered cytoplasmic eosinophilic globules with equivocal PAS-d staining. The case was referred with the question of whether this was alpha-1antitrypsin deficiency (AAT). PAT H O L OG I C A L F E AT U R E S
The liver biopsy specimens show a diffuse hepatocyte cytoplasmic clearing (Figure 17.1.1). No fatty change is seen. Rare apoptotic bodies are present, but there is no significant inflammation. Glycogenated nuclei are also easily found. No fibrosis is seen on trichrome stain. Scattered cytoplasmic eosinophilic globules are present (Figure 17.1.2), but there is no zonal distribution, and they appear smaller than typical alpha-1antitrypsin globules. A repeat PAS-d stain is negative. An iron stain is negative.
F I G U R E 1 7 . 1 . 2 Megamitochondria. The hepatocytes have numerous small red structures that are round to globoid in shape (arrows). These megamitochondria will be periodic acid-Schiff diastase negative and lack the zone 1 preference of alpha-1-antitrypsin. If you like, you can also stain the megamitochondria with a phosphotungstic acid-haematoxylin stain (see Figures 17.1.3 and 17.1.4).
DIAGNO SIS
Glycogenic Hepatopathy.
DISCUSSIO N
The biopsy findings are typical of glycogenic hepatopathy. Likewise, the clinical history is a perfect fit for this diagnosis. The pink globules that are seen represent mega-mitochondria. These large abnormal mitochondrial structures are not unique to glycogenic hepatopathy and are also commonly found in fatty liver disease as well as other chronic liver diseases. Sometimes they can resemble the globules of AAT. However, the globules typical of AAT tend to be larger and have a strong zone 1 distribution, especially in noncirrhotic livers. The globules of AAT are also strongly PAS-d positive in a properly stained section (Figure 17.1.3). If you are still uncertain, then a phosphotungstic acid-haematoxylin (PTAH) stain will nicely highlight the mega-mitochondria (Figure 17.1.4). Other Histologic Findings in Type I Diabetes FIGURE 17. 1. 1 Glycogenic hepatopathy. This is a more striking case of
glycogenic hepatopathy than shown in Figure 17.3. The hepatocytes are swollen with cleared cytoplasm.
Fatty liver
Although glycogenic hepatopathy is the most frequent histological finding in type I diabetic patients who have
255
256
CHAPTER
17:
G LY C O G E N I C
ABNORMALITIES
FIGURE 17. 1. 3 Periodic acid-Schiff diastase. Alpha-1-antitrypsin
globules. In this typical case of alpha-1-antitrypsin deficiency, numerous intracellular globules of alpha-1-antitrypsin protein are seen. In alpha-1-antitrypsin deficiency, the globules tend to have a zone 1 distribution, whereas megamitochondria are typically azonal.
hepatomegaly, fatty liver disease is also common in this setting. For example, in one study of 99 children with diabetes and hepatomegaly, glycogen accumulation was the most common finding associated with hepatomegaly. Moderate glycogen accumulation was seen in 22% of cases and marked glycogen accumulation in 19% of cases (1). However, fatty liver was also seen in nearly half of the total number of cases. Although the fatty change was usually mild, it did appear to explain the hepatomegaly in 8% of the children (1). Hepatosclerosis
Recently, an additional pattern of liver injury, termed “hepatosclerosis,” has been described in a series of liver biopsies from 12 diabetic patients (2). The liver biopsy specimens showed dense sinusoidal fibrosis, even though the livers were noncirrhotic. These findings were not accompanied by fatty liver or by glycogenic hepatopathy. Interestingly, the patients
ON
LIVER
BIOPSY
F I G U R E 1 7 . 1 . 4 Phosphotungstic acid–hematoxylin. megamito-
chondria. This stain highlights the usual mitochondria as numerous tiny blue dots in the hepatocyte cytoplasm. The megamitochondria can be seen as much larger structures with a thin clear halo around them. In the case shown, the megamitochondria are almost as big as some of the nuclei.
often had extensive histories of microangiopathic complications from their diabetes mellitus that involved multiple organ systems, suggesting that hepatosclerosis is a result of microangiopathic disease of the liver. Follow-up
The patients liver enzymes normalized with improved glycemic control. The abdominal pain also subsided.
References 1. Lorenz G, Barenwald G. Histologic and electron-microscopic liver changes in diabetic children. Acta Hepatogastroenterol (Stuttg). 1979;26:435–438. 2. Harrison SA, Brunt EM, Goodman ZD, et al. Diabetic Hepatosclerosis: perisinusoidal fibrosis without steatohepatitis amongst diabetics. Hepatology. 2003;38:666A.
Case 17.2
Glycogen in the Liver: Abnormal Versus Normal MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
A 61-year-old woman presented with headaches and spontaneous bleeding from oral mucosal membranes. Further workup showed mildly but persistently abnormal ALT levels (1.5–2.0 times the upper limit of normal). Her AST and alkaline phosphatase levels were within normal limits to slightly elevated. Her bilirubin was slightly elevated. She is non-obese, and viral studies and autoimmune studies were negative. She had a long history of Crohn disease but had good disease control for several decades. She had mild type II diabetes, currently managed by dietary considerations. A liver biopsy was performed. R E A SON F OR R E F E R R AL
The question raised on referral was whether the PAS positivity on the liver biopsy indicated glycogenic hepatopathy. F I G U R E 1 7 . 2 . 2 PAS. The same case as in Figure 17.1. The PAS stain is strongly and diffusely positive, but this finding alone is insufficient for the diagnosis of glycogenic hepatopathy.
PAT H OL OG I C F E AT U R E S
The liver biopsy showed no significant inflammation, fatty change, or fibrosis (Figure 17.2.1). The lobules showed mild hepatocyte disarray and rare apoptotic bodies. A reticulin stain suggested equivocal nodular regenerative hyperplasia type changes. An iron stain showed mild predominately Kupffer cell iron accumulation. A PAS stain was strongly positive (Figure 17.2.2).
DIAGNO SIS
Normal glycogen content.
DISCUSSIO N
Although the biopsy specimen shows subtle abnormalities, the findings do not strongly suggest a specific etiology. Strong hepatocyte PAS positivity is seen; however, the HE findings are not that of glycogenic hepatopathy. In glycogenic hepatopathy, the hepatocytes are diffusely enlarged and appear pale, often with accentuation of the cell membranes on routine HE stains. Overall, the PAS stain findings are within normal limits. PAS stains are often no longer part of the routine evaluation of liver biopsies (including within my practice). Because of this, many pathologists do not have an extended experience with evaluating PAS stains in various situations. Instead, PAS stains are often performed only when there is a history of diabetes and no other clear histological explanation for elevated liver enzymes. In such cases, strong but normal PAS staining can sometimes be misleading. This case illustrates the importance of having HE findings consistent with the diagnosis of glycogenic hepatopathy and not relying on a strong PAS stain for diagnosis. FIGURE 17. 2. 1 The liver showed mild lobular disarray with rare
apoptotic bodies and rare mitotic figures. The specimen does not show the typical hepatocyte clearing of glycogenic hepatopathy.
Follow-up
Subsequent studies found the patient had a hyperviscosity syndrome, which provides an explanation for the mild liver changes.
257
Case 17.3
Glycogenic Hepatopathy, Cause Uncertain MICHAEL TORBENSON
C L I N I C AL I N F OR M AT I ON
A 39-year-old malnourished man with a history of substantial, daily alcohol use and mental illness had normal ALT, and AST 1.5–2.5 times the upper limit of normal. He was a poor historian but claimed he had been sober for several months. He presented with chronic intermittent abdominal pain. He was not taking any medications. Alkaline phosphatase levels were normal. Clinically, fatty liver disease was suspected based on clinical history and ultrasound findings. A biopsy was performed to evaluate further the abdominal pain and to stage the liver fibrosis. R E A S ON F OR R E F E R R A L
The biopsy showed cytoplasmic clearing that suggested glycogenic hepatopathy, but there was no clinical history of type I diabetes, so the case was referred with the question as to whether this was still consistent with glycogenic hepatopathy. PAT H OL OG I C F E AT U R E S
On biopsy, hepatocytes are moderately enlarged and appear pale with accentuation of the cell membranes (Figure 17.3.1). The overall hepatic architecture is preserved, and there is no significant fibrosis. Glycogenated nuclei are occasionally seen. There is minimal patchy nonspecific portal chronic inflammation with no significant lobular inflammation. There is also mild (approximately 5%) macrovesicular steatosis. Minimal zone 3 pericellular fibrosis is present on trichrome stain.
DIAGNO SIS
Glycogenic hepatopathy.
DISCUSSIO N
The histological findings are entirely consistent with the diagnosis of glycogenic hepatopathy and do no not strongly suggest another etiology. What is unusual, however, is the lack of a clinical history to suggest diabetes. This case brings up the wider differential for cases of histologically typical glycogenic hepatopathy. Medication effect: Although this patient does not have any history of medication usage, glycogenic hepatopathy can be associated with medications. For example, short-term–high-dose steroid therapy can lead to glycogenic hepatopathy (1). In fact, the clinical presentation after high-dose steroid therapy (hepatomegaly and elevated transaminases elevations), as well as the histological findings, can be identical to that of glycogenic hepatopathy in the setting of diabetes mellitus. Alpha-glucosidase inhibitors are currently under study to treat individuals with type II diabetes mellitus (2). Alphaglucosidase inhibitors such as emiglitate and miglitol can induce hepatic glycogen accumulation in animal models when given at high dosages (3). However, it remains to be seen whether glycogenic hepatopathy will be a clinical side effect in patients treated with alpha-glucosidase inhibitors. It seems likely that the list of medications that can cause glycogenic hepatopathy will increase as this histological pattern of injury becomes increasingly recognized. Malnutrition: It is counterintuitive that malnutrition can lead to glycogenic hepatopathy. It has not been well described in the limited literature on the liver pathology of malnutrition and is probably not a typical finding. Nevertheless, glycogenic hepatopathy was observed in a 22-year-old patient with anorexia nervosa who presented with abnormal liver enzymes (4). The liver biopsy showed changes typical of glycogenic hepatopathy, and the patient’s liver enzyme abnormalities improved after nutritional therapy (4). In the case under consideration, malnourishment could potentially explain the glycogenic hepatopathy, but the precise etiology remains unresolved. Further evaluation to rule out subclinical diabetes or insulin resistance would also be important in this case.
FIGURE 17. 3. 1 The biopsy shows a milder case of glycogenic hepat-
opathy. The hepatocytes show moderate accumulations of glycogen.
Follow-up
None available. 258
CASE
17.3:
G LY C O G E N I C
H E PAT O PAT H Y,
References 1. Iancu TC, Shiloh H, Dembo L. Hepatomegaly following short-term high-dose steroid therapy. J Pediatr Gastroenterol Nutr. 1986;5:41–46. 2. van de Laar FA, Lucassen PL, Akkermans RP, van de Lisdonk EH, Rutten GE, van Weel C. Alpha-glucosidase inhibitors for patients with type 2 diabetes: results from a Cochrane systematic review and metaanalysis. Diabetes Care. 2005;28:154–163.
C AU S E
U N C E RTA I N
259
3. Lembcke B, Lamberts R, Wohler J, Creutzfeldt W. Lysosomal storage of glycogen as a sequel of alpha-glucosidase inhibition by the absorbed deoxynojirimycin derivative emiglitate (BAYo1248). A drug-induced pattern of hepatic glycogen storage mimicking Pompe’s disease (glycogenosis type II). Res Exp Med (Berl). 1991;191:389–404. 4. Komuta M, Harada M, Ueno T, et al. Unusual accumulation of glycogen in liver parenchymal cells in a patient with anorexia nervosa. Intern Med. 1998;37:678–682.
Case 17.4
Glycogenic Hepatopathy, Type II Diabetes MICHAEL TORBENSON
C L I N I C AL I N F OR M AT I ON
A 53-year-old obese male underwent bariatric surgery. He has type II diabetes mellitus and mild hypertension under good control. During the surgery, the liver was biopsied to stage and grade the fatty liver disease. R E A S ON F OR R E F E R R A L
a distinctive nodularity (Figure 17.4.1). The zone 3 hepatocytes do not appear to be atrophic as would be typical of nodular regenerative hyperplasia. This HE impression is confirmed by a reticulin stain. Only mild focal portal fibrosis is seen on trichrome, and advanced fibrosis is clearly not the cause of the nodularity. A PAS stain shows that the cleared hepatocytes and the vague nodularity are due to glycogen accumulation.
The biopsy shows a mild but distinctive nodularity in addition to the mild fatty change. The case was referred with the question of whether this nodularity represented nodular regenerative hyperplasia.
DIAGNO SIS
PAT H OL OG I C F E AT U R E S
Glycogenic hepatopathy, associated with diabetes type II.
The wedge biopsy shows mild (approximately 5%) macrovesicular steatosis. In addition, the zone 3 hepatocytes have a mild but distinctive cytoplasmic clearing, giving the biopsy DISCUSSIO N
Although glycogenic hepatopathy is most commonly seen in type I diabetic patients, it has also been reported in a type II diabetic patient after large doses of glucose were given to counteract an overdose of long-acting insulin (1). In addition, adults who are insulin dependent because of poorly controlled type II diabetes mellitus can develop glycogenic hepatopathy that closely resembles the changes seen in children with type I diabetes (2). Also of note, adults with type II diabetes can have milder forms of liver glycogenosis that only become evident when biopsies are performed to evaluate the liver for other disease processes, such as chronic hepatitis C or fatty liver disease (2). The clinical and pathological correlates of glycogenosis in this setting have not been well characterized to date, but this case appears to fit into this category. FIGURE 17. 4. 1 In this wedge biopsy from an individual undergoing
bariatric surgery, the liver shows only mild fatty change. A trichrome stain showed no significant fibrosis, and a reticulin stain showed no evidence of nodular regenerative hyperplasia. However, there is a distinct nodularity to the liver. In this case, the nodularity is a result of glycogen accumulation with a distinct zone 3 distribution. This glycogen accumulation has led the zone 3 hepatocytes having a more clear appearance to their cytoplasm on the H&E. A PAS stain confirmed the increased glycogen in the clear areas.
References
260
1. Tsujimoto T, Takano M, Nishiofuku M, et al. Rapid onset of glycogen storage hepatomegaly in a type-2 diabetic patient after a massive dose of long-acting insulin and large doses of glucose. Intern Med. 2006;45: 469–473. 2. Chatila R, West AB. Hepatomegaly and abnormal liver tests due to glycogenosis in adults with diabetes. Medicine (Baltimore). 1996;75: 327–333.
Case 17.5
Glycogen Pseudo–Ground-Glass MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON DIAGNO SIS
A 21-year-old man with stage IV Hodgkin’s lymphoma presented with fluctuating liver enzymes over a 7-month period, with AST and ALT levels ranging from 1 to 2 times the upper limit of normal. He had just finished chemotherapy. All viral studies were negative.
Glycogen pseudo–ground-glass inclusions.
DISCUSSIO N R E A SON F OR R E F E R R AL
The hepatocytes look like they contain hepatitis B groundglass cytoplasmic changes, but all viral serological and DNA testing has been negative. Thus, the etiology of the findings was unclear, and the case was sent for consultation.
PAT H OL OG I C F E AT U R E S
The biopsy shows a panlobular distribution of hepatocytes with distinctive, well-circumscribed, gray inclusions surrounded by a rim of normal cytoplasm (Figure 17.5.1). These changes are not accompanied by significant inflammation, fatty change, or fibrosis. No lymphoma is seen. Immunostains for hepatitis B surface antigen are negative.
FIGURE 17. 5. 1 Pseudo–ground-glass change is also associated with the abnormal accumulation of glycogen, but in this disease pattern the glycogen accumulation leads to discrete inclusion-like deposits. These deposits can closely mimic the ground-glass inclusions of chronic hepatitis B infection.
Glycogen pseudo–ground-glass change is a recently described entity that is seen in immunosuppressed individuals on numerous medications (1,2) and is associated with the accumulation of abnormal forms of glycogen (1). There does not appear to be a single drug or class of drugs that can consistently cause this change, but instead it appears that different drugs can lead to the same effect. The common denominator is immunosuppression combined with numerous medications. The differential includes ground glass changes that can be seen in later stages of chronic hepatitis B infection (Figure 17.5.2). Immunostains for hepatitis B surface antigen are helpful in ruling out hepatitis B infection, especially if serological testing has not been performed or is not available (Figure 17.5.3). The differential also includes drug effects such as cyanamide, Lafora bodies, fibrinogen, and uremia (Table 17.5.1). As is evident in Table 17.5.1, ground-glass type changes and their mimics tend to be associated with smooth endoplasmic reticulum proliferation, abnormal glycogen accumulation, or,
F I G U R E 1 7 . 5 . 2 Ground-glass inclusions of chronic hepatitis B are
shown. They can look identical to the inclusions of pseudo–ground glass.
261
262
CHAPTER
17:
G LY C O G E N I C
ABNORMALITIES
ON
LIVER
BIOPSY
TA BL E 1 7 . 5 . 1 Causes of pseudo–ground-glass type inclusions
in hepatocytes Staining Properties
Cyanamide
PAS diastase sensitivea
glycogen granules and dilated smooth endoplasmic reticulum
Fibrinogen
PAS Fibrinogen C3, C4 positive/
granular/fibrillar material within rough endoplasmic reticulum
Glycogen
PAS diastase sensitive
glycogen, occasional organellesb
Type IV glycogen storage disease
PAS partially diastasesensitivec colloidal iron negative
nonmembrane-bound fibrillar and granular material
Lafora
PAS diastase resistant colloidal iron positive
Fibril-like structures and electron-dense clumps
Uremia
PAS
smooth endoplasmic reticulum
FIGURE 17. 5. 3 Immunostain for hepatitis B surface antigen. In
some cases, clinical history and laboratory testing information is limited. HE findings alone are not sufficiently reliable in separating pseudo–ground glass from hepatitis-B-related changes. Immunostains for hepatitis B can be useful in separating hepatitis B ground-glass change from pseudo–ground glass change.
Electron Microscopy Findings
Type of Inclusion
Abbreviations: C3, C4, cytokeratin 3 and 4; PAS, periodic acid–Schiff diastase.
less commonly, protein accumulation. PAS diastase sensitivity can be of some help in this differential but should be used cautiously as the degree of digestion sensitivity or resistance can vary considerably between laboratories. In most cases, the distinctive clinical situations will clarify nature of the material.
References 1. Wisell J, Boitnott J, Haas M, et al. Glycogen pseudoground glass change in hepatocytes. Am J Surg Pathol. 2006;30:1085–1090. 2. Bejarano PA, Garcia MT, Rodriguez MM, Ruiz P, Tzakis AG. Liver glycogen bodies: ground-glass hepatocytes in transplanted patients. Virchows Arch. 2006;449:539–545. 3. Bruguera M, Lamar C, Bernet M, Rodés J. Hepatic disease associated with ground-glass inclusions in hepatocytes after cyanamide therapy. Arch Pathol Lab Med. 1986;110:906–910.
a Cyanamide pseudo–ground glass has been described as both diastase sensitive(3,4) and diastase resistant.(5) Most cases in the literature appear to be diastase sensitive. b Additional details of the organelles was obscured by poor preservation. c Pseudo–ground glass in type IV glycogen storage disease is variably diastase sensitive(6,7)
4. Hashimoto K, Hoshii Y, Takahashi M, et al. Use of a monoclonal antibody against Lafora bodies for the immunocytochemical study of groundglass inclusions in hepatocytes due to cyanamide. Histopathology. 2001; 39:60–65. 5. Zimmerman HF, Ishak KG. Hepatic injury due to drugs and toxins. In: MacSween RNM, Anthony PP, Scheuer PJ, Burt AD, Portmann BC, eds. Pathology of the liver. London, UK: Churchill Livingstone; 1995:563–634. 6. Sahoo S, Blumberg AK, Sengupta E, et al. Type IV glycogen storage disease. Arch Pathol Lab Med. 2002;126:630–631. 7. Vazquez JJ. Ground-glass hepatocytes: light and electron microscopy. Characterization of the different types. Histol Histopathol. 1990;5: 379–386.
Case 17.6
Smooth Endoplasmic Reticulum Proliferation MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
A 58-year-old man with chronic hepatitis C and human immunodeficiency virus (HIV) was biopsied for staging and grading of liver disease. He has no history of diabetes.
basophilic and slightly granular change to their cytoplasm (Figure 17.6.1). Distinctive inclusions with a rim of cytoplasm are not seen. There is marked lipofuscin but only mild patchy portal chronic inflammation and no significant lobular inflammatory activity.
R E A SON F OR R E F E R R AL
This case was referred for consultation as to whether the findings were those of glycogenic hepatopathy. DIAGNO SIS PAT H OL OG I C F E AT U R E S
Smooth endoplasmic reticulum proliferation.
The biopsy specimen shows a diffuse change affecting the hepatocyte cytoplasm. The cells appear swollen with a
DISCUSSIO N
The biopsy specimen shows a striking example of smooth endoplasmic reticulum proliferation. The change can mimic glycogenic hepatopathy at first, but further examination shows a lack of the distinctive cytoplasmic rarification of glycogenic hepatopathy. Instead, the cytoplasm has a diffuse gray stippled appearance that represents proliferation of the endoplasmic reticulum. In my practice, I see this most commonly in individuals with HIV/hepatitis C virus (HCV) coinfection, and it appears most likely to represent a drug effect. Others have also reported similar drug effects with phenytoin and barbiturates (1). The smooth endoplasmic reticulum proliferation in milder cases often does not involve the entire cell cytoplasm and often does not involve all hepatocytes. FIGURE 17. 6. 1 Diffuse smooth endoplasmic reticulin proliferation
in an individual with hepatitis C and human immunodeficiency virus coinfection. These changes can resemble glycogenic hepatopathy to some degree, but the smooth endoplasmic reticulin proliferation gives the hepatocyte cytoplasm a more basophilic and stippled appearance.
References
263
1. Chatila R, West AB. Hepatomegaly and abnormal liver tests due to glycogenosis in adults with diabetes. Medicine (Baltimore). 1996;75: 327–333.
This page intentionally left blank
18 Macrophage Infiltrate RYAN M. GILL, SANJAY KAKAR, AND LINDA D. FERRELL
I N T ROD U C T I ON
Macrophages represent an important component of the acute inflammatory response and within days often represent the predominant cell type at sites of injury. Macrophages are activated by cytokines (eg, interferon-gamma [IFN-]) and other nonimmune stimuli; the activated macrophage is larger, produces more cytokines, and has enhanced phagocytic capacity. In the setting of chronic inflammation, macrophages may accumulate and mediate significant tissue destruction and contribute to subsequent fibrosis. Macrophage activation can also occur in the setting of ischemia reperfusion, for example, following liver transplantation, and may contribute to injury through release of reactive oxygen species and cytokines, such as tumor necrosis factor alpha (1,2). Focal accumulation of activated macrophages, often with an epithelioid appearance, defines granulomatous inflammation (see Chapter 5); others may fuse to form multinucleate giant cells, or they may proliferate in a diffuse fashion, which is the focus of this chapter. In the liver, the macrophage is termed the Kupffer cell and is attached to the luminal surface of sinusoidal endothelial cells. Kupffer cells account for 80% of the mononuclear phagocytic system (3). Although still important in immune modulation, the Kupffer cell is less efficient at antigen presentation than other macrophages (4) and more readily utilizes its phagocytic and digestive abilities to efficiently ingest and remove particulate and soluble material, as well as micro-organisms/ endotoxin and degenerated cells from portal blood (5). Kupffer cells are more numerous in periportal sinusoids (6) and are derived at least in part from circulating monocytes (7); they proliferate in response to hepatic injury (8,9) and demonstrate functional heterogeneity (6,10), with report that periportal Kupffer cells are more phagocytically active (11). Kupffer cells can also migrate to areas of liver injury (12) and rapidly remove apoptotic hepatocytes. Once apoptotic hepatocytes have been phagocytosed, the macrophage (or clump of macrophages) may persist for weeks to months as evidence of previous injury (Figure 18.1). Kupffer cell–derived cytokine networks can even modulate hepatocyte function (13). Kupffer cells are recognized histologically as plump cells with abundant cytoplasm; they are easiest to appreciate when present in aggregates. The presence of phagocytosed periodic acid-Schiff (PAS)–positive, diastase (D)-resistant granular ceroid material allows them to be easily identified with a PASd stain (Figure 18.2). Kupffer cells can also be labeled using immunohistochemical lysosomal markers (CD68 and lysozyme) or the more specific macrophage marker, CD163. Historically, the term histiocyte has been used to denote macrophages in tissue sections. In the liver, a macrophage pattern of injury (in which abundant large macrophages are seen diffusely filling
F I G U R E 1 8 . 1 Persistent macrophages (arrows) after hepatocyte
injury.
F I G U R E 1 8 . 2 PASd-resistant ceroid material in macrophages.
sinusoids) likely represents an inherited enzyme defect (ie, a storage disorder resulting in accumulation of metabolic products), an infectious etiology, or an immunologic etiology. Distinction between these categories can often be made at high power through characterization of the phagocytosed material (Table 18.1). Drugs like amiodarone can lead to accumulation of phospholipids in the hepatocytes and occasionally in Kupffer cells (phospholipidosis) that can mimic storage diseases. In addition, rare neoplastic macrophage tumors or neoplastic mimics of a macrophage infiltrate may be encountered in the liver.
265
266
CHAPTER
18:
MACROPHAGE
TA B LE 18. 1 Classification of macrophage infiltrates by cytoplasmic
contents Metabolic Products Disorders of glycoprotein metabolism Disorders of amino acid metabolism Disorders of lipoprotein and lipid metabolism Disorders of iron metabolism Ceroid/lipofuscin-containing macrophages Phospholipidosis related to drugs like amiodarone Organisms Disseminated histoplasmosis Visceral leishmaniasis Disseminated cryptococcus Hematopoietic Cells Hemophagocytic lymphohistiocytosis Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease)
I N F I LT R AT E
seizures, and cherry-red macular spots to a more severe infantile form with gargoyle-like facies, peripheral neuropathy, impaired coordination, corneal opacities, impaired hearing, and hepatosplenomegaly. Types II and III are autosomal recessive diseases linked to 4q21–23 (14) with defective N-acetylglucosamine 1-phosphotransferase resulting in defective lysosomal enzyme targeting. Type II presents with early growth retardation, gingival hyperplasia, hepatomegaly, skeletal dysplasia, and psychomotor retardation (15). Type III presents in adulthood with less severe but similar features. Definitive diagnosis may require fibroblast culture and measurement of enzyme activity. Kupffer cells are enlarged and contain vacuolated cytoplasm. Hepatocytes, endothelial cells, stellate cells, and biliary epithelial cells demonstrate vacuoles at the ultrastructural level (16). Disorders of Amino Acid Metabolism
TA B LE 18. 2 Storage disorders with primarily Kupffer cell
involvement (by light microscopy) Disorders of Glycoprotein Metabolism Mucolipidosis I, II, & III Disorders of Amino Acid Metabolism Cystinosis Disorders of Lipoprotein and Lipid Metabolism Familial HDL deficiency Wolman disease Sulphatide lipidosis (metachromatic leukodystrophy) Ceramidase deficiency (Farber disease) Glycosyl ceramide lipidosis (Gaucher disease) Sphingomyelin-cholesterol lipidosis (Niemann-Pick disease) Disorders of Iron Metabolism Hemochromatosis, type IV (ferroportin disease) Unclassified Hermansky-Pudlak syndrome
M AC RO PH AG E I N F I LT R AT E — STOR AG E D I SORDER S
Cystinosis is an autosomal recessive disease with mutation in CTNS leading to accumulation of L-cystine crystals in lysosomes. In severe forms, crystal accumulation leads to renal tubular dysfunction and renal failure in childhood. Crystal deposition in the cornea and conjunctiva may lead to photophobia and allows for diagnosis through slit lamp evaluation. Crystals are also classically identified in neutrophils. Although most patients do not present with liver dysfunction, hepatomegaly is not uncommon. Liver biopsy may show perivenular hypertrophied Kupffer cells with intracytoplasmic cystine crystals (Figure 18.3), which demonstrate silver birefringence under polarized light. An association with portal hypertension has been noted (17,18) and a possible fibrogenic role for cystine-laden Kupffer cells has been proposed (19). Disorders of Lipoprotein and Lipid Metabolism
Disorders of lipoprotein and lipid metabolism account for the largest group of storage disorders with predominant macrophage involvement. Familial high-density lipoprotein deficiency (Tangier disease) is an autosomal recessive disease
Inherited storage disorders encountered by pathologists predominantly involve enzyme defects. Further distinction can be made between storage disorders involving hepatocytes as well as Kupffer cells (and occasionally other liver cell types) and those primarily presenting with macrophage involvement (Table 18.2). The latter typically involve defects in lysosomal enzymes important for breakdown of specific metabolic products. Disorders of Glycoprotein Metabolism
Mucolipidoses are rare lysosomal storage diseases with multiple combinatorial defects in the metabolism of mucopolysaccharide, lipid, and glycoprotein. Type I (termed sialidosis) is caused by mutation of NEU1 leading to a deficiency of lysosomal sialidase (neuraminidase). Clinical presentation ranges from late onset of visual defects, myoclonus, ataxia, hyperreflexia,
F I G U R E 1 8 . 3 Hypertrophied Kupffer cells with intracytoplasmic crystals in a case of cystinosis.
CHAPTER
18:
MACROPHAGE
involving mutation of ABCA1 on chromosome 9q31 (20). This leads to cholesterol accumulation in macrophages. Clinical presentation includes orange and enlarged tonsils, lymphadenopathy, hepatosplenomegaly, and peripheral neuropathy. Some patients have been reported to have hyperbilirubinemia (21,22). Liver biopsy shows abundant foamy Kupffer cells with birefringent cholesterol crystals (needle shaped and PAS negative) (23,24). In Wolman disease there is reduced lysosomal acid lipase (25), which results in abundant lipid and cholesterol crystals in macrophages (but also to some extent in hepatocytes) and a clinical presentation of diarrhea, emesis, and failure to thrive, with death typically by one year of life. Foamy macrophages are encountered in liver, spleen, adrenal glands, and intestines (as well as in bone marrow). Stains for lipid (eg, oil red O) on frozen sections may be helpful in identifying this entity. Ultrastructural analysis can also identify the predominant triglyceride droplets and cholesterol crystals. In sulphatide lipidosis (metachromatic leukodystrophy) the defect is in arylsulphatase A-mediated degradation of 3-O-sulphogalactosylceramide, which leads to demyelination and resultant neurologic symptoms. Presentation is variable but in its most severe form may manifest as progressive mental retardation, quadriplegia, and ataxia as well as hyperpyrexia. Liver biopsy may show periportal macrophages with metachromatic granules (26), which can be highlighted with a trichrome stain. Ceramidase deficiency (Farber disease) results in inability to degrade ceramide into sphingosine and fatty acid. Classic features of severe disease include early joint disease, subcutaneous wrist nodules, and a hoarse cry. Type IV ceramidase deficiency has been associated with neonatal hepatosplenomegaly and early death. Liver biopsy has reportedly demonstrated a diffuse expansion of Kupffer cells (27), which may be vacuolated. Glycosyl ceramide lipidosis, more commonly referred to as Gaucher disease, is the most common lysosomal storage disorder and is caused by a deficiency in acid-beta-glucosidase, which, along with Niemann-Pick disease (sphingomyelincholesterol lipidosis involving deficient acid sphingomyelinase), is described in the case studies.
I N F I LT R AT E
267
MACRO P H AGE INFILT R AT E—INFECT IO U S
Several infectious organisms are associated with a diffuse sinusoidal proliferation of macrophages (Table 18.1). Histoplasmosis is most often caused by Histoplasma capsulatum, especially in the Ohio River Valley. As an opportunistic infection, it is usually disseminated in the immune compromised. Hepatic involvement may result in hepatomegaly and small (3–4 μm) yeast can be identified in Kupffer cells. Although small granulomas may form, the disease may also present with diffuse macrophage infiltrate in the sinusoids. Distinction from leishmaniasis is possible since, of the 2, only H. capsulatum will stain with fungal stains, and H. capsulatum also does not have the kinetoplast identified in Leishmania sp. Disseminated Cryptococcus infection caused by Cryptococcus neoformans can be indistinguishable from histoplasmosis in the liver (when the organism is small and engulfed by Kupffer cells) and occurs in the same clinical settings. This is rarely a diagnostic dilemma since Cryptococcus more typically demonstrates variability in size, from 5 to 20 μm. Identification of a mucoid capsule is also characteristic, but if there is any doubt, culture allows for definitive classification. Visceral leishmaniasis (aka kala azar) is caused by several related protozoa (Leishmania sp.), which directly infect Kupffer cells and which can be diagnosed through morphologic identification of amastigotes (Figure 18.4). Visceral leishmaniasis is endemic in the Middle East, Asia, South America, Africa, and parts of Europe. Clinical presentation is with hepatosplenomegaly, fever, lymphadenopathy, and pancytopenia with hypergammaglobulinemia (30). The liver may be massively enlarged and will show Kupffer cell hyperplasia and hypertrophy with characteristic intracellular amastigotes (2–3 μm, ovoid clear organisms with a basophilic nucleus and paranuclear rod-shaped kinetoplast, often best seen on touch preparations). Note that in some cases, organisms may not be as readily identified. Steatosis may be encountered and fibrin ring granulomas have been described (31). In chronic adult cases fibrosis may develop. Before treatment with amphotericin B, visceral leishmaniasis
Disorders of Iron Metabolism
Type IV hemochromatosis, with its mutations in SLC40A1 (also known as the ferroportin-1 gene), which encodes an iron export protein abundant in Kupffer cells may first manifest as iron overload in Kupffer cells before appearing in hepatocytes (see Chapter 20). Unclassified
Hermansky-Pudlak syndrome is an autosomal recessive disorder without a well-defined etiology (28,29), which is characterized by oculocutaneous albinism and a defect in platelet aggregation. PASd-positive ceroid pigment deposition occurs in macrophages and may be appreciated on liver biopsy as tan granular material in hypertrophied Kupffer cells.
F I G U R E 1 8 . 4 Visceral leishmaniasis demonstrating a Kupffer cell
filled with amastigotes.
268
CHAPTER
18:
MACROPHAGE
I N F I LT R AT E
was usually fatal; with modern therapy (which in addition to liposomal amphotericin B includes sodium stibo-gluconate, and miltefosine), the prognosis is much improved in the immune competent. MAC RO PH AG E I N F I LT R AT E — I M M U N OL OGIC
“Immunologic disorders” include a spectrum of abnormalities leading to hemophagocytic lymphohistiocytosis (HLH), a clinicopathologic syndrome with proliferation of benignappearing histiocytes that may show phagocytosis of erythrocytes (hemophagocytosis). Bone marrow samples are more likely to demonstrate this finding than liver biopsy; hepatic involvement is characterized by Kupffer cell proliferation with variable hemophagocytosis (32,33) as well as a portal-based lymphohistiocytic infiltrate with a predominance of T cells. It is typical for Kupffer cells to stain less intensely with PAS in the setting of hemophagocytosis (34), and they frequently demonstrate siderosis (34), possibly secondary to preceding blood transfusions. In addition to morphologic demonstration of hemophagocytosis, HLH commonly presents with fever, splenomegaly, cytopenias (in at least 2 cell lines), hypertriglyceridemia, hypofibrinogenemia, and an elevated ferritin (>500 μg/L), as well as with sCD25 greater than 2400 U/mL and decreased NK-cell activity; 5 of these 8 criteria must be met to diagnose the syndrome (35). Although cytopenias may correlate with disease severity, they have not been found to correlate with hepatic involvement (34). In the setting of a child with acute liver failure and high ferritin, distinction from perinatal hemochromatosis can be critical, as the latter may be successfully treated with liver transplant (36,37). HLH may be inherited (inheritance is autosomal recessive with several described mutations), as in the fatal familial hemophagocytic lymphohistiocytosis (Farquhar disease) syndrome, which presents in the first year of life and involves a defect in granule-mediated cytotoxicity (eg, perforin) or natural killer (NK) cell function leading to uncontrolled macrophage activation. Inherited HLH may also manifest as part of an immune deficiency syndrome (eg, Chediak-Higashi syndrome, Girscelli syndrome, or X-linked lymphoproliferative syndrome). Also, severe combined immune deficiency (SCID) patients may develop “fatal infectious mononucleosis,” a terminal event involving massive hemophagocytosis, which may also involve the liver (Figures 18.5 and 18.6). HLH may be acquired in association with infection (usually Epstein-Barr virus [EBV]), toxins, malignancies (particularly in NK or T cell lymphomas, in which EBV is often still implicated), or rheumatic diseases with macrophage activation syndrome (38). Treatment involves immunosuppressive agents and cytotoxic drugs, but the prognosis is poor. Stem cell transplant is an option for patients with a genetic etiology. Among the reactive histiocytic proliferations, another consideration is sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease), first described by Drs. Rosai and Dorfman (39), which is of unknown etiology but may represent a reactive macrophage proliferation rather than a true neoplastic process. It is most common in
F I G U R E 1 8 . 5 Liver with hemophagocytosis in a patient with “fatal infectious mononucleosis.” (Courtesy of Dr. Elaine S. Jaffe.)
the pediatric population and most patients present with massive painless cervical lymph nodes; other lymph nodes can be involved and a significant subset will have extranodal involvement, including the liver in approximately 3% to 17% of cases (40). Leukocytosis and polyclonal hypergammaglobulinemia are associated with this diagnosis. Spontaneous remission is expected, but this may be protracted over months to years, and liver involvement is associated with a less optimal prognosis (40). When mass effect complicates the patient’s course, surgery may be indicated. Histologically, macrophages have abundant cytoplasm and an intermediate to large round nucleus with vesicular chromatin and 1 to several nucleoli. The slightly eosinophilic cytoplasm often contains phagocytosed lymphocytes or emperipolesis by a variety of cells (eg, neutrophils, red blood cells, or plasma cells). The proliferating histiocytes express CD68, CD163, and S-100 but are negative for CD1a and dendritic markers such as CD21. MACRO P H AGE INFILT R AT E— NEO P LA ST IC A ND MIMICS
Histiocytic sarcoma is a rare malignancy derived from macrophages, which has been described in the liver (41,42). Histologically, malignant histiocytes demonstrate a range of cytologic atypia, including possible multinucleate cells and stain with macrophage markers. Hemophagocytosis may also be present. In addition to CD68 and CD163, CD45, HLA-DR, CD43, and
CHAPTER
18:
MACROPHAGE
FIGURE 18. 6 CD163 immunostain highlights macrophages with
hemophagocytosis in this case of “fatal infectious mononucleosis.” (Courtesy of Dr. Elaine S. Jaffe.)
CD4 immunostains are usually positive and can be helpful in differentiating histiocytic sarcoma from other hematolymphoid neoplasms. S100 and langerin are negative or only weak and focally positive and histiocytic sarcoma does not express specific melanoma markers (allowing distinction from melanoma following staining with HMB-45, Melan-A, or tyrosinase). Clonal immunoglobulin or T-cell receptor gene rearrangements have rarely been reported. Histiocytic sarcoma is an aggressive malignancy, and patients may not respond well to treatment. Malignant tumors with a sinusoidal pattern of spread (such as hepatosplenic T-cell lymphoma or metastatic melanoma) can mimic macrophage infiltration. Morphologic observation of cytologic features should be helpful in ruling out metastatic tumor, but immunohistochemical stains can readily resolve indeterminate cases. Note that melanoma may express CD68 (and activated or neoplastic macrophages may label with S100, as described above) but should lack expression of the more specific histiocytic marker, CD163.
References 1. Bilzer M, Gerbes AL. Preservation injury of the liver: mechanisms and novel therapeutic strategies. J Hepatol. 2000;32:508–515. 2. Le Moine O, Louis H, Demols A, et al. Cold liver ischemia-reperfusion injury critically depends on liver T cells and is improved by donor pretreatment with interleukin 10 in mice. Hepatology. 2000;31:1266–1274.
I N F I LT R AT E
269
3. Saba TM. Physiology and physiopathology of the reticuloendothelial system. Arch Intern Med. 1970;126:1031–1052. 4. Rogoff TM, Lipsky PE. Role of the Kupffer cells in local and systemic immune responses. Gastroenterology. 1981;80:854–860. 5. Rifai A, Mannik M. Clearance of circulating IgA immune complexes is mediated by a specific receptor on Kupffer cells in mice. J Exp Med. 1984;160:125–137. 6. Wake K, Decker K, Kirn A, et al. Cell biology and kinetics of Kupffer cells in the liver. Int Rev Cytol. 1989;118:173–229. 7. Gale RP, Sparkes RS, Golde DW. Bone marrow origin of hepatic macrophages (Kupffer cells) in humans. Science. 1978;201:937–938. 8. Bouwens L, Baekeland M, Wisse E. Cytokinetic analysis of the expanding Kupffer-cell population in rat liver. Cell Tissue Kinet. 1986;19:217–226. 9. Johnson SJ, Hines JE, Burt AD. Macrophage and perisinusoidal cell kinetics in acute liver injury. J Pathol. 1992;166:351–358. 10. te Koppele JM, Thurman RG. Phagocytosis by Kupffer cells predominates in pericentral regions of the liver lobule. Am J Physiol. 1990;259: G814–G821. 11. Romert P, Quistorff B, Bhenke O. Histological evaluation of the zonation of colloidal gold uptake by the rat liver. Tissue Cell. 1993;25:19–32. 12. MacPhee PJ, Schmidt EE, Groom AC. Evidence for Kupffer cell migration along liver sinusoids, from high-resolution in vivo microscopy. Am J Physiol. 1992;263:G17-G23. 13. Maher JJ, Friedman SL. Parenchymal and nonparenchymal cell interactions in the liver. Semin Liver Dis. 1993;13:13–20. 14. Watanabe S, Phillips MJ. Ca2+ causes active contraction of bile canaliculi: direct evidence from microinjection studies. Proc Natl Acad Sci U S A. 1984;81:6164–6168. 15. Gumucio DL, Gumucio JJ, Wilson JA, et al. Albumin influences sulfobromophthalein transport by hepatocytes of each acinar zone. Am J Physiol. 1984;246:G86-G95. 16. Rappaport AM, Borowy ZJ, Lougheed WM, Lotto WN. Subdivision of hexagonal liver lobules into a structural and functional unit: role in hepatic physiology and pathology. Anat Rec. 1954;119:11–33. 17. DiDomenico P, Berry G, Bass D, Fridge J, Sarwal M. Noncirrhotic portal hypertension in association with juvenile nephropathic cystinosis: case presentation and review of the literature. J Inherit Metab Dis. 2004;27:693–699. 18. Rossi S, Herrine SK, Navarro VJ. Cystinosis as a cause of noncirrhotic portal hypertension. Dig Dis Sci. 2005;50:1372–1375. 19. Klenn PJ, Rubin R. Hepatic fibrosis associated with hereditary cystinosis: a novel form of noncirrhotic portal hypertension. Mod Pathol. 1994;7:879–882. 20. Rust S, Walter M, Funke H, et al. Assignment of Tangier disease to chromosome 9q31 by a graphical linkage exclusion strategy. Nat Genet. 1998;20:96–98. 21. Brook JG, Lees RS, Yules JH, Cusack B. Tangier disease (alphalipoprotein deficiency). JAMA. 1977;238:332–334. 22. Ferrans VJ, Fredrickson DS. The pathology of Tangier disease. A light and electron microscopic study. Am J Pathol. 1975;78:101–158. 23. Bale PM, Clifton-Bligh P, Benjamin BN, Whyte HM. Pathology of Tangier disease. J Clin Pathol. 1971;24:609–616. 24. Dechelotte P, Kantelip B, de Laguillaumie BV, Labbe A, Meyer M. Tangier disease. A histological and ultrastructural study. Pathol Res Pract. 1985;180:424–430. 25. Pagani F, Pariyarath R, Garcia R, et al. New lysosomal acid lipase gene mutants explain the phenotype of Wolman disease and cholesteryl ester storage disease. J Lipid Res. 1998;39:1382–1388. 26. Wolfe HJ, Pietra GG. The visceral lesions of metachromatic leukodystrophy. Am J Pathol. 1964;44:921–930. 27. Antonarakis SE, Valle D, Moser HW, Moser A, Qualman SJ, Zinkham WH. Phenotypic variability in siblings with Farber disease. J Pediatr. 1984;104:406–409. 28. Oh J, Ho L, Ala-Mello S, et al. Mutation analysis of patients with Hermansky-Pudlak syndrome: a frameshift hot spot in the HPS gene and apparent locus heterogeneity. Am J Hum Genet. 1998;62:593–598.
270
CHAPTER
18:
MACROPHAGE
29. Schallreuter KU, Frenk E, Wolfe LS, Witkop CJ, Wood JM. HermanskyPudlak syndrome in a Swiss population. Dermatology. 1993;187:248–256. 30. Badaro R, Jones TC, Carvalho EM, Sampaio D, Reed SG, Barral A, Teixeira R, et al. New perspectives on a subclinical form of visceral leishmaniasis. J Infect Dis. 1986;154:1003–1011. 31. Moreno A, Marazuela M, Yebra M, et al. Hepatic fibrin-ring granulomas in visceral leishmaniasis. Gastroenterology. 1988;95:1123–1126. 32. Hsu TS, Komp DM. Clinical features of familial histiocytosis. Am J Pediatr Hematol Oncol. 1981;3:61–65. 33. Favara BE. Histopathology of the liver in histiocytosis syndromes. Pediatr Pathol Lab Med. 1996;16:413–433. 34. Tsui WM, Wong KF, Tse CC. Liver changes in reactive haemophagocytic syndrome. Liver. 1992;12:363–367. 35. Janka GE, Schneider EM. Modern management of children with haemophagocytic lymphohistiocytosis. Br J Haematol. 2004;124:4–14. 36. Parizhskaya M, Reyes J, Jaffe R. Hemophagocytic syndrome presenting as acute hepatic failure in 2 infants: clinical overlap with neonatal hemochromatosis. Pediatr Dev Pathol. 1999;2:360–366.
I N F I LT R AT E
37. Natsheh SE, Roberts EA, Ngan B, Chait P, Ng VL. Liver failure with marked hyperferritinemia: “ironing out” the diagnosis. Can J Gastroenterol. 2001;15:537–540. 38. Sawhney S, Woo P, Murray KJ. Macrophage activation syndrome: a potentially fatal complication of rheumatic disorders. Arch Dis Child. 2001;85:421–426. 39. Rosai J, Dorfman RF. Sinus histiocytosis with massive lymphadenopathy. A newly recognized benign clinicopathological entity. Arch Pathol. 1969;87:63–70. 40. Foucar E, Rosai J, Dorfman R. Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): review of the entity. Semin Diagn Pathol. 1990;7:19–73. 41. Akishima Y, Akasaka Y, Yih-Chang G, et al. Histiocytic sarcoma with fatal duodenal ulcers. Pathol Res Pract. 2004;200:473–478. 42. Hornick JL, Jaffe ES, Fletcher CD. Extranodal histiocytic sarcoma: clinicopathologic analysis of 14 cases of a rare epithelioid malignancy. Am J Surg Pathol. 2004;28:1133–1144.
Case 18.1
Gaucher Disease RYAN M. GILL, SANJAY KAKAR, AND LINDA D. FERRELL
C L I N IC AL I N F OR M AT I ON
DISCUSSIO N
A 17-year-old male presented to his primary care physician with a history of prolonged epistaxis. Physical examination documented moderate to severe hepatosplenomegaly, and laboratory values were significant for mild thrombocytopenia (50 10ˆ12/L). The patient was of Ashkenazi Jewish descent. A bone marrow biopsy showed mild megakaryocytic hyperplasia and was followed by a liver biopsy. R E A SON F OR R E F E R R AL
Macrophage infiltrate, suspicious for a storage disorder. PAT H OL OG I C F E AT U R E S
Liver biopsy demonstrated perivenular hypertrophy and hyperplasia of Kupffer cells, including some very large forms with lightly eosinophilic “crinkled” cytoplasm and central or eccentric nuclei (Figure 18.1.1). A PAS stain highlighted macrophage “striations” (Figure 18.1.2), typical of Gaucher cells. Hepatic plates were focally disrupted, but the architecture was otherwise intact and there was no fibrosis.
D I AG N OS I S
Perivenular Gaucher cells, no evidence of fibrosis.
FIGURE 18. 1. 1 Kupffer cells with a crinkled fibrillated appearance
in Gaucher disease.
Gaucher disease is an autosomal recessive disease in which patients are deficient in acid beta-glucosidase (glucocerebrosidase), which results in accumulation of glucocerebroside in macrophages (1,2). The most common form is type I, which is prevalent in the Ashkenazi Jewish population and which has variable presentation at any age. Patients with type I Gaucher disease typically have hepatosplenomegaly and skeletal disease (metaphyseal aseptic necrosis, bone pain, spontaneous fractures, degenerative hip disease, and possibly an “Erlenmeyer flask” deformity) but lack neurodegeneration. Type II disease presents in infancy with predominant neurodegeneration and moderate hepatosplenomegaly as well as possible ichthyosis (3). Type II disease does not have an ethnic predilection, and death usually occurs before 3 years of age. Type III disease is similar to type I disease, except that it is usually detected before age 50 and has associated terminal neurodegeneration. This form is more common in patients of Swedish ancestry. Hepatosplenomegaly can lead to platelet sequestration and mild thrombocytopenia (4). Diagnosis may be confirmed through assay of acid beta-glucosidase enzyme in white blood cells or fibroblasts. Recombinant enzyme replacement therapy is available, which results in rapid reversal of hepatomegaly (5–8) but does not mitigate neurologic disease. Morphologic findings in the liver include focal or perivenular macrophage hyperplasia and hypertrophy with characteristic striated, crinkled, or fibrillary cytoplasm, which can be highlighted with a PAS or trichrome stain. Hepatocyte atrophy, fibrosis, and portal hypertension may occur, and rare
FIGURE 18.1.2 Gaucher cells in a liver biopsy from a patient with Gaucher disease demonstrating characteristic PAS-positive striations.
271
272
CHAPTER
18:
MACROPHAGE
cases of cirrhosis have been described (9,10). Ultrastructural analysis demonstrates Kupffer cell cytoplasm packed with tubules. Macrophages resembling Gaucher cells have been described in several hematological disorders like chronic myelogenous leukemia, myelodysplastic syndrome, acute lymphoblastic leukemia, and plasma cell neoplasms. These have been referred to as pseudo-Gaucher cells and are morphologically and immunohistochemically indistinguishable from Gaucher cells. These cells usually occur in the bone marrow and rarely in the liver. On electron microscopy, pseudo-Gaucher cells have heterogenous inclusions and dense fibrillary bodies compared with the tubular inclusions of Gaucher cells (11,12).
References 1. Brady RO, Kanfer JN, Bradley RM, Shapiro D. Demonstration of a deficiency of glucocerebroside-cleaving enzyme in Gaucher’s disease. J Clin Invest. 1966;45:1112–1115. 2. Cox TM, Schofield JP. Gaucher’s disease: clinical features and natural history. Baillieres Clin Haematol. 1997;10:657–689. 3. Sidransky E, Ginns EI. Clinical heterogeneity among patients with Gaucher’s disease. JAMA. 1993;269:1154–1157.
I N F I LT R AT E
4. Zimran A, Kay A, Gelbart T, Garver P, Thurston D, Saven A, Beutler E. Gaucher disease. Clinical, laboratory, radiologic, and genetic features of 53 patients. Medicine (Baltimore). 1992;71:337–353. 5. Barton NW, Brady RO, Dambrosia JM, et al. Replacement therapy for inherited enzyme deficiency-Macrophage-targeted glucocerebrosidase for Gaucher’s disease. N Engl J Med. 1991;324:1464–1470. 6. Grabowski GA, Barton NW, Pastores G, et al. Enzyme therapy in type 1 Gaucher disease: comparative efficacy of mannose-terminated glucocerebrosidase from natural and recombinant sources. Ann Intern Med. 1995;122:33–39. 7. Germain DP. Gaucher’s disease: a paradigm for interventional genetics. Clin Genet. 2004;65:77–86. 8. Grabowski GA, Leslie N, Wenstrup R. Enzyme therapy for Gaucher disease: the first 5 years. Blood Rev. 1998;12:115–133. 9. Lachmann RH, Wight DG, Lomas DJ, Fisher NC, Schofield JP, Elias E, Cox TM. Massive hepatic fibrosis in Gaucher’s disease: clinicopathological and radiological features. QJM. 2000;93:237–244. 10. Cajaiba MM, Reyes-Múgica M. Gaucher or pseudo-Gaucher? The challenge of several diseases colliding in a pediatric patient. Hum Pathol. 2009;40:594–598. 11. Zidar BL, Hartsock RJ, Lee RE, et al. Pseudo-Gaucher cells in the bone marrow of a patient with Hodgkin’s disease. Am J Clin Pathol. 1987;87:533–536. 12. James SP, Stromeyer FW, Chang C, Barranger JA. Liver abnormalities in patients with Gaucher’s disease. Gastroenterology. 1981;80:126–133.
Case 18.2
Niemann-Pick Disease RYAN M. GILL, SANJAY KAKAR, AND LINDA D. FERRELL
C L I N IC AL I N F OR M AT I ON
A 25-year-old woman was noted to have hepatosplenomegaly and elevated total bilirubin (8 mg/dL). Abdominal computed tomography (CT) scan suggested the possibility of cirrhosis and a liver biopsy was performed. R E A SON F OR R E F E R R AL
Cryptogenic cirrhosis. PAT H OL OG I C F E AT U R E S
Liver biopsy shows focal or diffuse aggregates of hypertrophic vacuolated foamy periportal and sinusoidal macrophages (Figure 18.2.1). In some areas, the hepatocytes also show vacuolization and are pale, making distinction from macrophages difficult. On PAS stain the foamy cells are pale staining and lack striations (Figure 18.2.2). A Gomori trichrome stain demonstrates cirrhosis (Figure 18.2.3). Ultrastructural analysis demonstrated macrophages with densely packed laminated cytoplasmic inclusions (Figure 18.2.4).
F I G U R E 1 8 . 2 . 2 Niemann-Pick disease with PAS-negative foamy
macrophages.
D I AG N OS I S
Niemann-Pick disease with cirrhosis. D I S C U S S I ON
Niemann-Pick disease is a group of autosomal recessive sphingomyelin-cholesterol lipidoses that are separated into
F I G U R E 1 8 . 2 . 3 Niemann-Pick disease with cirrhosis on Gomori
trichrome stain.
FIGURE 18.2.1 Niemann-Pick disease with sinusoidal and periportal foamy macrophages.
types A, B, C, and most recently D (1). Types A and B are caused by mutation in SMPD1, which leads to sphingomyelin accumulation in histiocytes. Type A presents shortly after birth with hepatosplenomegaly, lymphadenopathy, and possibly hydrops fetalis (2–4). Macular cherry- red spots may be identified on physical examination. This is followed by neurologic impairment during infancy and leads to seizure and death by 5 years of age. Type B may be diagnosed in infancy, childhood, or adulthood. Although type B is characterized by hepatosplenomegaly (and patients may develop cirrhosis), pulmonary function typically
273
274
CHAPTER
18:
MACROPHAGE
I N F I LT R AT E
On biopsy, the presence of foamy vacuolated macrophages in the liver, bone marrow, or other sites is often an early clue to the diagnosis, which may be assisted by functional and molecular studies. Ultrastructurally, laminated cytoplasmic inclusions ranging from 1 to 5 micrometers in diameter (7) are characteristic. Progressive liver fibrosis may develop in Niemann-Pick disease, and can progress to cirrhosis in both children and adults (8,9). Therapeutic options are limited at present.
References
FIGURE 18. 2. 4 Transmission electron micrograph of Kupffer cell
with characteristic laminated inclusions in Niemann-Pick disease.
deteriorates more prominently than hepatic function and can be terminal. Splenic rupture is also a significant cause of death in type B patients. Types C and D (“Nova Scotia type”) are distinct from A and B in that they are caused by a defect that leads to accumulation of cholesterol and lipid in lysosomes and involve different candidate genes. They may present at any age, and patients develop neurologic complications at a slower rate; there is also often a history of neonatal cholestasis, and liver failure can develop in a minority of patients (5). There are also rare reports of associated hepatocellular carcinoma (6).
1. James SP, Stromeyer FW, Chang C, Barranger JA. Liver abnormalities in patients with Gaucher’s disease. Gastroenterology. 1981;80:126–133. 2. Greer WL, Riddell DC, Gillan TL, et al. The Nova Scotia (type D) form of Niemann-Pick disease is caused by a G3097-->T transversion in NPC1. Am J Hum Genet. 1998;63:52–54. 3. Crocker AC, Farber S. Niemann-Pick disease: a review of eighteen patients. Medicine (Baltimore). 1958;37:1–95. 4. Meizner I, Levy A, Carmi R, Robinsin C. Niemann-Pick disease associated with nonimmune hydrops fetalis. Am J Obstet Gynecol. 1990;163:128–129. 5. Kelly DA, Portmann B, Mowat AP, Sherlock S, Lake BD. Niemann-Pick disease type C: diagnosis and outcome in children, with particular reference to liver disease. J Pediatr. 1993;123:242–247. 6. Gartner JC Jr, Bergman I, Malatack JJ, et al. Progression of neurovisceral storage disease with supranuclear ophthalmoplegia following orthotopic liver transplantation. Pediatrics. 1986;77:104–106. 7. Lynn R, Terry RD. Lipid histochemistry and electron microscopy in adult Niemann-Pick disease. Am J Med. 1964;37:987–994. 8. Takahashi T, Akiyama K, Tomihara M, et al. Heterogeneity of liver disorder in type B Niemann-Pick disease. Hum Pathol. 1997;28:385–388. 9. Tassoni JP Jr, Fawaz KA, Johnston DE. Cirrhosis and portal hypertension in a patient with adult Niemann-Pick disease. Gastroenterology. 1991;100:567–569.274
19 Approach to Liver Biopsy With Minimal or Nonspecific Histologic Findings DHANPAT JAIN AND SANJAY KAKAR
It is often frustrating in clinical practice to get a liver biopsy with minimal or nonspecific changes in the setting of liver enzyme abnormalities. Most commonly it occurs in the setting of a milder form of an already known disease, and this seldom causes clinical problems. An example of this is a liver biopsy from a patient with chronic hepatitis C infection showing no fibrosis (stage 0) and minimal inflammatory activity (grade 0–1). However, there are certain disorders that typically present with subtle or nonspecific histopathologic changes that can be easily overlooked. A list of such disorders seen commonly in practice and their corresponding clinical patterns are listed in Table 19.1. These are briefly discussed here; details can be found in the respective chapters. Despite an extensive search in some cases, the etiology of the liver enzyme abnormalities still remains unexplained. Sampling error, exposure to an unidentified toxin/drug, and hepatic manifestations of an unidentified or subclinical systemic disorder are likely causes in such situations. It cannot be overemphasized that the spectrum of drug-induced liver disease is wide, and it is not unfair to say that for every liver disorder there is a drug toxicity that can mimic it. Hence, for any liver injury of unclear etiology, drug-induced liver injury should always be carefully evaluated and excluded. H E PAT I T I C D I SE ASE S Resolving Hepatitis
If the biopsy is performed late in the course of acute hepatitis, the clinical and biochemical picture can resemble acute hepatitis with mild elevations of transaminases, but the biopsy shows mild changes. These include no or mild portal/ acinar inflammation with or without mild hepatocellular TA B LE 19. 1 Differential diagnosis in liver biopsy
with mild or nonspecific changes Hepatitic Pattern Resolving hepatitis Nonspecific reactive hepatitis Viral hepatitis (false-negative serology) Connective tissue diseases Celiac disease Biliary Diseases Primary biliary cirrhosis Primary sclerosing cholangitis Mast cell disorders Vanishing bile duct syndrome
Metabolic Disorders Glycogenic hepatopathy Hypervitaminosis A Wilson disease Alpha-1-antitrypsin deficiency Portal Hypertension Idiopathic portal hypertension Venous outflow obstruction Nodular regenerative hyperplasia Other Amyloidosis
damage. The most striking finding is the presence of macrophages in the sinusoids. Cytoplasmic pigment (lipofuscin) is often present and can be highlighted with (periodic acidSchiff diastase) PASd stain. This picture reflects the resolving phase of acute hepatitis. Adverse drug reaction is generally the most common cause. Viral Hepatitis
In most cases of hepatitis B and C, the diagnosis is established by serology, and the biopsy is performed for grading and staging. In some instances, especially with hepatitis C, serological tests can give false-negative results. Polymerase chain reaction (PCR) for hepatitis C should always be considered before the biopsy findings are regarded as nonspecific. Nonspecific Reactive Hepatitis
This represents changes in the liver in response to systemic inflammatory processes, febrile illnesses, or inflammation somewhere in the splanchnic bed. Mild elevation of alanine transaminase (ALT) and aspartate transaminase (AST) can be present. On biopsy, the changes tend to be mild and focal. Mild lymphocytic infiltrate may be seen in the portal tracts; plasma cells and eosinophils may rarely be present. Lymphoid aggregates may be present in older patients. Interface hepatitis is minimal, if present. Few foci of lymphocytic inflammation and focal necrosis can be seen in the hepatic parenchyma. Small macrophage collections can occur in the sinusoids. Nonspecific reactive hepatitis needs to be distinguished both from resolving hepatitis and hepatitis C as described above. Collagen Vascular Diseases
Liver involvement can also occur in systemic lupus erythematosus (SLE), rheumatoid arthritis, mixed connective tissue disease, and Sjogren syndrome. The presence of autoantibodies and derangement of liver enzymes raises the possibility of autoimmune hepatitis (AIH). However, the inflammation and hepatocellular damage in these diseases is mild. Apart from mild nonspecific inflammation, other histological changes that can be seen are steatosis, sinusoidal dilatation, and nodular regenerative hyperplasia. Celiac Disease
The most frequent manifestation of liver involvement by celiac disease is mild elevations of transaminases. This can be observed in 40% to 50% of untreated patients. Histologically, 275
276
CHAPTER
19:
LIVER
BIOPSY
WITH
these patients show nonspecific reactive hepatitis and revert to normal within 6 to 12 months of gluten-free diet. Other histological manifestations include acute hepatitis, chronic hepatitis, nodular regenerative hyperplasia, and, rarely, cirrhosis. In addition, coexisting celiac disease is seen in 3% to 6% of AIH, primary biliary cirrhosis (PBC), and PSC. The relationship of celiac disease and autoimmune liver disease is not clear. Autoimmune liver dysfunction often does not respond to gluten-free diet. Serological tests for celiac disease should be done in all cases of unexplained liver dysfunction. B I L I A RY D I SE ASE S Primary Biliary Cirrhosis
The majority of the patients (50%–60%) are asymptomatic at diagnosis and the biopsy is often performed due to incidental hepatomegaly, elevated alkaline phosphatase, or antimitochondrial antibodies (AMA). The distribution of bile duct injury is heterogeneous early in the course of the disease, and hence typical findings may not be seen on biopsy. The morphological findings include nonspecific portal inflammation, minimal bile duct injury, or even no pathological changes. The majority of asymptomatic patients with normal alkaline phosphatase and AMA titer 1:40 or higher, become symptomatic during follow-up, even if the biopsy does not show diagnostic changes. Hence in the setting of AMA positivity, the possibility of PBC should always be raised even if the biopsy findings are normal or nonspecific. Primary Sclerosing Cholangitis
Like PBC, histological changes can be patchy in early disease and may not be represented in the biopsy. If histological changes are nonspecific, PSC cannot be excluded on histologic grounds. Visualization of the biliary tree by cholangiography is the gold standard for diagnosis of PSC. Mast Cell Disease
Liver involvement manifests as portal infiltrates, often with mild bile duct damage and can mimic biliary disease. High index of suspicion is necessary to obtain immunohistochemical stains like tryptase, CD25, and/or CD117 to conclusively identify mast cells.
MINIMAL
HISTOLOGIC
CHANGES
and drugs are common underlying causes; however, the etiology may not be obvious in some cases. The clinical course is unpredictable with spontaneous resolution in some cases and progression to end-stage liver disease requiring transplantation in others. META BO LIC DISEA SES
The histologic findings in several metabolic disorders can be subtle and overlooked on hematoxylin and eosin (HE) sections. Glycogenic hepatopathy is characterized by diffusely swollen hepatocytes due to excess accumulation of glycogen. This commonly occurs in poorly controlled type 1 diabetes. There is no significant inflammation, steatosis, or hepatocellular damage. Hepatic stellate cells (Ito cells) are modified fibroblasts that store lipids and vitamin A in the normal liver. They are located in the space of Disse between the sinusoidal endothelium and the hepatocytes but are generally not easily visible. In certain conditions, especially hypervitaminosis A, excessive lipid gets stored in the stellate cells (stellate cell lipidosis). These lipid-laden cells can easily be mistaken for hepatocytes with steatosis. Their characteristic morphology and location along the sinusoids between the hepatic plates distinguishes them from steatotic hepatocytes. Wilson disease should always be considered in young patients (50 years) who present with unexplained liver dysfunction. Histological findings can include acute hepatitis, chronic hepatitis, steatosis, and cirrhosis. Nonspecific changes like mild inflammation and mild steatosis can be the only histological changes. The diagnosis can be established by serum ceruloplasmin levels and/or copper estimation from the paraffin block. Alpha-1-antitrypsin (AAT) deficiency is characterized by cytoplasmic globules that are typically seen in periportal or periseptal hepatocytes. In homozygous disease, the globules are large and numerous. However, in heterozygous disease, the globules can be small and inconspicuous. Hence PASd stain should be done in all cases with unexplained liver dysfunction. Immunohistochemistry may reveal globules in some cases that were not detected on PASd stain. The presence of globules is not specific for AAT deficiency and can occur with acute inflammation and alcoholic liver disease. Serum levels of AAT are not reliable for ruling out either homozygous or heterozygous disease; the diagnosis is confirmed by typing using isoelectric focusing or molecular techniques like single-strand conformational polymorphism (SSCP) and DNA sequencing.
Vanishing Bile Duct Syndrome and Idiopathic Adult Ductopenia
P O RTA L H Y P ERT ENSIO N
Loss of bile ducts can be patchy and may not involve all portal areas equally. Bile duct damage, ductular reaction, inflammation, or hepatocellular damage may not be seen, and the loss of bile ducts can be overlooked. Cytokeratin (CK) 7 and cytokeratin 19 stains may be helpful to highlight the missing ducts. CK7 stain, in addition, may show strong staining of the periportal hepatocytes. Biliary disorders like PBC or PSC
Some diseases can lead to portal hypertension without significant histologic changes in the liver, especially on needle biopsies. These include idiopathic portal hypertension (also called noncirrhotic portal fibrosis and hepatoportal sclerosis), nodular regenerative hyperplasia, and portal vein thrombosis. Venous outflow obstruction manifests as sinusoidal dilatation and congestion that can be overlooked on the biopsy.
CHAPTER
19:
LIVER
BIOPSY
WITH
OT H E R
Amyloid deposits in the liver can be seen in vessel walls in the portal tracts or in the sinusoids. Some cases show nodular deposits in the sinusoids. Amyloidosis can be missed if the deposits are sparse and not accompanied by liver parenchymal changes like atrophy. The pale blue homogenous appearance of amyloid deposits on the trichrome stain can serve as a useful trigger to obtain specific stains like Congo red.
MINIMAL
HISTOLOGIC
CHANGES
277
Diabetes-associated hepatosclerosis is related to microangiopathy in long-standing diabetes. The hepatic arterioles in the portal tracts show mild thickening and hyalinization similar to diabetic hyaline arteriosclerosis in kidneys and other organs. In addition, they may also show variable amount of perisinusoidal, perivenular, or portal fibrosis. The bile ducts appear histologically normal, although alkaline phosphatase can be elevated.
Case 19.1
Mild Hepatic Steatosis Versus Ito Cell Lipidosis KATHARINE VAN PATTEN, SANJAY KAKAR, AND DHANPAT JAIN
C L I N I C AL I N F OR M AT I ON
A 62-year-old asymptomatic woman was found to have mildly elevated liver transaminases (AST 49 U/L, ALT 60 U/L). Viral hepatitis and autoimmune serologic studies were negative. The physical examination was unremarkable and the patient had no contributory past medical history. R E A S ON F OR R E F E R R A L
Unclear reasons for elevated transaminases in light of negative viral and autoimmune serologies and near normal appearance of the liver biopsy. PAT H OL OG I C F E AT U R E S
The liver biopsy appeared normal on low magnification without any inflammatory changes or fibrosis (Figure 19.1.1). On closer examination, the only significant finding noted was the presence of multivacuolated cells with peripheral crescentshaped nuclei within the hepatic lobules (Figure 19.1.2). The vacuoles appeared to indent the nuclei in many of the cells producing a scalloped nuclear outline (Figure 19.1.2). These cells lay outside the hepatic cords as highlighted by the reticulin stain (Figure 19.1.3). Electron microscopy (EM) showed that the cells are located in the perisinusoidal space of Disse. The ultrastructure of the lipid vacuoles, nuclear scalloping, and the location of the cells were consistent with hepatic
F I G U R E 1 9 . 1 . 2 Higher magnification showing typical morphology
of lipidotic Ito cells with multiple lipid vacuoles in the cells with scalloped nuclei located adjacent to the hepatic cords.
F I G U R E 1 9 . 1 . 3 Reticulin stain shows empty spaces corresponding
to the lipidotic Ito cells.
Ito (stellate) cells (Figures 19.1.3 and 19.1.4). On further inquiry, the patient admitted to having used excessive multivitamin pills daily. The serum vitamin A levels were not available.
FIGURE 19. 1. 1 (A) Low magnification of the liver biopsy showing
near normal histology. (B) Higher magnification showing few vacuolated cells that simulate steatotic hepatocytes.
278
DIAGNO SIS
Ito (stellate) cell lipidosis.
CASE
19.1:
MILD
H E PAT I C
S T E AT O S I S
Electronmicrograph showing multivacuolated Ito cells in the space of Disse.
FIGURE 19. 1. 4
D I S C U S S I ON
Hepatic Ito cells are also known as stellate cells, perisinusoidal cells, or fat storing cells. They are located in the space of Disse between the hepatocytes and the sinusoidal endothelium. Under normal conditions, the cells store the majority of the body’s vitamin A in small lipid droplets in their cytoplasm (1). Ito cells in their quiescent state are difficult to recognize on HE-stained sections of liver. In chronic liver disease, the Ito cells can lose their fat droplets and transform into an activated myofibroblastic state (2). These cells play an important role in fibrogenesis, in part by producing extracellular matrix. Ito cells express a variety of markers including alpha–smooth muscle actin when activated as myofibroblasts but not in the quiescent or lipidotic state (1). In the setting of excess vitamin A ingestion, the Ito cells’ lipid droplets increase dramatically in size, and the cells become easily visible in histologic sections as fat-laden cells, and the condition is termed “Ito cell lipidosis” (ICL) (1). Most commonly, Ito cells in this setting are mistaken for steatotic hepatocytes. They are relatively easy to identify in nonsteatotic liver, and it becomes increasingly difficult to recognize them in a steatotic liver. The Ito cells typically are smaller than hepatocytes, have wispy cytoplasmic strands separating the lipid vacuoles, and the lipid vacuoles indent a peripherally placed nucleus (Figure 19.1.2). The Ito cells may be differentiated from hepatocytes by their location outside the hepatic cords as highlighted by reticulin stain (Figure 19.1.3), smaller size, and most importantly their scalloped nuclear outline similar to a lipoblast. The perisinusoidal location in the space of Disse and morphologic details of Ito cells are best appreciated on EM (Figure 19.1.4). However, EM is not necessary to make a
VERSUS
ITO
CELL
LIPIDOSIS
279
diagnosis of ICL in routine practice. One of the biggest limitations in this area of diagnostic pathology is the lack of a reliable immunohistochemical marker for these lipidotic cells applicable in paraffin sections. The presence of vitamin A in these cells by fluorescence can be demonstrated on frozen tissue or biochemical quantitation on wet liver tissue but is seldom possible in clinical practice. Ito cell lipidosis may also result from conditions other than excess vitamin A intake (Table 19.1.1). ICL has been reported in patients with cholestasis, chronic pancreatitis, diabetes, human immunodeficiency virus (HIV) infection, PBC, early alcoholic disease, and certain drugs (3–6). Fish oil and vitamin E consumption are also believed to enhance the risk for vitamin A toxicity. It is controversial whether topical retinoids can also cause ICL (7,8). In a study by Levin et al of 14 cases of ICL, 5 patients (35.7%) had a history of oral vitamin A intake, while 2 patients (14.2%) were using only topical retinoids (1). Rare congenital disorders of vitamin A metabolism can result in vitamin A toxicity, even with normal vitamin A intake (9). It has also been suggested that decreased synthesis of retinol-binding protein by the liver or defective transport of vitamin A from Ito cells to the hepatocytes may also result in ICL (5,10). In general, patients with ICL do not show features of systemic vitamin A toxicity, which requires a much higher level of vitamin A excess. It is important to recognize that the serum vitamin A levels may be normal in ICL and cannot be used to exclude this diagnosis. In the study by Levine et al, 2 of the 7 tested patients showed normal serum vitamin A levels (1). The most common liver enzyme abnormality in these patients is elevation of alkaline phosphatase, although in some cases the liver enzymes are within normal limits. Regardless of the cause, the distribution of these cells has been shown to be focal, zonal, or sometimes diffuse (1). Significant ICL has been reported to occur in about 1.1% of all nontransplant liver biopsies; however, we believe this to be an underestimate. In our experience, finding a few scattered lipidotic Ito cells in liver biopsies with other disorders is not uncommon. The clinical significance of this finding has not been systematically studied. Lipidotic Ito cells are most likely
TA BL E 1 9 . 1 . 1 Sources of excess vitamin A and other etiologies
of Ito cell lipidosis Sources of excess vitamin A
Vitamin supplements Fish oil Fresh liver Topical retinoids (controversial)
Other diseases associated with Ito cell lipidosis
Cholestasis Chronic pancreatitis Diabetes AIDS Alcoholic liver disease Primary biliary cirrhosis Drugs (methotrexate, steroids, valproic acid)
280
CHAPTER
19:
LIVER
BIOPSY
WITH
MINIMAL
HISTOLOGIC
CHANGES
would be appropriate if clinical and laboratory abnormalities persist despite stopping vitamin A. In conclusion, underrecognition of ICL is common in routine practice and is largely due to a lack of familiarity with the entity.
References
FIGURE 19. 1. 5 A case of chronic hepatitis C with Ito cell lipidosis
where these cells can be easily confused with steatotic hepatocytes.
to be overlooked when a concomitant more obvious disorder is present in the liver biopsy (Figure 19.1.5). Vitamin A toxicity has been reported to result in progressive liver injury leading to cirrhosis, and withdrawal in most cases results in either stable disease or regression of the fibrosis. The consequences of a milder form of the disease remain unclear at present. Clinically, consideration for reducing or stopping the vitamin A intake and follow-up with serial liver transaminase levels should be undertaken. The role of follow-up biopsies is yet undefined in this setting; however, it
1. Levine PH, Delgado Y, Theise ND, West AB. Stellate-cell lipidosis in liver biopsy specimens. Recognition and significance. Am J Clin Pathol. 2003;19(2):254–258. 2. Friedman SL, Wei S, Blaner WS. Retinol release by activated rat hepatic lipocytes: regulation by Kupffer cell-conditioned medium and PDGF. Am J Physiol. 1993;264(5 pt 1):G947–G952. 3. Bronfenmajer S, Schaffner F, Popper H. Fat-storing cells (lipocytes) in human liver. Arch Pathol. 1966;82(5):447–453. 4. Latry P, Bioulac-Sage P, Echinard E, et al. Perisinusoidal fibrosis and basement membrane-like material in the livers of diabetic patients. Hum Pathol. 1987;18(8):775–780. 5. Dupon M, Kosaifi T, Le Bail B, Lacut Y, Balabaud C, Bioulac-Sage P. Lipid-laden perisinusoidal cells in patients with acquired immunodeficiency syndrome. Liver. 1991;11(4):211–219. 6. Hautekeete ML, Geerts A, Seynaeve C, Lazou JM, Kloppel G, Wisse E. Contributions of light and transmission electron microscopy to the study of the human fat-storing cell. Eur J Morphol. 1993; 31(1–2):72–76. 7. Fallon MB, Boyer JL. Hepatic toxicity of vitamin A and synthetic retinoids. J Gastroenterol Hepatol. 1990;5(3):334–342. 8. Vahlquist A. Long-term safety of retinoid therapy. J Am Acad Dermatol. 1992;27(6 pt 2):S29–S33. 9. Carpenter TO, Pettifor JM, Russell RM, et al. Severe hypervitaminosis A in siblings: evidence of variable tolerance to retinol intake. J Pediatr. 1987;111:507–512. 10. Nyberg A, Berne B, Nordlinder H, et al. Impaired release of vitamin A from liver in primary biliary cirrhosis. Hepatology. 1988;8(1): 136–141.
20 Interpreting Iron in Liver Specimens MICHAEL TORBENSON
Significant progress has been made over the past several decades in understanding the causes and significance of iron accumulation in the liver. There has also been an increased effort to standardize terminology used when discussing iron-related disease. We shall use the following commonly used definitions: the term “hemochromatosis” indicates hepatic iron accumulation in the setting of a genetic mutation; the term “siderosis” indicates hepatic iron accumulation without genetic mutations; the term “genetic nonhemochromatotic iron overload disorder” will be used to refer to a range of rare genetic disorders that lead to iron accumulation that is primarily deposited in Kupffer cells and macrophages. M A J O R PROT E I N S A N D C E L L S I N VOLV E D I N I RON M E TAB OL I S M
There are many proteins and cells involved in iron metabolism that are relevant to iron overload in liver pathology. The major ones are listed below as a quick reference.
OV ERV IEW O F NO R MA L IRO N META BO LIS M
The healthy adult body contains a total of approximately 3 to 5 grams of iron (as a point of reference, a U.S. nickel weighs 5 grams). About 20 mg of iron is needed each day for normal physiological functions, largely heme synthesis, but the majority of this daily need is met through recycling of damaged and obsolescent red blood cells. Because of the efficiency of this red blood cell recycling, only 1 to 2 mg per day is needed in a healthy diet. Iron is important in several metabolic processes outside of heme synthesis, including oxidative phosphorylation and DNA synthesis. Despite this importance, iron can be toxic at high levels. Thus, iron levels are tightly regulated within the body. The human body has no physiological way to excrete iron. Instead, regulatory mechanisms are focused on iron absorption from the intestine. Separate, but tightly integrated, controls also regulate blood iron levels. Iron Absorption
Proteins DMT: Ferritin:
Ferroportin:
Hemojuvelin:
Hemosiderin: Transferrin:
Dimetal transporter-1. Transports iron from gut lumen to enterocyte cytoplasm Protein located in the cell cytoplasm that has an enormous capacity to bind iron; major physiological storage form of iron Transports iron out of cells into the blood stream (principally enterocytes and macrophages, hepatocytes) The precise role of this membrane-bound protein is not clear. However, it appears to interact with important signaling pathways (bone morphogenic protein [BMP], SMAD) that have hepcidin as a downstream target. Without hemojuvelin, these signaling pathways are not able to activate hepcidin in a normal fashion. Abnormal deposits of iron Transports iron in blood
Iron is absorbed primarily in the duodenum and proximal jejunum. Heme-iron (about 10% of a typical diet) is taken up by the enterocytes after disassociation from globin, whereas nonheme iron (about 90% of a typical diet) is first reduced from a ferric to a ferrous state and then transported by a protein called dimetal transporter-1 (DMT-1) across the cell membrane into the enterocytes (Figure 20.1). There are several additional iron transport mechanisms for getting luminal iron into the enterocyte cytoplasm, but they appear to be less important. Once iron is within the enterocytes, it has two main possible fates. If the body is iron replete, then the iron remains within the cytoplasm of the enterocytes and is stored as ferritin. When the enterocyte eventually dies and is sloughed, the iron within the cell’s cytoplasm is lost within the fecal stream. This is a key control mechanism to prevent iron overload. In contrast, if the body needs iron, then the iron is transported out of the enterocytes by ferroportin with some help by accessory proteins, including ceruloplasmin and hephaestin, and enters the blood stream where it is bound by transferrin and circulates within the blood.
Cells Enterocytes: Hepatocytes:
Macrophages:
Absorption and short-term storage of iron Major producer of ferritin, hepcidin Major organ for storage of iron in the form of ferritin, hemosiderin Main recycler of old/damaged red blood cells Major cell type for storage of iron in the form of ferritin, hemosiderin
Getting Iron From the Blood to Cells Throughout the Body
All cells have mechanisms to determine whether they have sufficient iron stores within their cytoplasm to meet their needs. If cells need more iron, they increase their expression of transferrin receptors. Currently, there are 2 known receptors: transferrin receptor 1 and transferrin receptor 2. Transferrin receptor 1 is on all nucleated cells, whereas receptor 2 is primarily found in the liver. These receptors are located in the cell 281
282
CHAPTER
20:
INTERPRETING
IRON
IN
LIVER
SPECIMENS
B A
C FIGURE 20. 1 (A) Overview of iron absorption; (B) When the body needs iron in its tissues, iron is shunted out of the enterocytes and
exported into the blood; (C) When the body does not need iron, then the enterocytes will store the iron in their cytoplasm. Eventually, the enterocyte dies and is lost into the intestinal lumen, at which time the iron will be lost with it. Abbreviations: Dcytb, duodenal cytochrome B; DMT, dimetal transporter; HCP, haem carrier protein.
membrane and can bind to and take in transferrin-bound iron from the blood. In healthy individuals, the blood contains much more transferrin protein than iron and the transferrin levels are approximately 30% saturated with iron. As blood iron levels increase, the excess transferrin proteins serve as a sort of buffer and will bind more iron to prevent excess free iron in the blood. Thus, increased serum transferrin levels can serve as a sensitive early indicator of excess iron absorption. All nucleated cells have transferrin receptors that can uptake transferrin-bound iron to meet the cells’ needs. Hepatocytes, with their abundant transferrin receptors, take up any excess iron, which then can be stored in the form of ferritin and, in times of great excess, as hemosiderin.
Iron Storage
If there is excess iron in the body, it can be incorporated into ferritin molecules for storage, largely in hepatocytes and macrophages. Ferritin is produced principally by the liver and is found in the liver cytoplasm, where it can hold up to 4500 atoms of iron per ferritin protein complex. Ferritin is typically not observed on Perls’ Prussian Blue stain, but occasionally it can be seen as a diffuse blush of blue in hepatocyte cytoplasm. The iron in ferritin can be rapidly accessed for physiological needs. If ferritin levels are excessive over a sufficiently long period of time, hemosiderin deposits can then develop. Hemosiderin is typically granular and golden brown on HE staining and is composed of iron and various proteins, principally degraded ferritin. The vast majority of the metal
CHAPTER
20:
INTERPRETING
in hemosiderin is iron but small amounts of copper and calcium can also be detected. Despite an identical HE appearance of hemosiderin in both genetic and nongenetic causes of iron overload, there are differences in both the metallic as well as the organic components at the molecular level (1). In contrast to ferritin, the iron in hemosiderin is not as readily available for biological needs. In sum, there are two important reservoirs of iron that can both be tapped to keep iron levels in the blood at physiologically correct levels: (1) iron stored within enterocytes and (2) iron stored as ferritin, principally in hepatocytes and macrophages. Both reservoirs have separate but interconnected control mechanisms that serve to regulate iron flow into the blood. If they both are unable to meet the demands for iron, then iron deficiency develops; if they have dysregulated (mutated) control mechanisms, then hemochromatosis can develop.
IRON
IN
LIVER
SPECIMENS
283
A
H E PC I D I N : C E NT R A L R E G U L ATOR OF I RO N TRAFFICKING Hepcidin
Hepcidin is the central regulator of iron stores in the body. Hepcidin is a secreted protein whose major source is hepatocytes. Peripheral adipose tissue can also be an important source, particularly with obesity. Hepcidin functions as a hormone and lowers blood iron levels. Hepcidin tracks iron levels similar to how insulin tracks blood sugar levels; in other words, if iron levels are high, hepcidin levels will be high, but if iron levels are low, hepcidin levels will be low (Figure 20.2). Control of Iron Release from Stores in the Enterocytes, Liver, and Macrophages
When blood levels are low, iron is released from enterocytes into the blood. In addition, iron is released from hepatocytes and macrophages where it has been stored as ferritin. However, when blood iron levels are adequate, then iron is blocked from being released from these two compartments. Hepcidin is the main regulator of this process: it blocks iron release from hepatocytes, macrophages, and enterocytes. Hepcidin accomplishes this by causing degradation of ferroportin, the protein that regulates transport of iron from enterocytes, macrophages, and hepatocytes into the blood. When hepcidin levels are low, there is increased iron absorption from the gut and increased release of iron into the blood from the iron stored in macrophages and hepatocytes. Hepcidin is an acute phase reactant and levels can be elevated in a variety of inflammatory and infectious conditions. In addition to inflammation, hepcidin levels are also increased by excess body iron stores. Many other factors are known to influence hepcidin levels (Figure 20.2). Recent findings have shown the central role of hepcidin in hemochromatosis. In fact, many of the mutations that
B F I G U R E 2 0 . 2 (A) Hepcidin blocks iron absorption from the intestine. Hepcidin also blocks iron that is stored in hepatocytes and macrophages from being released; (B) The regulation of hepcidin is complicated, but there are many known factors that can increase or decrease hepcidin levels.
lead to hemochromatosis, whether in hemochromatosis type 1 (HFE), juvenile hemochromatosis (HAMP), hemojuvelin (HJV), hemochromatosis type 3 (TfR2), all lead to decreased hepcidin production or impaired hepcidin function (2). A deficiency of hepcidin function first manifests as increased serum transferrin saturation levels. Later, increased serum ferritin levels are found and eventually increased serum iron levels are seen. This chronic excess of blood iron levels eventually leads to the accumulation of hemosiderin deposits in the liver and other organs.
References 1. Ward RJ, O’Connell MJ, Dickson DP, et al. Biochemical studies of the iron cores and polypeptide shells of haemosiderin isolated from patients with primary or secondary haemochromatosis. Biochim Biophys Acta. 1989;993:131–133. 2. Pietrangelo A. Hemochromatosis: an endocrine liver disease. Hepatology. 2007;46:1291–1301.
Case 20.1
Genetic Hemochromatosis MICHAEL TORBENSON
C L I N I C AL I N F OR M AT I ON
A 56-year-old woman underwent liver transplantation for alcohol-related cirrhosis and a 4 cm hepatocellular carcinoma. On histological examination, the liver showed marked iron accumulation including iron in the bile ducts. R E A S ON F OR R E F E R R A L
The liver contained large amounts of iron and was referred with the question of whether these histological findings were diagnostic of genetic hemochromatosis. PAT H O L OG I C AL F E AT U R E S
On Perls’ iron stain, the liver shows marked iron deposition with iron deposited fairly uniformly throughout the nodules. Heavy iron deposits are present in both the hepatocytes and the Kupffer cells (Figure 20.1.1). Overall, the iron is heaviest at the periphery of the nodules and tapers somewhat toward the center (Figure 20.1.2). Iron is also seen in septal-sized bile ducts (Figure 20.1.3). Follow-up genetic studies on this patient confirmed C282Y homozygosity.
F I G U R E 2 0 . 1 . 2 Perls’ iron stain; a zone 1 pattern of iron deposition was still evident in some of the cirrhotic nodules.
D I AG N OS I S
Cirrhosis in the setting of alcoholic liver disease and genetic hemochromatosis.
F I G U R E 2 0 . 1 . 3 Perls’ iron stain; iron was also present within the
biliary epithelium of septal-sized bile ducts. DISCUSSIO N
FIGURE 20. 1. 1 Perls’ iron stain; the liver showed marked iron accumulation with iron present in both hepatocytes and Kupffer cells.
The histological findings are certainly consistent with genetic hemochromatosis. However, histological findings are not diagnostic for genetic mutations per se. In genetic hemochromatosis, iron classically accumulates initially within zone 1 hepatocytes, and a clear gradient can often be seen in the amount of iron between zone 1 and zone 3 hepatocytes, even with advanced iron accumulation. In addition, the iron distribution often has a distinctive clustering around the bile canaliculi (Figure 20.1.4). With time, injury and death of hepatocytes will
284
CASE
20.1:
GENETIC
FIGURE 20. 1. 4 Perls’ iron stain; iron can be found in some cases deposited in hepatocytes immediately around the canaliculi.
FIGURE 20. 1. 5 Perls’ iron stain; iron can be found in small proliferation bile ductules in many cases that do not have genetic hemochromatosis.
lead to a redistribution of some iron into Kupffer cells and portal macrophages. However, a zone 1 distribution of iron can be seen in other non-hemochromatosis conditions, particularly once a liver is cirrhotic, and a diagnosis of hemochromatosis should not be based on recognizing a zonal pattern alone. The zone 1 predominant pattern of iron deposition is most likely a pattern seen whenever there is dysregulation of hepcidin, either through mutations or through reduced hepcidin production due to cirrhosis or other causes. Iron can also be seen in biliary epithelium on iron stain. Of note, iron is commonly seen in reactive bile ductules in areas of subacute parenchymal collapse in cirrhotic livers (Figure 20.1.5). This finding appears to have no association with hemochromatosis. Iron can also be deposited in the epithelium of the bile duct proper, as was seen in this case. In general, this pattern of iron deposition tracks better with the
H E M O C H R O M AT O S I S
285
F I G U R E 2 0 . 1 . 6 Perls’ iron stain; in some cases, iron can also be
found in endothelial cells.
overall severity of deposition within the liver and less so with HFE mutations per se. However, there is very little data that examines this specific question. With iron overload due to transfusion dependent anemias and similar causes, iron is classically first deposited in Kupffer cells, and with time there is involvement of the hepatocytes. Iron can also be seen in some cases either exclusively in portal endothelial cells (Figure 20.1.6) or in a combination of endothelial, hepatocyte, and Kupffer cell iron accumulation. At this time, there has not been any specific linkage of endothelial iron accumulation to a disease process or genetic mutation. In 1 study, endothelial iron positivity was linked to decreased interferon response in individuals with chronic hepatitis C infection (1). However, this finding has not been replicated and currently the etiology and significance of endothelial iron accumulation remains unclear. In regard to this particular consulting case, there is no reliable way to precisely parse out the roles of genetic versus environmental factors in the iron accumulation: they both contributed and their relative contributions cannot be discerned by histology. However, we discuss genetic cause of iron overload in more detail below. MUTAT IO NS IN IRO N- R ELAT ED GENES
There are a number of mutations that lead to hemochromatosis. The number will probably continue to grow with time. Despite this, these conditions share a core set of common findings as listed below: 1. As noted previously, a common mechanism is that all mutations, at least in part, involve abnormally low levels or dysfunction of hepcidin. 2. Most mutations are inherited—new sporadic mutations appear to be very rare. 3. Most mutations are recessively inherited.
286
CHAPTER
20:
INTERPRETING
4. The clinical consequences include iron deposition in the liver, heart, joints, and endocrine organs. There is an established increased risk for hepatocellular carcinoma and possibly increased risks for cholangiocarcinoma as well as other nonhepatic malignancies. 5. Blood findings progress in severity from elevated transferrin saturation levels, to elevated ferritin levels, to elevated iron levels. 6. Histologically, iron is deposited primarily in hepatocytes. Classic findings on iron stains for hemochromatosis include a zone 1 distribution of iron deposits, a pericanalicular pattern of iron deposits within the hepatocyte cytoplasm, and iron deposits in bile duct epithelial cells. However, these features are not specific for hemochromatosis. 7. Clinical management revolves around phlebotomy, which can be life saving as it can prevent the clinical sequelae listed above (No. 4). Individuals have intact erythropoiesis, and so they tolerate phlebotomy well.
HFE Mutations
HFE mutations were first linked to hereditary hemochromatosis in 1996. Since that time, over 37 mutations in this gene have been reported (16), but by far the most numerically and clinically important are C282Y and H63D mutations. C282Y mutations are strongly linked to northern European genetic ancestry (16) (Table 20.1.1), whereas H63D mutations have a wider ethnic distribution (3). Overall, the C282Y mutation accounts for 80% to 90% of hereditary hemochromatosis cases, whereas H63D accounts for approximately 60% of non-C282Y hereditary hemochromatosis cases (2,4). Other mutations, such as S65C, have also been linked to iron accumulation, but these mutations are significantly less common and data on their clinicopathological significance is limited. Gene penetrance is variable for all HFE mutations and to accommodate this, 4 clinical stages of the disease have been defined: genetic predisposition without abnormality, asymptomatic iron overload, iron overload with early symptoms, and iron overload with organ damage most commonly seen in the liver, heart, joints, pancreas, and other endocrine organs (5).
TABLE 20.1.1 HFE mutations in northern European populations
Genetic Status
Population Frequency
C282Y heterozygote
9.2
C282Y homozygote
0.4
H63D heterozygote
21.6
H63D homozygote
2.0
C282Y/H63D compound heterozygote
1.8
Wild/Wild From Ref. 16.
65.1
IRON
IN
LIVER
SPECIMENS
Individuals with C282Y mutations are at greater risk for iron accumulation than those with H63D mutations. Not surprisingly, C282Y homozygotes are at higher risk for iron accumulation than are C282Y heterozygotes. However, there is great phenotypic variation, even in individuals with C282Y homozygosity, underscoring the importance of other factors such as polymorphisms or mutations in other genes, environmental influences, and demographics such as age and gender. For example, in a major population-based study from Australia, 203 individuals who were homozygous for C282Y mutations were followed for 12 years. Twenty eight percent of the men, but only 1% of women, developed ironoverload–related diseases (6). This same research group also examined C282Y/H63D compound heterozygotes and found that only 1/82 men and none of the 95 women developed iron-overload–related disease over a 12-year study interval (7). This and other data argue for a strong protective effect for female gender. However, this does not appear to be solely due to physiological blood loss, and other gender-associated polymorphisms appear likely (reviewed in Wood et al) (8). The mechanism by which HFE mutations lead to iron accumulation are incompletely understood. At this time there are 2 major theories. The first suggests that the HFE protein is critical in determining the enterocytes’ internal “set point” for determining its cellular iron state. With HFE mutations, the enterocyte set point incorrectly indicates the cell is iron deficient, leading to increased enterocyte absorption of iron. The second theory focuses on the observation that, for incompletely understood reasons, individuals with HFE mutations have abnormally low plasma hepcidin levels. These low levels of hepcidin then lead to gradual excess iron absorption and deposition in the hepatocytes and other organ tissues. Both theories have supporting data from both animal models and human observations, suggesting both will be at least partially correct in the end. Causes of Death in HFE -Related Hemochromatosis
Clinical follow-up studies have consistently identified cirrhosis and liver decompensation as well as hepatocellular carcinoma as leading causes of death in individuals who are untreated or incompletely treated for HFE hemochromatosis (Table 20.1.2). However, there is also an increased risk for morbidity from heart failure and complications of diabetes. An increased risk for nonliver cancer has also been identified in some but not all studies. Treatment by phlebotomy can substantially lower the risk of death. A single unit of blood can safely remove 200 to 450 mg of iron, and over a period of time, usually a year or 2, phlebotomy can restore safe levels of iron within the blood. Liver Transplantation for HFE Iron Overload
Overall, hereditary hemochromatosis is an uncommon indication for an orthotopic liver transplantation (OLT). An early study of liver transplant outcomes that examined 5180 liver transplantations reported that only 56 (1%) of the
CASE
20.1:
GENETIC
transplantations were for hemochromatosis (9). This and other early studies reported an overall decreased posttransplant survival rate for patients with hereditary hemochromatosis compared with those transplanted for other causes of chronic liver disease, with major causes of mortality including TA B LE 20. 1. 2 Morbidity in individuals with clinical HFE
hemochromatosis
287
H E M O C H R O M AT O S I S
infection, cardiac failure, and cancer development (9–11). However, a more recent study has shown great improvement over the last decade in the survival of individuals transplanted for hemochromatosis (12). This increased survival likely reflects better patient selection and pre- and posttransplant management. Despite this, cardiovascular disease continues to be an important cause of morbidity and mortality (12). When examining an explanted liver with iron overload, any foci (even subcentimeter foci) with decreased iron deposition should be targeted for sectioning to evaluate for carcinoma. These “iron-free foci” are often associated with dysplastic nodules or with frank carcinoma. They can rarely be seen on needle biopsies also, and, when present, should be indicated in the report evaluated for malignancy.
Data
Milman et al (13)
Niederau et al (14)
Fargion et al (15)
Number of deaths
147
69
44
Length of follow-up
8.5 years, median
14 years, mean
4 years, median
Ethnicity
Danish
German
Italian
Cirrhosis, no cancer
32
20
23
Hepatocellular carcinoma
23
28
45
Non-liver cancer
11
12
14
Cardiovascular disease
11
20
7
Cerebrovascular disease
5
Not stated
Not stated
Hemojuvelin mutations are the most common cause of juvenile hemochromatosis. Nevertheless, this remains a relatively rare disease. There can be marked hepatocellular iron overload, and the disease typically runs a severe clinical course.
Respiratory disease
5
Not stated
Not stated
Hepcidin (Usually Children/Early Onset)
Sepsis
3
Not stated
Not stated
OT H ER MUTAT IO NS C AUSING GENET I C H EMO CH RO MATO SIS
Causes of death (%)
Other mutations in key iron-related genes can also lead to genetic hemochromatosis. These mutations are summarized in Table 20.1.3 as well as briefly described below. Hemojuvelin Mutations (Usually Children/Early Onset)
This rare form of genetic iron overload has marked hepatocellular iron overload and typically runs a severe clinical course.
TA B LE 20. 1. 3 Overview of genetic iron diseases involving the liver Gene
Also Known as (Some Names Used in the Literature)
Chromosome
Transmission
Onset
Iron Location
HFE
Hemochromatosis type 1
6p21.3
Recessive
Late
Hepatocytes Kupffer cells
HJV (hemojuvelin)
Juvenile hemochromatosis type 2A
1p21
Recessive
Early
Hepatocytes Kupffer cells
HAMP (hepcidin)
Juvenile hemochromatosis type 2B
19q13.1
Recessive
Early
Hepatocytes Kupffer cells
TfR2
Hemochromatosis type 3
7q22
Recessive
Late
Hepatocytes Kupffer cells
SCL11A2 (DMT-1)
None yet
12q13
Recessive
Early
Hepatocytes Kupffer cells
SLC40A1 (ferroportin)
Ferroportin disease type B
2q32
Dominant
Late
Hepatocytes Kupffer cells
Diseases With Iron Deposited Primarily in Mesenchymal Cells SLC40A1 (ferroportin)
Ferroportin disease type A (hemochromatosis type 4)
2q32
Dominant
Late
Kupffer cells hepatocytes
Tf (transferrin)
Hypotransferrinemia
3q21
Recessive
Early
Kupffer cells hepatocytes
CP (ceruloplasmin)
Hypoceruloplasminemia
3q23-35
Recessive
Late
288
CHAPTER
20:
INTERPRETING
Hypogonadism and cardiac disease are also prominent clinical manifestations. Transferrin Receptor Gene 2 (Usually Adults/Late Onset)
This rare form of genetic iron overload has marked hepatocellular iron overload and has a variable clinical course. DMT-1 Mutations (Usually Older Children)
This very rare disease has very few reported cases (about 4 to date), so data is quite limited. Children present with severe microcytic anemia. Iron accumulation is primarily in hepatocytes but, in very young children, biopsies can be negative for iron. N O N H E MOC H ROM ATOT I C I RON OV E R L OA D D I S E A S E (I E , P R E D OM I N AT E LY M E S E N C Y M A L I RON AC C U M U L AT I ON )
Ferroportin disease is a classic example of hereditary iron overload where the iron accumulation can be predominately in Kupffer cells. In contrast to the causes of hemochromatosis discussed above, all of which have elevated transferrin saturation levels early in the disease course, transferrin saturation levels in ferroportin disease do not become elevated until much later in the disease course. Ferroportin disease also stands out for its dominant inheritance pattern. Of note, there is substantial phenotypic variability and the disease is divided into 2 subtypes with different disease manifestations. Several other rare forms of genetic nonhemochromatotic iron overload disorder are also listed in Table 20.1.3. Clinically, ferroportin disease is characterized by increased serum ferritin but low or normal transferrin saturation
FIGURE 20. 1. 7 Perls’ iron stain; although the specificity of this find-
ing is uncertain, clumpy foci of iron deposits have been reported in ferroportin disease.
IRON
IN
LIVER
SPECIMENS
(type A). The inheritance pattern is autosomal dominant, and, histologically, the iron is deposited in Kupffer cells more than hepatocytes (type A), though for ferroportin disease type B the data is somewhat more mixed, and hepatocytes can have similar or more iron than Kupffer cells. Histologically, iron deposits can also be found in larger clumped foci (Figure 20.1.7). Clinically, both types are milder than HFE-mutation–associated hemochromatosis. Type A ferroportin disease tends to be milder than type B. Also of note, ferroportin disease does not respond well to phlebotomy due to an impaired erythropoiesis response.
References 1. Kaji K, Nakanuma Y, Harada K, Sakai A, Kaneko S, Kobayashi K. Hemosiderin deposition in portal endothelial cells is a histologic marker predicting poor response to interferon-alpha therapy in chronic hepatitis C. Pathol Int. 1997;47:347–352. 2. Mura C, Raguenes O, Ferec C. HFE mutations analysis in 711 hemochromatosis probands: evidence for S65C implication in mild form of hemochromatosis. Blood. 1999;93:2502–2505. 3. Settin A, El-Bendary M, Abo-Al-Kassem R, El Baz R. Molecular analysis of A1AT (S and Z) and HFE (C282Y and H63D) gene mutations in Egyptian cases with HCV liver cirrhosis. J Gastrointestin Liver Dis. 2006;15:131–135. 4. Limdi JK, Crampton JR. Hereditary haemochromatosis. Qjm. 2004;97:315–324. 5. Pietrangelo A. Hereditary hemochromatosis—a new look at an old disease. N Engl J Med. 2004;350:2383–2397. 6. Allen KJ, Gurrin LC, Constantine CC, et al. Iron-overload-related disease in HFE hereditary hemochromatosis. N Engl J Med. 2008;358: 221–230. 7. Gurrin LC, Bertalli NA, Dalton GW, et al. HFE C282Y/H63D compound heterozygotes are at low risk of hemochromatosis-related morbidity. Hepatology. 2009;50:94–101. 8. Wood MJ, Powell LW, Ramm GA. Environmental and genetic modifiers of the progression to fibrosis and cirrhosis in hemochromatosis. Blood. 2008;111:4456–4462. 9. Kilpe VE, Krakauer H, Wren RE. An analysis of liver transplant experience from 37 transplant centers as reported to Medicare. Transplantation. 1993;56:554–561. 10. Farrell FJ, Nguyen M, Woodley S, et al. Outcome of liver transplantation in patients with hemochromatosis. Hepatology. 1994;20:404–410. 11. Brandhagen DJ, Alvarez W, Therneau TM, et al. Iron overload in cirrhosis-HFE genotypes and outcome after liver transplantation. Hepatology. 2000;31:456–460. 12. Yu L, Ioannou GN. Survival of liver transplant recipients with hemochromatosis in the United States. Gastroenterology. 2007;133: 489–495. 13. Milman N, Pedersen P, á Steig T, Byg KE, Graudal N, Fenger K. Clinically overt hereditary hemochromatosis in Denmark 1948–1985: epidemiology, factors of significance for long-term survival, and causes of death in 179 patients. Ann Hematol. 2001;80:737–744. 14. Niederau C, Fischer R, Purschel A, Stremmel W, Häussinger D, Strohmeyer G. Long-term survival in patients with hereditary hemochromatosis. Gastroenterology. 1996;110:1107–1119. 15. Fargion S, Mandelli C, Piperno A, et al. Survival and prognostic factors in 212 Italian patients with genetic hemochromatosis. Hepatology. 1992;15:655–659. 16. Hanson EH, Imperatore G, Burke W. HFE gene and hereditary hemochromatosis: a HuGE review. Human Genome Epidemiology. Am J Epidemiol. 2001;154:193–206.
Case 20.2
Grading Iron MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
A 46-year-old man was biopsied to stage and grade chronic hepatitis C. An iron stain shows moderate iron accumulation.
and density of blue staining correlates, albeit imperfectly, with tissue iron concentrations. The stain is not as sensitive for very low levels of iron but is easier and more reproducible than other methods such as the Tirmann-Schmeltzers method, which can identify both ferric and ferrous forms of iron.
R E A SON F OR R E F E R R AL
Ferritin: Normally no ferritin will be seen. However, in cases of elevated serum ferritin levels, ferritin may be seen as a light diffuse blue blush of the hepatocyte or Kupffer cell cytoplasm (Figure 20.2.2). Hemosiderin: Hemosiderin can be seen as brown granular deposits on HE stains and as a bright blue granular staining on iron stain. Residual brown granular material is often seen on iron stain and represents lipofuscin in most cases.
The case was referred for recommendations for the best grading system for iron overload. PAT H OL OG I C F E AT U R E S
The biopsy shows moderate zone 1 hepatocellular iron accumulation (Figure 20.2.1).
D I AG N OS I S
Moderate hepatocellular iron accumulation in the setting of chronic hepatitis C .
D I S C U S S I ON
The major histochemical stain used to detect iron in the liver is Perls’ Prussian Blue (note that the most correct spelling is Perls or Perls’ Prussian Blue, not Perl’s Prussian Blue). This stain is named after Max Perls, a German pathologist who first suggested the stain. The basic chemistry of Perls’ Prussian Blue is that iron in the ferric state will react with hydrochloric acid to form ferric ferrocyanide, an insoluble blue compound (Prussian Blue) that can be seen histologically. The distribution
FIGURE 20. 2. 1 Perls’ iron stain; moderate iron accumulation is seen
within periportal hepatocytes.
When evaluating a surgical pathology liver specimen, it is prudent patient care to provide information on the amount of iron accumulation in the hepatocellular and Kupffer cell compartments that is sufficiently detailed to be clinically actionable when appropriate. A description is sufficient for this purpose, and there is no data to support an additional need to provide a formal number based on a specific scoring system. However, if the pathologists or clinicians prefer to provide a formal numerical assessment, that is fine. Sufficient scoring system detail should then be provided to allow a reader of the report to determine what the numbers mean. For example, a statement of the sort “iron grade 2” is in itself fairly useless and is strongly discouraged as neither the magnitude of the scale nor the location of the iron is apparent from this statement. As discussed above, there is no reliable way to determine whether the iron seen in a typical hepatitis C biopsy, such as
F I G U R E 2 0 . 2 . 2 Perls’ iron stain; a light blue blush on iron stain can be seen in some cases and represents ferritin.
289
290
CHAPTER
20:
INTERPRETING
the case under consideration, is because of genetic versus nongenetic causes. In these situations, my practice is to report the iron as described above. When iron levels reach the moderate to marked levels in noncirrhotic specimens, then I indicate in a note that the iron accumulation is more than is seen in typical cases of chronic hepatitis C and that further genetic testing may be of interest as clinically indicated. Iron-Grading Systems
There are many iron-grading systems that have been proposed over the years. They vary considerably in their approach: some are based on zonation of iron distribution, some on the lowest magnification that discernable granules can be seen, some on the percentage of hepatocytes positive for iron. There is a nice summary of these iron-grading systems by Dr. Randall G Lee in Diagnostic Liver Pathology, also available online (http://tpis1.upmc.com:81/tpis/dlp/DLPHome.html; click on Chapter 9 and find Table 9-3). This book chapter is somewhat dated and does not cover several newer systems but is still very useful. The system by Deugnia and Turlin (1) has the advantage of having been validated, but it is too complex to be readily adopted for routine diagnostic use. Is any system clearly the best? Probably not, but I personally use a schema (Table 20.2.1) based on the percentage of hepatocytes positive for iron, similar to that described by LeSage et al (2). For routine diagnostic purposes, I include the descriptor (eg, “mild,” etc) in the pathology report but do not routinely provide the corresponding numerical grade. I believe that this simple-to-use classification system provides sufficient clinical information for patient care. But there are many reasonable alternatives to consider if you prefer a different approach. A modified Scheuer’s system (shown in Table 20.2.2) is also a very useful and popular system. If employed, separate numbers should be given for hepatocellular and the reticuloendothelial iron.
IRON
IN
LIVER
SPECIMENS
TABLE 20.2.1 My iron scoring system (similar to that of LeSage)
Grade
Description
Hepatocytes (%)
Lobular Kupffer Cells (%)
None
None
0
None
1
Minimal
5
5
2
Mild
5–30
5–30
3
Moderate
31–60
31–60
4
Marked
60
60
Note: For studies, I also record the zonal pattern of iron and whether the distribution is homogeneous. For some studies, I also record endothelial iron and portal macrophage iron. From Ref. 2.
TA BL E 2 0 . 2 . 2 Modified Scheuer’s system Grade
Description
0
Iron granules absent or iron granules barely seen at 400
1
Iron granules resolved at 250
2
Iron granules resolved at 100
3
Iron granules resolved at 25
4
Iron deposits resolved at 10 or iron deposits visible without magnification
From Ref. 3.
References 1. Deugnier Y, Turlin B. Pathology of hepatic iron overload. World J Gastroenterol. 2007;13:4755–4760. 2. LeSage GD, Baldus WP, Fairbanks VF, et al. Hemochromatosis: genetic or alcohol-induced? Gastroenterology. 1983;84:1471–1477. 3. Turlin B, Deugnier Y. Evaluation and interpretation of iron in the liver. Semin Diagn Pathol. 1998;15:237–245.
Case 20.3
Hepatic Iron Index MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
A 43-year-old man with cryptogenic cirrhosis undergoes liver transplantation. R E A SON F OR R E F E R R AL
The liver shows marked iron accumulation. It was unclear to the referring pathologist what role the hepatic iron index currently plays in evaluating biopsies, as compared with genetic testing. PAT H OL OG I C F E AT U R E S
The liver shows established cirrhosis with moderate to marked iron accumulation (Figure 20.3.1). There is no other strong histological clue to the underlying liver disease.
Quantitative Measurement of Hepatic Iron Concentrations
The normal adult liver has between 10 to 36 μmol iron/g dry weight of liver. Hepatic iron concentrations measured in fresh liver tissue or in paraffin-embedded tissue are equivalent. Thus, paraffin-embedded tissues are preferred over fresh tissues in most cases because they allow direct visualization of the tissue and assure the tissue is representative. This prevents submission of tissue that is largely composed of collapsed/ fibrotic stroma or a nodule that is either unusually high or low in stainable iron compared with the rest of the tissue. Excess iron accumulation has been classified as mild (up to 150 μmol iron/g dry weight of liver), moderate (151–300), and marked (>301) (1). Iron levels greater than 400 μmol are the most strongly associated with cirrhosis but lower levels of iron also contribute to fibrosis progression in the setting of other liver diseases.
D I AG N OS I S
Hepatic Iron Index
Cirrhosis with marked iron overload.
D I S C U S S I ON
Given the routine availability of HFE genetic testing, the hepatic iron index has little role for diagnosing hemochromatosis in most cases. However, clinical requests for iron quantitation are not uncommon and can still be useful in patient care in some situations.
Historically, the hepatic iron index was calculated as an aid to interpreting quantitative tissue iron levels. The hepatic iron index adjusts the total iron concentration for age, based on the observation that hepatic iron concentrations tend to increase steadily with age in individuals with genetic hemochromatosis, but not in individuals with secondary iron overload. In a noncirrhotic liver, a hepatic iron index greater than 1.9 was considered suggestive of genetic hemochromatosis. Given the advances in understanding the causes of hemochromatosis and the readily available genetic testing for HFE mutations in many parts of the world, the diagnostic role of the hepatic iron index has somewhat diminished in importance, but direct measurement of hepatic iron concentration remains useful in guiding therapy, and we still get many requests for blocks to be submitted for quantitative iron analysis. The formula for the hepatic iron index is as follows: μg iron per gram dry weight of liver/55.846 Patient’s Age The value of 55.846 represents the atomic weight of iron. Noninvasive Measurements of Hepatic Iron
FIGURE 20. 3. 1 Perls’ iron stain; established cirrhosis with moderate
to marked iron accumulation is seen.
Magnetic resonance imaging (MRI)-based imaging studies have advanced in recent years to the point that they can reasonably assess iron accumulation and can also distinguish hepatic from reticuloendothelial iron deposits. There have been multiple validation studies, and MRI has established for itself an important role in measuring iron in the liver. Recent expert opinion review articles on hemochromatosis
291
292
CHAPTER
20:
INTERPRETING
have highlighted the changing role of the biopsy in managing patients with HFE hemochromatosis (2,3). Biopsies continue to be important in determining the fibrosis stage and to search for any associated lesions (eg, evaluation of mass lesion). Some experts (3) foresee a further diminution of the role of liver biopsies with the advent of noninvasive markers of liver fibrosis.
IRON
IN
LIVER
SPECIMENS
References 1. Deugnier Y, Turlin B. Pathology of hepatic iron overload. World J Gastroenterol. 2007;13:4755–4760. 2. Pietrangelo A. Hemochromatosis: an endocrine liver disease. Hepatology. 2007;46:1291–1301. 3. Deugnier Y, Brissot P, Loreal O. Iron and the liver: update 2008. J Hepatol. 2008;(48)(suppl 1):S113–S123.
Case 20.4
Marked Hepatic Iron but No Genetic Mutation MICHAEL TORBENSON
C L I N IC AL I N F OR M AT I ON
DISCUSSIO N
A 63-year-old woman with cryptogenic cirrhosis underwent liver transplantation. Genetic testing for HFE mutations were negative. Although fatty liver has not been documented by biopsy, she is obese with type II diabetes, hypertension, and hyperlipidemia.
Iron can accumulate in cirrhotic livers of individuals who do not have clinical findings of genetic hemochromatosis. In a classic study by Ludwig et al, iron stains were positive in 32% of 447 liver explants with varying underlying liver diseases. For those diseases with at least 5 cases in this study, the proportions of cases with any degree of positivity by iron stain were as follows: hereditary hemochromatosis (100%), cryptogenic cirrhosis (65%), alcohol cirrhosis (63%), chronic hepatitis B cirrhosis (65%), AAT cirrhosis (56%), chronic hepatitis C cirrhosis (42%), primary biliary cirrhosis (PBC) (10%), and primary sclerosing cholangitis (PSC) (7%). In this same study, the number of cases with a hepatic iron index of greater than 1.9 were as follows: hereditary hemochromatosis (HH) (100%), alpha-1-antitrypsin (AAT) (28%), cryptogenic cirrhosis (19%), alcohol cirrhosis (14%), chronic hepatitis B cirrhosis (18%), chronic hepatitis C cirrhosis (7%), PBC (1%), and cirrhosis from PSC (1%). This and other data sets document that other diseases can have iron deposition within the liver and that in AAT deficiency and in cryptogenic cirrhotic livers, 20% or more of cases can have hepatic iron indexes greater than 1.9. Another important observation from these data is that biliary cirrhosis is only rarely associated with iron overload. An important study by Kowdley et al (1) found that patients with significant hepatic iron accumulation had decreased survival following transplantation regardless of whether they had an HFE mutation. The reason(s) for this are unclear, but at least in a subset of these individuals, there can be significant extrahepatic stores of iron at the time of transplantation, often clinically unrecognized (2). In cases such as these, the stress of surgery or other posttransplant factors may place this group of patients at increased risk for heart failure.
R E A SON F OR R E F E R R AL
Based on the clinical findings, she most likely has fatty liver disease–associated cirrhosis. But there is also marked iron accumulation. PAT H OL OG I C F E AT U R E S
The liver shows established cirrhosis with moderated to marked and somewhat patchy iron accumulation (Figure 20.4.1). No specific findings to suggest an alternative cause for her cirrhosis is seen and a diagnosis of cryptogenic cirrhosis, most likely fatty liver disease, based on clinical history is rendered (see Case 4.5 for further discussion on fatty liver and iron stores). Subsequent HFE mutational studies were positive for H63D heterozygosity.
D I AG N OS I S
Cryptogenic cirrhosis, most likely due to nonalcoholic fatty liver disease. Moderate iron accumulation with H63D heterozygosity.
References 1. Kowdley KV, Brandhagen DJ, Gish RG, et al. Survival after liver transplantation in patients with hepatic iron overload: the national hemochromatosis transplant registry. Gastroenterology. 2005;129:494–503. 2. Fenton H, Torbenson M, Vivekanandan P, Yeh MM, Hart J, Ferrell L. Marked iron in liver explants in the absence of major hereditary hemochromatosis gene defects: a risk factor for cardiac failure. Transplantation. 2009;87:1256–1260.
FIGURE 20. 4. 1 Perls’ iron stain. Sections of the liver showed established cirrhosis with moderate patchy hepatocellular and Kupffer cell iron accumulation.
293
Case 20.5
Iron in the Setting of Chronic Hepatitis C MICHAEL TORBENSON
C L I N I C AL I N F OR M AT I ON
A 46-year-old man was biopsied to stage and grade chronic hepatitis C. An iron stain shows mild iron accumulation. R E A S ON F OR R E F E R R A L
The referring pathologist raised the question as to the significance of the iron in this liver biopsy. PAT H OL OG I C F E AT U R E S
The biopsy showed moderate portal chronic inflammation and mild lobular activity with mild portal fibrosis on trichrome stain. Overall the histological findings were that of typical chronic hepatitis C infection. An iron stain showed mild hepatocellular iron accumulation in zone 1 hepatocytes (Figure 20.5.1).
D I AG N OS I S
Chronic hepatitis C with mild hemosiderosis.
D I SC U SSI ON
Iron deposits including both hepatocellular as well as reticuloendothelial are seen in liver biopsies of individuals with chronic hepatitis C virus (HCV), with a reported range of 5% to 48% (1–6). Overall, the median is approximately 30% for
FIGURE 20. 5. 1 Perls’ iron stain. The liver shows mild hepatocellular
iron accumulation.
these studies, and the variation presumably reflects differences in gender, viral genotypes, and the proportion of cirrhotics in the cohort. Livers with genotype 3 infection tend to have more hepatocellular iron than other genotypes (4). In the majority of cases, the iron deposits are mild, occasionally moderate, and only very rarely severe. A large body of literature has been published on the question of the significance of HFE mutations in chronic HCV. Unfortunately, despite all of the work, the literature is substantially mixed on the question of whether HFE mutations increase the risk for fibrosis progression. This current state of confusion likely reflects the many different study populations, study designs, as well as variable penetration of genetic hemochromatosis. Many studies also do not adequately control for potentially confounding variables such as gender, viral genotypes, duration of HCV infection, and so on. Nevertheless, one reasonable way to synthesize the data on HFE mutations and HCV is as follows: (1) individuals with chronic HCV do not have an increased risk for HFE mutations (1,7–9); (2) once an individual has chronic HCV infection, HFE mutations may increase the rate of fibrosis progression (10) and the presence of HFE mutations is associated with higher fibrosis stages in many (1,7–12) but not all studies (2,3). The strength of the association between HFE mutations and fibrosis has been measured by both relative risks, where a relative risk of 4.6 has been reported (9) and the odds ratio, where an odds ratio for C282Y heterozygosity has been reported ranging from 2.5 to 30 (1,7,10). Overall, C282Y alleles appear to have a stronger risk for fibrosis than H63D alleles (10). With a sufficiently long duration of chronic HCV infection, the risk of cirrhosis is high regardless of HFE mutational status, and the effect of HFE mutations may be harder to discern (10). Interestingly, for unclear reasons, HFE mutations have also been linked to increased inflammation on liver biopsy in some studies (8,9). Despite the observations linking HFE mutations to increased fibrosis and less consistently to increased inflammation, HFE mutation status has typically not been associated with increased iron deposits by histochemical analysis (3,8,9). In contrast, H63D, but not C282Y mutations, were associated with increased hepatic iron concentrations in 1 study (10). Most of the data discussed above is from studies that looked at HFE mutations. The question then naturally arises of the meaning of mild to moderate iron deposits in individuals with chronic HCV who lack HFE mutations. Unfortunately, the data are no more clear on this point and the same “take home message” as above appears to apply: most likely there is either no role or a very limited role in terms of fibrosis progression for minimal or very mild iron on a liver biopsy; for moderate iron there is likely a modest role. For marked iron accumulation, a role in fibrosis progression seems very likely even if it has not yet been specifically demonstrated.
294
CASE
20.5:
IRON
IN
THE
SETTING
Of note, there is only limited longitudinal data or paired biopsy studies on the role of iron in fibrosis progression, and it is hoped that future studies will permit a more accurate and nuanced understanding of the role of iron in fibrosis progression. One of the few paired biopsy studies that specifically analyzed the role for iron found no role for fibrosis progression in 214 individuals, but the time interval between biopsies was only 2.5 years, which limits the findings applicability (13).
6.
7.
8.
References 1. Gehrke SG, Stremmel W, Mathes I, Riedel HD, Bents K, Kallinowski B. Hemochromatosis and transferrin receptor gene polymorphisms in chronic hepatitis C: impact on iron status, liver injury and HCV genotype. J Mol Med. 2003;81:780–787. 2. Thorburn D, Curry G, Spooner R, et al. The role of iron and haemochromatosis gene mutations in the progression of liver disease in chronic hepatitis C. Gut. 2002;50:248–252. 3. Negro F, Samii K, Rubbia-Brandt L, et al. Hemochromatosis gene mutations in chronic hepatitis C patients with and without liver siderosis. J Med Virol. 2000;60:21–27. 4. Sebastiani G, Vario A, Ferrari A, et al. Hepatic iron, liver steatosis and viral genotypes in patients with chronic hepatitis C. J Viral Hepat. 2006;13:199–205. 5. Pirisi M, Scott CA, Avellini C, et al. Iron deposition and progression of disease in chronic hepatitis C. Role of interface hepatitis, portal
9.
10.
11.
12.
13.
OF
CHRONIC
H E PAT I T I S
C
295
inflammation, and HFE missense mutations. Am J Clin Pathol. 2000;113: 546–554. Valenti L, Pulixi EA, Arosio P, et al. Relative contribution of iron genes, dysmetabolism and hepatitis C virus (HCV) in the pathogenesis of altered iron regulation in HCV chronic hepatitis. Haematologica. 2007;92:1037–1042. Erhardt A, Maschner-Olberg A, Mellenthin C, et al. HFE mutations and chronic hepatitis C: H63D and C282Y heterozygosity are independent risk factors for liver fibrosis and cirrhosis. J Hepatol. 2003; 38:335–342. Martinelli AL, Franco RF, Villanova MG, et al. Are haemochromatosis mutations related to the severity of liver disease in hepatitis C virus infection? Acta Haematol. 2000;102:152–156. Geier A, Reugels M, Weiskirchen R, et al. Common heterozygous hemochromatosis gene mutations are risk factors for inflammation and fibrosis in chronic hepatitis C. Liver Int. 2004;24:285–294. Tung BY, Emond MJ, Bronner MP, Raaka SD, Cotler SJ, Kowdley KV. Hepatitis C, iron status, and disease severity: relationship with HFE mutations. Gastroenterology. 2003;124:318–326. Smith BC, Gorve J, Guzail MA, et al. Heterozygosity for hereditary hemochromatosis is associated with more fibrosis in chronic hepatitis C. Hepatology. 1998;27:1695–1699. Bonkovsky HL, Troy N, McNeal K, et al. Iron and HFE or TfR1 mutations as comorbid factors for development and progression of chronic hepatitis C. J Hepatol. 2002;37:848–854. Ryder SD, Irving WL, Jones DA, et al. Progression of hepatic fibrosis in patients with hepatitis C: a prospective repeat liver biopsy study. Gut. 2004;53:451–455.
Case 20.6
Neonatal Hemochromatosis LINDA D. FERRELL
C L I N I C AL I N F OR M AT I ON
A female infant, born at term to a 30-year-old healthy mother, developed hypoglycemia on the first day of life. The baby became lethargic and bradycardic by day 3, requiring ventilatory assistance. The patient was also anemic, requiring transfusion, and received fresh frozen plasma for what was thought to be disseminated intravascular coagulopathy. The patient was also treated with ampicillin and gentamicin for possible sepsis. Abdominal ultrasound revealed ascites and reduced portal vein blood flow to the liver. Laboratory tests included markedly elevated total bilirubin at 22 mg/dL, with mild transaminase elevations. TORCH titer (toxoplasma, other infections, rubella, cytomegalovirus, herpes virus) determinations for infectious processes were unremarkable, screening tests for galactosemia and tyrosinemia were negative, and serum AAT level was normal. No maternal risks for growth or congenital abnormalities were found, and the infant had not received fructose. The patient died at 5 weeks of age, and an autopsy was performed.
acini (Figure 20.6.2). Abundant iron was diffusely present in hepatocytes (Figure 20.6.3), 4 on a scale of 0 to 4 and diffuse throughout the liver, with much less iron present in stromal or other mesenchymal cells (1). No significant inflammatory changes were present. Some multinucleate hepatocytes were present but were not a prominent finding. Bile ducts were present in portal zones. There was no fatty change or extramedullary hematopoiesis. Prominent macrophage infiltrates typical of storage or metabolic disorders were not present. Other organs demonstrated pigment in parenchymal
R E A S ON F OR R E F E R R A L
The liver showed extensive parenchymal destruction and intraparenchymal fibrosis but without nodularity, suggesting a possible inherited or metabolic disorder. PAT H OL OG I C F E AT U R E S
The liver was firm and shrunken and microscopically demonstrated a diffuse fibrosis in the parenchyma, with both sinusoidal and pericellular patterns and rare small nodule formation (Figure 20.6.1). Bile plugs were noted in cholangiolar-like structures as well as within hepatocyte rosettes/
FIGURE 20. 6. 1 Trichrome stain; liver parenchyma is severely
distorted by extensive fibrosis with limited nodule formation.
F I G U R E 2 0 . 6 . 2 Hepatocytes, mostly in small clusters, are surrounded
by extensive fibrous tissue. Cholestasis, few pigmented macrophages, and ductular reaction are present.
F I G U R E 2 0 . 6 . 3 Iron stain; iron is moderately prominent in hepatocytes but is also present in stromal cells. Small ductular structures with bile are present.
296
CASE
20.6:
N E O N ATA L
FIGURE 20. 6. 4 Iron stain; iron is present in cardiac muscle.
cells, including the myocardium (Figure 20.6.4), exocrine and endocrine pancreas, thyroid follicular cells, bronchial glands, and gastric antral mucosal epithelial cells. The placenta was not siderotic, and the spleen showed minimal (1) iron in scattered macrophages.
D I AG N OS I S
Neonatal hemochromatosis.
H E M O C H R O M AT O S I S
297
failure, similar to what is seen in end-stage cirrhosis. (8,9). In the case of secondary iron overload with cirrhosis, the lack of hepcidin production by the severely damaged liver may result in the hepatocyte iron overload and also may shift iron from the reticuloendothelial system (macrophages) to organs and thus cause the extrahepatic iron overload (9). The differential diagnosis typically includes other inherited disorders, as the pattern of fibrosis can be similar in metabolic disorders. Since almost all inheritable metabolic disorders do not present at this early age with liver failure due to extensive fibrosis, the list is quite short. For example, in tyrosinemia, which can occasionally present early in life, the presentation is typically not earlier than 3 to 4 months of age. Galactosemia and fructosemia do not cause this kind of fibrosis until the patient has been exposed to these sugars. Zellweger syndrome (due to absence of peroxisomes) can also present in the perinatal period but typically does not show the massive liver parenchymal damage as seen here so early in the course, and instead tends to show a more neonatal hepatitis– like pattern of injury with possible ductopenia. The neonatal hepatitis–like pattern of injury has more giant cells and lacks the extensive and well-established fibrosis of neonatal hemochromatosis (see Chapter 10). Thus, overall, this change is somewhat unique for this age group in that the patient is born with advanced fibrosis and liver failure and may represent a secondary iron overload rather than primary iron-related injury.
References D I S C U S S I ON
Neonatal, or perinatal, hemochromatosis (NH) (1,2) is one of the most common causes of cirrhosis and death in the perinatal period. This entity typically presents in the first few days of life as severe liver disease manifested by ascites, hypoglycemia, hyperbilirubinemia, and/or other symptoms of liver failure. Anemia and bleeding problems are also typically present. If NH is suspected clinically, some have suggested that MRI might be helpful to demonstrate the iron in the liver and other organs (other than the spleen), such as pancreas (3,4). Early treatment with antioxidant-chelation is not very successful, so transplantation is often the treatment of choice (5,6). No inheritable defect has been demonstrated, but siblings can be affected, which suggests an inheritable or maternal link. Other inheritable or congenital associations have also been noted including Down syndrome. However, the disease in siblings may also suggest that liver injury is not due to iron but other causes like intrauterine ischemia or various forms of intrauterine hepatitis, including an alloimmune form of hepatitis (7). The iron overload may be secondary to the liver
1. Whitington PF. Fetal and infantile hemochromatosis. Hepatology. 2006;43(4):654–660. 2. Witzleben CL, Uri A. Perinatal hemochromatosis: entity or end result? Hum Pathol. 1989;20:335–340. 3. Udell IW, Barshes NR, Voloyiannis T, et al. Neonatal hemochromatosis: radiographical and histological signs. Liver Transpl. 2005;11(8): 998–1000. 4. Williams H, McKiernan P, Kelly D, Baumann U. Magnetic resonance imaging in neonatal hemochromatosis; Are we there yet? Liver Transpl. 2006;12:1725. 5. Heffron T, Pillen T, Welch D, et al. Medical and surgical treatment of neonatal hemochromatosis: single center experience. Pediatr Transplant. 2008;11(4):347–348. 6. Grabhorn E, Richter A, Burdelski M, Rogiers X, Ganschow R. Neonatal hemochromatosis: long-term experience with favorable outcome. Pediatrics. 2006;118(5):2060–2065. 7. Whitington PF. Neonatal hemochromatosis: a congenital alloimmune hepatitis. Semin Liver Dis. 2007;27(3):243–250. 8. Deugnier Y, Brissot P, Loreal O. Iron and the liver: Update 2008. J Hepatol. 2008;48:S113-S123. 9. Fenton H, Torbenson M, Vivekanandan P, Yeh MM, Hart J, Ferrell L. Marked iron accumulation in liver explants in the absence of major gene defects of hereditary hemochromatosis: a risk factor for cardiac failure. Transplantation. 2007;87:1256–1260.
This page intentionally left blank
21 Wilson Disease LINDA D. FERRELL
Wilson disease (WD) is an autosomal recessive disorder that results from the accumulation of copper in various tissues due to the lack of synthesis of a membrane-associated copper transport protein. Regulation of copper balance involves the liver, with excess copper excreted in the bile and eventually in the stool. Modern diets typically contain about 1 mg/day of copper, which is about 25% more than needed; thus, about 0.25 mg/day needs to be excreted. Wilson disease patients have a defect in this excretion (1). The gene responsible for WD has been identified on chromosome 13 and designated as ATP7B. Numerous mutations in this gene have been identified, and no one mutation accounts for more than 30% of the total group of mutations; so most patients are compound heterozygotes (2). Thus, it has not proven practical to develop a DNA test for diagnosis to date. It is possible that the large number of mutations may account for some of the variability in clinical and pathological presentation. C L I N I C AL P R E S E N TAT I ON S
The patients generally present as teens or young adults and almost never before the age of 5 years. The disease may present with either hepatic disease (more frequently seen in children, young adults) or neurological symptoms (older patients). The hepatic disease may present in a variety of forms, including acute and chronic patterns (Table 21.1). Presentation with liver disease after 50 years of age is very rare (3). The fulminant pattern may be the first clinical presentation of the disease and mimics fulminant viral hepatitis. The associated severe hemolysis points toward WD. Fulminant WD can be associated with normal ceruloplasmin as the serum levels become elevated during the hepatitic flare as a “nonspecific acute phase reactant” response. Kayser-Fleischer (KF) rings tend to be found in the later stages, more typically in patients with neurological syndromes (2). In asymptomatic WD, the earliest changes include fatty change, glycogenated nuclei, and rare hepatocyte spotty necrosis (4). The acute hepatitic and fulminant forms have the
microscopic pattern of acute hepatitis of variable severity (5), with the fulminant form typically showing severe hepatocyte damage and dropout, with confluent, bridging, and/or panacinar necrosis. A variable degree of inflammatory infiltrate is typically present. This form may be the first clinical presentation of the disease, but in many cases periportal fibrosis or even a background of cirrhosis may be evident indicating previously undetected disease. Copper may be difficult to identify on sections in this form, and, if present, may be seen in macrophages rather than hepatocytes. The chronic hepatitic form usually has the features of active chronic hepatitis with interface hepatitis and variable degree of lobular inflammation and necrosis; bridging necrosis may be present (4,6). Chronic hepatitic changes may be associated with bridging fibrosis or cirrhosis and the amount of inflammatory infiltrate may vary from none to marked. Cirrhosis, when present, tends to be a macronodular, but micronodular form may also be seen, especially in the young patients with superimposed fulminant failure (7). Electron microscopy in WD shows characteristic mitochondrial changes, including variation in size and vacuolar defect in the cristae that is thought to be fairly pathognomonic for WD (5). Several recommended methods for screening for WD are available, as noted by Brewer et al.
TA B LE 21. 1 Various presentations in Wilson disease Hepatic
Acute hepatitis, chronic hepatitis, cirrhosis, liver failure
Neurological
Tremor, Parkinsonism, dysarthria, dystonia, other movement disorders
Behavioral
Loss of ability to focus on tasks, emotionality, depression, bizarre behaviors, insomnia, other behavioral disturbances
Adapted from Brewer et al (1).
299
1. Serum ceruloplasmin. This test is easy to do, and low serum ceruloplasmin will point to diagnosis in 75% of cases. However, ceruloplasmin levels can be normal in WD, particularly in the hepatitic variants, and many heterozygotes have values in low range. 2. Twenty-four–hour urine copper. This value will be less than 100 μg in symptomatic WD. A lower value in symptomatic, untreated patients typically excludes the diagnosis. However, the urine copper can be elevated in the absence of WD, if the liver disease is long-standing and has an obstructive component. 3. Quantitative copper assay on liver biopsy. This is the gold standard for diagnosis. The value is more than 200 μg/g dry weight of liver (normal 20–50 μg/g) in most cases. Patients with obstructive liver disease can have copper levels similar to those seen in WD. Heterozygotes may also show high hepatic copper but is typically not over 125 μg/g. The liver biopsy may show high copper content even when staining for copper is negative. It is thought that the copper stains only identify certain types of bound copper. 4. Slit-lamp examination for KF rings. This test is close to 100% diagnostic for patients with neurological or psychiatric symptoms but not nearly as helpful for hepatic presentations. Patients with obstructive liver disease may also develop KF rings.
300
CHAPTER
21:
WILSON
The diagnosis depends on the finding of low serum ceruloplasmin (although 5%–10% of patients may have normal ceruloplasmin, especially those with a fulminant course with hepatic necrosis as noted above), increased 24-hour urine copper, and increased hepatic copper content (greater than 200– 250 μg/g dry weight). The early diagnosis of WD is important as therapy will prevent the complications of copper overload. D I FFE R E N T I A L D I AG N OSI S
The histopathology of the various forms of WD may be indistinguishable from other causes of hepatitis or cirrhosis by light microscopy, including viral, autoimmune, and drug etiologies, but features such as fatty change, glycogenated hepatocytic nuclei, Mallory-Denk bodies in periportal hepatocytes, and moderate to marked copper deposits may help to make the diagnosis of WD. Unfortunately, triad of changes (fat, glycogenated nuclei, and Mallory-Denk bodies) are also features of steatohepatitis, and with the increase of obesity in children and young adults, there is potential to miss the diagnosis of WD. AS S E S S M E N T O F C OP P E R D E P OS I T S B Y L I GH T M I C ROSC OP Y
Copper staining may be helpful but tends to be patchy. It is typically seen in the hepatocytes but can also be present in portal macrophages and Kupffer cells. In cirrhotic livers, many nodules may have no copper, whereas it may be abundant in others. In chronic obstructive liver disease, the copper is seen in periportal hepatocytes; this pattern can usually be distinguished from that of WD. Commonly used stains for identifying copper deposits include orcein, rhodanine, rubeanic acid, and crystal violet. The intensity of copper staining by histochemical stains may not correlate with the dry weight copper results. This may be related to the patchy distribution or inability of histochemical stains to highlight all the bound or unbound hepatic copper. A N T I C O P P E R D RU G T R E AT M E N T
Several agents have been used for the treatment of WD (8). Zinc is the most recently FDA-approved drug for treatment of WD. Zinc acts by induction of intestinal cell metallothionein, which has a high affinity for copper. Once induced, it binds copper from food and endogenous secretions such as in saliva, gastric, and intestinal juices and holds it within the intestinal cell, which sloughs into the lumen of the bowel at a 6-day
DISEASE
turnover rate, taking the copper with it for excretion in the stool. Zinc thus produces a mucosal block of copper absorption. Because of the substantial secretion of copper in saliva and gastric juices, this mechanism produces a negative copper balance and produces a sustained loss of copper. A recent longterm study of zinc monotherapy demonstrated better response in patients with neurologic disease and a less satisfactory response in patients with liver disease, perhaps due to the liver dysfunction in the latter group of patients (9). Other agents are available as well (10). Penicillamine was the first orally effective drug developed as a chelator for copper, but it is more toxic than zinc therapy, and is now mostly limited to use in patients with acute, relatively severe disease when one is trying to avoid liver transplantation. Toxicity includes hypersensitivity reactions, suppression of bone marrow, proteinuria, development of autoimmune disorders such as lupus or Goodpasture syndrome, reduced immune response resulting in infections, skin disorders such as elsastosis perforans serpiginosa, and collagen disorders such as facial wrinkling. Trientine is also a chelator, and is thought to be safer than penicillamine. Tetrathiomolybdate is an anticopper agent that prevents absorption of dietary and endogenously secreted copper.
References 1. Brewer GJ, Fink JK, Hedera P. Diagnosis and treatment of Wilson’s disease. Sem Neurol. 1999;19:261–270. 2. Ferenci P. Wilson’s Disease. Clin Gastroenterol Hepatol. 2005;3(8): 726–733. 3. Ala A, Borjigin J, Rochwarger A, Schilsky M. Wilson disease in septuagenarian siblings: raising the bar for diagnosis. Hepatology. 2005;41: 668–670. 4. Stromeyer FW, Ishak KG. Histology of the liver in Wilson’s disease: a study of 34 cases. Am J Clin Pathol. 1980;73:12–24. 5. Portmann B, Thompson R, Roberts E, Paterson A. Genetic and metabolic liver disease. In: Burt AD, Portmann BC, Ferrell LD, eds. MacSween’s Pathology of the Liver. 5th ed. Edinburgh, UK: Elsevier; 2007:249–254. 6. Scott J. Wilson’s disease presenting as chronic active hepatitis. Gastroenterol. 1978;74:645–651. 7. Davies SE, Williams R, Portmann B. Hepatic morphology and histochemistry of Wilson’s disease presenting as fulminant hepatic failure: a study of 11 cases. Histopath. 1989;15:385–394. 8. Aaseth J, Flaten TP, Anderson O. Hereditary iron and copper deposition: diagnostics, pathogenesis, and therapeutics. Scand J Gastroenterol. 2007;42:673–681. 9. Linn FH, Houwen RH, van Hattum J, van der Kleij S, van Erpecum KJ. Long-term exclusive zinc monotherapy in symptomatic Wilson disease: experience in 17 patients. Hepatology. 2009;50:1442–1452. 10. Brewer GJ, Askari F, Dick RB, et al. Treatment of Wilson’s disease with tetrathiomolybdate: V. Control of free copper by tetrathiomolybdate and a comparison with trientine. Transl Res. 2009;154:70–77.
Case 21.1
Fulminant Form of Wilson Disease LINDA D. FERRELL
C L I N IC AL I N F OR M AT I ON
A 19-year-old woman was referred for fulminant liver failure of unknown etiology. The patient had been well until about 2 weeks prior to her hospital admission when she first experienced vomiting, fatigue, and dark-colored urine. Over a period of several days she became more lethargic and experienced 2 seizures with loss of consciousness. Viral serologies were negative, and there was no history of toxic exposure or acetaminophen use. There were no KF rings or neurologic involvement. Studies for quantitative copper on the liver were then submitted as well, which revealed a value of 711 μg/g (normal 10–15 μg/g). The serum copper values received after her demise were 394 mg/dL (normal 75–150 mg/dL) and serum ceruloplasmin 11 mg/dL (normal 18–45 mg/dL). The patient soon developed sepsis followed by cardiac arrest and death within 24 hours of admission. R E A SON F OR R E F E R R AL
F I G U R E 2 1 . 1 . 2 Trichrome stain of liver at lower magnification shows pale zones of necrosis consistent with necrosis surrounding regenerative nodules. The darker blue areas are residual portal zones and central veins.
To determine the etiology of liver failure of unknown etiology. PAT H OL OG I C F E AT U R E S
The liver at autopsy was smaller than normal and had a grossly nodular appearance (Figure 21.1.1) mimicking cirrhosis. On microscopy, the liver parenchyma demonstrated numerous regenerative nodules with mildly thickened hepatocyte plates (Figure 21.1.2). The nodules were separated by loose connective tissue stroma admixed with prominent ductular reaction, which is likely to be ductular transformation of hepatocytes to a more duct-like phenotype due to the extensive injury. Necrosis was prominent both in central and periportal areas (Figures 21.1.3–21.1.5). Mild to moderate inflammatory infiltrates were present throughout the parenchyma and portal F I G U R E 2 1 . 1 . 3 Extensive necrosis near a larger hepatic (central) vein. Note the loose back ground tissue in the areas of hepatocyte dropout and necrosis. A small regenerative nodule is present (center) with thicker hepatic plates and acinar change and minimal fatty change. Ductular reaction is present in necrotic zones, associated with a mostly lymphocytic infiltrate and congestion. Lymphocytic infiltrates are present within the large vein consistent with a venulitis.
FIGURE 21. 1. 1 The liver has a nodular appearance secondary to regenerative changes (not cirrhosis).
areas. Trichrome stain confirmed that these ductular areas were not associated with dense scarring of established cirrhosis but instead showed the two-tone pattern of staining seen after necrosis (Figures 21.1.2, 21.1.6 and 21.1.7). In contrast, residual portal zones and central veins demonstrated intense darker
301
302
CHAPTER
21:
WILSON
FIGURE 21. 1. 4 Smaller hepatic (central) vein and perivenular zone
with extensive hepatocyte dropout, ductular reaction (transformation of hepatocytes to ductular pattern), mild inflammatory infiltrates, and a few residual hepatocytes with minimal fatty change.
FIGURE 21. 1. 5 Small portal zone with duct and artery containing mild mononuclear infiltrates. Surrounding areas show extensive hepatocyte dropout, with congestion and ductular reaction (transformation of hepatocytes to ductular pattern).
blue staining of established collagen fibers (Figures 21.1.6 and 21.1.7). Reticulin stain also demonstrated collapsed framework of subacute severe hepatitis rather than the more dense organized scar of cirrhosis (Figure 21.1.8). Copper stain (rubeanic acid) demonstrated patchy copper deposits, both in hepatocytes and Kupffer cells within the necrotic zones.
D I AG N OS I S
Severe hepatitis with massive necrosis, consistent with Wilson disease.
DISEASE
FIGURE 21.1.6 Trichrome stain at higher magnification demonstrates regenerative nodule on the left and extensive necrosis to the right of the residual portal zone (center). Note the two-tone staining pattern: established collagen in the portal zone is dark blue, and the necrosis is pale blue. Ductular reaction is prominent in necrotic zones.
F I G U R E 2 1 . 1 . 7 Trichrome stain demonstrates zones of previous hepatocyte necrosis and regenerative hepatocytes. Note the widened hepatic plates in the regenerative nodule (right). A residual central vein (upper center) and portal zone (lower center) are highlighted by the darker blue staining of the collagen.
DISCUSSIO N
Severe acute hepatitic presentation of WD can be confused with other more common etiologies, such as acute viral hepatitis, autoimmune hepatitis (AIH), idiosyncratic drug reactions, and idiopathic forms of acute severe hepatitis, the latter accounting for about 15% of acute hepatitis that progresses to liver failure. Although serum ceruloplasmin and copper staining were helpful in this case, serum ceruloplasmin may be elevated as an acute phase reactant, and copper stain may be negative due to extensive necrosis. The clinical observation of hemolysis can be key in alerting to the possibility of WD.
CASE
21.1
:
FULMINANT
FORM
OF
WILSON
DISEASE
303
A carefully done trichrome stain can help to differentiate the necrotic areas from background fibrosis, the latter evidence for previous chronic disease, a feature not seen in typical idiosyncratic drug reactions, hepatitis A, or idiopathic forms of acute hepatitis. Autoantibodies including antinuclear antibody (ANA), smooth muscle antibody (SMA), liver kidney microsomal-1 (LKM-1), and soluble liver antigen (SLA) are necessary to evaluate for AIH. A careful history is necessary to exclude unusual drug reactions, including any history of exposure to over-the-counter drugs, herbal agents, or other nutritional supplements, including ones used sporadically or that have recently been changed to a new “brand.” Travel history should be sought to exclude the possibility of hepatitis E.
FIGURE 21. 1. 8 Reticulin stain shows collapsed framework consist-
ent with acute/subacute process.
Case 21.2
Chronic Hepatitis Due to Wilson Disease LINDA D. FERRELL
C L I N I C AL I N F OR M AT I ON
The patient described in Case 21.1 also had an identical twin sister, who was in a “normal state of health” at the time her twin presented with liver failure. However, after her twin’s death, the sister was also examined, was found to have mildly elevated transaminases, and liver biopsy was performed. Quantitative copper on that sample revealed a value of 812 μg/g. Serum ceruloplasmin was very low at 4 mg/dL.
interface activity (Figure 21.2.1). Fatty change of mild to moderate degree was also present, but there were no ballooned hepatocytes, significant inflammation, or centrizonal fibrosis to suggest steatohepatitis.
DIAGNO SIS
Mild chronic hepatitis, and early fibrosis (stage 1–2), consistent with early stage Wilson disease.
R E A S ON F OR R E F E R R A L
Evaluation for possible Wilson disease. PAT H OL OG I C F E AT U R E S
The twin sister’s liver biopsy showed mild chronic hepatitislike pattern with mild portal mononuclear infiltrate and mild
DISCUSSIO N
This case demonstrates the variability in presentation that can occur in WD, even in identical twins, where one would presume that the genetic defect is the same. Thus, there may be environmental factors or other modulators that influence the clinical presentation. The fatty change in this case can be mistaken for fatty liver disease (FLD); however, the patient did not have typical risk factors. Given the rising incidence of childhood obesity, however, it may become more difficult to use the clinical parameters alone to exclude FLD as etiology for fatty change, and many recommend the routine evaluation for WD using 1 or more of the recommended screening tests like serum ceruloplasmin, 24-hour urine copper, and/or quantitative copper values on liver tissue to help exclude this possibility.
FIGURE 21.2.1. Twin sister, liver biopsy. Note the mild portal hepati-
tis with interface activity and periportal fibrosis, as well as fatty change.
304
Case 21.3
Cirrhosis With Chronic Hepatitis Consistent With Wilson Disease LINDA D. FERRELL
C L I N IC A L P R E SE N TAT I ON
A 24-year-old woman presented with fatigue and was found to have elevated liver transaminases. Further workup revealed low serum ceruloplasmin and markedly elevated 24-hour urine copper, leading to diagnosis of WD. Over the next 2 years, the patient’s condition deteriorated, and it was found that she had not been compliant with her therapy for copper reduction. Transaminases at this time were elevated to 3 times normal, alkaline phosphatase 2 times normal, and bilirubin was over 30 mg/dL. She received liver transplantation. R E A SON F OR R E F E R R AL
To confirm the suspicion of WD and undergo transplantation for end-stage liver disease. PAT H OL OG I C F E AT U R E S
F I G U R E 2 1 . 3 . 2 Glycogenated nuclei can be seen in Wilson disease in any stage and were notable within hepatocytes in cirrhotic nodules in this case.
This case demonstrates end-stage cirrhosis due to WD (Figure 21.3.1). Typical features of WD are present, including copper deposits and glycogenated nuclei (Figure 21.3.2), Mallory-Denk bodies (Figures 21.3.3 and 21.3.4), and chronic inflammatory infiltrates with a chronic hepatitis pattern of injury (Figure 21.3.3). Copper stain (rubeanic acid) was focally positive in isolated nodules. Rubeanic acid stains the copper as dark green granules in hepatocytes (Figure 21.3.5).
F I G U R E 2 1 . 3 . 3 Giant (multinucleate) hepatocytes, moderate chronic inflammation, and mild focal fatty change are also present. Mallory-Denk bodies are also present (large cell bottom center).
DIAGNO SIS FIGURE 21. 3. 1 The cirrhosis of Wilson disease may have variable
amounts of inflammation and ductular reaction. Note mild degree of scattered fatty change.
305
Cirrhosis with moderate activity consistent with Wilson disease.
306
CHAPTER
21:
WILSON
DISEASE
F I G U R E 2 1 . 3 . 5 Rubeanic acid copper stain demonstrated patchy, focal positivity (green granules) in hepatocytes and some Kupffer cells. 21. 3. 4 Mallory-Denk bodies as darker eosinophilic, ropy, cytoplamic inclusions were present in some of the damaged hepatocytes.
FIGURE
D I SC U SSI ON
This case demonstrates chronic hepatitic form of WD with established cirrhosis and illustrates some of the helpful features of late stage WD, including the Mallory-Denk bodies and copper deposits. In this age group, other causes of chronic hepatitis such as AIH, viral hepatitis B or C, and alpha-1-antitrypsin
(AAT) deficiency should be considered. Viral and autoimmune markers, and periodic acid–Schiff diastase (PASd) and/ or immunostain for AAT can help exclude these etiologies. The pattern of portal-based inflammation with activity of this extent, in combination with the absence of ballooned hepatocytes, pericellular scarring, and clinical risk factors of metabolic syndrome or alcohol help to exclude late-stage chronic steatohepatitis.
22 Liver Transplant Pathology OYEDELE ADEYI
Although attempts to develop liver replacement regimens similar to renal dialysis in end-stage renal disease patients are ongoing, liver transplant remains the only treatment for endstage liver disease. Liver failure following chronic diseases or resulting from an acute fulminant injury are the usual indications for transplant. There are usually well-defined selection criteria that employ extensive clinical and psychosocial evaluations to confirm appropriateness of such a procedure. Most institutions have locally approved selection criteria for recruiting patients into the transplant system, often fashioned after the Milan or some other (eg, UCSF) criteria (1,2). In a field hampered by severe organ shortages, these criteria are useful for managing organ allocation and ensuring patients with measurable benefits are allocated the available resources. However, as summarized in Table 22.1, pretransplant selection is only one of several factors affecting outcome in transplanted patients (3). In broad terms other factors include operative complications, immunologic injury, infections, recurrent disease, neoplasms, and adverse reaction to medication. It is in managing the impact of these factors that the pathologist plays an important role. The scope of liver allograft pathology has recently been reviewed (4). Pathologists are called upon to guide treatment decisions by recognizing and grading the severity of the pathologic process(es), differentiating between 2 conditions requiring different therapies with potentially damaging outcome if incorrectly classified, determining the extent of injury by staging, as well as determining the rate of progression of a previously documented injury. Sometimes new diseases are identified in an allograft biopsy that call for new clinical parameters to be investigated and/or the course and type of immunosuppression to be altered. Recognizing morphologic pattern (or patterns when 2 or more processes overlap) is as important in liver allograft
biopsy as it is in the nontransplant biopsy. Interpretation of liver allograft biopsy therefore requires the understanding of nontransplant liver pathology as both involve pattern recognition, understanding of underlying clinical context, and ability to correlate morphology with clinical scenario. There are, however, certain instances that are unique to liver allografts and pose diagnostic dilemmas, especially when the contending differential diagnoses require diametrically opposing interventions, (eg, cellular rejection versus recurrent viral hepatitis). Some other sources of dilemma are: 1. Poor understanding of pathogenesis and/or natural history of a specific pattern of injury (eg, plasma-cell-rich hepatitis/zone 3 perivenulitis) 2. Defining the more prominent entity when 2 or more truly exist in the same biopsy 3. Determining the cause of cholestatic injury especially late in the graft when chronic rejection, stricture, and recurrent sclerosing cholangitis are in contention Figure 22.1 summarizes the timeline when some of the discussed problems encountered in liver transplant patients are likely to present and helps to understand the importance of overlaps. The aim of this chapter is to illustrate selected problem areas with clinical examples and suggest approaches to resolving these dilemmas. The first 2 examples are, however, not much of a dilemma but serve as the basis for understanding and discussing the subsequently illustrated problems.
TA B LE 22. 1 Factors affecting outcome in liver transplant recipients Donor factors
Recipient Factors
Management Factors
Age
Primary disease
Organ quality
Rejection episodes
Immunosuppression regimen
ABO blood group
ABO blood group
Surgery type
Cardiac death versus brain death
Immediate pretransplant clinical status, eg, fulminant failure
Surgeon
Warm and cold ischemia time
Neoplasms
Medications, eg, Interferon, Septra F I G U R E 2 2 . 1 This chart shows the timeline when the commonly encountered causes of liver allograft dysfunction are likely to emerge.
307
308
CHAPTER
22:
LIVER
TRANSPLANT
References 1. De Carlis L, Belli LS, Romani F, et al. Selection criteria for liver transplantation: preliminary experience of Niguarda Hospital, Milan. Transplant Proc. 1989;21(1, pt 2):2415–2416. 2. Yao FY, Ferrell L, Bass NM, et al. Liver transplantation for hepatocellular carcinoma: expansion of the tumor size limits does not adversely impact survival. Hepatology. 2001;33(6):1394–1403.
PAT H O L O G Y
3. Busuttil RW, Farmer DG, Yersiz H, et al. Analysis of long-term outcomes of 3200 liver transplantations over two decades: a single-center experience. Ann Surg. 2005;241(6):905–916, discussion 916–908. 4. Adeyi O, Fischer SE, Guindi M. Liver allograft pathology: Approach to interpretation of needle biopsies with clinico-pathological correlation. J Clin Pathol. 2010;63(1):47–74.
Case 22.1
Acute Cellular Rejection OYEDELE ADEYI
4. Virtually quiescent lobule between zone 1 and zone 3 indicating that the inflammatory targets are in the portal area and hepatic vein endothelium (Figures 22.1.1 and 22.1.3).
C L I N IC AL I N F OR M AT I ON
A 52-year-old man transplanted for hepatitis C–induced endstage liver disease 2 weeks prior to biopsy. He achieved postoperative troughs of alanine aminotransferase (ALT) 35 U/L, aspartate transaminase (AST) 33 U/L, alkaline phosphatase (ALP) 115 U/L by 1 week posttransplant but now shows new rise in ALT to 97, AST 88, ALP 150; bilirubin remains unchanged.
DIAGNO SIS
Acute cellular rejection, rejection activity index (RAI) 7/9.
R E A SON F OR R E F E R R AL
Looks like rejection, but interface and parenchymal inflammation are present; is there hepatitis C as well? PAT H OL OG I C F E AT U R E S
Adequate needle biopsy shows inflammatory infiltrates restricted to portal tracts and perivenular areas of zone 3 (Figure 22.1.1). Close-up views show characteristic features of acute cellular rejection (ACR), namely: 1. Mixed infiltrates including lymphocytes (with activated/ blast forms), neutrophils, eosinophils, and few plasma cells. The infiltrates expand portal tracts and spill over into the hepatic plate that is interface activity (Figure 22.1.2) 2. Duct injury with inflammation within epithelial basement membrane (Figure 22.1.3) 3. Portal and hepatic vein phlebitis, characterized by inflammatory cells in the subendothelial space, resulting in lifting of the endothelial cells; the inflammation, though not always, but as exemplified by this case of severe rejection, has spilled into the perivenular lobule in zone 3 resulting in hepatocellular necrosis (Figure 22.1.4).
FIGURE 22. 1. 1 Acute cellular rejection low power view showing
portal and central distribution of inflammation. (H&E 20 ).
F I G U R E 2 2 . 1 . 2 Close-up view of two portal tracts from Figure 22.1 showing mixed infiltrate that includes activated lymphocytes, eosinophils, and neutrophils.
F I G U R E 2 2 . 1 . 3 Inflammation of bile duct is accompanied by evidence of duct epithelial injury, including irregular spacing and loss of polarity (H&E 200 ).
309
310
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
TA BL E 2 2 . 1 . 1 Banff 1997 criteria for acute cellular rejection
in liver allografts Parameter Scored Portal inflammation (PI)
Bile duct injury/ inflammation FIGURE 22. 1. 4 Acute cellular rejection: a close-up view of one of
the terminal hepatic venules from Figure 22.1.1 shows phlebitis and perivenular inflammation; the latter results in hepatocellular necrosis around the venule and justifies a score of 3 out of 3 for phlebitis on the Banff scale (H&E 200 ). D I SC U SSI ON
This case highlights the characteristic features of ACR: it occurs within the first few weeks posttransplant, shows a steep rise in enzymes, the enzyme pattern could be hepatitic, cholestatic, or (as in this case) mixed. The histological features of duct, endothelial, and sometimes parenchymal injury help in understanding the reason for the range in the pattern of enzyme elevation and underlie the Banff scoring system for RAI used in assessing rejection severity (1). ACR is an immunologic injury implying there are specific targets in the liver. These targets are bile duct epithelium and vascular endothelium. Hence, the infiltrates, which are often mixed comprising small and activated lymphocytes, eosinophils, and neutrophils, are typically centered where these targets are located, that is portal tract and/or zone 3 around hepatic venules. When severe, portal and zone 3 infiltrates could spill into the zonal hepatocytes and should not be interpreted as hepatitis. ACR (with some exceptions discussed later) should not have significant lobular infiltrate away from the immediate vicinity of the portal tract or hepatic venule. When lobular inflammation is present, it should raise the questions of alternative or overlapping cause(s) of graft injury. Late-occurring, “atypical” cellular rejection is the exception, and this is illustrated later in this chapter. The 3 parameters graded in ACR are illustrated in Figures 22.1.1 and 22.1.4 and include portal infiltrates (PI), bile duct injury (DI), and subendothelial inflammation/endothelial injury (EI) (aka phlebitis, endotheliitis). Each of these parameters is scored using the Banff criteria summarized in Table 22.1.1
Venous phlebitis/ endothelial injury (EI)
Total (RAI)
Criteria
RAI Score
Inflammation in minority of portal tracts not expanding and mostly lymphocytic
1
Mixed lymphocytic inflammation in majority or all portal tracts, and expanding portal tracts
2
Mixed lymphocytic inflammation in majority or all portal tracts and expanding portal tracts with spillover to interface/periportal hepatocytes
3
Bile duct epithelium inflammation/mild injury in minority of portal tracts
1
PI of 2 or 3 with marked evidence of epithelial injury in few ducts
2
PI of 2 or 3 with marked evidence of epithelial injury in most ducts ⴙ/− outright duct necrosis in some ducts
3
Subendothelial lymphocytes in some but not the majority of portal and/or hepatic venules
1
Subendothelial lymphocytes in most portal and/or hepatic venules
2
EI of 2 with perivenular inflammation and hepatocellular dropout necrosis
3
Sum of scores in each of the three parameters
0–9
Abbreviation: RAI, rejection activity index.
to obtain an RAI of 0 to 3, which when added up would give a total RAI on a scale of 0 to 9. The example presented here was given a total score of 7 (PI 3 DI 1 EI 3). Few lymphocytes in the bile duct epithelium do not necessarily imply duct injury, as this is frequently seen in nonrejection-related portal inflammation including hepatitis C. Features for recognizing duct injury include inflammation as well as any or all of the following: increased nucleocytoplasmic ratio, irregular spacing in epithelium, cytoplasmic vacuolization, disordered polarity, or outright necrosis. Also, the occasional subendothelial lymphocyte could be seen outside of rejection; in ACR, subendothelial inflammation/phlebitis is characterized by endothelial lifting with or without reactive appearance to the nucleus or perivenular inflammation (Figure 22.1.4).
Reference 1. Banff schema for grading liver allograft rejection: an international consensus document. Hepatology. 1997;25(3):658–663.
Case 22.2
Recurrent Hepatitis C OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
This 57-year-old man was transplanted 17 months ago for hepatitis C–related end-stage liver disease. Baseline enzyme levels at 1 year posttransplant were ALT 33–40 U/L, AST 33 U/L, ALP 90–105 U/L, Bilirubin less than 1 mg/dL (17.1 μmol/L). In the last 3 to 6 months there has been a gradual rise in enzymes, and now enzyme levels are ALT 75, AST 69, ALP 100, and bilirubin is normal. R E A SON F OR R E F E R R AL
There is hepatitis but also some lymphocytes in the subendothelial space; is there some rejection as well? PAT H OL OG I C F E AT U R E S
Adequate biopsy shows mostly lymphocytic inflammation, but also with (very) few other cells that include neutrophils and plasma cells. The infiltrates are located mostly in the portal tracts but also patchily within the lobule, the latter associated with small foci of hepatocellular necrosis and occasional apoptoses (Figures 22.2.1 and 22.2.3). Although a portal vein shows occasional subendothelial lymphocytes, it is minimal and not associated with true phlebitis or endothelial injury.
F I G U R E 2 2 . 2 . 2 Higher magnification of one of the portal tracts in figure 22.2.1 showing predominantly lymphocytic infiltrates, and although the infiltrate extends to the subendothelium of the portal vein, the overall features do not support cellular rejection. Such changes are not infrequent in viral hepatitis, including in the non-transplant population (H&E 100 ).
FIGURE 22. 2. 1 Recurrent viral hepatitis C in a recently transplanted
F I G U R E 2 2 . 2 . 3 The lobule is mildly active as seen here, with clusters of lymphocytes and few apoptotic hepatocytes; hepatitis is mild going by the degree of lobular changes (H&E 200 ).
patient is illustrated at low magnification. There is predominantly portal inflammation but also scattered cells in the lobules (H&E 50 ).
311
312
CHAPTER
22:
LIVER
TRANSPLANT
D I AG N OS I S
Recurrent viral hepatitis C posttransplant, mild activity; fibrosis stage 0 (Metavir); negative for ACR.
D I SC U SSI ON
This case illustrates histopathological features of recurrent viral hepatitis C posttransplant. Infection with hepatitis C virus (HCV) remains a major cause of hepatic and extrahepatic morbidity (1,2). Differentiating recurrent hepatitis C from ACR is therefore a common, and unfortunately sometimes difficult, task the pathologist is confronted with (3). In this case, however, the features despite the subendothelial lymphocytes are purely of hepatitis and consistent with recurrent HCV. Features indicative of hepatitis and not ACR are the mostly monotonous nature of infiltrates, the absence of duct epithelial or endothelial injury, and presence of lobular inflammation with single cell necrosis and apoptosis. These features are typical of HCV similar to what could be seen in the nontransplant setting, including the “passenger” lymphocytes in the subendothelial space. Occasional bile duct lymphocytes follow the same rule and are sometimes present in HCV biopsies but with no evidence of epithelial injury as described in the first case.
PAT H O L O G Y
Understanding the implication of making a diagnosis of ACR in HCV helps in forming a pathologic diagnosis. Transplant hepatologists are likely to employ steroid therapy or increase other forms of immunosuppression if the pathologist suggests there is significant ACR: an action that could be damaging to the graft (4). The features of rejection in the context of HCV patients should therefore be prominent and convincing to make it the primary diagnosis in a biopsy. In cases where ACR is being considered as overlapping with what is otherwise a hepatitic process, but in which the features suggestive of ACR are mild and less than RAI of 3, HCV should always be favored as the primary process. Nevertheless, in these cases the features indicative of coexisting (or indeterminate for) ACR should still be mentioned in the report to assist clinicians in making a decision on immunosuppression adjustment as part of recurrent HCV management.
References 1. Brown RS Jr, Gaglio PJ. Scope of worldwide hepatitis C problem. Liver Transpl. 2003;9(11):S10–S13. 2. Adeyi OA. Vascular and glomerular manifestations of viral hepatitis B and C: a review. Semin Diagn Pathol. 2009;26(2):116–121. 3. Demetris AJ, Eghtesad B, Marcos A, et al. Recurrent hepatitis C in liver allografts: prospective assessment of diagnostic accuracy, identification of pitfalls, and observations about pathogenesis. Am J Surg Pathol. 2004;28(5):658–669. 4. McCaughan GW, Zekry A. Impact of immunosuppression on immunopathogenesis of liver damage in hepatitis C virus-infected recipients following liver transplantation. Liver Transpl. 2003;9(11):S21–S27.
Case 22.3
Acute Cellular Rejection Versus Recurrent Hepatitis C OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
A 48-year-old man received a living donor liver 9 months prior due to hepatitis C cirrhosis and hepatocellular carcinoma. Since his last blood tests 5 weeks earlier, his liver enzymes have increased from ALT 65 to 112 U/L, AST 60 to 107 U/L, ALP 116 to 122 U/L, bilirubin is unchanged at 1.2 mg/dL (20.5 μmol/L). He is on cyclosporin (CyA) immunosuppression; he admitted missing “a few doses” a week ago but has not missed any since; serum level of CyA is within acceptable range. A prior biopsy performed at 6 months (ie, 3 months before the current biopsy) had shown recurrent viral hepatitis C. R E A SON F OR R E F E R R AL
There is recurrent viral hepatitis C, but this biopsy appears different from prior biopsy; is there rejection? F I G U R E 2 2 . 3 . 2 Higher magnification of Figure 22.3.1 shows
inflammation is both portal and lobular (H&E 50 ).
PAT H OL OG I C F E AT U R E S
There is a dense infiltrate on low power that at first appears limited to the portal tracts (Figure 22.3.1). Closer view, however, shows active lobular infiltrates with apoptotic hepatocytes (Figures 22.3.2 and 22.3.3). The portal infiltrate varies, being predominantly lymphocytic in some (Figure 22.3.2) and mixed in others (Figure 22.3.4) where the mixture includes activated lymphocytes and eosinophils. Duct injury and subendothelial inflammation is present (Figure 22.3.4).
F I G U R E 2 2 . 3 . 3 Active inflammation in the lobule typical of hepatitis is seen. This portion of the biopsy recapitulates the findings in an earlier biopsy in this patient (H&E 100 ).
FIGURE 22. 3. 1 Low power view shows what appears to be inflammation restricted to portal tracts in this case of co-existing cellular rejection and recurrent viral hepatitis C, but as one looks more closely (see Figures 22.3.2 and 22.3.3) lobular inflammation is also appreciated (H&E 25 ).
313
DIAGNO SIS
Acute cellular rejection superimposed on recurrent hepatitis C (coexisting ACR and HCV).
314
CHAPTER
22:
LIVER
TRANSPLANT
FIGURE 22. 3. 4 Two of the portal tracts shown in Figure 22.3.1 are
illustrated showing a mixed pattern to the inflammation as well as injury to bile ducts and portal vein (H&E; upper panel 100 ; lower panel 200 ). D I SC U SSI ON
Although differentiating viral hepatitis C from ACR could be challenging in liver allograft biopsy, this case illustrates a not unusual occurrence of both entities in the same patient. Hepatitis with active lobular inflammation is usually not due to ACR, although, as illustrated later, this could occur in some late rejections. In this case, the patient had a prior
PAT H O L O G Y
biopsy documenting recurrent HCV without features of ACR. Comparing this biopsy with the previous one, there is a clear difference in that though hepatitis persists, the infiltrates in some portal tracts are heavier, more mixed, and associated with convincing duct epithelial injury. Two helpful practices in reviewing liver allograft biopsies are also highlighted by this case: one is to incorporate information gleaned from previous biopsies when available; the other is to consider the clinical scenario carefully, including the pattern of changes in liver enzymes, for example sudden and steep versus slow but persistent rise. Most ACR occur within 1 to 3 months posttransplant and rarely beyond the first year (see Figure 22.1.1 in Case 22.1). Cellular rejection beyond this period is usually (though not always) in the context of a recognizable factor, such as poor compliance and/or therapeutic dose reduction due to infections or neoplasms, such as posttransplant lymphoproliferative disease (PTLD) (1–3). The history of skipped CyA at 9 months posttransplant is therefore very important in this patient, although having resumed the serum levels were at therapeutic levels at presentation. As in nontransplant liver biopsies, pattern recognition is essential as it prompts one to recognize when 2 or more overlapping patterns coexist. When this is the case, the pathologist should try as much as possible to communicate to the clinician which entity is the more prominent. In this case, despite the risks associated with increased immunosuppression in the context of HCV, the pathologist has no choice than to communicate the obvious rejection features in this case. Because some of the portal infiltrate is clearly due to HCV, appropriately applying the Banff score could be difficult and potentially misleading. My approach is not to give a score but to have a verbal discussion with the hepatologist and give a descriptive comment on the relative significance of each lesion, that is bile duct injury; phlebitis, especially when associated with perivenular necrosis; and underlying hepatitis. In this case, both ACR and HCV are mild and communicated as such. The patient’s rejection resolved without any aggressive treatment, and he was encouraged to not “skip a few doses” again.
References 1. Wiesner RH. Advances in diagnosis, prevention, and management of hepatic allograft rejection. Clin Chem. 1994;40(11, pt 2): 2174–2185. 2. Anand AC, Hubscher SG, Gunson BK, McMaster P, Neuberger JM. Timing, significance, and prognosis of late acute liver allograft rejection. Transplantation. 1995;60(10):1098–1103. 3. D’Antiga L, Dhawan A, Portmann B, et al. Late cellular rejection in paediatric liver transplantation: aetiology and outcome. Transplantation. 2002;73(1):80–84.
Case 22.4
Late Cellular Rejection Versus Autoimmune Hepatitis Versus Recurrent Hepatitis C OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
The best way to illustrate this problem area is to present 4 patients with different background yet comparable histology. Patient I: A 62-year-old man transplanted 1 year earlier for hepatitis C cirrhosis had protocol biopsy with corresponding ALT 44 U/L; AST 52 U/L; ALP 140 U/L; normal bilirubin (Figures 22.4.1A; 22.4.2A; 22.4.3A). Patient II: A 43-year-old woman transplanted 5 years earlier for cirrhosis due to autoimmune hepatitis (AIH); her
baseline enzymes were ALT 13, AST 24, ALP 105, but these suddenly went up to ALT 288 U/L, AST 489 U/L, ALP 275 U/L, with normal bilirubin, prompting biopsy. This represents a third episode with similar presentation since transplant, each previous episode having responded well to steroid therapies (Figures 22.4.1B, 22.4.2B, and 22.4.3B). Patient III: A 19-year-old woman transplanted as a child for biliary atresia was biopsied because her ALT went up from a baseline of 33 to 261 U/L, AST from 26 to 168 U/L; ALP remained unchanged at 62 U/L (Figures 22.4.1C, 22.4.2C, and 22.4.3C).
FIGURE 22. 4. 1 Low power views of biopsies from four different patients with different background but similar histopathologic features that include hepatitic pattern of injury with portal and zone 3 accentuation of inflammation. Panels A–D, respectively, from patients I, II, III, and IV; (H&E 50 .)
315
316
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
v FIGURE 22. 4. 2 Higher magnification of representative portal tract from each of patients I, II, II, & IV (panels A–D, respectively) shows the infiltrate to contain many plasma cells, although to variable degree. In addition panel D from patient IV shows duct injury typical of acute cellular rejection (blue arrow), a feature not seen in panels A–C (H&E 100 ).
Patient IV: A 55-year-old man transplanted for alcohol cirrhosis 4 years earlier; he recently had immunosuppression (Tacrolimus-based) reduced because of an infection; subsequently, ALT and AST rose 2 to 3 times from his baseline of 45–52 U/L (Figures 22.4.1D, 22.4.2D, and 22.4.3D).
cell populations, are otherwise less mixed compared with typical cellular rejection discussed earlier. Only patient IV has associated bile duct epithelial injury (Figure 22.4.2D; blue arrow).
R E A S ON F OR R E F E R R A L
Biopsy in patient I was performed as a surveillance/protocol biopsy; the rest were performed to diagnose a cause for elevated liver enzymes, which had happened rather suddenly. PAT H OL OG I C F E AT U R E S
These 4 cases have hepatitic features including portal and lobular inflammation, as well as marked accentuation of inflammation around the hepatic veins and venules, resulting in perivenular hepatocellular dropout necroses. The infiltrates, although to varying degrees have significant plasma
DIAGNO SIS
For reasons discussed below, the respective diagnoses rendered were as follows: Patient I: hepatitis with zone 3 accentuation of inflammation, consistent with recurrent viral hepatitis C. Patient II: recurrent AIH. Patient III:immune-mediated injury—hepatitic/“atypical” acute cellular rejection versus de novo AIH (see comment). Patient IV: cellular rejection with hepatitic features, late occurring.
CASE
22.4:
POSTTRANSPLANT
H E PAT I T I S
317
FIGURE 22. 4. 3 The accentuation of inflammation in zone 3 around the central vein (cv) with hepatic vein phlebitis is seen in patients I–IV (panels A–D, respectively) (H&E 100 ).
D I S C U S S I ON
This area of liver allograft pathology remains a major topic for discussion, especially in more recent times. This pattern of injury has been given various names depending on the context, presumed pathogenesis, or personal preference. This appearance is similar to the features seen in AIH in the nontransplant setting, and, therefore, many experts believe this pattern has an immune-mediated basis. The fact that it has been reported as a feature of recurrent hepatitis C virus (HCV) supports its immunologically mediated nature of the injury, but also raises the problem of autoimmunity versus alloimmunity. The difficulty arising from these issues is significant for 2 main reasons: 1. What should be the right response-increased immunosuppression versus reduction in immunosuppression? 2. What is the natural history and prognosis? As illustrated in the 4 cases presented, this injury is best defined in 2 broad categories: viral hepatitis (infective) versus
nonviral hepatitis (noninfective). Whether or not the nonviral hepatitis causes represent late-onset cellular rejection, de novo, or recurrent AIH is important but not overly critical in terms of clinical intervention. Even so some experts would argue against the term “autoimmune” since there was no autoantigen in the allograft per se. It is becoming clear therefore that (at least some if not all of) these cases represent a form of rejection without significant duct injury and have been variously named hepatitic variant of ACR, plasma cell-rich ACR, late-occurring or atypical rejection with zone 3 perivenulitis, and so on. These descriptive terms highlight their differences from the typical, early occurring ACR(1,2). Apart from the morphological differences with typical ACR, these late occurring rejections are also more likely to be resistant to treatment and have a propensity to predict and/or progress to chronic rejection, fibrosis, and graft loss (2–4). The presence of these injuries in an HCV patient raises other problems that could hardly be solved on morphological grounds alone but should be presented as a possibility in
318
CHAPTER
22:
LIVER
TRANSPLANT
the appropriate clinical context. For example, the biopsy from patient I was read as being consistent with recurrent HCV after correlation with viral load, the slow rise in enzymes, the timing posttransplant when recurrent HCV is the rule rather the exception, and absence of suggestive autoimmune serum markers. In summary, hepatitis with zone 3 inflammation (perivenulitis) with or without plasma-cell-rich infiltrate, with or without duct injury, tends to occur late in transplant and probably represents an “atypical” form of rejection that is likely to be relatively resistant to treatment compared with typical ACR. Current knowledge indicates progression to chronic rejection (CR) and/or significant fibrosis and therefore requires immunosuppressive treatment. Alternatively recurrent HCV could produce similar injury and should remain in contention to be supported by relevant clinical and biochemical data. A curious observation is that when these cases are due to HCV there is a tendency toward lower levels of transaminases expected for the degree of inflammation and zone 3 necrosis, as exemplified by patient I. Yet it seems that HCV-transplanted patients presenting this way have a significantly higher risk for poorer outcome
PAT H O L O G Y
(ie, death, graft loss, and fibrosis) than matched controls with more typical features of recurrent HCV (5).
References 1. Krasinskas AM, Demetris AJ, Poterucha JJ, Abraham SC. The prevalence and natural history of untreated isolated central perivenulitis in adult allograft livers. Liver Transpl. 2008;14(5):625–632. 2. Pappo O, Ramos H, Starzl TE, Fung JJ, Demetris AJ. Structural integrity and identification of causes of liver allograft dysfunction occurring more than 5 years after transplantation. Am J Surg Pathol. 1995;19(2): 192–206. 3. Fiel MI, Agarwal K, Stanca C, et al. Posttransplant plasma cell hepatitis (de novo autoimmune hepatitis) is a variant of rejection and may lead to a negative outcome in patients with hepatitis C virus. Liver Transpl. 2008;14(6):861–871. 4. Anand AC, Hubscher SG, Gunson BK, McMaster P, Neuberger JM. Timing, significance, and prognosis of late acute liver allograft rejection. Transplantation. 1995;60(10):1098–1103. 5. Ward SC, Schiano TD, Thung SN, Fiel MI. Plasma cell hepatitis in hepatitis C virus patients post-liver transplantation: case-control study showing poor outcome and predictive features in the liver explant. Liver Transpl. 2009;15(12):1826–1833.
Case 22.5
Fibrosing Cholestatic Hepatitis C Versus Biliary Obstruction Versus Adverse Reaction to Medication OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
PAT H O LO GIC FEAT UR ES
A 45-year-old man transplanted for hepatitis C (genotype 1b) cirrhosis; his best numbers posttransplant was at week 1: ALT 67 U/L, AST 47 U/L, ALP 96 U/L, bilirubin 7 mg/dL (120 μmol/L). The numbers began to rise and Septra which he was initially placed on was preemptively discontinued. However, enzyme rise continued and by the third week the following numbers were recorded: ALT 68; AST 77; ALP 189; bilirubin 14.7 mg/dL (252 μmol/L). He was biopsied because of the rising bilirubin on day 25 (Figures 22.5.1 and 22.5.2). By 20 weeks posttransplant, his liver enzyme numbers were ALT 45, AST 150, ALP and 129; 35.1 mg/dL (bilirubin 600 μmol/L), and he was rebiopsied (Figures 22.5.3 and 22.5.4).
The biopsy at the third week posttransplant (Figures 22.5.1 and 22.5.2) shows cholestasis, mild hepatocellular disarray, and swelling but otherwise preserved liver architecture with no significant inflammation or fibrosis. There are no features of biliary obstruction or duct necrosis or atrophy. However, isolated apoptotic hepatocytes are present, giving an overall nonspecific pattern of injury and raising the differential diagnoses of adverse reaction to medication (he was on Septra 1), early stages of cholestatic hepatitis C, and delayed recovery from ischemic/reperfusion injury. Septra had been withdrawn, and by the following week with worsening jaundice, antiviral therapy was commenced. Follow-up biopsy 17 weeks later (Figures 22.5.3 and 22.5.4) shows more cholestasis with centrizonal accentuation, (still) very minimal inflammation, progressive periportal and bridging fibrosis, mild ductular reaction, hepatocellular swelling and disarray, and single cell necrosis. Although there is some ductular reaction, it is not associated with other features
R E A SON F OR R E F E R R AL
Worsening jaundice despite adequate immunosuppression and negative magnetic resonance cholangiopancreatography (MRCP).
FIGURE 22. 5. 1 There is hepatocellular cholestasis 3 weeks posttransplant with no significant inflammation. Mild hepatocellular swelling is present but no fibrosis (see also Figure 22.5.2). Subsequent biopsy shown in Figures 22.5.3 and 22.5.4 showed diagnostic features of fibrosing cholestatic hepatitis (left panel: H&E 25 ; right panel: Masson’s Trichrome stain 25 ).
319
320
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22. 5. 2 Higher magnification of 3 weeks posttransplant biopsy showing cholestasis, absent fibrosis, hepatocellular swelling; these features had raised the possibilities of early features of cholestatic recurrent viral hepatitis C versus adverse reaction to Septra. Follow-up biopsy shown in Figures 22.5.3 and 22.5.4 showed diagnostic features of fibrosing cholestatic hepatitis (left panel: H&E 200 ; right panel: Masson’s Trichrome stain 200 ).
of biliary obstruction, namely portal edema, or increased copper accumulation (copper stain not shown), coupled with negative imaging of the biliary tree.
both hepatitis B and C patients who are immunocompromised for different reasons, including stem cell and solid organ transplantation and HIV infection (3–6). In addition to immunosuppression other clinicopathological parameters include
D I AG N OS I S
• High viral load (except in some human immunodeficiency virus [HIV] patients) • Periportal fibrosis • Swollen hepatocytes • Prominent cholestasis • Paucity of inflammation • / ductular reaction • Elevated bilirubin • Sometimes only modest elevation of transaminases • Rapidly progressive fibrosis and cholestasis to liver failure
Fibrosing cholestatic hepatitis C.
D I SC U SSI ON
Fibrosing cholestatic viral hepatitis (FCVH) was first described in 1991 by Davies et al in patients transplanted on account of chronic viral hepatitis B (2). It has since been recognized in
CASE
22.5:
POSTTRANSPLANT
C H O L E S TA S I S
WITH
OR
WITHOUT
FIBROSIS
321
FIGURE 22. 5. 3 Follow-up to the biopsy shown in Figures 22.5.1 and 22.5.2 at 20 weeks posttransplant shows typical features of fibrosing
cholestatic viral hepatitis: cholestasis, fibrosis, minimal inflammation, mild ductular proliferation, and hepatocellular swelling (left and right panels: Masson’s Trichrome stain 100 ).
All these features are identified in the illustrated case. This patient soon succumbed to liver failure, and autopsy showed fibrosis had progressed to full-blown cirrhosis, all within a span of 6 months. Stricturing/obstruction is a fairly common complication of liver transplant, and although MRCP is a very sensitive modality for diagnosing strictures, in our experience development of characteristic features seems to lag behind clinical and pathological parameters. The first biopsy showed no features of obstruction, although in the latter biopsy, periportal fibrosis and ductular reaction would ordinarily have raised some concerns. Arguing against obstruction is the low level of ALP for the degree of cholestasis and ductular reaction. Also, although imaging lags behind clinical parameters, by 5 months one would have expected there to be some features. When uncertain, a copper stain could be helpful (see Case 22.6), which was negative in this case.
The first biopsy at 3 weeks raised the usual differential diagnoses that include adverse drug reaction, particularly important in that the patient was on Septra. Septra is a commonly used prophylactic agent posttransplant. It is a known cause of cholestatic liver injury with or without ductopenia, although this complication is fortunately rare (3). The exclusion of a drug-induced liver injury in any context is a formidable challenge. The fact that this patient’s Septra was stopped in the first week posttransplant did not exclude it as the cause of cholestasis. As such the 3 weeks biopsy, though helpful in excluding biliary disease, could not have excluded Septra as the etiology, except that the modest elevation of liver enzymes would somehow be unusual. The follow-up biopsy showing rapidly progressive fibrosis and worsening cholestasis weeks after Septra withdrawal, and documented high viral load, however, point toward FCVH that failed antiviral therapy.
322
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22. 5. 4 Hematoxylin and eosin stain showing swollen hepatocytes and absent to very minimal inflammation in this case of fibrosing
cholestatic hepatitis recurrence at 20 weeks in this patient; note the absence of portal edema or other features of obstruction, which is a major consideration in the differential diagnosis (left and right panels: H&E 100 ).
References 1. Thies PW, Dull WL. Trimethoprim-sulfamethoxazole-induced cholestatic hepatitis. Inadvertent rechallenge. Arch Intern Med. 1984;144(8): 1691–1692. 2. Davies SE, Portmann BC, O’Grady JG, et al. Hepatic histological findings after transplantation for chronic hepatitis B virus infection, including a unique pattern of fibrosing cholestatic hepatitis. Hepatology. 1991; 13(1):150–157. 3. Cooksley WG, McIvor CA. Fibrosing cholestatic hepatitis and HBV after bone marrow transplantation. Biomed Pharmacother. 1995;49(3): 117–124.
4. Furuta K, Takahashi T, Aso K, Hoshino H, Sato K, Kakita A. Fibrosing cholestatic hepatitis in a liver transplant recipient with hepatitis C virus infection: a case report. Transplant Proc. 2003;35(1):389–391. 5. Rosenberg PM, Farrell JJ, Abraczinskas DR, Graeme-Cook FM, Dienstag JL, Chung RT. Rapidly progressive fibrosing cholestatic hepatitis— hepatitis C virus in HIV coinfection. Am J Gastroenterol. 2002;97(2): 478–483. 6. Suresh RL, Merican I, Chang KM, Yong SM, Purusothaman V. Cholestatic fibrosing hepatitis and hepatitis B after bone marrow transplantation. Med J Malaysia. 2001;56(4):508–511.
Case 22.6
Mechanical Biliary Obstruction Versus Chronic Rejection OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
A 46-year-old woman transplanted 13 years ago for AIH has “recently” elevated liver enzymes, although she was last followed up a year prior when her numbers were ALT 55 U/L, AST 54 U/L, ALP 161 U/L, and normal bilirubin. Her immunosuppression is deemed adequate. At the time of biopsy her liver enzymes were ALT 64, AST 70, and ALP 323, whereas serum bilirubin remained normal. Imaging studies show normal appearing liver and no evidence suggestive of obstruction. R E A SON F OR R E F E R R AL
Hepatologists think this could be chronic rejection partly because of negative imaging; biopsy shows some senescent bile ducts but it otherwise looks like obstruction; could chronic rejection be ruled out? PAT H OL OG I C F E AT U R E S
The biopsy (Figures 22.6.1–22.6.5) shows expanded portal tracts with mild inflammation, associated with ductular reaction, and increased copper at the expanded periportal areas. Routine stains show preserved and healthy-appearing ducts in most portal tracts, except in 1 or 2 portal tracts where senescent changes are seen, characterized by increased cytoplasmic eosinophilia, irregular spacing of duct epithelial cells, and distorted appearance of the duct shape (shown in Figure 22.6.5A; Figure 22.6.5B is another example of obstruction-induced senescence but in a different patient). One portal tract shows a duct with suggestive but not convincing periductal laminar fibrosis (Figure 22.6.4B).
F I G U R E 2 2 . 6 . 2 Higher magnification of one of the portal tracts in Figure 22.6.1, showing ductular proliferation in addition to portal widening (H&E 200 ).
F I G U R E 2 2 . 6 . 3 Positive rhodanine stain for periportal copper (arrows and inset) is helpful in identifying obstruction/stricture as this is typically absent in chronic rejection (rhodanine stain 100 ; inset 200 ).
DIAGNO SIS FIGURE 22. 6. 1 Low power view of a case of biliary stricture that
shows portal-centered mild inflammation with portal widening (H&E 25 ).
323
Biliary pattern changes consistent with obstruction/ stricture.
324
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22. 6. 4 Another helpful feature in distinguishing obstruction from chronic rejection is the absence of significant duct loss; as shown
here the ducts for the most parts are present and look normal (panels A&B). The laminar fibrosis in panel B is subtle at best, but when compared with a similar-sized nearby duct it is suggestive and also in keeping with the findings in the rest of the biopsy. Such findings could, however, be patchy, and its absence therefore does not rule out an obstructive process (H&E 200 ).
D I S C U S S I ON
Obstruction in liver allograft has features similar to nontransplant liver—portal expansion and edema, ductular reaction, mild, usually neutrophilic, inflammation and retention of copper in the periportal hepatocytes. These are all seen in this biopsy and support the diagnosis of biliary obstruction, despite negative imaging. Nevertheless, MRCP remains a very sensitive way of diagnosing posttransplant strictures, although its sensitivity for nonstricture- forming obstructions appears to be in the 60% range (1,2). We have seen examples of biopsies with histopathologic features of mechanical obstruction and initially negative MRCP studies but in which the obstructive nature was later confirmed either by follow-up MRCP or endoscopic retrograde cholangiopancreatography (ERCP). The suspicion for CR both clinically and histologically is reasonable, especially given the history of AIH as the primary disease 13 years posttransplant. Senescence of duct epithelial cells in majority of portal tracts and/or duct loss in at least 50% of portal tracts is the criteria for diagnosing CR (3). However, in this case only a minority of portal tracts show senescence (or atrophy) and in the context of an obstructive
process. One should bear in mind that senescent appearance could be seen in obstruction even in nontransplant patients and should therefore be interpreted with caution in the absence of significant (ie, up to 50%) ductopenia while evaluating for CR. It is important not to use senescent bile duct appearance in patients with features of obstruction as the sole criteria for diagnosing chronic rejection (4,5). In addition to not fulfilling the criteria for CR as related to the number of senescent bile ducts, the illustrated case has other features not characteristic of CR, including this degree and nature of inflammation, copper accumulation, and ductular reaction. Although ductular reaction could be seen as part of the features of CR reversal in patients treated early for CR, it is otherwise not a feature of (untreated) CR (4). The features in this biopsy therefore favor an obstruction, and follow-up studies confirmed the presence of multiple segmental (mostly intrahepatic) strictures. Many late occurring strictures are non-anastomotic, the exact pathogenesis is unclear, but are a well-documented complication of liver allografts, with reported cumulative incidence of 16% at 10 years in one large series (6).
CASE
22.6:
BILIARY
OBSTRUCTION
VERSUS
CHRONIC
REJECTION
325
FIGURE 22. 6. 5 Duct epithelial senescence, exemplified here in these ducts (arrows, panels A&B) with irregular epithelial spacing, loss of
polarity, nuclear atypia, and increased cytoplasmic eosinophilia should be interpreted with caution in cases with an obstructive process. In the absence of more than 50% ductopenia, epithelial senescence should not be the basis for diagnosing chronic rejection in patients with obstruction (H&E 200 ).
References 1. Aufort S, Molina E, Assenat E, et al. Value of MRCP for diagnosis of biliary complications after liver transplantation. J Radiol. 2008;89(2): 221–227. 2. Kitazono MT, Qayyum A, Yeh BM, Chard PS, Ostroff JW, Coakley FV. Magnetic resonance cholangiography of biliary strictures after liver transplantation: a prospective double-blind study. J Magn Reson Imaging. 2007;25(6):1168–1173. 3. Demetris A, Adams D, Bellamy C, et al. Update of the international banff schema for liver allograft rejection: working recommendations for
the histopathologic staging and reporting of chronic rejection. An international panel. Hepatology. 2000;31(3):792–799. 4. Demetris AJ. Distinguishing between recurrent primary sclerosing cholangitis and chronic rejection. Liver Transpl. 2006;12(11 suppl 2): S68–S72. 5. Demetris AJ, Adeyi O, Bellamy CO, et al. Liver biopsy interpretation for causes of late liver allograft dysfunction. Hepatology. 2006;44(2): 489–501. 6. Buis CI, Verdonk RC, Van der Jagt EJ, et al. Nonanastomotic biliary strictures after liver transplantation, part 1: Radiological features and risk factors for early vs. late presentation. Liver Transpl. 2007;13(5):708–718.
Case 22.7
Chronic Rejection Versus Recurrent Primary Sclerosing Cholangitis Versus Non-PSC Stricture OYEDELE ADEYI
C L I N I C AL I N F OR M AT I ON
A 36-year-old man was transplanted 27 months earlier for primary sclerosing cholangitis (PSC). This was his second graft having lost the first one to “chronic rejection and recurrent PSC” after 10 years. His retransplant ALP was always high between 260 U/L and 300 U/L. Four months prior to this biopsy, his ALT was 83 U/L, AST 62 U/L, ALP 277 U/L, and bilirubin 1.1 mg/dL (17.2 μmol/L); but these numbers gradually rose and at the time of biopsy were ALT 187, AST 206, ALP 500, bilirubin. 15.6 mg/dL (266 μmol/L). He had been on Tacrolimus-based immunosuppression with satisfactory levels, and there have been no recent changes in medication.
R E A S ON F OR R E F E R R A L
Explain reason for persistent mixed cholestatichepatitic pattern of injury; is there PSC recurrence or chronic rejection or other cause of obstruction?
PAT H OL OG I C F E AT U R E S
From Figures 22.7.1–22.7.3 similar features to those seen in the previous case can be appreciated: portal widening, ductular reaction, epithelial senescence (arrow in Figure 22.7.3), and periportal copper. The present case, however, shows more fibrosis. These features as discussed earlier favor an obstructive and stricturing process. In addition, however, there is ductopenia, albeit in less than 50% of sampled portal tracts.
DIAGNO SIS
Biliary pattern with ductular reaction, cholestasis, and ductopenia, most consistent with recurrent sclerosing cholangitis.
DISCUSSIO N
There is no evidence of an overlapping hepatitic process; the elevated transaminases in this patient are most probably due to hepatotoxicity resulting from intracellular retention of high amounts of bile salts and hydrophobic bile acid in this very cholestatic liver (1). Although it is important to correlate morphology with liver enzyme profile, certain exceptions, as in this case, should be borne in mind when elevated transaminases are not from an active hepatitic injury. The main question here is recurrent PSC versus CR, and it is not always possible to make a separation. PSC is a stricturing process and therefore produces similar findings to the previous case. However, unlike in the previous case there is higher degree of epithelial senescence and small bile duct loss albeit less than 50%, which brings CR as a valid differential diagnosis. Criteria for CR include senescence in the majority, or duct loss in greater than 50%, of portal tracts. Even if duct loss had been up to 50%, in this case it would still be difficult to diagnose CR, partly because PSC could also produce significant ductopenia, and partly because there are other features
FIGURE 22. 7. 1 Recurrent primary sclerosing cholangitis showing portal expansion and portal-centered fibrosis (left panel: H&E 25 ; right panel: Masson’s Trichrome stain 25 ).
326
CASE
22.7:
CHRONIC
REJECTION
VERSUS
RECURRENT
PSC
327
FIGURE 22. 7. 2 Higher magnifications of Figure 22.7.1 in a case of primary sclerosing cholangitis (PSC) show ductular proliferation and
other features shared with non-PSC biliary obstruction earlier illustrated in Figures 22.6.1, 22.6.2 and 22.6.5 (right and left panels: Masson’s trichrome 100 ).
one would ordinarily not expect with CR. These include fibrosis, which is uncommon and usually a very late feature in CR. Also, ductular reaction or significant copper retention would be unusual findings in CR (2). CR is best viewed as a “serious problem with little noise.” In addition to the duct findings, other features to look for in CR include perivenular fibrosis, minimal to mild lymphoplasmacytic portal inflammation (unlike that mixed with neutrophils in PSC/strictures), and (hardly seen in needle biopsy, but when available) hepatic arterial intimal foam cell accumulation with or without concentric fibrointimal hyperplasia (2). In needle biopsies, employing cytokeratin 7 (CK7) immunostain to support routine stain findings should be considered. CK7 could often add new information, especially in cases where there is a question as to the exact degree of duct loss, in which case it might outline hitherto difficult to appreciate small portal tracts with absent ducts. Figure 22.7.4 compares the expected CK7 findings in this case of recurrent PSC with a biopsy from another patient with CR. Lastly most patients with CR have documented
prior episodes of ACR, suboptimum immunosuppression, or other clinical indicators in favor of an immunologic injury. In the case described, immunosuppression had been closely monitored especially in view of the patient’s prior history. Figure 22.7.5 is presented from a CR patient to further highlight important differences from PSC. A more challenging situation is trying to differentiate PSC from non-anastomotic strictures of other causes. The overlapping features between this and the previous case highlight why this differentiation would be a challenge. Generally speaking, “de novo” PSC though reported (3), is at best extremely rare to be practically nonexistent. Therefore, multiple stricturing occurring in a patient transplanted for non-PSC cause could hardly be regarded as PSC. Also, although some degree of ductopenia is possible in mechanical obstruction as in non-PSC strictures, they are less likely to be to the degree one would expect in PSC or CR. In summary differentiating PSC from CR is possible with careful evaluation for the features here discussed and
328
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22. 7. 3 Same case of PSC as in earlier figure shows bile duct epithelial senescence (left panel: H&E 100 ; right panel: H&E 200 ).
FIGURE 22. 7. 4 Cytokeratin 7 (CK7) staining pattern in the primary sclerosing cholangitis (PSC; left panel) shows prominent ductular proliferation with no visible duct. The right panel compares the CK7 pattern in a chronic rejection (CR) patient, where a very atrophic/ senescent duct is seen but not associated with ductular proliferation. CR characteristically lacks ductular proliferation, unless in the recovery phase (CK7 immunohistochemistry 100 ).
CASE
22.7:
CHRONIC
REJECTION
VERSUS
RECURRENT
PSC
329
FIGURE 22. 7. 5 An example of chronic rejection (CR) is presented with features that help distinguish it from the obstructive/stricturing fea-
tures of PSC. As shown, CR features include duct loss, minimal fibrosis, absent ductular proliferation, negative rhodanine satin for copper, and (when available) graft arteriopathy including intimal thickening and accumulation of foamy macrophages (A and D, H&E 100 ; B, Masson’s Trichrome stain 100 ; C, rhodanine stain 100 ).
previously published (2). However, until new insights into the pathogenesis of late non-anastomotic strictures emerge, recurrent PSC should only be considered in patients with PSC as the original disease, and alternative cause(s) for strictures in other patients should be investigated.
References 1. Attili AF, Angelico M, Cantafora A, Alvaro D, Capocaccia L. Bile acidinduced liver toxicity: relation to the hydrophobic-hydrophilic balance of bile acids. Med Hypotheses. 1986;19(1):57–69.
2. Demetris AJ. Distinguishing between recurrent primary sclerosing cholangitis and chronic rejection. Liver Transpl. 2006;12(11 suppl 2): S68–S72. 3. McPartland KJ, Lewis WD, Gordon FD, et al. Post-liver transplant cholestatic disorder with biliary strictures: de novo versus recurrent primary sclerosing cholangitis. Pathol Int. 2009;59(5):312–316.
Case 22.8
Zone 3 (Centrilobular) Necrosis OYEDELE ADEYI
C L I N I C AL I N F OR M AT I ON
The cases of three patients are presented here to discuss this problem. Patient I: A 67-year-old man transplanted 8 months earlier for end-stage liver disease due to chronic hepatitis B. He was biopsied as a result of recent elevation of liver enzymes. He has had on and off problems with rejection, but his baseline prior to recent rise in enzymes was ALT 98 U/L, AST 56 U/L, and ALP 97 U/L; at the time of biopsy these were ALT 306, AST 184, and ALP 262.
Patient II: A 48-year-old man was transplanted 2 weeks prior to this biopsy for hepatitis B cirrhosis complicated by hepatocellular carcinoma. His transaminases very slowly reached a nadir of ALT 89 U/L and AST 43 U/L on day 9. The ALP, however, which was initially 71 U/L began to rise earlier than the ALT and AST. By day 14 posttransplant, the ALT was 1401 U/L, AST 3029 U/L, ALP 445 U/L, and total bilirubin 14 mg/dL (241 μmol/L). Due to worsening liver function, he was imaged and taken back to the operating room for exploration, at which time a biopsy was also performed.
FIGURE 22. 8. 1 Zone 3 necrosis around central veins (cv) in a patient recently treated for severe acute cellular rejection. The inflammatory
infiltrates previously documented (see Figure 22.8.2) have disappeared from zone 3 and portal tracts (pv) following treatment, but the resulting zone 3 necrosis is yet to heal (upper panels: 50 ; lower panels: 100 ; upper right panel: Masson’s Trichrome stain; others: H&E).
330
CASE
22.8:
ZONE
3
(CENTRILOBULAR)
Patient III: A 54-year-old woman transplanted for primary biliary cirrhosis and developed biliary complications which kept her ALP high and for which she was stented. Her baseline liver enzymes were ALT 32 U/L, AST 29 U/L, ALP 924 U/L, and total bilirubin 1.3 mg/dL (23 μmol/L). However, by 15 weeks posttransplant, the ALT was 51, AST 64, ALP 841, and bilirubin 1.0 mg/dL (17.1 μmol/L). Repeated imaging showed normal liver with draining stent. A biopsy was performed. R E A SON F OR R E F E R R AL
Patient I: This patient has been treated for rejection, but his enzymes have not normalized; is there something else going on? Patient II: There is poor flow through the parenchyma with hepatic artery thrombus; assess for extent of liver damage. Patient III: Rule out rejection or other cause for slowly rising transaminases.
NECROSIS
331
PAT H O LO GIC FEAT UR ES
Patient I: Sections of needle biopsy (Figure 22.8.1) show a core with preserved architecture, and little inflammation, limited to occasional portal tracts. There is, however, significant necrosis around central veins (cv). Duct injury or phlebitis to suggest active rejection is not appreciated. The explanation for zone 3 necrosis becomes apparent after reviewing the most recent previous biopsy, obtained a week earlier, and after 2 days steroid bolus for presumed acute cellular rejection. The prior biopsy (Figure 22.8.2) shows features of severe acute cellular rejection (even after 2 days of steroid boluses). The degree of hepatic vein phlebitis with perivenular hepatocyte necrosis is severe justifying the maximum score on the Banff scale as described earlier in Case 22.1 and summarized in Table 22.1.1, Case 22.1. After steroids, this patient was further treated more aggressively for “steroid-resistant” rejection before the biopsy illustrated in Figure 22.8.1, with the result
FIGURE 22. 8. 2 Severe acute cellular rejection with zone 3 inflammation and necrosis were present in the pre-treatment biopsy preceding the one illustrated in Figure 22.8.1. Abbreviations: cv, central vein; pv, portal vein. (H&E; upper left panel 50 ; other panels 100 .)
332
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
that the cellular infiltrate cleared but the necrosis was yet to heal/resolve.
DIAGNO SIS
Patient II: The wedge biopsy shown in Figure 22.8.3 shows geographic necrosis that appears to be bridging zone 3 (cv) areas in the left panel; the periportal areas are less necrotic. Higher magnifications in the right panels show some hemorrhage in addition to necrosis but no significant inflammation, sinusoidal dilatation, or fibrosis.
Patient I: Residual perivenular (zone 3) necrosis due to recent severe acute cellular rejection. Patient II: Severe ischemic necrosis secondary to hepatic artery thrombosis. Patient III: Zone 3 sinusoidal congestion and perivenular hepatocyte dropout consistent with outflow problems.
Patient III: Sections highlighted in Figure 22.8.4 show preserved alternation between portal (pv) and central (cv) veins with no fibrosis in the trichrome stain at the right lower panel. However, sinusoidal dilatation in zone 3 (cv) areas as well as perivenular hepatocellular necrosis are identified. Away from the central veins toward portal tracts, the sinusoidal dilatation and necrosis decrease.
DISCUSSIO N
The differential diagnosis of zone 3 necrosis is the same as in the nontransplant liver and includes ischemia/shock, drug toxicity (especially acetaminophen), outflow obstruction
FIGURE 22. 8. 3 Severe necrosis centered on zone 3 (central vein [cv]) with relative preservation of zone 1/periportal hepatocytes (portal vein [pv]). The necrosis is extensive in this patient with hepatic artery thrombosis (left and right upper panels: H&E; right lower panel: Masson’s Trichrome stain 100 ).
CASE
22.8:
ZONE
3
(CENTRILOBULAR)
NECROSIS
333
FIGURE 22. 8. 4 Zone 3 (central veins [cv]) necrosis, congestion, and sinusoidal dilatation is seen in this example of outflow obstruction posttransplant. These features are absent or minimal in zone 1 areas (portal veins [pv]) (left panels and right upper panel: H&E; right lower panel: Masson’s Trichrome stain; left panels: 25 ; right panels: 50 ).
(including cardiac causes), and alcohol-induced acute liver injury. In addition, however, (as summarized in Table 22.8.1), severe cellular rejection (typical and atypical forms) and preservation and reperfusion injury, both of which are not relevant in the native liver, should be considered when reviewing liver allograft biopsies with zone 3 necrosis. Although certain histologic parameters could point to a likely cause for zone 3/perivenular necrosis, knowledge of the clinical context including timing of event relative to transplant date, imaging results, recent therapeutic intervention, and, when available, review of previous biopsies are important in the evaluation of these biopsies. This way necrosis from a recent rejection (or other process) could be easily placed in the right context. Causes of ischemic-type necrosis posttransplant include thrombosis, outflow obstruction, preservation injury (comprising of “down-time,” and the cold and warm ischemic periods);
TA BL E 2 2 . 8 . 1 Causes of zone 3 necrosis in liver allografts Ischemia (including hepatic artery thrombosis, shock, for example, from sepsis or hypovolemia) Transport/preservation (ischemic) injury Reperfusion injury Outflow obstruction Severe acute cellular rejection Hepatitis with central perivenulitis Possible but rarely encountered: Drug or other toxic injury Alcohol
reperfusion injury, ironically caused by the re-establishment of vascular flow modulated by the release of reactive free radicals and pro-inflammatory cytokines (1,2); and ischemia resulting from poor vascular flow (either reduced flow into, or
334
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22.8.5 Zone 3 necrosis from preservation/reperfusion injury is milder than that seen in Figure 22.8.3, but the injured hepatocytes have
released fat into the sinusoids where they are partly engulfed by Kupffer cells (arrows). Lipopeliosis occurs in steatotic donor liver and could further comprise graft recovery (left panels: H&E; right panels: Masson’s Trichrome stain. Left upper panel 25 ; others 50 ).
obstructed flow out of, the graft). Ischemic and (as in acetaminophen toxicity) free-radical-induced liver injuries have a more pronounced effect on zone 3 hepatocytes due to metabolic zonation (3,4), and underlie the reasons for zone 3 predisposition to these injuries. The case of patient II is an example of a severe case of hepatic artery thrombosis. Hepatic artery thrombosis (HAT) could be early or late, although the outcome seems to be better in terms of graft loss in late than early cases. HAT remains a serious complication of liver transplant (5). The true incidence is uncertain, but overall it would seem to be less than 10% (probably approximately 4% in adults and 8% in children) (5). The definition of what should constitute early HAT also varies, but most cases occurred within the first few posttransplant days to weeks. The pathologist’s involvement in these cases is likely to be in reviewing intraoperative wedge biopsies (as in this illustration), or reviewing of failed graft, but occasional cases could also come as needle biopsy especially when Doppler evaluations have been inconclusive. Because bile ducts remain
absolutely dependent on hepatic arterial flow, ischemic duct injury is often a major issue; and in milder cases of HAT, parenchymal injury could be absent, the only findings being those of ischemic “cholangitis” associated with elevated ALP. However, in more catastrophic cases, differential enzyme rise is not always apparent. The features of ischemic necrosis are the same in the native and allograft liver and include zone 3 hepatocellular necrosis with or without hemorrhage. Sinusoidal dilatation seen in outflow obstruction cases are absent in ischemia. An important differential diagnosis is acetaminophen toxicity, but, fortunately, this complication is rare in the transplant population. Another differential diagnosis is severe rejection. Zone 3 necrosis with associated inflammation should raise the possibility of rejection or other entities discussed earlier in Case 22.4. A potential problem, however, arises when ischemic injury predisposes the graft to, or otherwise coexists with, cellular rejection, in which case the pathology review should be performed in the context of underlying clinical and radiologic factors, in order not to miss
CASE
22.8:
ZONE
3
(CENTRILOBULAR)
NECROSIS
335
FIGURE 22.8.6 Zone 3 injury without outright necrosis, characterized by hepatocellular swelling and intracellular cholestasis is seen in this example of preservation/reperfusion injury (left panels: H&E; right panels: Masson’s Trichrome stain; upper panels: 25 ; lower panels: 50 ).
the opportunity to flag the ischemia problem while emphasizing rejection. A useful clue is to weigh the severity of necrosis with the degree of inflammation in patients not yet treated for rejection. A significant problem could result when a patient had been treated for rejection before biopsy (the case of patient I), and in these cases matching the degree of inflammation with necrosis could be potentially misleading. Preservation and reperfusion injury (PRI) is not illustrated but is another differential diagnosis of hepatic artery thrombosis. PRI-induced zone 3 necrosis is usually (and fortunately) not as severe as the illustrated HAT case. The necrosis tends to be less pronounced and hardly bridging to produce the geographic pattern seen here. Figure 22.8.5 shows an example of a patient with reperfusion injury, as well as some other problems that could be associated. This is a patient who received cadaveric liver; the liver enzymes were observed to be unsatisfactorily creeping very slowly toward normalization prompting a biopsy. The zonal necrosis is less severe than in the patient with hepatic artery thrombosis, but the dead hepatocytes also
appear to have been steatotic, such that released fat was left freely floating in the sinusoids or within sinusoidal Kupffer cells (black arrows); hence, the term lipopeliosis. Lipopeliosis could further hamper the rate of recovery from reperfusion/preservation injury and should always be part of the pathology report whenever present. Also, some biopsies with preservation/reperfusion injury could show only subtle changes such as zone 3 hepatocellular swelling with or without cholestasis but no outright necrosis, as illustrated in Figure 22.8.6. In these cases, it is important to consider intra-abdominal or systemic sepsis and (again) drug reaction on the differential diagnosis list. Lastly outflow obstruction remains a major, although no longer frequent, posttransplant complication. The third illustrated case exemplifies this. Following this biopsy, additional imaging revealed a hepatic vein blockage hitherto undetected and was subsequently fixed. The absence of characteristic radiologic findings of hepatic vein blockage should not deter one from flagging a biopsy with features such as seen in this case and discussed in more detail in Case 22.10.
336
CHAPTER
22:
LIVER
TRANSPLANT
References 1. Carden DL, Granger DN. Pathophysiology of ischaemia-reperfusion injury. J Pathol. 2000;190(3):255–266. 2. Teoh NC, Farrell GC. Hepatic ischemia reperfusion injury: pathogenic mechanisms and basis for hepatoprotection. J Gastroenterol Hepatol. 2003;18(8):891–902. 3. Sastre J, Rodriguez JV, Pallardo FV, et al. Effect of aging on metabolic zonation in rat liver: acinar distribution of GSH metabolism. Mech Ageing Dev. 1992;62(2):181–190.
PAT H O L O G Y
4. Gebhardt R. Metabolic zonation of the liver: regulation and implications for liver function. Pharmacol Ther. 1992;53(3):275–354. 5. Bekker J, Ploem S, de Jong KP. Early hepatic artery thrombosis after liver transplantation: a systematic review of the incidence, outcome and risk factors. Am J Transplant. 2009;9(4):746–757.
Case 22.9
Cytomegalovirus Hepatitis OYEDELE ADEYI
C L I N IC AL I N F OR M AT I ON
PAT H O LO GIC FEAT UR ES
A 42-year-old woman was transplanted 7 months prior to biopsy for fulminant hepatitic failure. Despite adequate Tacrolimus-based immunosuppression, her liver enzymes continued to rise; at the time of biopsy these were ALT 65 U/L, AST 313 U/L, and ALP 271 U/L.
Sections (Figures 22.9.1 and 22.9.2) show liver tissue with preserved architecture and only minimal lymphocytic inflammation limited to 1 or 2 portal tracts. Higher magnification shows preserved ducts and absence of phlebitis or other features of cellular rejection. However, in the sinusoid and in at least 1 portal vein endothelium, large cells with nonclassic but suggestive nuclear 6 cytoplasmic inclusions of the cytomegalovirus (CMV) type are observed. Also seen are few foci of small aggregates of neutrophils (neutrophilic microabscess— blue arrow in Figure 22.9.2). Ancillary immunohistochemistry
R E A SON F OR R E F E R R AL
Patient is adequately immunosuppressed but could there still be rejection?
FIGURE 22. 9. 1 Cytomegalovirus hepatitis in a liver allograft biopsy shows large cells with nuclear and cytoplasmic inclusions in sinusoidal and portal vein endothelia (black arrows). Note minimal portal inflammation a small cluster of lobular neutrophils (blue arrow) (all panels: H&E; left upper panel, 12.5 ; left lower panel, 25 ; right upper and lower panels, 50 ).
337
338
CHAPTER
22:
LIVER
TRANSPLANT
PAT H O L O G Y
FIGURE 22. 9. 2 Cytomegalovirus (CMV) hepatitis showing more sinusoidal cells with nuclear and cytoplasmic inclusions in the upper panels (arrows); the lower panels are immunohistochemistry stains for CMV in which occasional normal-sized cells also contain viral particles (upper panels: H&E, 50 ; lower panels: CMV immunohistochemistry, 25 ).
confirms CMV in many cells (sinusoidal, portal vein endothelium, and occasional hepatocytes).
D I AG N OS I S
Cytomegalovirus hepatitis. D I SC U SSI ON
Although the overall incidence appears to be falling compared with the earlier days of solid organ transplant, probably because of new immunosuppressives, antigenic monitoring, and use of prophylaxis or pre-emptive treatment by some programs, CMV infection nevertheless remains an important complication of organ transplantation. In liver transplant patients who received no prophylactic treatment, this complication when it occurs is usually in the first 3 months posttransplant, peaking at 4 weeks (1). It could, however, occur later after
discontinuation of treatment or as a de novo infection. CMV infection could be organ-specific or systemic, but the liver is the most commonly involved organ in liver transplant patients (2). Cholestatic enzyme elevation is characteristic, although the pattern could be mixed with significant transaminase elevation. Risk factors for developing posttransplant CMV hepatitis include seropositive donor and/or recipient; induction with lymphocyte depleting agents; mycophenolate-based immunosuppression; retransplantation; fulminant hepatic failure as the primary disease, hepatic artery thrombosis, and hepatitis C infection (1). Diagnosis employs blood sampling for antigenemia assays, nucleic acid amplification, or Hybrid Capture CMV DNA Assay, and liver biopsy. Liver biopsy depends on the identification of infected cells that are large (cytomegaly) and show characteristic cytoplasmic and nuclear inclusions of the Cowdry type A. It is important to note that because of the patients’ immunosuppressive states, inflammatory response could be modest or completely absent. Also, expecting “classic” inclusions seen in
CASE
22.9:
CYTOMEGALOVIRUS
pathology atlases could make for false negativity. The illustrated case shows few big cells but none could really be described as “classic” in the regular sense, and as shown by immunohistochemistry few cells also harbor the viral material without cytomegaly or demonstrable inclusions. Immunohistochemical and/or in situ hybridization methods are frequently helpful, being more sensitive and specific than routine histology, and it is possible to demonstrate nuclear staining in normal size cells by these special methods. In immunosuppressed patients with unexplained elevation of liver enzymes, ancillary staining for CMV should always be considered, even in the absence of characteristic inclusions and/or significant inflammation on routine stains. The illustrated patient had at least 3 risk factors including positive donor serology, fulminant hepatitis as the
H E PAT I T I S
339
original disease, and exposure to immunosuppressive therapy. Although CMV hepatitis could be successfully treated the virus cannot be eliminated, and there is a risk for recurrence, overall reduced patient survival, biliary complications, and increased risk for acute and chronic rejections (1,3).
References 1. Sampathkumar P, Paya CV. Management of cytomegalovirus infection after liver transplantation. Liver Transpl. 2000;6(2):144–156. 2. Paya CV, Hermans PE, Wiesner RH, et al. Cytomegalovirus hepatitis in liver transplantation: prospective analysis of 93 consecutive orthotopic liver transplantations. J Infect Dis. 1989;160(5):752–758. 3. Razonable RR. Cytomegalovirus infection after liver transplantation: current concepts and challenges. World J Gastroenterol. 2008;14(31): 4849–4860.
Case 22.10
Graft Versus Host Disease OYEDELE ADEYI
C L I N I C AL I N F OR M AT I ON
R EA SO N FO R R EFER R A L
A 51-year-old man had nonmyeloablative stem cell transplant for chronic lymphocytic leukemia (CLL) after failing multiple cycles of chemotherapy. The graft failed at 3 months, and he was given donor leukocytes infusion. The following weeks, however, witnessed progressive rise in liver enzymes and worsening skin rash. Liver and skin biopsies were performed at a time when ALP was 808 U/L, AST 114 U/L, ALT 606, and total bilirubin of 17.2 mg/dL (294 μmol/L).
Increasing liver enzymes of mixed pattern: rule out graft versus host disease versus drug reaction or infiltrative disease. PAT H O LO GIC FEAT UR ES
The biopsy shows an adequate core of liver tissue with preserved architecture and very minimal portal lymphocytes (Figure 2.10.1). The only positive findings are absence of
F I G U R E 2 2 . 1 0 . 1 Chronic graft versus host disease in a patient who received bone marrow transplantation a little more than
3 months earlier. Liver biopsy shows minimal portal and no lobular inflammation, as well as absent bile ducts in the portal tracts (left panels: H&E; right panels: CK7 immunohistochemistry; left upper panel: 12.5 ; right upper and left lower panels: 25 ; right lower panel: 50 ).
340
CASE
22.10:
GRAFT
VERSUS
interlobular bile ducts in all but 1 of 11 sampled portal tracts and cholestasis. The absent bile ducts are further confirmed on deeper levels that include CK7 stain (Figure 22.10.1, right panels).
D I AG N OS I S
Consistent with chronic graft versus host disease (GVHD).
D I S C U S S I ON
GVHD affecting the liver is not an issue in liver transplant recipients. GVHD is a major consideration in bone marrow (BMT)/stem cell (SCT) transplant and small bowel transplant recipients, although it has been reported as rare events in other solid organ transplants including liver allograft recipients (1–3).
HOST
DISEASE
341
The liver pathologist’s role mirrors what it is in other aspects of liver pathology: diagnose GVHD and confirm or exclude other pathological process(es) that could be responsible for abnormal liver enzyme. It is important to note that because BMT/SCT patients are on multiple medications including antibiotics, the diagnosis of (chronic) GVHD in this context should only be made definitively in the context of supporting clinical parameters (eg, extrahepatic features of GVHD, absence of potential drug etiology). This is necessary because the histopathological hallmarks of GVHD could be mirrored by some medications especially antimicrobials like the macrolide antibiotics, Bactrim® (Septra®), and some antifungals, among several others that BMT/SCT patients are almost as a rule exposed to (4–7). GVHD is either acute or chronic, depending on the timing posttransplant, 3 months being the commonly quoted timeline. Histologically acute GVHD shows more portal infiltrate and evidence of duct injury by infiltrating lymphocytes.
FIGURE 22. 10. 2 Chronic graft versus host disease in a patient who received bone marrow transplantation a little more than 3 months earlier. Liver biopsy shows no inflammation but significant cholestasis with a small focus of bile infarct (arrow) (all panels: H&E; left panels: 25 ; right panels: 50 ).
342
CHAPTER
22:
LIVER
TRANSPLANT
Chronic GVHD occurs more than 3 months posttransplant, represents a different disease rather than progression of acute GVHD, and is characterized by no significant or very minimal infiltrate but significant loss of small bile ducts, (vanishing ducts), which is essentially “chronic ductopenic rejection in the native liver.” The illustrated case therefore, by virtue of timing and histopathological parameters is consistent with chronic GVHD. At the time of biopsy, this patient was also on fluconazole and vancomycin and other drugs with no known concern for hepatotoxicity. fluconazole could potentially produce hepatotoxicity, but the reported incidence is low (compared with other azoles in clinical use). Also, the concurrent skin biopsy showed features compatible with GVHD, such that the best interpretation in this context was chronic GVHD. An explanation for the ALT of 606 U/L was not immediately apparent, although it is likely due to secondary hepatocellular injury with focal bile infarcts from severe cholestasis, shown in Figure 22.10.2 (8).
References 1. Barton-Burke M, Dwinell DM, Kafkas L, et al. Graft-versus-host disease: a complex long-term side effect of hematopoietic stem cell transplant. Oncology (Williston Park). 2008;22(11 suppl Nurse Ed):31–45.
PAT H O L O G Y
2. Abu-Elmagd K, Bond G. The current status and future outlook of intestinal transplantation. Minerva Chir. 2002;57(5):543–560. 3. Kohler S, Pascher A, Junge G, et al. Graft versus host disease after liver transplantation—a single center experience and review of literature. Transpl Int. 2008;21(5):441–451. 4. Rosenberg PM, Farrell JJ, Abraczinskas DR, Graeme-Cook FM, Dienstag JL, Chung RT. Rapidly progressive fibrosing cholestatic hepatitis—hepatitis C virus in HIV coinfection. Am J Gastroenterol. 2002;97(2):478–483. 5. Adriaenssens B, Roskams T, Steger P, Van Steenbergen W. Hepatotoxicity related to itraconazole: report of three cases. Acta Clin Belg. 2001;56(6):364–369. 6. Altraif I, Lilly L, Wanless IR, Heathcote J. Cholestatic liver disease with ductopenia (vanishing bile duct syndrome) after administration of clindamycin and trimethoprim-sulfamethoxazole. Am J Gastroenterol. 1994;89(8):1230–1234. 7. Ramachandran R, Kakar S. Histological patterns in drug-induced liver disease. J Clin Pathol. 2009;62(6):481–492. 8. Attili AF, Angelico M, Cantafora A, Alvaro D, Capocaccia L. Bile acidinduced liver toxicity: relation to the hydrophobic-hydrophilic balance of bile acids. Med Hypotheses. 1986;19(1):57–69.
23 Benign Hepatocellular Lesions VALÉRIE PARADIS
I N T ROD U C T I ON
Benign hepatocellular lesions encompass 2 distinct groups of hepatocellular proliferations (focal nodular hyperplasias [FNH] and hepatocellular adenomas [HCA]) in terms of pathogenesis, clinical presentation, and behavior. Both groups of lesions are mostly observed in young women with normal liver, usually in the context of oral contraception (OC). Pathological diagnosis of benign hepatocellular proliferations has become more and more challenging for several reasons: (1) the increased detection of asymptomatic liver tumors by extensive use of abdominal imaging, (2) the necessity to obtain accurate diagnosis for subsequent adequate management, and (3) the requirement for tumor biopsy, especially in atypical lesions on imaging. FO C A L N OD U L A R H Y P E R P L A S I A
FNH accounts for the second most common benign liver process, following hemangioma. It is predominantly diagnosed in women of 30 to 50 years of age, not influenced by OC (1–3). FNH is considered as a hyperplastic focal reaction resulting from an arterial malformation (4). This hypothesis that suggests increased arterial blood flow hyperperfuses the local liver parenchyma, leading to secondary hepatocellular hyperplasia, has been reinforced by molecular data showing that FNH are polyclonal regenerative processes (5,6). The vast majority of FNH are asymptomatic and discovered incidentally during liver ultrasound examination. In addition, complications of FNH, such as rupture or bleeding, are exceptional, and no evidence of malignant transformation has been reported so far. Histologic Patterns
In most cases, FNH displays a typical morphological pattern for both radiologists and pathologists. Grossly, FNH is a well-circumscribed, unencapsulated, usually solitary mass, characterized by a central fibrous scar that radiates into the liver parenchyma. Histologically, FNH is composed of benignappearing hepatocytes arranged in nodules that are partly or completely delineated by fibrous septa originating from the central scar. In the fibrous septa, large and dystrophic vessels are observed, associated with ductular reaction and inflammatory cells in varied intensity. The hepatocytes are hyperplastic, arranged in liver plates of normal or slightly increased thickness with a well-preserved reticulin framework. Hepatocytes may be hydropic, related to cholestatic changes. Steatosis, usually of mild intensity, may also be observed. Besides the classic FNH, several variants are described with increased frequency, usually “atypical” on imaging, potentially leading to misdiagnosis. These variants include FNH without central fibrous scar, FNH with prominent steatosis,
and heterogeneous lesions displaying telangiectatic or adenomatous changes (7). On histological examination, FNH without macroscopic central fibrous scar is usually of small size and exhibits all the pathological features of classic FNH but with few, thin and short fibrous septa. Pathological diagnosis of FNH, even in its classic form, may be difficult on biopsy specimen where fibrous septa and thick abnormal arteries are usually missing (8). In addition to the presence of ductular reaction at the border between fibrous septa and hepatocellular component, immunostaining with glutamine synthetase showing focal positive hepatocellular areas usually centered around by hepatic veins, described as a map-like pattern, is highly consistent with the diagnosis of FNH (8,9). Given that FNH are regenerative lesions with only rare occurrence of complications, in cases of a definitive diagnosis, no follow-up or treatment for asymptomatic FNH is required, no matter the size and the number of lesions. However, surgical resection is indicated in doubtful or symptomatic cases, such as large FNH located in the left liver and pedunculated lesions. H EPATO CELLULA R A DENO MA S
HCA is a rare, benign liver neoplasm strongly associated with OC use in females and androgen steroid therapy in males (10,11). Its incidence is estimated to be 0.1 per year per 100 000 in non-OC users, and reaches 3 to 4 per 100 000 in long-term OC users (12). HCA can also occur spontaneously or be associated with underlying metabolic diseases, including type I glycogen storage disease, iron overload related to betathalassemia and diabetes mellitus (13). Pathologic Findings
HCAs are usually solitary, sometimes pedunculated, with a diameter that can reach 30 cm. Large subcapsular vessels are commonly found on macroscopic examination. On cut sections, the tumor is well delineated, sometimes encapsulated, of fleshy appearance ranging in color from white to brown. Heterogeneous areas of necrosis and/or hemorrhage may be observed, preferentially in tumors of large size. Histologically, HCA consists of a proliferation of benign hepatocytes arranged in a trabecular pattern, without any residual portal tracts. Small, thin, and unpaired vessels are usually found throughout the tumor. Hepatocytes may have intracellular fat or increased glycogen. A certain degree of cellular atypia can be detected, especially in patients who have taken steroids for many years. In that context, differential diagnosis with hepatocellular carcinoma (HCC) may be difficult.
343
344
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
Compared with FNH, patients with HCA are more likely to present with symptoms, especially if they have large tumors. Moreover, potential serious complications may occur, including spontaneous bleeding and hemorrhage, especially for tumors larger than 5 cm in diameter (14,15). In addition, HCA, as monoclonal process, may progress to malignancy with a risk ranging from 4% to 10% (13,15,16). Recent studies confirmed that this risk is higher in male patients and in large HCA (15). It also recently appears that metabolic syndrome, the incidence of which is increasing worldwide, may favor development of HCC in a pre-existing HCA (17). Pathomolecular Classification of Hepatocellular Adenomas
Molecular comprehensive studies have recently gained further insights into the knowledge of HCA showing a great heterogeneity in that group of tumors, resulting in the description of 3 main subtypes associated with specific phenotypical and molecular features (18). The first group of HCA displays biallelic mutations of the TCF1 gene inactivating the hepatocyte nuclear factor 1 (HNF1 ) transcription factor and is phenotypically characterized by marked steatosis, absence of cytological abnormalities, and inflammatory infiltrates (18,19). Although, HNF1 mutations are somatic in most cases of these HCAs, patients with inherited mutation in 1 allele of HNF1 may develop maturity-onset diabetes of the young type 3 (MODY3) and are predisposed to have familial liver adenomatosis, classically defined by the presence of at least 10 HCA, when the second allele is inactivated in hepatocytes by somatic mutation or chromosome deletion (20). The second group of HCA displays -catenin–activating mutations and is characterized by increased risk for malignant transformation into HCC. These HCA are mostly encountered in male patients and frequently show significant cellular atypia and pseudo-glandular formation. The third group of HCA corresponds mainly to the initially called “telangiectatic form of FNH,” appearing as well-delineated, unencapsulated tumors showing significant areas of vascular changes (foci of sinusoidal dilation and/ or peliotic changes), without evident fibrous scar (21). The hepatocellular proliferation contains few and short fibrous septa around clusters of small vessels, sometimes accompanied by inflammatory infiltrates (mainly composed of lymphocytes and macrophages) and a relatively low degree of ductular reaction. Notably, significant steatosis may be observed inside and outside the lesion, with various degrees of intensity (22). Although commonly observed in women using OP contraception, telangiectatic/inflammatory (Tel/Infl) HCA are reported in patients with increased BMI and are associated with inflammatory syndrome (increased C reactive protein [CRP] or fibrinogen serum levels) (22). Interestingly, in approximately 60% of Tel/Infl HCA, IL6 signaling pathway has been shown to be activated in relation with mutations in the IL6ST gene that encodes the signaling coreceptor gp130 (23). As a matter of fact, the gain-of-function somatic mutations in gp130 may result in the inflammatory phenotype of HCA and explain activation of the acute inflammatory phase observed
LESIONS
in tumoral hepatocytes. Finally, a small group of HCA remain “unclassified” since they do not display any specific morphological or genotypical features. In contrast to FNH, and due to the potential risk for complications, surgical resection is required for HCA larger than 5 cm in diameter and all HCAs in males whatever their size. Small lesions with a low risk of complication could be initially observed after cessation of OC. Importantly, long-term follow-up of patients with unresected HCA showed a relative stability for most of them and even a significant regression in a small proportion of cases after interruption of OC (15,16). Immunophenotypical features of hepatocellular adenomas
Surrogate immunophenotypical markers related to the genetic abnormalities may be used in the classification of the 3 main subtypes of HCAs (24). Indeed, expression of liver fatty acid binding protein (LFABP), a protein positively regulated by HNF1 , is absent in steatotic HNF1 -mutated HCA, whereas it is highly expressed in non-tumoral liver. Tel/ Infl HCA display positive immunostaining with acute phase inflammatory proteins such as serum amyloid A (SAA) and CRP. Most of -catenin–mutated HCA present abnormal and nuclear staining of -catenin in tumoral hepatocytes, usually with a focal positivity restricted to few isolated tumoral hepatocytes. Immunostaining with glutamine synthetase, a
-catenin–targeted gene, showing a strong homogeneous or heterogeneous cytoplasmic staining, increases the sensibility for diagnosis of -catenin–mutated HCA. Lastly, -catenin mutations may be observed in some SAA- positive Tel/ Infl HCA, whereas gp130 activation (Tel/Infl subtype) and HNF1 inactivation (steatotic subtype) are mutually exclusive (23,24). Overall, although the diagnostic value of the different phenotypical markers is very good to excellent on paraffin sections, it is essential to compare their tissue expression in the tumor and its respective nontumoral liver in parallel. Finally, besides the 3 well-characterized subtypes, a fourth group of HCA includes the lesions without any specific morphological features nor the genetic abnormalities previously described. In surgical series of HCA, steatotic and Tel/Inf subtypes appear to be equally distributed, accounting for 85% of overall HCAs, when -catenin–mutated HCA are reported in 10% to 15% of cases. Table 23.1 summarizes the immunostaining pattern of the panel of markers used for HCA subtyping. Multiple Adenomas and Liver Adenomatosis
Review of the literature reporting patients with multiple HCA, including the so-called adenomatosis, does not support an arbitrary classification based on clinical and imaging characteristics (13,25,26). Patients with multiple HCA are predominantly females, but the use of OC appears to be less prevalent (25). Patients with glycogen storage disease type I are also at risk of developing multiple HCAs (27). Recent study confirmed that risk of complications, including bleeding and malignant transformation, is similar to that in patients with solitary HCA and is not influenced by the number of
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
TA B LE 23. 1 Immunophenotypical characteristics of benign
hepatocellular tumors Tumor
-Catenin
Glutamine Synthetase
LFABP
SAA
Focal nodular hyperplasia
/Membranous
Perivascular
Steatotic HCA
/Membranous
Focal
Tel/Infl HCA
/Membranous
Focal
HCA with cellular atypia
Nuclear/ cytoplasmic
Diffuse
Unclassified HCA
/Membranous
Focal
Normal liver
/Membranous
Centrolobular
Abbreviations: HCA, hepatocellular adenoma; LFABP, liver fatty acid binding protein; SAA, serum amyloid A; Tel/Infl HCA, telangiectactic/ inflammatory hepatocellular adenoma.
tumors (13,15,26). Except for the number of lesions, no difference is observed between imaging features of adenomatosis and solitary HCA. Interestingly, on imaging, 3 main morphologic patterns of liver adenomatosis have been described: the steatotic form, the peliotic/telangiectatic form and the mixed form (28). To note, the proportion of steatotic HCA is higher and presence of microadenomatous foci in the “nontumoral liver” is more often observed in patients with liver adenomatosis (15). As proposed for patients with solitary HCA, surgical treatment should be restricted to HCA associated with higher risk for complications. C H A L L E N GE S I N B I OP SY D I AG N OS I S
The specific biologic behavior of the benign hepatocellular lesions requires an accurate diagnosis in order to define the appropriate management consisting of no follow-up or resection for FNHs and follow-up and/or ablation for HCAs according to their characteristics. Although diagnosis and HCA subtyping may be reached in most cases by clinical and imaging approaches in experienced centers, liver biopsy is required for atypical lesions (either FNH or HCA) and in HCAs without specific radiological characteristics (mainly the -catenin–mutated HCA and the “unclassified” subtypes of HCA). On biopsy, immunophenotypical markers may be useful, especially glutamine synthetase for diagnosis of FNH and the panel (LFABP/SAA or CRP/ -catenin) for HCA subtyping. At last, 1 major issue remains: the differential diagnosis between HCA and well-differentiated HCC, which may be very difficult on biopsy. In these situations, reticulin pattern and surrogate markers, including glypican-3, can be useful tools (29,30).
References 1. Edmondson HA. Tumors of the liver and intrahepatic bile ducts. In: Atlas of Tumor Pathology. Fascicle 25 First serie, Washington, DC: Armed Forces Institute of Pathology; 1958.
LESIONS
345
2. International Working Party. Terminology of nodular hepatocellular lesions. Hepatology. 1995;22:983–993. 3. Ishak KG. Hepatic neoplasms associated with contraceptive and anabolic steroids. Carcinogenetic hormones. In: Lingeman CH ed. Recent Results in Cancer Research. New York, NY: Springer-Verlag; 1979:72–128. 4. Wanless IR, Mawdsley C, Adams R. On the pathogenesis of focal nodular hyperplasia. Hepatology. 1985;5:1194–1200. 5. Gaffey MJ, Iezzoni JC, Weiss LM. Clonal analysis of focal nodular hyperplasia of the liver. Am J Pathol. 1996;148:1089–1096. 6. Paradis V, Laurent A, Fléjou JF, Vidaud M, Bedossa P. Evidence for the polyclonal nature of focal nodular hyperplasia of the liver by the study of X chromosome inactivation. Hepatology. 1997;26:891–895. 7. Nguyen BN, Fléjou JF, Terris B, Belghiti J, Degott C. Focal nodular hyperplasia of the liver: a comprehensive pathologic study of 305 lesions and recognition of new histologic forms. Am J Surg Pathol. 1999;23: 1441–1454. 8. Makhlouf HR, Abdul-Al HM, Goodman ZD. Diagnosis of focal nodular hyperplasia of the liver by needle biopsy. Human Pathology. 2005;36: 1210–1216. 9. Bioulac-Sage P, Laumonier H, Rullier A, et al. Over-expression of glutamine synthetase in focal nodular hyperplasia: a novel easy diagnostic tool in surgical pathology. Liver Int. 2009;29:459–465. 10. Nime F, Pickren JW, Vana J, Aronoff BL, Baker HW, Murphy GP. The histology of liver tumors in oral contraceptive users observed during a national survey by the American College of Surgeons Commission on Cancer. Cancer. 1979;44:1481–1489. 11. Coombes GB, Reiser J, Paradinas FJ, Burn I. An androgen associated hepatic adenoma in a trans-sexual. Br J Surg. 1978;65:869–870. 12. Wittekind C. Hepatocellular carcinoma and cholangiocarcinoma. In: Hermanek P, Gospodarowicz MK, Henson DE, et al, eds. Prognostic Factors in Cancer. Berlin: Springer; 1995:88–93. 13. Barthelmes L, Tait IS. Liver cell adenomas and liver cell adenomatosis. HBP. 2005;7:186–196. 14. Terkivatan T, de Wilt JH, de Man RA, van Rijn RR, Tilanus HW, IJzermans JN. Treatment of ruptured hepatocellular adenoma. Br J Surg. 2001;88:207–209. 15. Dokmak S, Paradis V, Vilgrain V, et al. A single-center surgical experience of 122 patients with single and multiple hepatocellular adenomas. Gastroenterology. 2009;137:1698–1705. 16. Bioulac-Sage P, Laumonier H, Couchy G, et al. Hepatocellular adenoma management and phenotypic classification: the Bordeaux experience. Hepatology. 2009;50:481–489. 17. Paradis V, Zalinski S, Chelbi E, et al. Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant fibrosis: a pathological analysis. Hepatology. 2009;49:851–859. 18. Zucman-Rossi J, Jeannot E, Van Nhieu JT, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology. 2006; 43:515–524. 19. Bluteau O, Jeannot E, Bioulac-Sage P, et al. Bi-allellic inactivation of TCF1 in hepatic adenomas. Nat Genet. 2002;32:312–315. 20. Bacq Y, Jacquemin E, Balabaud C, et al. Familial liver adenomatosis associated with hepatocyte nuclear factor 1 alpha inactivation. Gastroenterology. 2003;125:1470–1475. 21. Wanless IR, Albrecht S, Bilbao J. Multiple focal nodular hyperplasia of the liver associated with vascular malformations of various organs and neoplasia of the brain. Mod Pathol. 1989;2:456–462. 22. Paradis V, Champault A, Ronot M, et al. Telangiectatic adenomas: an entity associated with increased body mass index and inflammation. Hepatology. 2007;46:140–146. 23. Rebouissou S, Amessou M, Thomas C, et al. Frequent in-frame somatic deletions activate gp130 in inflammatory hepatocellular tumors. Nature. 2009;457:2000–2004. 24. Bioulac-Sage P, Rebouissou S, Thomas C, et al. Hepatocellular adenoma subtype classification using molecular markers and immunohistochemistry. Hepatology. 2007;46:740–748.
346
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
25. Flejou JF, Barges J, Menu Y, et al. Liver adenomatosis: an entity distinct from liver adenoma? Gastroenterology. 1985;89:1132–1138. 26. Ribeiro A, Burgart LJ, Nagorney DM, Gores GJ. Management of liver adenomatosis: results with a conservative surgical approach. Liver Transpl Surg. 1998;4:388–398. 27. Labrune P, Trioche P, Duvaltier I, Chevalier P, Odièvre M. Hepatocellular adenomas in glycogen storage disease type I and III: a series of 43 patients and review of the literature. J Ped Gastroenterol Nutr. 1997;24: 276–279.
LESIONS
28. Lewin M, Handra-Luca A, Arrivé L, et al. Liver adenomatosis: classification pf MR imaging features and comparison with pathologic findings. Radiology. 2006;241:433–440. 29. Wang XY, Degos F, Dubois S, et al. Glypican-3 expression in hepatocellular tumors: diagnostic value for preneoplastic lesions and hepatocellular carcinomas. Hum Pathol. 2006;37:1435–1441. 30. Shafizadeh N, Ferrell LD, Kakar S. Utility and limitations of glypican-3 expression for the diagnosis of hepatocellular carcinoma at both ends of the differentiation spectrum. Mod Pathol. 2008;21:1011–1018.
Case 23.1
Atypical Focal Nodular Hyperplasias on Imaging VALÉRIE PARADIS
C L I N IC AL I N F OR M AT I ON
Asymptomatic liver nodule discovered in the United States in a 42-year-old woman. Liver function tests were normal. On imaging, the nodule of 30 mm in diameter was hypervascular, partly steatotic, and occurred in a normal background liver. A biopsy was performed for differential diagnosis between FNH and HCA. PAT H OL OG I C F E AT U R E S
The biopsy of the nodule shows hepatocellular areas crossed by fibrous bands of various thickness (Figure 23.1.1). Fibrous septa contain multiple vascular channels with few ductules and inflammatory cells (Figure 23.1.2). Some dystrophic vessels with thickened wall are present (Figure 23.1.3). Hepatocellular proliferation is made of normal looking hepatocytes arranged in liver plates of normal thickness with presence of moderate steatosis (Figure 23.1.3). Glutamine synthetase immunostaining shows a typical patchy pattern with foci of positive- stained hepatocytes (Figure 23.1.4).
F I G U R E 2 3 . 1 . 2 Several small vessels are present in the fibrous septa, as well as scattered inflammatory cells. A few small ductules are present near the interface of the fibrous tissue and hepatocytes.
D I AG N OS I S
Steatotic focal nodular hyperplasias.
D I S C U S S I ON
The pathological features observed on the biopsy, especially the presence of fibrous bands with dystrophic vessels, are
F I G U R E 2 3 . 1 . 3 Focal thick-walled muscular vessel (higher magnifi-
cation of Figure 23.1.1) is present with dystrophic morphology; note muscle bundles extending into fibrous septa to the right of the vessel. Hepatocytes are steatotic but otherwise unremarkable. Ductular reaction is very scant.
FIGURE 23. 1. 1 Fibrous band with muscular vessel separates hepato-
cellular parechyma of the lesion. Hepatocytes are steatotic.
strongly concordant for the diagnosis of FNH. Such diagnosis could be difficult on biopsy, especially for pathologists with limited experience with the lesion, where thick abnormal arteries are usually missing (1). Presence of ductular reaction at the border between fibrous septa and hepatocellular component, which can be highlighted by immunostaining with cytokeratins 7 or 19, was shown to be the most consistent
347
348
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
LESIONS
centered by hepatic veins, described as a map-like pattern, is highly suggestive of FNH (2). In spite of the presence of fibrous scar, this case was considered as atypical by radiologists, mainly because of the detection of significant steatosis on imaging modalities. Indeed, liver steatotic nodules are more often considered as HCA than FNH. However, diagnosis of FNH should not be excluded by the presence of intratumoral steatosis. Indeed, in a series of 100 consecutive needle biopsies of FNH, steatosis was observed in 25% of cases, with a moderate to severe intensity in 10% (1). Given that management of FNH and HCA are radically different, with abstention for FNH and resection for most HCA, the impact of the biopsy in the positive diagnosis of FNH is highly valuable and of clinical importance. FIGURE 23. 1. 4 Glutamine synthetase immunostain demonstrates a prominent, but patchy, pattern of staining in much of the lesion on the biopsy sample.
diagnostic histological feature for FNH (1). In addition, and as illustrated in this case, immunostaining with glutamine synthetase showing focal- positive hepatocellular areas usually
References 1. Makhlouf HR, Abdul-Al HM, Goodman ZD. Diagnosis of focal nodular hyperplasia of the liver by needle biopsy. Hum Pathol. 2005;36: 1210–1216. 2. Bioulac-Sage P, Laumonier H, Rullier A, et al. Over-expression of glutamine synthetase in focal nodular hyperplasia: a novel easy diagnostic tool in surgical pathology. Liver Int. 2009;29:459–465.
Case 23.2
Focal Nodular Hyperplasia Versus Inflammatory/Telangiectatic Hepatocellular Adenoma VALÉRIE PARADIS
C L I N IC AL I N F OR M AT I ON
A 36-year-old woman presented with a 6-cm nodule in the right liver. There were no symptoms, and liver function tests were normal. On imaging, lesion was atypical for classic FNH due to the absence of fibrous scar. The biopsy of the nodule was performed in order to exclude the possibility of an HCA. PAT H OL OG I C F E AT U R E S
The biopsy of the nodule shows a hepatocellular proliferation with few small vessels included and discrete sinusoidal dilatation (Figure 23.2.1). The hepatocellular proliferation is vaguely nodular with thin liver cell plates at the periphery, and focally steatotic (Figure 23.2.2). The vessels are thin, usually unpaired, slightly surrounded by extracellular matrix component (Figure 23.2.3). Very few foci of ductular reaction are visible (Figure 23.2.3). The tumoral hepatocytes are regular, arranged in normal thickness plates with very few lymphocytes observed (Figure 23.2.3). Glutamine synthetase immunostaining shows the typical patchy pattern (Figure 23.2.4) without any SAA staining (Figure 23.2.5).
F I G U R E 2 3 . 2 . 2 The lesion has a vague nodularity, with thinner
hepatic plates at the periphery of the “nodules.” Some hepatocytes contain fat.
D I AG N OS I S
FNH without fibrous scar.
F I G U R E 2 3 . 2 . 3 Higher magnification of Figure 23.2.5 showing “un-
paired” vessel (no accompanying bile duct). Note lack of ductular reaction at interface of fibrous tissue and hepatocytes.
DISCUSSIO N
FIGURE 23. 2. 1 Hepatocytic lesion with a focus of a few small ves-
sels without accompanying duct. Focal mild sinusoidal dilation is also present.
This case illustrates one variant of FNH that lacks the classic fibrous central scar. Indeed, presence of fibrous scar is a major diagnostic feature of FNH on imaging. Pathological analysis on the surgical specimen of FNH without a macroscopic central scar may be able to show several thin and short fibrous septa that are
349
350
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
FIGURE 23. 2. 4 Glutamine synthetase immunostain. Prominent, but
patchy, pattern of staining as seen in previous case (Figure 23.2.4). Note cores on the left are nonlesional tissue with only limited perivenous positivity.
LESIONS
most of them, have been initially considered as atypical forms of FNH, so-called telangiectatic form of FNH (3). Interestingly, molecular studies, based on clonal analysis and gene expression analysis, demonstrated that these kinds of lesions were clonal processes with gene expression pattern similar to HCA (4). These results led to reconsideration of “telangiectatic form of FNH” as Tel/Infl HCA, a distinct subtype of HCA (4,5). Later on, immunohistochemical classification of the different subtypes of HCA, based on the correlation between genotype and phenotype, was developed (6,7). Tel/Infl HCA, which are morphologically characterized by intratumoral vascular changes and inflammatory foci, display overexpression of markers of the acute phase inflammatory reaction, including SAA and CRP (7). Interestingly, we observed a systemic biological inflammatory syndrome in patients with Tel/Infl HCA that has resolved after tumor resection (8). Therefore, in addition to glutamine synthetase, immunohistochemical analysis with inflammatory markers may be helpful, especially on biopsy samples, in the differential diagnosis between FNH and Tel/Infl HCA. In Tel/ Infl HCA, SAA immunostaining is restricted to the tumoral hepatocytes, with high specificity and sensitivity, even though focal staining has been described in non-tumoral liver in a few cases. Finally, this case perfectly illustrates the significant contribution of immunophenotyping in the diagnosis of atypical FNH.
References
FIGURE 23. 2. 5 Serum amyloid A immunostain. Negative for granular cytoplasmic staining to exclude Tel/Infl HCA.
usually lacking on biopsy samples. Nevertheless, and as discussed above, additional key features have to be carefully screened, such as the ductular reaction and the map-like pattern of glutamine synthetase immunostaining (1,2). The main differential diagnosis of FNH without macroscopic central fibrous scar, for radiologists, and pathologists as well, is the Tel/Infl subtype of HCA. Tel/Infl HCA, at least
1. Makhlouf HR, Abdul-Al HM, Goodman ZD. Diagnosis of focal nodular hyperplasia of the liver by needle biopsy. Hum Pathol. 2005;36: 1210–1216. 2. Bioulac-Sage P, Laumonier H, Rullier A, et al. Over-expression of glutamine synthetase in focal nodular hyperplasia: a novel easy diagnostic tool in surgical pathology. Liver Int. 2009;29:459–465. 3. Wanless IR, Albrecht S, Bilbao J. Multiple focal nodular hyperplasia of the liver associated with vascular malformations of various organs and neoplasia of the brain. Mod Pathol. 1989;2:456–462. 4. Paradis V, Benzekri A, Dargère D, et al. Telangiectatic focal nodular hyperplasia: a variant of hepatocellular adenoma. Gastroenterology. 2004;126:1323–1329. 5. Bioulac-Sage P, Rebouissou S, Sa Cunha A, et al. Clinical morphologic, and molecular features defining so-called telangiectatic focal nodular hyperplasias of the liver. Gastroenterology. 2005;128:1211–1218. 6. Zucman-Rossi J, Jeannot E, Van Nhieu JT, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology. 2006;43:515–524. 7. Bioulac-Sage P, Rebouissou S, Thomas C, et al. Hepatocellular adenoma subtype classification using molecular markers and immunohistochemistry. Hepatology. 2007;46:740–748. 8. Paradis V, Champault A, Ronot M, et al. Telangiectatic adenomas: an entity associated with increased body mass index and inflammation. Hepatology. 2007; 46:140–146.
Case 23.3
Hepatocellular Adenoma Subtyping: Inflammatory/Telangiectatic Versus Steatotic Adenoma VALÉRIE PARADIS
C L I N IC AL I N F OR M AT I ON
A 23-year-old woman presented with several liver nodules (>5 on imaging), one of which was 10 cm in diameter in the right lobe of an otherwise normal liver. Imaging features of all nodules was consistent with steatotic HCA. The largest nodule was biopsied before resection.
PAT H OL OG I C F E AT U R E S
The biopsy shows a very-well differentiated hepatocellular proliferation with some sinusoidal dilatation (Figure 23.3.1). Some mononuclear infiltrates are focally noted (Figure 23.3.2). The tumoral hepatocytes are regular with a trabecular pattern, without significant steatosis observed (Figure 23.3.2). SAA immunostaining demonstrates a cytoplasmic granular positivity of the tumoral hepatocytes (Figure 23.3.3). The liver proliferation also displays a positive staining with LFABP (cytoplasmic and nuclear) (Figure 23.3.4). Resection specimen (right hepatectomy) shows multiple nodules ranging from 1 to 10 cm. The largest nodule is unencapsulated, nodular shaped, yellowish with telangiectatic changes (Figure 23.3.5). On hematoxylin and eosin staining, the liver tumor is characterized by regular hepatocellular proliferation containing clusters of vascular channels surrounded by extracellular matrix with few inflammatory cells. Note the presence of steatotic hepatocytes (Figure 23.3.6).
F I G U R E 2 3 . 3 . 2 Focal mononuclear inflammation, a feature that is commonly present in this lesion.
F I G U R E 2 3 . 3 . 3 Serum amyloid A immunostain demonstrates posi-
tive staining as represented by relatively diffuse and prominent cytoplasmic granules in lesional hepatocytes.
DIAGNO SIS FIGURE 23. 3. 1 Biopsy sample of the lesion demonstrates relatively
normal-appearing hepatocytic plate architecture, but sinusoids are significantly dilated.
351
Tel/Infl HCA in the context of liver adenomatosis.
352
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
FIGURE 23. 3. 4 Liver fatty acid binding protein immunostain
demonstrates positivity in both cytoplasm and nuclei in lesional hepatocytes.
FIGURE 23. 3. 5 Resection specimen demonstrates a large pale, yellow tan lesion with somewhat irregular, but rounded, borders and mottled vascular/telangiectatic pattern.
D I SC U SSI ON
Liver adenomatosis was described as a separate entity characterized by the presence of multiple HCA (arbitrarily more than 10), observed both in men and women with no relationship to the use of OC (1). Later, it appeared that such classification was no longer relevant on the basis of demographics and clinical evolution, especially regarding rate of complications (2). Germline mutations of HNF1 were initially reported in cases of familial liver adenomatosis (3). Although a steatotic subtype of HCA is more prevalent in
LESIONS
F I G U R E 2 3 . 3 . 6 The resected tumor contains clusters of vascular channels in fibrous matrix with few inflammatory cells. Some fat is also present in tumoral hepatocytes.
patients with liver adenomatosis, other subtypes may be observed, especially Tel/Infl HCA, as illustrated in our case. Thus, imaging analysis of liver adenomatosis has described 3 different patterns: steatotic, peliotic (or telangiectatic), and mixed forms (4). Interestingly, the main pathological feature associated with the presence of multiple HCAs is the presence of microadenomas in the adjacent liver as observed in our case (5). Lastly, the rate of significant complications (hemorrhage and malignancy) is not related to the number of HCA but to its size and its pathological subtype with a lower risk for steatotic HCA. Although not present on the biopsy, significant intratumoral steatosis was observed both on macroscopy and histology of the surgical resection, highlighting that steatosis is not restricted to the steatotic LFABP-negative subtype of HCA. Therefore, significant steatosis (>10%) has been reported in at least half of Tel/Infl HCA, with moderate to severe intensity in about 30% of cases (6). In cases of Tel/Infl HCA, presence of intratumoral steatosis may be related to the clinical context of the patient since this HCA subtype is more prevalent in patients with increased body mass index and metabolic syndrome (6). On the other hand, the degree of steatosis may be highly variable in steatotic HNF1 -mutated HCA, ranging from less than 10% to more than 60% (7). Indeed, in the original study correlating genotype and phenotype of HCA, only 36% of HNF1 -mutated HCA displayed severe steatosis, as much as 60% (7). Lastly, this case also raises the major issue of biopsy sampling of benign hepatocellular tumors, especially for HCA subtyping. Indeed, whereas imaging diagnosis was steatotic HCA, no significant steatosis was observed on the biopsy of the tumoral nodule despite its representativity. Altogether, these limitations emphasize the importance of systematic immunohistochemistry for subtyping HCA on biopsy specimens.
CASE
23.3:
HCA
SUBTYPING:
INF/TEL
References 1. Flejou JF, Barges J, Menu Y, et al. Liver adenomatosis: an entity distinct from liver adenoma? Gastroenterology. 1985;89:1132–1138. 2. Veteläinen R, Erdogan D, de Graaf W, et al. Liver adenomatosis: reevaluation of aetiology and management. Liver Int. 2008;28:499–508. 3. Bacq Y, Jaquemin E, Balabaud C, et al. Familial liver adenomatosis associated with hepatocyte nuclear factor 1 alpha inactivation. Gastroenterology. 2003;125:1470–1475. 4. Lewin M, Handra-Luca A, Arrivé L, et al. Liver adenomatosis: classification pf MR imaging features and comparison with pathologic findings. Radiology. 2006;241:433–440.
VERSUS
S T E AT O T I C
ADENOMA
353
5. Dokmak S, Paradis V, Vilgrain V, et al. A single-center surgical experience of 122 patients with single and multiple hepatocellular adenomas. Gastroenterology. 2009;137:1698–1705. 6. Paradis V, Champault A, Ronot M, et al. Telangiectatic adenomas: an entity associated with increased body mass index and inflammation. Hepatology. 2007;46:140–146. 7. Zucman-Rossi J, Jeannot E, Van Nhieu JT, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology. 2006;43:515–324.
Case 23.4
Hepatocellular Adenoma Subtyping: Associated Liver Nodules VALÉRIE PARADIS
C L I N I C AL I N F OR M AT I ON
A 68-year-old woman presented with 3-fold increase in gamma glutamyl transpeptidase (GGT). Liver imaging revealed 3 nodules, including 2 centimetric lesions (segments II and IV) suggestive of FNH and 1 larger nodule (5 cm in its largest axis) appearing steatotic. Nontumoral liver was possibly slightly dysmorphic. Biopsy of the largest nodule and nontumoral liver was performed before resection.
PAT H OL OG I C F E AT U R E S
The biopsy demonstrates both nodule and adjacent nontumoral liver. The nodule corresponds to a well-differentiated hepatocellular proliferation mainly composed of steatotic hepatocytes (Figure 23.4.1). The steatotic proliferation contains small, unpaired arteries (Figure 23.4.2). The tumor is composed of more than 80% of steatotic hepatocytes, with large and small droplet patterns of steatosis (Figure 23.4.2). Nontumoral liver displays normal architecture without any steatosis (Figure 23.4.3). LFABP immunostaining demonstrates absent labeling of the hepatocellular proliferation compared with the positivity of the nontumoral hepatocytes (cytoplasmic and nuclear labeling) (Figure 23.4.4). Macroscopic view of the resected nodule shows a multinodular tan nodule of 5 cm in its largest axis (Figure 23.4.5). Hematoxylin and eosin (H&E) staining of the resected specimen confirms the diagnosis of steatotic HCA (Figure 23.4.6).
F I G U R E 2 3 . 4 . 2 Biopsy sample of lesion shows that small unpaired
arteries and hepatocytes contain both large and small droplet fat. Glycogenated nuclei are also present.
F I G U R E 2 3 . 4 . 3 Biopsy sample of nonlesional tissue (note portal
zone on right) shows no fat in hepatocytes.
DIAGNO SIS
Steatotic LFABP-negative HCA.
DISCUSSIO N FIGURE 23. 4. 1 Biopsy of the lesion demonstrates fatty hepatocytes.
This case illustrates a typical case of steatotic HNF1 -mutated HCA. This HCA subtype is the most frequently encountered
354
CASE
23.4:
HCA
SUBTYPING:
FIGURE 23. 4. 4 LFABP immunostain is negative in lesional hepatocytes (with fat) as compared with the positive nuclear and cytoplasmic staining of adjacent normal liver without fat.
FIGURE 23. 4. 5 Resection specimen demonstrates a multinodular
tan tumor.
with the Tel/Infl subtype, accounting for approximately 40% to 45% of HCA, respectively. As far as complications are concerned, it appears that steatotic HNF1 -mutated HCA are less prone to progress to hepatocellular carcinoma (HCC) compared with Tel/Infl and -catenin–activated subtypes (1,2). As a matter of fact, almost all of them occur in women, and, interestingly, it has been shown that HNF1 and gp130 mutations (characterizing Tel/Infl HCA) are mutually exclusive (3). Accordingly, a more conservative clinical approach would be
A S S O C I AT E D
LIVER
NODULES
355
F I G U R E 2 3 . 4 . 6 Resected tumor shows extensive fatty change.
considered for steatotic HNF1 -mutated HCA except for the very large tumors, as size remains one of the main risk factors for malignancy. Imaging and pathological diagnosis of steatotic HNF1 mutated HCA should be, in theory, the easiest to do, given that steatosis, a well-recognized feature, is the main morphological hallmark. Nevertheless, and as discussed above (Chapter 23.3), steatosis may be not so prominent. In these cases, immunophenotypical analysis is helpful, showing absence of staining in tumoral hepatocytes contrasting with nontumoral hepatocytes that display cytoplasmic and sometimes additional nuclear positivity for LFABP. Comparison of immunostaining between tumor and nontumoral counterpart is required. In the screening of patients with HCA, additional tumors of the liver may be detected. Among them, hemangiomas, the most frequent benign tumors of the liver and FNH are commonly reported. As illustrated in our case, the presence of steatotic HNF1 -mutated HCA and FNH is quite a frequent association as well, reaching 24% of cases (1).
References 1. Bioulac-Sage P, Laumonier H, Couchy G, et al. Hepatocellular adenoma management and phenotypic classification: the Bordeaux experience. Hepatology. 2009;50:481–489. 2. Paradis V, Champault A, Ronot M, et al. Telangiectatic adenomas: an entity associated with increased body mass index and inflammation. Hepatology. 2007; 46:140–146. 3. Rebouissou S, Amessou M, Thomas C, et al. Frequent in-frame somatic deletions activate gp130 in inflammatory hepatocellular tumors. Nature. 2009;457:2000–2004.
Case 23.5
Hepatocellular Adenoma Subtyping: Inflammatory/Telangiectatic Adenoma VALÉRIE PARADIS
C L I N I C AL I N F OR M AT I ON
A 23-year-old woman on oral contraceptives for 8 years presented with back pain. Imaging revealed a liver nodule of 5 cm in segment VI with steatotic features suggestive of an HCA. Liver function tests were normal. A liver biopsy was performed followed by resection. PAT H OL OG I C F E AT U R E S
Biopsy of the nodule shows a well-differentiated hepatocellular proliferation with a vaguely nodular pattern, containing quite numerous thin vessels (Figure 23.5.1). Some areas demonstrate steatotic hepatocytes (Figure 23.5.2). In other places, hepatocellular proliferation is composed of clear hepatocytes arranged in liver plates slightly increased or with pseudoglandular formation (Figure 23.5.3). Presence of few inflammatory infiltrates throughout the nodule is noted (Figure 23.5.4). There is no significant cellular atypia. -catenin immunostaining demonstrates only a few tumoral hepatocytes with a nuclear positivity (Figure 23.5.5). Large areas of hepatocellular proliferation are strongly reactive with glutamine synthetase (Figure 23.5.6). SAA immunostaining shows very few positive tumoral hepatocytes (Figure 23.5.7). Macroscopic view of the resected nodule shows a 5 cm nonencapsulated yellowish nodule with few congestive changes (Figure 23.5.8). Pathological analysis of the surgical specimen clearly demonstrates that the hepatocellular proliferation contains the presence of small clusters of slightly dystrophic vessels surrounded by extracellular matrix and inflammatory infiltrates (Figure 23.5.9).
F I G U R E 2 3 . 5 . 2 Biopsy of the lesion at higher magnification high-
lights isolated unpaired vessels and fatty change of tumoral hepatocytes. No cellular atypia is present.
F I G U R E 2 3 . 5 . 3 Biopsy sample demonstrates that hepatic plates are mildly thickened and show focal pseudoglandular formation.
DIAGNO SIS FIGURE 23. 5. 1 Biopsy shows a hepatocellular proliferation with vaguely nodular pattern and numerous thin-walled vessels.
356
Tel/Infl HCA with -catenin activation.
CASE
23.5:
HCA
SUBTYPING:
INF/TEL
ADENOMA
357
F I G U R E 2 3 . 5 . 6 -catenin immunostain of biopsy demonstrates a few
FIGURE 23. 5. 4 Biopsy sample. The lesion also contains a few inflammatory infiltrates. In addition, note again the small vascular channels.
positive nuclei.
FIGURE 23. 5. 5 Glutamine synthetase immunostain of biopsy shows
F I G U R E 2 3 . 5 . 7 Serum amyloid A immunostain shows only minimal, focal cytoplasmic staining of lesional hepatocytes.
large zone of strong positivity.
D I S C U S S I ON
Tel/Infl with steatotic HNF1 -mutated tumors represent the most prevalent subtypes of HCA. As already discussed, Tel/Infl HCA are predominantly observed in young women with a long history of OC use (1). In addition, this subtype is frequently observed in patients with increased body mass index that may explain presence of steatosis (at least >30%) in adjacent nontumoral liver in one-third of cases (1,2). Although diagnosis of HCA was confident on the biopsy, subtyping into Tel/Infl HCA is not so evident and could be suggested based on the vaguely nodular architecture of the proliferation associated with the presence of few inflammatory foci. Indeed, the clusters of vessels surrounded by extracellular matrix,
observed on the surgical specimen, were lacking on the biopsy sample. Additional immunohistochemical analysis searching for SAA staining in tumoral cells was of potential value in this case, showing focal positivity in one area. More importantly, surrogate histological features have to be taken into account, especially the presence of pseudoglandular formation. Indeed, this feature has been described in -catenin–activated liver tumors, including HCA and HCC as well (3,4). Therefore, and independently of the presence of cellular atypia, -catenin pattern has to be checked. Immunoprofiles of -catenin and glutamine synthetase were positive and so consistent with diagnosis of
-catenin–activated HCA. In fact, glutamine synthetase is useful to improve diagnostic accuracy of -catenin activation given that -catenin staining may be difficult to interpret (focal
358
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
LESIONS
FIGURE 23. 5. 8 Resection sample reveals a white to yellow tan, vaguely nodular tumor with irregular mottled vascular pattern.
F I G U R E 2 3 . 5 . 9 Resected tumor shows slightly dystrophic vessels and
nuclear staining) and has much lower sensitivity compared with glutamine synthetase (2). Among Tel/Infl HCAs, a proportion may harbor
-catenin activation (5). Interestingly, no major differences appear between patients with Tel/Infl HCA according to the presence of -catenin activation, except for the presence of HCC (5). In the analysis of a series of HCC that developed on a pre-existing HCA, we observed that the Tel/Infl subtype was the more frequent (52%), and that approximately half of them displayed -catenin activation (personal data). Altogether, these data support that Tel/Infl subtype and presence of
-catenin activation are the main risk factors for progression to HCC.
References
inflammatory infiltrates scattered throughout the lesion.
1. Paradis V, Champault A, Ronot M, et al. Telangiectatic adenomas: an entity associated with increased body mass index and inflammation. Hepatology. 2007;46:140–146. 2. Bioulac-Sage P, Rebouissou S, Thomas C, et al. Hepatocellular adenoma subtype classification using molecular markers and immunohistochemistry. Hepatology. 2007;46:740–748. 3. Zucman-Rossi J, Jeannot E, Van Nhieu JT, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology. 2006;43:515–524. 4. Audard V, Grimber G, Elie C, et al. Cholestasis is a marker for hepatocellular carcinomas displaying beta-catenin mutations. J Pathol. 2007;212:345–352. 5. Bioulac-Sage P, Laumonnier H, Couchy G, et al. Hepatocellular adenoma management and phenotypic classification: the Bordeaux experience. Hepatology. 2009;50:481–489.
Case 23.6
Hepatocellular Adenoma Subtyping: Adenoma With Atypical Features VALÉRIE PARADIS
C L I N IC AL I N F OR M AT I ON
A 52-year-old man referred for a unique nodule of 12 cm in diameter of the right lobe developed in a background normal liver. Alpha-fetoprotein (AFP) is normal. Right hepatectomy was performed. PAT H OL OG I C F E AT U R E S
Macroscopic analysis of the right hepatectomy identifies a well-limited, unencapsulated nodule of 12 cm in its largest axis. The nodule is lobulated, heterogeneous with cholestatic changes. Adjacent liver parenchyma was macroscopically normal (Figure 23.6.1). On HE staining, the hepatocellular proliferation displays a very well-differentiated trabecular pattern with pseudoglandular formation (Figure 23.6.2). The tumor contains numerous arteries, usually either thin and unpaired (Figure 23.6.3). Significant cellular atypia is focally observed with cholestasis as well (Figure 23.6.4). Reticulin staining shows a preserved reticulin framework (Figure 23.6.5).
-catenin immunostaining demonstrates aberrant nuclear positivity in some tumoral hepatocytes (Figure 23.6.6). SAA immunostaining was negative.
F I G U R E 2 3 . 6 . 2 Tumoral hepatocytes are arranged in thin trabecularlike pattern with focal pseudoglandular formation.
D I AG N OS I S
-catenin–activated HCA with cell atypia.
F I G U R E 2 3 . 6 . 3 Higher magnification of pseudoglandular formation.
Note the isolated unpaired artery.
DISCUSSIO N
FIGURE 23. 6. 1 Resection specimen demonstrates a mass with heterogeneous brown, hemorrhagic appearance in a background of normal liver.
This case reports a hepatocellular proliferation occurring in a male patient. Specific context, including androgen treatment, Fanconi’s anemia, or use of recreational steroids are considered as significant risk factors in men for developing hepatocellular tumors (1,2). More recently, several cases of HCC arising in a pre-existing HCA have been reported in patients with metabolic syndrome (3).
359
360
CHAPTER
23:
BENIGN
H E PAT O C E L L U L A R
LESIONS
FIGURE 23. 6. 4 Note significant nuclear atypia as compared with
F I G U R E 2 3 . 6 . 6 b-catenin immunostain shows nuclear positivity in
that seen in the previous cases. Cholestasis is also present.
some tumoral hepatocytes consistent with b-catenin–mutated variant of HCA.
FIGURE 23. 6. 5 Reticulin stain demonstrates intact framework, including around the pseudoglands.
morphological features on imaging, and the final diagnosis relies exclusively on the pathological analysis. Differential diagnosis between well-differentiated HCC and HCA with cellular atypia remains a major issue, especially on biopsy specimen. Additional features, including the reticulin pattern and surrogate immunophenotypical markers such as glypican-3, may be helpful for that purpose. Further molecular screening for chromosomal abnormalities has been recently shown to be helpful in the diagnosis of atypical hepatocellular neoplasms, defined as “HCA with focal atypical morphological features or unusual clinical settings, including HCA occurring in men” (5). Indeed, those atypical HCA share similar chromosomal alterations to HCC. Given all these considerations, a biopsy diagnosis has poor predictive value and due to the high potential for these HCA to transform to HCC, systematic resection should be considered in all cases, independent of their size.
References According to the genotype-phenotype classification of HCA, when morphologically characterized by presence of cellular atypia and pseudo-glandular formation, as illustrated in our case, HCA displayed -catenin mutations and were predominantly described in male patients (4). In this subgroup of tumors, several were considered borderline, difficult to classify between HCA and HCC or associated with HCC. The link between -catenin mutations and the higher risk for malignant transformation into HCC is now well established and reinforced by recent data showing -catenin activation in 64% of HCA with superimposed HCC (personal data). By contrast to the other subtypes (Tel/Infl and steatotic), -catenin–activated HCA do not present specific
1. Velazquez I, Alter BP. Androgens and liver tumors: Fanconi’s anemia and non-Fanconi’s conditions. Am J Hematol. 2004;77:257–267 2. Gorayski P, Thompson CH, Subhash HS, Thomas AC. Hepatocellular carcinoma associated with recreational anabolic steroid use. Br J Sports Med. 2008;42:74–75. 3. Paradis V, Zalinski S, Chelbi E, et al. Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant fibrosis: a pathological analysis. Hepatology. 2009;49:851–859. 4. Zucman-Rossi J, Jeannot E, Van Nhieu JT, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classification and relationship with HCC. Hepatology. 2006;43:515–524. 5. Kakar S, Chen X, Ho C, et al. Chromosomal abnormalities determined by comparative genomic hybridization are helpful in the diagnosis of atypical hepatocellular neoplasms. Histopathology. 2009;55:197–205.
24 Biliary Neoplasms KISHA MITCHELL AND DHANPAT JAIN
Most of the benign biliary tumors remain asymptomatic and are discovered incidentally. The malignant tumors also remain asymptomatic in the early stages leading to late detection and poor prognosis. Adenocarcinoma of the bile duct, also referred to as cholangiocarcinoma (CC), is the second most common primary epithelial malignancy of the liver after hepatocellular carcinoma (HCC), accounting for about 15% of all primary malignant hepatic neoplasms (1). Based on the location, it is classified into intrahepatic (so-called
peripheral CC), hilar (Klatskin tumor), and extra hepatic. Despite many common features, there are important differences with regard to etiopathogenesis and management such that the classification by the site is clinically important. Of these, extrahepatic CC constitutes about 40% of all tumors, whereas hilar and intrahepatic tumors constitute 30% each (2). Some use the term CC purely for the intrahepatic tumors, whereas others use this term to designate all bile duct adenocarcinomas.
FIGURE 24. 1 Mass-forming cholangiocarcinoma (CC): (A) solid, unencapsulated but well-defined white, intrahepatic mass. (B) Periductal,
infiltrating CC: The white tumor grows along bile ducts resulting in obstruction and dilatation of the distal bile ducts. The tumor extends into the parenchyma to form a fairly well-defined mass. (C) Periductal stricturing type CC involving the hilar bile ducts (Klatskin’s tumor). No mass is demonstrably present. On the left side, the bile cut is dilated distal to the ill-defined tumor that is present in the region of the suture. These patients often present with obstructive jaundice. (D) Intraductal papillary CC: fleshy polypoid masses within the biliary system. In this case, the tumor is entirely intraductal without parenchymal invasion.
361
362
CHAPTER
24:
BILIARY
NEOPLASMS
E P I D E M I OL OG Y
R ISK FACTO R S A ND ET IO LO GY
The incidence of CC is variable worldwide with highest incidence reported from Southeast Asia, Chile, and Japan, largely secondary to parasitic infections of the liver (3,4). In Thailand, CC occurs at an incidence rate of 88 per 100 000 in men and 37 per 100 000 in women (1). The parasitic infections associated with CC are Opisthorchis viverrini and Clonorchis sinensis, which are endemic in these parts of the world. Even in United States, the incidence of CC is increasing and has reportedly risen by 165% between 1975 and 1999 (4). CC typically occurs in the elderly, presenting most commonly in the seventh decade of life and is more common in men than women (M:F ratio is 1.5:1). In the United States, the highest incidence occurs in Hispanics and the lowest in African Americans (4). This increase in incidence has been similarly accompanied by an increase in mortality. In Italy, a 40-fold increase in mortality was reported between 1980 and 2003 (4). Overall, the prognosis of CC remains poor.
The vast majority of CCs (80–85%) arise in patients without any known risk factor or background of cirrhosis (4,5). Known risk factors include chronic biliary inflammation, liver disease, and congenital disorders (6). Primary sclerosing cholangitis (PSC) is the most consistent risk factor associated with CC in Western countries. Patients with PSC have a cumulative annual risk of 1.5% per year for CC after the development of jaundice (5). The majority of PSC-associated CC cases are diagnosed within a year following diagnosis of PSC, and 27% of patients are diagnosed following liver transplantation (7). Additional risk factors in the setting of PSC include older age at diagnosis (though PSC patients develop CC younger than the general population) (8), history of inflammatory bowel disease with dysplasia or malignancy, smoking, and alcohol use (4). Infestation with liver parasites (Opisthorchis viverrini, Clonorchis sinensis, and Schistosoma japonicum) is also a significant risk factor with increased prevalence in endemic areas
FIGURE 24. 2 (A) Well-differentiated, conspicuous gland-forming cholangiocarcinoma (CC) with focally cribriform glands. (B) Moderately differentiated CC with narrow tubules and ducts embedded in dense, collagenous stroma. (C) CC with extensive perineural infiltration. (D) Undifferentiated CC composed of bizarre cells without glandular differentiation.
CHAPTER
24:
BILIARY
(9). Presumably, the ensuing chronic inflammation following colonization by the parasite predisposes to risk of malignancy. Similarly, hepatolithiasis is also thought to result in increased risk of CC. Patients who have late resection (>20 years) of a choledochal cyst are also at increased risk of CC, as are patients who have abnormal pancreaticobiliary junctions and various other biliary malformations including Caroli disease. Exposure to toxic compounds such as Thorotrast (thorium dioxide) also predisposes to CC, although this agent has been out of use for decades and new cases attributed to it are unlikely to be encountered. Cirrhosis and, in particular, chronic hepatitis C virus (HCV) are also risk factors for CC with 10-fold and 4-fold higher risks reported respectively (5). PAT H OL OG Y
Though cirrhosis is a risk factor, CC most comonly arises in a noncirrhotic liver. Based on the gross appearance, it is often
NEOPLASMS
363
classified into 3 types (Figures 24.1A–D). The mass-forming type is the commonest and is represented by a gray-white, firm, solid intrahepatic mass (Figure 24.1A). The periductal infiltrating-type rarely forms a big mass. It spreads along portal tracts resulting in biliary strictures of involved ducts (Figures 24.1B and 24.1C). The intraductal-type is typically a polypoid mass that grows predominantly within the duct lumen (Figure 24.1D). Any combination of the three may be seen in a single case. Histologically, CC is an adenocarcinoma and may have many different histologic patterns with many recognized histologic subtypes (Figures 24.2 and 24.3) (2). CC may be very well differentiated (Figure 24.2A); however, the most common pattern is infiltrating, well to moderately differentiated with tubular or glandular structures dispersed within a very dense fibrotic stroma (desmoplasia) (Figure 24.2B), often with conspicuous perineural invasion (Figure 24.2C). The glandular component can also show cribriform, nesting, cord-like, or
FIGURE 24. 3 (A) Intraductal papillary cholangiocarcinoma (CC): The tumor is entirely within the lumen of the bile duct with a papillary
configuration. (B) Clear cell CC: extensive clear cytoplasmic change with small, often basally located nuclei; areas of gland formation are present. (C) von Meyenburg complex (VMC)-like pattern: dilated and angulated tubules with eosinophilic luminal material within a fibrotic stroma. (D) Oncocytic CC: abundant deeply eosinophilic cytoplasm and only mildly atypical nuclei.
364
CHAPTER
24:
BILIARY
NEOPLASMS
Variant
Morphology
Typical
Tubular or papillary structures, often within desmoplastic stroma cribriforming mucin
Adenosquamous and squamous
Significant squamous component with keratin and/or intercellular bridges, often advanced stage at diagnosis
Mucinous
Abundant extracellular mucin with floating tumor clusters
Signet-ring cell
Predominant component is signet-ring cells within mucin lakes
immunohistochemistry. CC can be differentiated from HCC in most cases without significant difficulty, largely based solely on morphology or aided by a small panel of immunohistochemical markers. The greater difficulty often lies in differentiating CC from other metastatic carcinomas. Specific immunohistochemical markers for CC are lacking, and its separation from metastatic carcinomas requires the demonstration of a specific differentiation marker for other tumor types (eg, thyroid transcription factor, prostate specific antigen, gross cystic disease fluid protein, etc.). In practice, the tumor morphology and the clinical setting should dictate the choice of the immunohistochemical panel.
Sarcomatous
Predominant spindle cell component that is malignant fibrous histiocytoma or fibrosarcoma-like
T R EAT MENT A ND P RO GNO SIS
Lymphoepithelioma-like
Identical to extrahepatic variant with rich lymphoplasmacytic stroma
Clear cell
Predominantly abundant clear cytoplasm
Mucoepidermoid
Resembles salivary gland tumors with epidermoid, mucous secreting, and intermediate cells
TA B LE 24. 1 Variants of cholangiocarcinoma
Complete surgical resection offers the only chance of cure for these aggressive tumors, although the majority of cases at presentation are unresectable. Liver transplant for CC is offered only for a subset of patients that fulfill very strict criteria at limited centers and still remains controversial.
From Ref. 1.
papillary patterns. Poor differentiation is not uncommon, and some tumors may even be undifferentiated (Figure 24.2D). Other variants described in CC include mucinous, squamous, clear cell, papillary, oncocytic, or spindle/sarcomatoid (Figure 24.3A–D) (Table 24.1) (1). The immunophenotype of CC is not very distinctive. CC is positive for cytokeratins (CK) 7 and 19 and mucin core (MUC) proteins 1, 2, and 3 by immunohistochemistry. Positivity for CK20 may also be seen in a subset of CCs. They are also immunoreactive for MOC 31 and carcinoembryonic antigen (CEA). The positivity for polyclonal CEA (pCEA) tends to be cytoplasmic and surface, which is different from canalicular positivity of HCC. CC is negative for hepatocyte antigen (Hep Par 1), glypican-3, albumin, and AFP. DIF FE R E N T I A L D I AG N OSI S A N D P R AC T I C A L I SSUES
Most commonly, CC needs to be distinguished from benign biliary tumors on one hand and metastatic carcinoma and HCC on the other. The differentiation of CC from benign lesions is largely dependent on histomorphology, whereas differentiation from other malignant tumors often requires
References 1. Nakanuma Y, Curado M, Franceschi S, Gores G, Pradis V, Sripa B. Intrahepatic Cholangiocarcinoma. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. WHO Classification of tumors of the digestive system. Lyon: IARC Press; 2010. P217-224. 2. Goodman ZD. Neoplasms of the liver. Mod Pathol. 2007;(20 suppl 1): S49–S60. 3. Shaib Y, El-Serag HB. The epidemiology of cholangiocarcinoma. Semin Liver Dis. 2004;24(2):115–125. 4. Gatto M, Bragazzi MC, Semeraro R, et al. Cholangiocarcinoma: update and future perspectives. Dig Liver Dis. 2010;42(4):253–260. 5. Aljiffry M, Walsh MJ, Molinari M. Advances in diagnosis, treatment and palliation of cholangiocarcinoma: 1990–2009. World J Gastroenterol. 2009;15(34):4240–4262. 6. Bosch FX, Ribes J, Diaz M, Cleries R. Primary liver cancer: worldwide incidence and trends. Gastroenterology. 2004;127(5 suppl 1):S5–S16. 7. Boberg KM, Bergquist A, Mitchell S, et al. Cholangiocarcinoma in primary sclerosing cholangitis: risk factors and clinical presentation. Scand J Gastroenterol. 2002;37(10):1205–1211. 8. Bergquist A, Broome U. Hepatobiliary and extra-hepatic malignancies in primary sclerosing cholangitis. Best Pract Res Clin Gastroenterol. 2001;15(4):643–656. 9. Poomphakwaen K, Promthet S, Kamsa-Ard S, et al. Risk factors for cholangiocarcinoma in Khon Kaen, Thailand: a nested case-control study. Asian Pac J Cancer Prev. 2009;10(2):251–258.
Case 24.1
Bile Duct Adenoma Versus Biliary Hamartoma KISHA MITCHELL AND DHANPAT JAIN
C L I N I C AL I N F OR M AT I ON
A 46-year-old woman was found to have a 1.5 cm white nodule in the right lobe of her liver during abdominal imaging undertaken for uterine fibroids. She had never used oral contraceptives and denied alcohol or other drug use.
PAT H OL OG I C F E AT U R E S
The liver biopsy showed a proliferation of small angulated and round ducts, densely hyalinized stroma, and scattered chronic inflammatory cells (Figures 24.1.1A,B). The ducts have bland cuboidal to low columnar biliary-type epithelium with minimal pleomorphism. No mitotic figures or necrosis were seen. Few benign lymphoid aggregates were seen at the periphery. Scant amount of adjacent benign liver parenchyma present appeared normal.
R E A S ON F OR R E F E R R A L
To distinguish between bile duct adenoma (BDA), VMC, welldifferentiated CC, and metastatic adenocarcinoma (MA).
FIGURE 24.1.1 (A) Core biopsy with variable architecture, some areas
are extensively hyalinized and sparsely cellular, others with a proliferation of small angulated and round ducts with less stroma. The poor definition of the rounded border renders distinction from carcinoma difficult. (B) High power view showing an area of relatively greater cellularity with minimal atypia of individual cells.
FIGURE 24. 1. 2 (A) Resection specimen with small tan nodule in right lobe of liver. (B) Well-circumscribed lesion with conspicuous zonal pattern due to marked cellularity peripherally and sclerosis centrally. (C) Zonal pattern of bile duct adenoma highlighted by Masson’s Trichrome stain.
365
366
CHAPTER
24:
BILIARY
D I AG N OS I S
Bile duct adenoma.
D I SC U SSI ON
Differentiation from VMC is relatively easy, since the lesion lacks cystic dilatation of the glands, insipissated bile, and is relatively large for a VMC (1.5 cm). On the contrary, differentiation from a well-differentiated CC is extremely difficult. Lack of mitosis, significant nuclear atypia, and tumor type desmoplasia favors BDA, although none of these features by themselves is sufficiently specific to exclude a CC on needle biopsy. Since the possibility of well-differentiated CC could not be completely excluded, a wedge resection of the liver was subsequently undertaken (Figures 24.1.2A–C), which confirmed the diagnosis of BDA.
NEOPLASMS
BDAs rarely result in a needle biopsy; however, when a biopsy is undertaken as in this case, the diagnosis can be challenging. The lesions can show features that overlap with VMC on one hand and CC/MA on the other. BDAs are usually 0.5cm to 1.5cm in size, tend to be larger than VMC, but smaller than most CCs (1). They may occur in any area of the liver but, similar to VMCs, are most easily and frequently recognized during gross examination in the subcapsular location. The mean age of affected patients is 55 years (range 1–99 years), and there is no gender predilection. They appear as grey-white, tan, or yellow nodules that are round or oval in shape (Figure 24.1.2A). Though unencapsulated, they are typically well circumscribed (Figure 24.1.2B,C). Extension into the immediately adjacent portal tracts may impart an irregular outline on low magnification. The lesion consists of irregular and somewhat angulated ductular structures in variably collagenized stroma. In most lesions, especially those that are relatively large, the center of the lesion tends to be more
FIGURE 24. 1. 3 (A) Unencapsulated lesion that is distinctly demarcated from the adjacent parenchyma. (B) Central area of prominent hya-
linized sclerosis, devoid of ducts or ductules. (C) Peripheral area of increased cellularity with prominent ducts, less stroma, and no significant atypia. (D) Nodular lymphoid aggregate at periphery of bile duct adenoma.
CASE
24.1:
BILE
DUCT
ADENOMA
VERSUS
BILIARY
HAMARTOMA
367
FIGUR E 24. 1. 4 (A) Cut surfaces of liver with multiple, inconspicu-
ous small white nodules (arrows), characteristic of VMC. (B) The nodules (arrows) are also poorly defined from the capsular surface but may be identified with careful observation. F I G U R E 2 4 . 1 . 5 (A) Needle core biopsy of liver with dilated duc-
tular proliferation of VMC, more prominent stroma than in the cellular areas of bile duct adenoma (see Figure 24.1.1). (B) High power highlighting luminal insipissated eosinophilic material.
FIGURE 24. 1. 6 (A) Circumscribed bile duct adenoma with surrounding lymphoid aggregates, occurring with a VMC (arrow). (B) VMC-like area of BDA with mildly dilated glands with luminal material.
368
CHAPTER
24:
BILIARY
sclerotic with densely eosinophilic hyalinized collagen, and fewer and often compressed ductular structures lined by low cuboidal to flattened epithelium (Figure 24.1.3A–C). At the periphery, BDAs are more cellular with less stroma and the ductular structures are composed of cuboidal to low columnar epithelium. The interface between the lesion and adjacent hepatic parenchyma is well demarcated, a feature that is very helpful in distinguishing BDAs from CCs. BDA lacks infiltrative growth or an expansile growth pattern, and hence the adjacent hepatic parenchyma lacks features of compression or
NEOPLASMS
reactive changes. Rare examples of BDA with oncocytic or clear cells have also been described (2,3). Some lesions may show microcalcification, or even more rarely, noncaseating granulomas. Portal tracts with normal bile ducts can often be identified in the lesion. Variable lymphocytic infiltrate, sometimes forming nodular aggregates, can be seen at the periphery (Figure 24.1.3D). In contrast, VMC are usually smaller (200U) are present, these also suggest a component of HCC, and this should be included in the biopsy report.
FIGURE 24. 2. 2 (A) cytokeratin (CK) 7 positivity in CC (CK19 has a similar pattern). (B) Cytoplasmic polyclonal CEA positivity in CC.
372
CHAPTER
24:
BILIARY
The second situation is that in which some HCCs show positivity for CK19 and CK7, markers that are typically associated with CC, though retaining other hepatocytic markers in varying combinations (HepPar-1, canalicular pCEA, and AFP). This phenotype may be seen in HCC arising in a liver with less fibrosis or without cirrhosis and is associated with poorer prognosis; thus, sharing many characteristics with CC (2). These tumors have now been reclassified by WHO as mixed HCC-CC with stem cell features. Differentiation of CC from a metastasis is often difficult since there is no specific marker for bile duct origin. The clinical situation, serum markers, imaging studies, and tumor morphology should be carefully evaluated to come up with likely
NEOPLASMS
primary sites. It is not necessary to throw the entire panel of immunohistochemical markers at all cases. Evaluation of the genotype of a tumor based on molecular techniques may eventually be helpful; however, the current data are insufficient to advocate their routine use.
References 1. Nakajima T, Kondo Y, Miyazaki M, Okui K. A histopathologic study of 102 cases of intrahepatic cholangiocarcinoma: histologic classification and modes of spreading. Hum Pathol. 1988;19(10):1228–1234. 2. Durnez A, Verslype C, Nevens F, et al. The clinicopathological and prognostic relevance of cytokeratin 7 and 19 expression in hepatocellular carcinoma. A possible progenitor cell origin. Histopathology. 2006;49(2):138–151.
Case 24.3
Cholangiocarcinoma in Association With Von Meyenburg Complexes KISHA MITCHELL AND DHANPAT JAIN
C L IN I C AL F E AT U R E S
R EA SO N FO R R EFER R A L
A 61-year-old woman presented with a 3-week history of abdominal pain and weight loss. On examination, she was jaundiced with a mildly enlarged liver. Laboratory investigation revealed minimal liver enzyme abnormalities, normal serum albumin, and normal PT. Serum CA 19-9 was moderately elevated, whereas serum AFP, CA 125, and carcinoembryonic antigen (CEA) were normal. Abdominal CT scan revealed multiple small hypodense lesions through the left lobe of the liver with a dominant 4 cm left lobe liver mass.
The tumor had features of CC; however, it was unclear whether the entire biliary proliferation was malignant.
DIAGNO SIS
Cholangiocarcinoma in association with VMC/bile duct hamartomas.
DISCUSSIO N PAT H OL OG I C F E AT U R E S
The left lobectomy showed a 4 cm adenocarcinoma in a noncirrhotic liver. The tumor was composed of variably sized but predominantly small atypical glands in a desmoplastic stroma. The tumor had an infiltrative and destructive growth pattern with replacement of hepatic parenchyma and normal portal structures (Figures 24.3.1A,B). Based on the histology, the features are typical of a well- to moderately differentiated intrahepatic CC. Multiple separate and adjacent nodules (0.2–1.5 cm) were seen, raising a concern for multifocal CC or intrahepatic metastasis. Some of these nodules were