CAMERON Current Surgical Therapy, Thirteenth Edition [PDF]

Zitiervorschau

           13

th

e d iti o n

John L. Cameron

MD, FACS, FRCS(Eng)(hon), FRCS(Ed)(hon), FRCSI(hon) The Alfred Blalock Distinguished Service Professor Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

Andrew M. Cameron MD, PhD, FACS

Professor of Surgery Chief, Division of Transplatation Surgical Director, Liver Transplantation Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

1600 John F. Kennedy Blvd. Ste. 1600 Philadelphia, PA 19103-­2899 CURRENT SURGICAL THERAPY, THIRTEENTH EDITION Copyright © 2020 by Elsevier, Inc. All rights reserved.

ISBN: 978-­0-­323-­64059-­6

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notice Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds or experiments described herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. To the fullest extent of the law, no responsibility is assumed by Elsevier, authors, editors or contributors for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Previous editions copyrighted 2017 by Elsevier, Inc., and 2014, 2011, 2008, 2004, 2001, 1998, 1995, 1992, 1989, 1986, 1984 by Saunders, Inc., an affiliate of Elsevier, Inc. Library of Congress Control Number: 2019940501

Content Strategist: Jessica McCool Senior Content Development Manager: Kathryn DeFrancesco Publishing Services Manager: Catherine Jackson Senior Project Manager/Specialist: Carrie Stetz Design Direction: Brian Salisbury Printed in Canada Last digit is the print number: 9 8 7 6 5 4 3 2 1

Contributors

Maher A. Abbas, MD, FACS, FASCRS Professor of Surgery Dubai Colorectal and Digestive Clinic Dubai, United Arab Emirates ENTERAL STENTS IN THE TREATMENT OF COLONIC OBSTRUCTION Christopher J. Abularrage, MD, FACS Associate Professor Division of Vascular Surgery and Endovascular Surgery Johns Hopkins Hospital Baltimore, Maryland CAROTID ENDARTERECTOMY Ali F. AbuRahma, MD Professor of Surgery Chief, Vascular & Endovascular Surgery Director, Vascular Fellowship & Residency Programs Department of Surgery Robert C. Byrd Health Sciences Center; Medical Director, Vascular Laboratory Charleston Area Medical Center West Virginia University Charleston, West Virginia

David B. Adams, MD Distinguished University Professor Emeritus Department of Surgery Medical University of South Carolina Charleston, South Carolina PANCREAS DIVISUM AND OTHER VARIANTS OF ­DOMINANT DORSAL DUCT ANATOMY MANAGEMENT OF PANCREATIC NECROSIS MANAGEMENT OF CHRONIC PANCREATITIS Reid B. Adams, MD Chair, Department of Surgery Claude A. Jessup Professor of Surgery University of Virginia Charlottesville, Virginia CYSTIC DISEASE OF THE LIVER Gina L. Adrales, MD, MPH, FACS Chief, Division of Minimally Invasive Surgery Director, Minimally Invasive Surgery Training and Innovation Center Johns Hopkins University School of Medicine Baltimore, Maryland

FEMOROPOPLITEAL OCCLUSIVE DISEASE

MANAGEMENT OF INGUINAL HERNIA

Zachary T. AbuRahma, DO Vascular Surgery Resident Charleston Area Medical Center West Virginia University Charleston, West Virginia

Suresh K. Agarwal, MD Professor and Division Chief Trauma, Acute & Critical Care Surgery Department of Surgery Duke University Durham, North Carolina

FEMOROPOPLITEAL OCCLUSIVE DISEASE Charles A. Adams Jr, MD Chief, Division of Trauma and Surgical Critical Care Department of Surgery Rhode Island Hospital; Associate Professor Department of Surgery Warren Alpert School of Medicine at Brown University Providence, Rhode Island SEPTIC RESPONSE AND MANAGEMENT

TRACHEOSTOMY Anant Agarwalla, MD Gastroenterology Fellow Johns Hopkins University Baltimore, Maryland ENDOSCOPIC THERAPY FOR ESOPHAGEAL VARICEAL HEMORRHAGE

Steven A. Ahrendt, MD Associate Professor Department of Surgery Division of Surgical Oncology University of Pittsburgh Medical Center Pittsburgh, Pennsylvania MANAGEMENT OF BENIGN BILIARY STRICTURES Essa M. Aleassa, MD Fellow Bariatric and Metabolic Institute Cleveland Clinic Foundation Cleveland, Ohio; Assistant Professor Department of Surgery College of Medicine and Health Sciences United Arab Emirates University Al Ain, Abu Dhabi, United Arab Emirates CARDIOVASCULAR DISEASE RISK REDUCTION AFTER BARIATRIC SURGERY Mohammad Al Efishat, MD Chief Resident Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF INTRAHEPATIC, HILAR, AND DISTAL CHOLANGIOCARCINOMAS Aamna Ali, MD Resident Department of General Surgery Arrowhead Regional Medical Center/Kaiser Permanente Fontana, California ABDOMINAL AORTIC ANEURYSM AND UNEXPECTED ABDOMINAL PATHOLOGY John C. Alverdy, MD, FACS Sara and Harold Lincoln Thompson Professor of Surgery Executive Vice Chair Department of Surgery University of Chicago Chicago, Illinois GALLSTONE PANCREATITIS v

vi

CONTRIBUTORS

Robert Amajoyi, MD, FACS, FICS, FASCRS Assistant Professor Department of Surgery University of Missouri-­Kansas City School of Medicine Kansas City, Missouri

Elliot A. Asare, MD, MS Chief Resident, General Surgery Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin

MANAGEMENT OF HEMORRHOIDS

Roland Assi, MD, MMS Cardiothoracic Surgery Fellow Department of Surgery University of Pennsylvania Philadelphia, Pennsylvania

Kwame S. Amankwah, MD, MSc, FACS Professor and Chief, Vascular and Endovascular Surgery Department of Surgery Division of Vascular Surgery University of Connecticut (UCONN Health) Farmington, Connecticut VENOUS THROMBOEMBOLISM: PREVENTION, DIAGNOSIS, AND TREATMENT Ali Aminian, MD Associate Professor Department of General Surgery Cleveland Clinic Cleveland, Ohio CARDIOVASCULAR DISEASE RISK REDUCTION AFTER BARIATRIC SURGERY Ciro Andolfi, MD Fellow Department of Surgery Center for Medical and Surgical Simulation MacLean Center for Clinical Medical Ethics University of Chicago Medicine & Biological Sciences Division Chicago, Illinois SPLENECTOMY FOR HEMATOLOGIC DISORDERS Gary Anthone, MD, FACS Assistant Professor Department of Surgery Creighton University School of Medicine Omaha, Nebraska MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION Maggie Arnold, MD, FACS Assistant Professor of Vascular Surgery Department of General Surgery Johns Hopkins School of Medicine Baltimore, Maryland TREATMENT OF CLAUDICATION Amanda K. Arrington, MD Assistant Professor Department of Surgery University of Arizona Tucson, Arizona PROPER USE OF CHOLECYSTOSTOMY TUBES

UNUSUAL PANCREATIC TUMORS

MANAGEMENT OF DESCENDING THORACIC AND THORACOABDOMINAL AORTIC ANEURYSMS Hugh G. Auchincloss, MD, MPH Assistant Professor of Surgery Massachusetts General Hospital Boston, Massachusetts MEDIASTINAL MASSES Dan E. Azagury, MD Assistant Professor Department of Surgery Stanford University School of Medicine Stanford, California SURGICAL MANAGEMENT OF GASTROESOPHAGEAL REFLUX DISEASE Faris K. Azar, MD Assistant Professor of Surgery Florida Atlantic University St Mary’s Medical Center West Palm Beach, Florida PANCREATIC AND DUODENAL INJURIES Ali Azizzadeh, MD, FACS Professor and Director, Division of Vascular Surgery Vice-­Chair, Department of Surgery Associate Director, Smidt Heart Institute Cedars-­Sinai Medical Center Los Angeles, California BUERGER’S DISEASE (THROMBOANGIITIS OBLITERANS) Sudeep Banerjee, MD, MAS Postdoctoral Fellow Department of Surgery University of California, San Diego La Jolla, California; General Surgery Resident Department of Surgery University of California, Los Angeles Los Angeles, California MANAGEMENT OF CYSTIC DISORDERS OF THE BILE DUCTS

Philip S. Barie, MD, MBA, FIDSA, FACS, FCCM Professor of Surgery Division of Trauma, Burns, Critical and Acute Care Professor of Public Health in Medicine Division of Medical Ethics Weill Cornell Medical College New York, New York MEDICAL MANAGEMENT OF THE BURN PATIENT Richard J. Barth Jr, MD Professor of Surgery Geisel School of Medicine at Dartmouth Dartmouth-­Hitchcock Medical Center Lebanon, New Hampshire USE OF OPIOIDS IN THE POSTOPERATIVE PERIOD Richard J. Battafarano, MD, PhD Associate Professor and Chief Division of Thoracic Surgery Johns Hopkins Medical Institutions Baltimore, Maryland ESOPHAGEAL FUNCTION TESTS Joel M. Baumgartner, MD, MAS Associate Professor Department of Surgery University of California, San Diego La Jolla, California MANAGEMENT OF PERITONEAL SURFACE MALIGNANCIES Robert J. Beaulieu, MD Resident Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF ACUTE AORTIC DISSECTION MANAGEMENT OF TIBIOPERONEAL ARTERIAL OCCLUSIVE DISEASE Cigdem Benlice, MD Resident Department of General Surgery University of Ankara Ankara, Turkey MANAGEMENT OF TOXIC MEGACOLON Mitchell A. Bernstein, MD, FACS, FASCRS Director Division of Colon & Rectal Surgery NYU Langone Health New York, New York RECTOVAGINAL FISTULA

C O N T R I B U TO R S

Cherisse Berry, MD, FACS Assistant Professor of Surgery Associate Trauma Medical Director NYC Health+Hospitals/Bellevue New York University School of Medicine New York, New York MANAGEMENT OF DIAPHRAGMATIC INJURIES Saveri Bhattacharya, DO Assistant Professor Department of Medical Oncology Sidney Kimmel Cancer Center Thomas Jefferson University Philadelphia, Pennsylvania MANAGEMENT OF RECURRENT AND METASTATIC BREAST CANCER James H. Black III, MD, FACS David Goldfarb MD Associate Professor of Surgery Vascular Surgery and Endovascular Surgery Johns Hopkins School of Medicine Baltimore, Maryland MANAGEMENT OF ACUTE AORTIC DISSECTION Alex B. Blair, MD Surgical Resident Department of Surgery Johns Hopkins Hospital Baltimore, Maryland FAMILIAL GASTRIC CANCER Kirby I. Bland, MD, FACS Professor and Chair Emeritus Department of Surgery University of Alabama at Birmingham Birmingham, Alabama MOLECULAR TARGETS IN BREAST CANCER Joseph-­Vincent V. Blas, MD Associate Program Director Vascular Surgery Fellowship Division of Vascular Surgery Prisma Health-Upstate Greenville, South Carolina AORTOILIAC OCCLUSIVE DISEASE James N. Bogert, MD, FACS Assistant Professor Department of Surgery Creighton University School of Medicine St. Joseph’s Hospital and Medical Center Phoenix, Arizona USE OF RESUSCITATIVE ENDOVASCULAR BALLOON OCCLUSION OF THE AORTA IN RESUSCITATION OF THE TRAUMA PATIENT

vii

Judy C. Boughey, MD, FACS Professor of Surgery Department of Surgery Mayo Clinic Rochester, Minnesota

Benjamin N. Breyer, MD, MAS Associate Professor Department of Urology University of California, San Francisco San Francisco, California

CONTRALATERAL PROPHYLACTIC MASTECTOMY

RETROPERITONEAL INJURIES: KIDNEY AND URETER

Kelly A. Boyle, MD Resident Department of General Surgery Medical College of Wisconsin Milwaukee, Wisconsin

William M. Brigode, MD Resident Department of Surgery University of California, San Francisco East Bay Oakland, California

ANTIBIOTICS FOR CRITICALLY ILL PATIENTS Rachel F. Brem, MD, FACR, FSBI Professor and Vice Chair Director, Breast Imaging and Interventional Center Department of Radiology George Washington University Program Leader, Breast Cancer George Washington University Cancer Center Washington, DC A SURGEON’S PRACTICAL GUIDE TO BREAST IMAGING Megan Brenner, MD, MS, FACS Professor of Surgery University of California Riverside School of Medicine Director of Surgical Research Comparative Effectiveness and Clinical Outcomes Research Center (CECORC) Riverside University Health Systems Moreno Valley, California USE OF RESUSCITATIVE ENDOVASCULAR BALLOON OCCLUSION OF THE AORTA IN RESUSCITATION OF THE TRAUMA PATIENT J. Michael Brewer, DO, MS Assistant Professor Department of Medicine University of Mississippi Medical Center Jackson, Mississippi CARDIOVASCULAR PHARMACOLOGY Olaya I. Brewer Gutierrez, MD Department of Gastroenterology and Hepatology Johns Hopkins Hospital Baltimore, Maryland OBSTRUCTIVE JAUNDICE: ENDOSCOPIC THERAPY

MANAGEMENT OF PHEOCHROMOCYTOMA L.D. Britt, MD, MPH, FACS Henry Ford Professor and Edward J. Brickhouse Chairman Department of Surgery Eastern Virginia Medical School Norfolk, Virginia BLUNT ABDOMINAL TRAUMA Malcolm V. Brock, MD, FACS Professor of Surgery Division of Thoracic Surgery Johns Hopkins Medical Institutions Baltimore, Maryland USE OF ESOPHAGEAL STENTS MANAGEMENT OF PRIMARY CHEST WALL TUMORS Carlos V.R. Brown, MD Professor of Surgery Chief, Division of Acute Care Surgery Dell Medical School University of Texas at Austin Austin, Texas MANAGEMENT OF PANCREATIC PSEUDOCYST F. Charles Brunicardi, MD Moss Foundation Professor and Vice Chairman Department of Surgery University of California, Los Angeles Los Angeles, California MANAGEMENT OF DUODENAL ULCERS Keely L. Buesing, MD, FACS Associate Professor Department of General Surgery University of Nebraska Medical Center Omaha, Nebraska ACUTE KIDNEY INJURY IN THE INJURED AND CRITICALLY ILL

viii

CONTRIBUTORS

Marko Bukur, MD, FACS Associate Professor of Surgery Trauma Medical Director Division of Trauma, Emergency Surgery, and Surgical Critical Care Bellevue Hospital Center New York University School of Medicine New York, New York ACID-­BASE PROBLEMS Robert Bulat, MSc, MD, PhD Assistant Professor of Medicine Division of Gastroenterology Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF MOTILITY DISORDERS OF THE STOMACH AND SMALL BOWEL Richard A. Burkhart, MD Assistant Professor Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF INTRAHEPATIC, HILAR, AND DISTAL CHOLANGIOCARCINOMAS Ronald W. Busuttil, MD, PhD Distinguished Professor and Executive Chair Department of Surgery David Geffen School of Medicine at the University of California, Los Angeles Los Angeles, California ROLE OF LIVER TRANSPLANTATION IN PORTAL HYPERTENSION Christian Cain, MD Assistant Professor Department of Surgery Division of Trauma, Critical Care, and Acute Care Surgery R. Adams Cowley Shock Trauma Center University of Maryland Medical Center Baltimore, Maryland GAS GANGRENE OF THE EXTREMITY Rachael A. Callcut, MD, MSPH, FACS Associate Professor of Surgery Department of Surgery University of California, San Francisco San Francisco, California POSTOPERATIVE RESPIRATORY FAILURE

Nicholas A. Calotta, MD Resident Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland SURGICAL INFECTIONS OF THE HAND Richard P. Cambria, MD Chief, Division of Vascular & Endovascular Surgery Department of Surgery Massachusetts General Hospital Boston, Massachusetts ENDOVASCULAR TREATMENT OF ABDOMINAL AORTIC ANEURYSMS Andrew M. Cameron, MD, PhD Professor of Surgery Division Chief of Abdominal Transplantation Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland HEPATIC MALIGNANCY: RESECTION VERSUS TRANSPLANTATION Melissa S. Camp, MD, MPH Assistant Professor Department of Surgery Johns Hopkins Hospital Baltimore, Maryland BREAST CANCER: SURGICAL THERAPY Karel D. Capek, MD Physician Burns and Surgical Critical Care Galveston, Texas BURN WOUND MANAGEMENT Jonathan A. Cardella, MSc, MD, FRCS Assistant Professor and Program Director Division of Vascular Surgery Yale University New Haven, Connecticut VENOUS THROMBOEMBOLISM: PREVENTION, DIAGNOSIS, AND TREATMENT Paul Cartwright, MD Resident Department of Surgery Medical Center of Central Georgia Macon, Georgia SCREENING FOR BREAST CANCER

Cassius Iyad Ochoa Chaar, MD, MS Assistant Professor Department of Surgery Section of Vascular Surgery Yale School of Medicine New Haven, Connecticut VENA CAVA FILTERS Elliot L. Chaikof, MD, PhD Johnson and Johnson Professor of Surgery Harvard Medical School Chair Roberta and Stephen R. Weiner Department of Surgery Beth Israel Deaconess Medical Center Boston, Massachusetts POPLITEAL AND FEMORAL ARTERY ANEURYSM Akhil Chawla, MD Complex Surgical Oncology Fellow Department of Surgery Massachusetts General Hospital Brigham and Women’s Hospital Dana-­Farber Cancer Institute Harvard Medical School Boston, Massachusetts MANAGEMENT OF GALLBLADDER CANCER Po-­Hung Chen, MD Assistant Professor of Medicine Division of Gastroenterology and Hepatology Johns Hopkins School of Medicine Baltimore, Maryland MANAGEMENT OF REFRACTORY ASCITES Jill R. Cherry-­Bukowiec, MD, MS Associate Professor of Surgery Director, Trauma Burn ICU Assistant Director, Scientific Trunk Department of Surgery University of Michigan Ann Arbor, Michigan CATHETER SEPSIS IN THE INTENSIVE CARE UNIT Rachel L. Choron, MD Assistant Professor Department of Surgery Division of Acute Care Surgery Robert Wood Johnson University Hospital Rutgers Robert Wood Johnson Medical School New Brunswick, New Jersey MANAGEMENT OF CLOSTRIDIUM DIFFICILE COLITIS PANCREATIC AND DUODENAL INJURIES

C O N T R I B U TO R S

Michael A. Choti, MD, MBA, FACS Chief, Department of Surgery Division of Surgical Oncology Banner MD Anderson Medical Center Phoenix, Arizona MANAGEMENT OF RECTAL CANCER Kathryn Ely Pierce Chuquin, MD Resident Department of Surgery Mount Sinai Hospital New York, New York MANAGEMENT OF RECURRENT INGUINAL HERNIA Jose R. Cintron, MD Chairman, Division of Colon and Rectal Surgery Department of Surgery John H. Stroger Hospital of Cook County; Associate Professor Department of Surgery University of Illinois College of Medicine at Chicago Chicago, Illinois MANAGEMENT OF PRURITUS ANI William G. Cioffi, MD J. Murray Beardsley Professor & Chairman Department of Surgery Alpert Medical School of Brown University; Chief of Surgery Rhode Island Hospital and The Miriam Hospital Providence, Rhode Island SEPTIC RESPONSE AND MANAGEMENT Bryan Clary, MD, MBA Professor and Chair of Surgery Surgeon-­in-­Chief University of California, San Diego San Diego, California MANAGEMENT OF CYSTIC DISORDERS OF THE BILE DUCTS

Hiram S. Cody III, MD Attending Surgeon Breast Service, Department of Surgery Memorial Sloane Kettering Cancer Center; Professor of Surgery Weill Cornell Medical College New York, New York MARGINS: HOW TO AND HOW BIG? Thomas H. Cogbill, MD, FACS Attending Surgeon Department of General and Vascular Surgery Gundersen Health System La Crosse, Wisconsin MANAGEMENT OF PERIPHERAL ARTERIAL THROMBOEMBOLISM Andrew J. Cohen, MD Assistant Professor of Urology Director of Trauma and Reconstructive Urologic Surgery Brady Urologic Institute at Johns Hopkins School of Medicine Baltimore, Maryland RETROPERITONEAL INJURIES: KIDNEY AND URETER Raul Coimbra, MD, PhD, FACS Surgeon-­in-­Chief, Riverside University Health System Medical Center Director, Comparative Effectiveness and Clinical Outcomes Research Center (CECORC) Riverside University Health System Moreno Valley, California; Professor of Surgery Loma Linda University School of Medicine Loma Linda, California PREHOSPITAL MANAGEMENT OF THE TRAUMA PATIENT

Jordan M. Cloyd, MD Assistant Professor Department of Surgery The Ohio State University Columbus, Ohio

Kyle G. Cologne, MD Associate Professor Department of Surgery Division of Colorectal Surgery University of Southern California Keck School of Medicine Los Angeles, California

MANAGEMENT OF MALIGNANT LIVER TUMORS

MANAGEMENT OF ANORECTAL ABSCESS AND FISTULA

ix

Paul M. Colombani, MD, MBA, FACS, FAAP Chair Emeritus, Department of Surgery Johns Hopkins All Children’s Hospital St. Petersburg, Florida; Professor of Surgery, Pediatrics, and Oncology Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland REPAIR OF PECTUS EXCAVATUM Jonathan Cools-­Lartigue, MD, PhD Resident Department of Surgery McGill University Montreal, Quebec, Canada MULTIMODALITY THERAPY IN ESOPHAGEAL CANCER Robert N. Cooney, MD, FACS, FCCM Professor and Chairman Department of Surgery SUNY Upstate Medical University Syracuse, New York METABOLIC CHANGES FOLLOWING BARIATRIC SURGERY Gregory A. Coté, MD, MS Professor Department of Medicine Division of Gastroenterology and Hepatology Medical University of South Carolina Charleston, South Carolina PANCREAS DIVISUM AND OTHER VARIANTS OF DOMINANT DORSAL DUCT ANATOMY Bryan A. Cotton, MD, MPH Professor Department of Surgery University of Texas Health Science Center; Senior Researcher Center for Translational Injury Research Houston, Texas COAGULATION ISSUES AND THE TRAUMA PATIENT

Kathryn E. Coan, MD Assistant Professor Department of General Surgery Creighton Medical School Phoenix, Arizona

Martin A. Croce, MD Professor Department of Surgery University of Tennessee Health Science Center Senior Vice-­President and CMO Regional One Health Memphis, Tennessee

PRIMARY HYPERPARATHYROIDISM

THE INJURED SPLEEN

x

CONTRIBUTORS

Alisa Cross, MD Assistant Professor Department of Surgery University of Oklahoma Oklahoma City, Oklahoma SURGICAL USE OF ULTRASOUND IN THE TRAUMA AND CRITICAL CARE SETTINGS Alfred Croteau, MD Acute Care Surgeon Department of Surgery Hartford Healthcare Director, Surgical Stimulation (ACS-­AEI) Center for Education, Simulation, and Innovation Hartford Hospital Hartford, Connecticut; Assistant Professor Department of Surgery University of Connecticut Farmington, Connecticut CATHETER SEPSIS IN THE INTENSIVE CARE UNIT Gabriel Crowl, MD Resident Division of Vascular Surgery and Endovascular Therapy University Hospitals Cleveland Medical Center Cleveland, Ohio MANAGEMENT OF ANEURYSMS OF THE EXTRACRANIAL CAROTID AND VERTEBRAL ARTERIES Steven C. Cunningham, MD, FACS Director of Pancreatic and Hepatobiliary Surgery Direct of Research Department of Surgery Saint Agnes Hospital and Cancer Institute Baltimore, Maryland MANAGEMENT OF PRIMARY SCLEROSING CHOLANGITIS Leigh Anne Dageforde, MD, MPH Assistant Surgeon Department of Surgery Division of Transplantation Massachusetts General Hospital; Assistant Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ACUTE CHOLANGITIS

Paul D. Danielson, MD, FACS, FAAP Chief, Division of Pediatric Surgery Johns Hopkins All Children’s Hospital St. Petersburg, Florida; Associate Professor of Surgery Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland REPAIR OF PECTUS EXCAVATUM Alan Dardik, MD, PhD Professor and Vice-­Chairman (Faculty Affairs) Department of Surgery Interim Division Chief Division of Vascular Surgery Yale University School of Medicine New Haven, Connecticut; Attending Surgeon Department of Vascular Surgery VA Connecticut Healthcare Systems West Haven, Connecticut RAYNAUD’S PHENOMENON R. Clement Darling III, MD Professor Department of Surgery Albany Medical College Chief, Division of Vascular Surgery Albany Medical Center Hospital Albany, New York MANAGEMENT OF RUPTURED ABDOMINAL AORTIC ANEURYSMS Halley Darrach, BS Medical Student Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

Paula Denoya, MD, FACS, FASCRS Associate Professor of Surgery Program Director, Colorectal Surgery Residency Program Co-­Director, Stony Brook Inflammatory Bowel Disease Center Clinical Liaison Physician, Cancer Committee Stony Brook Medicine Stony Brook, New York USE OF STRICTUREPLASTY IN CROHN’S DISEASE E. Gene Deune, MD Associate Professor Department of Orthopedic Surgery Johns Hopkins University School of Medicine Baltimore, Maryland SURGICAL INFECTIONS OF THE HAND Navpreet K. Dhillon, MD Resident Physician Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California VENTILATOR-­ASSOCIATED PNEUMONIA Sandra R. DiBrito, MD, PhD Resident Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF SMALL BOWEL TUMORS MANAGEMENT OF TRAUMATIC LIVER INJURY Ellen Dillavou, MD, FACS Associate Professor of Surgery Department of Vascular Surgery Duke University Durham, North Carolina

LYMPHEDEMA

TREATMENT OF VARICOSE VEINS

Daniel T. Dempsey, MD, MBA Professor Department of Surgery University of Pennsylvania Philadelphia, Pennsylvania

Mary E. Dillhoff, MD, MS, FACS Associate Professor Department of Surgery The Ohio State University College of Medicine Columbus, Ohio

BENIGN GASTRIC ULCER

MANAGEMENT OF ZOLLINGER-­ELLISON SYNDROME

C O N T R I B U TO R S

Joseph DiNorcia, MD Assistant Professor of Surgery Department of Surgery David Geffen School of Medicine at the University of California, Los Angeles Los Angeles, California

Geoffrey P. Dunn, MD, FACS Department of Surgery, Emeritus UPMC Hamot Erie, Pennsylvania

ROLE OF LIVER TRANSPLANTATION IN PORTAL HYPERTENSION

Brett L. Ecker, MD Assistant Instructor Department of Surgery University of Pennsylvania Philadelphia, Pennsylvania

Gerard M. Doherty, MD Moseley Professor and Chair of Surgery Harvard Medical School Surgeon-­in-­Chief and Crowley Family Chair Brigham and Women’s Hospital Boston, Massachusetts NONTOXIC GOITER Jay Doucet, MD, MSc, FACS, FRCRC Professor and Chief Division of Trauma, Surgical Critical Care, Burns, and Acute Care Surgery Department of Surgery University of California, San Diego San Diego, California

SURGICAL PALLIATIVE CARE

NEOADJUVANT AND ADJUVANT THERAPY FOR PANCREATIC CANCER David T. Efron, MD, FACS Professor Department of Surgery Johns Hopkins School of Medicine; Vice Chair, Acute Care Surgery and Clinical Care Integration Chief, Division of Acute Care Surgery Johns Hopkins Hospital Baltimore, Maryland SPLENIC SALVAGE PROCEDURES

PREHOSPITAL MANAGEMENT OF THE TRAUMA PATIENT

EMERGENCY DEPARTMENT RESUSCITATIVE THORACOTOMY

Joseph Dubose, MD, FACS Professor of Surgery University of Maryland Medical System R. Adams Cowley Shock Trauma Center Baltimore, Maryland

Jonathan Efron, MD Professor of Surgery and Urology Department of Surgery Johns Hopkins Medical Institutions Baltimore, Maryland

PNEUMATOSIS INTESTINALIS AND THE IMPORTANCE FOR THE SURGEON

ENHANCED RECOVERY AFTER SURGERY

Quan-­Yang Duh, MD Professor and Chief Section of Endocrine Surgery University of California, San Francisco Attending Surgeon VA Medical Center San Francisco, California ADRENAL INCIDENTALOMA Mark D. Duncan, MD, FACS Vice Chair; Associate Professor of Surgery and Oncology Department of Surgery Johns Hopkins Medical Institutions Baltimore, Maryland FAMILIAL GASTRIC CANCER MANAGEMENT OF SMALL BOWEL TUMORS

Anne P. Ehlers, MD, MPH Senior Fellow Department of Surgery University of Washington Seattle, Washington ENDOSCOPIC TREATMENT OF BARRETT’S ESOPHAGUS Aslam Ejaz, MD, MPH Assistant Professor Department of Surgery The Ohio State University Wexner Medical Center Columbus, Ohio MANAGEMENT OF GASTRIC ADENOCARCINOMA Alexandra W. Elias, MD Department of General Surgery Mayo Clinic Jacksonville, Florida NEOADJUVANT AND ADJUVANT THERAPY FOR COLORECTAL CANCER

xi

Mustapha El Lakis, MD Clinical Assistant Professor Department of Surgery University of Pittsburgh Medical Center Pittsburgh, Pennsylvania MANAGEMENT OF THYROIDITIS E. Christopher Ellison, MD, FACS Robert M. Zollinger Professor Emeritus Department of Surgery The Ohio State University Columbus, Ohio MANAGEMENT OF ZOLLINGER-­ELLISON SYNDROME Adham Elmously, MD Resident Department of Surgery New York-­Presbyterian Weill Cornell Medical Center New York, New York MANAGEMENT OF SMALL BOWEL OBSTRUCTION Michael J. Englesbe, MD Professor Department of Surgery–Transplantation University of Michigan Ann Arbor, Michigan PERIOPERATIVE OPTIMIZATION Derek J. Erstad, MD Chief Resident in Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF BENIGN LIVER TUMORS Jairo A. Espinosa, MD Resident Department of General Surgery Western Michigan University Homer Stryker MD School of Medicine Kalamazoo, Michigan SURGICAL SITE INFECTIONS Eric W. Etchill, MD, MPH Resident Department of Surgery Johns Hopkins School of Medicine Baltimore, Maryland ACHALASIA OF THE ESOPHAGUS David Euhus, MD Professor Department of Surgery Johns Hopkins University Baltimore, Maryland LYMPHATIC MAPPING AND SENTINEL LYMPHADENECTOMY

xii

CONTRIBUTORS

Cory R. Evans, MD Associate Professor Department of General Surgery University of Tennessee Health Science Center Memphis, Tennessee

Marissa Famularo, DO Integrated Vascular Resident Department of Vascular Surgery Cooper University Hospital Camden, New Jersey

PENETRATING NECK TRAUMA

ENDOVASCULAR MANAGEMENT OF VASCULAR INJURIES

Douglas B. Evans, MD Professor and Chair Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin

Sandy H. Fang, MD Assistant Professor Department of Surgery Johns Hopkins Medical Institutions Baltimore, Maryland

UNUSUAL PANCREATIC TUMORS

MANAGEMENT OF DIVERTICULAR DISEASE OF THE COLON

Heather L. Evans, MD, MS Professor and Vice-­Chair of Clinical Research and Applied Informatics Department of Surgery Medical University of South Carolina Charleston, South Carolina ANTIFUNGAL THERAPY IN THE SURGICAL PATIENT Timothy C. Fabian, MD Professor Emeritus Department of Surgery University of Tennessee Health Science Center Memphis, Tennessee PENETRATING NECK TRAUMA Nadege T. Fackche, MD Postdoctoral Research Fellow General Surgery Division of Surgical Oncology Johns Hopkins University Baltimore, Maryland USE OF ESOPHAGEAL STENTS TOTAL PANCREATECTOMY WITH ISLET AUTOTRANSPLANTATION MANAGEMENT OF PRIMARY CHEST WALL TUMORS Peter J. Fagenholz, MD Department of Surgery Division of Trauma, Emergency Surgery, and Critical Care Massachusetts General Hospital; Assistant Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ACUTE CHOLECYSTITIS MANAGEMENT OF ACUTE NECROTIZING PANCREATITIS

Alik Farber, MD Chief, Division of Vascular and Endovascular Surgery Associate Chair for Clinical Operations Department of Surgery Boston Medical Center; Professor of Surgery and Radiology Boston University School of Medicine Boston, Massachusetts CHRONIC MESENTERIC ISCHEMIA David V. Feliciano, MD Clinical Professor Department of Surgery University of Maryland School of Medicine; Attending Surgeon Shock Trauma Center University of Maryland Medical Center Baltimore, Maryland PENETRATING ABDOMINAL TRAUMA Carlos Fernandez-­del Castillo, MD Professor Department of Surgery Massachusetts General Hospital Boston, Massachusetts INTRADUCTAL PAPILLARY MUCINOUS NEOPLASMS OF THE PANCREAS Paula Ferrada, MD, FACS Professor Department of Surgery Virginia Commonwealth University Richmond, Virginia PNEUMATOSIS INTESTINALIS AND THE IMPORTANCE FOR THE SURGEON

Marco Ferrara, MD Assistant Professor Department of Surgery University of Central Florida College of Medicine; Associate Colon and Rectal Surgery Colon and Rectal Clinic of Orlando Orlando, Florida DIAGNOSIS, TREATMENT, AND SURGICAL MANAGEMENT OF FISSURES-­IN-­ANO Cristina R. Ferrone, MD Associate Professor of Surgery Department of Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF GALLBLADDER CANCER Alessandro Fichera, MD, FACS, FASCRS Professor and Division Chief, Gastrointestinal Surgery Department of Surgery University of North Carolina Medical Center Chapel Hill, North Carolina MANAGEMENT OF LOWER GASTROINTESTINAL BLEEDING FLUID AND ELECTROLYTE THERAPY Filippo Filicori, MD Attending Surgeon Lenox Hill Hospital Assistant Professor of Surgery Zucker School of Medicine at Hofstra/ Northwell New York, New York MANAGEMENT OF ESOPHAGEAL CANCER John Filtes, MD Resident Department of Radiology Division of Interventional Radiology New York-­Presbyterian/Columbia University Irving Medical Center New York, New York TRANSHEPATIC INTERVENTIONS FOR OBSTRUCTIVE JAUNDICE Celeste C. Finnerty, PhD Professor Department of Surgery University of Texas Medical Branch Galveston, Texas BURN WOUND MANAGEMENT

C O N T R I B U TO R S

Josef E. Fischer, MD William V. McDermott Distinguished Professor of Surgery Department of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ENTEROCUTANEOUS FISTULAS Thomas M. Fishbein, MD Professor of Surgery and Executive Director MedStar Georgetown Transplant Institute MedStar Georgetown University Hospital Washington, DC MANAGEMENT OF INTESTINAL FAILURE Marco P. Fisichella, MD Associate Professor of Surgery Department of Surgery Harvard Medical School; Associate Chief of Surgery Department of Surgery West Roxbury VA Boston, Massachusetts SPLENECTOMY FOR HEMATOLOGIC DISORDERS Michael Fitzgerald, MD Surgical Critical Care Fellow University of Tennessee Health Science Center Memphis, Tennessee THE INJURED SPLEEN Robert J. Fitzgibbons Jr, MD, FACS Harry E. Stuckenhoff Professor and Chairman Department of Surgery Creighton University School of Medicine Omaha, Nebraska

Yuman Fong, MD Sangiacomo Chair and Chairman Department of Surgery City of Hope Medical Center Duarte, California

Ryan B. Fransman, MSc, MD Surgical Resident Department of Surgery Johns Hopkins University Baltimore, Maryland

ABLATION OF COLORECTAL LIVER METASTASES

MANAGEMENT OF DIVERTICULOSIS OF THE SMALL BOWEL

Zhi Ven Fong, MD, MPH General Surgery Resident Department of Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF ACUTE NECROTIZING PANCREATITIS Deshka S. Foster, MD, MA General Surgery Resident Department of Surgery Stanford University Stanford, California MANAGEMENT OF PANCREATIC ISLET CELL TUMORS EXCLUDING GASTRINOMA Spiros G. Frangos, MD, MPH, FACS Professor of Surgery Chief of Surgery, NYC Health+Hospitals/ Bellevue Chief of Trauma, Emergency Surgery, and Surgical Critical Care New York University School of Medicine New York, New York MANAGEMENT OF DIAPHRAGMATIC INJURIES

James Fleshman, MD Chairman Department of Surgery Baylor University Medical Center Dallas, Texas

Todd D. Francone, MD, FACS, FASCRS Staff Surgeon Colon and Rectal Surgery Massachusetts General Hospital; Assistant Professor of Surgery Tufts Medical School Boston, Massachusetts; Director, Division of Colon and Rectal Surgery Director, Robotic Surgery Associate Chair, Department of Surgery Newton-­Wellesley Hospital Newton, Massachusetts

MANAGEMENT OF RECTAL PROLAPSE

MANAGEMENT OF PILONIDAL DISEASE

Guillermo Foncerrada, MD, MMS Research Scientist Department of Surgery Shriners Hospital for Children University of Texas Medical Branch Galveston, Texas

Adam Franklin, MD Fellow Section of Plastic Surgery Oklahoma University Health Science Center Oklahoma City, Oklahoma

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

BURN WOUND MANAGEMENT

xiii

ABDOMINAL WALL RECONSTRUCTION

EMERGENCY DEPARTMENT RESUSCITATIVE THORACOTOMY Jonathan Friedstat, MD, MPH, FACS Assistant Surgeon Department of Surgery Massachusetts General Hospital; Instructor of Surgery Harvard Medical School Shriners Hospitals for Children Boston, Massachusetts ELECTRICAL INJURY AND LIGHTNING INJURIES Julie A. Freischlag, MD Chief Executive Officer Wake Forest Baptist Health Wake Forest Baptist Medical Center Dean Wake Forest School of Medicine Winston-­Salem, North Carolina THORACIC OUTLET SYNDROME Christopher Frost, MD Resident Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland NERVE INJURY AND REPAIR John Futchko, MD Resident Department of Cardiothoracic and Vascular Surgery Montefiore Medical Center New York, New York AXILLOFEMORAL BYPASS GRAFTING IN THE TWENTY-­FIRST CENTURY Alodia Gabre-­Kidan, MD Assistant Professor Department of Surgery Johns Hopkins School of Medicine Baltimore, Maryland ENHANCED RECOVERY AFTER SURGERY

xiv

CONTRIBUTORS

Michele A. Gadd, MD Assistant Professor of Surgery Department of Surgical Oncology Massachusetts General Hospital Boston, Massachusetts

Danon E. Garrido, MD Assistant Professor Vascular and Endovascular Surgery University of Mississippi Medical Center Jackson, Mississippi

GENETIC COUNSELING AND TESTING

BALLOON ANGIOPLASTY AND STENTS IN CAROTID ARTERY OCCLUSIVE DISEASE

Michele M. Gage, MD Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF INTRAHEPATIC, HILAR, AND DISTAL CHOLANGIOCARCINOMAS MANAGEMENT OF CUTANEOUS MELANOMA Susan Galandiuk, MD Professor of Surgery Program Director, Section of Colon & Rectal Surgery The Hiram C. Polk Jr, MD, Department of Surgery University of Louisville School of Medicine Louisville, Kentucky MANAGEMENT OF CHRONIC ULCERATIVE COLITIS Samuel M. Galvagno Jr, DO, PhD, FCCM Professor Associate Director, Maryland Critical Care Network Interim Medical Director, Lung Rescue Unit (ECMO unit) Deputy Director, Shock Trauma Go-­Team Department of Anesthesiology Program in Trauma R. Adams Cowley Shock Trauma Center University of Maryland School of Medicine Baltimore, Maryland AIRWAY MANAGEMENT IN THE TRAUMA PATIENT Matthew Garner, MD, MPH Research Fellow Surgical Oncology Johns Hopkins Hospital Baltimore, Maryland; General Surgery Resident Albany Medical Center Albany, New York USE OF ESOPHAGEAL STENTS MANAGEMENT OF PRIMARY CHEST WALL TUMORS

Susan L. Gearhart, MD Associate Professor of Colorectal Surgery Director, Colorectal Surgery Fellowship Program Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF LARGE BOWEL OBSTRUCTION Thomas Genuit, MD, MBA, FACS, FCCM Professor and Chair Program Director, General Surgery Department of Surgery Florida Atlantic University Charles E. Schmidt College of Medicine Boca Raton, Florida INITIAL ASSESSMENT AND RESUSCITATION OF THE TRAUMA PATIENT Christos Georgiades, MD, PhD Professor Department of Radiology & Radiological Sciences Johns Hopkins University Baltimore, Maryland

Richard C. Gilmore, MD Resident Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland BENIGN BREAST DISEASE SCREENING FOR BREAST CANCER Armando E. Giuliano, MD, FACS, FRCSEd Executive Vice Chair, Surgery Chief of Surgical Oncology Clinical Professor of Surgery Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California ABLATIVE TECHNIQUES IN THE TREATMENT OF BENIGN AND MALIGNANT BREAST DISEASE Natalia Glebova, MD, PhD Assistant Professor Section of Vascular and Endovascular Surgery University of Colorado Denver, Colorado ABDOMINAL ANEURYSM AND UNEXPECTED ABDOMINAL PATHOLOGY Jaime M. Glorioso, MD Resident Department of General Surgery Mayo Clinic Rochester, Minnesota

TRANSARTERIAL CHEMOEMBOLIZATION FOR LIVER METASASES

HEPATIC MALIGNANCY: RESECTION VERSUS TRANSPLANTATION

Bruce L. Gewertz, MD, FACS Chair and Vice-­President, Interventional Services Department of Surgery Vice-­Dean, Academic Affairs Cedars-­Sinai Medical Center Los Angeles, California

Naeem Goussous, MD Resident Department of Surgery Saint Agnes Hospital; Transplantation Fellow Department of Surgery University of Maryland Baltimore, Maryland

BUERGER’S DISEASE (THROMBOANGIITIS OBLITERANS) Joseph S. Giglia, MD Associate Professor Department of Surgery University of Cincinnati College of Medicine Cincinnati, Ohio ATHEROSCLEROTIC RENAL ARTERY STENOSIS

MANAGEMENT OF PRIMARY SCLEROSING CHOLANGITIS Jeremy Goverman, MD Assistant Professor Department of Surgery Harvard Medical School MGH Trustee’s Fellow in Burns Division of Burns Massachusetts General Hospital Boston, Massachusetts ELECTRICAL INJURY AND LIGHTNING INJURIES

C O N T R I B U TO R S

Michael D. Grant, MD Section Chief of Breast Surgery Baylor University Medical Center Dallas, Texas ROLE OF STEREOTACTIC BREAST BIOPSY IN THE MANAGEMENT OF BREAST DISEASE Michael P. Grant, MD, PhD, FACS Paul N. Manson Distinguished Service Professor Chief, Plastic and Reconstructive Surgery Director, Maxillofacial Trauma Service R. Adams Cowley Shock Trauma Center University of Maryland School of Medicine Baltimore, Maryland NONMELANOMA SKIN CANCERS EVALUATION AND MANAGEMENT OF FACIAL INJURIES Laura Grimsley, MD, MBA Resident Department of Surgery University of Tennessee Medical Center Knoxville, Tennessee ATHLETIC PUBALGIA: THE “SPORTS HERNIA” Alessandro Gronchi, MD Chair, Sarcoma Service Department of Surgery Fondazione IRCCS Istituto Nazionale dei Tumori Milan, Italy MANAGEMENT OF SOFT TISSUE SARCOMA José G. Guillem, MD, MPH Attending Surgeon Memorial Sloan-­Kettering Cancer Center Professor of Surgery Cornell University Medical College New York, New York SURGICAL MANAGEMENT OF COLON CANCER Aakanksha Gupta, MD Trauma Research Fellow Division of Trauma, Burns, Critical and Acute Care Weill Cornell Medicine New York, New York MEDICAL MANAGEMENT OF THE BURN PATIENT NavYash Gupta, MD, FACS Associate Professor Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California BUERGER’S DISEASE (THROMBOANGIITIS OBLITERANS)

John W. Harmon, MD, FACS Professor of Surgery Department of General Surgery Johns Hopkins University Baltimore, Maryland MANAGEMENT OF DIVERTICULOSIS OF THE SMALL BOWEL

John B. Hanks, MD C. Bruce Morton Professor and Chief Department of Surgery Division of General Surgery University of Virginia Health Systems Charlottesville, Virginia MANAGEMENT OF SOLITARY NECK MASS

Mehran Habibi, MD, MBA Assistant Professor of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

Andrew Harris, BS Research Fellow Department of Orthopedic Spine Surgery Johns Hopkins University Baltimore, Maryland

MANAGEMENT OF MALE BREAST CANCER

SPINE AND SPINAL CORD INJURIES

David J. Hackam, MD, PhD Chief, Pediatric Surgery Department of Surgery Johns Hopkins University Pediatric Surgeon in Chief and Co-­Director Johns Hopkins Children’s Center Baltimore, Maryland

Jennifer L. Hartwell, MD Assistant Professor Department of Surgery Indiana University Indianapolis, Indiana

APPENDICITIS Adil H. Haider, MD, MPH, FACS Dean Medical College Aga Khan University Karachi, Pakistan MANAGEMENT OF INTRAABDOMINAL INFECTIONS Krista L. Haines, DO, MA Assistant Professor Department of Surgery Duke University School of Medicine Durham, North Carolina TRACHEOSTOMY Wissam J. Halabi, MD Assistant Professor Department of Surgery University of California, Davis Sacramento, California ENTERAL STENTS IN THE TREATMENT OF COLONIC OBSTRUCTION Christopher L. Hallemeier, MD Assistant Professor Department of Radiation Oncology Mayo Clinic Rochester, Minnesota MANAGEMENT OF RADIATION INJURY TO THE LARGE AND SMALL BOWEL

xv

ABDOMINAL COMPARTMENT SYNDROME AND MANAGEMENT OF THE OPEN ABDOMEN Gabrielle E. Hatton, MD Resident Department of Surgery McGovern Medical School University of Texas Health Science Center at Houston Houston, Texas ENDOCRINE CHANGES IN CRITICAL ILLNESS Elliott R. Haut, MD, PhD, FACS Associate Professor of Surgery, Anesthesiology/Critical Care Medicine (ACCM) and Emergency Medicine Department of Surgery Division of Acute Care Surgery Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF TRAUMATIC LIVER INJURY Jason S. Hawksworth, MD Associate Professor of Surgery MedStar Georgetown Transplant Institute MedStar Georgetown University Hospital Washington, DC MANAGEMENT OF INTESTINAL FAILURE Jin He, MD, PhD, FACS Assistant Professor Departments of Surgery and Oncology Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF PERIAMPULLARY CANCER

xvi

CONTRIBUTORS

Sharon Henry, MD, FACS Professor of Trauma Surgery University of Maryland School of Medicine Director, Wound Healing and Metabolism R. Adams Cowley Shock Trauma Center University of Maryland Medical Center Baltimore, Maryland GAS GANGRENE OF THE EXTREMITY Joseph M. Herman, MD, MSc Department of Radiation Oncology & Molecular Radiation Services Johns Hopkins University School of Medicine Baltimore, Maryland INTRAOPERATIVE RADIATION FOR PANCREATIC CANCER David N. Herndon, MD Retired Galveston, Texas BURN WOUND MANAGEMENT Caitlin W. Hicks, MD, MS Assistant Professor of Surgery Division of Vascular Surgery and Endovascular Therapy Johns Hopkins University School of Medicine Baltimore, Maryland BRACHIOCEPHALIC RECONSTRUCTION O. Joe Hines, MD Professor and Chief Department of Surgery Division of General Surgery David Geffen School of Medicine University of California, Los Angeles Los Angeles, California PANCREATIC DISRUPTIONS LEADING TO PANCREATIC FISTULA, PANCREATIC ASCITES, OR PANCREATIC PLEURAL EFFUSIONS Julie L. Holihan, MD, MS Clinical Instructor Department of Surgery McGovern Medical School Houston, Texas INCISIONAL, EPIGASTRIC, AND UMBILICAL HERNIAS Courtenay M. Holscher, MD Halsted General Surgery Resident Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland CAROTID ENDARTERECTOMY TREATMENT OF CLAUDICATION ACUTE MESENTERIC ISCHEMIA

Kelvin Hong, MD Division Chief Interventional Radiology Johns Hopkins University Baltimore, Maryland

Ashley I. Huppe, MD Assistant Professor Department of Radiology, Breast Imaging University of Kansas Health System Kansas City, Kansas

ACUTE PERIPHERAL ARTERIAL AND BYPASS GRAFT OCCLUSION: THROMBOLYTIC THERAPY

A SURGEON’S PRACTICAL GUIDE TO BREAST IMAGING

Michael G. House, MD, FACS Associate Professor Department of Surgery Indiana University School of Medicine Indianapolis, Indiana MANAGEMENT OF GASTROINTESTINAL STROMAL TUMORS Cary Hsu, MD, FACS Assistant Professor of Surgery Michael E. DeBakey Department of Surgery Division of Surgical Oncology Baylor College of Medicine Houston, Texas DUCTAL AND LOBULAR CARCINOMA IN SITU OF THE BREAST Renee Huang, MD Assistant Professor Department of Surgery Montefiore Medical Center New York, New York MANAGEMENT OF COLONIC VOLVULUS Kevin Hughes, MD Co-­Director, Avon Comprehensive Breast Evaluation Center Division of Surgical Oncology Massachusetts General Hospital; Associate Professor Department of Surgery Harvard Medical School Boston, Massachusetts; Medical Director Bermuda Cancer Genetics and Risk Assessment Clinic Hamilton, Bermuda GENETIC COUNSELING AND TESTING Travis Hull, MD, PhD General Surgery Resident Department of Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF ACQUIRED ESOPHAGEAL RESPIRATORY TRACT FISTULA

David F. Hutcheon, MD Assistant Professor of Medicine Department of Medicine Johns Hopkins Medical Institutions Baltimore, Maryland ACUTE COLONIC PSEUDO-­OBSTRUCTION (OGILVIE’S SYNDROME) Irada Ibrahim-­zada, MD, PhD Surgical Oncology Fellow Roswell Park Cancer Institute Buffalo, New York MANAGEMENT OF ECHINOCOCCAL CYST DISEASE OF THE LIVER MANAGEMENT OF BENIGN BILIARY STRICTURES Kamal M.F. Itani, MD Chief of Surgery VA Boston Health Care System Professor of Surgery Boston University Boston, Massachusetts NECROTIZING SKIN AND SOFT TISSUE INFECTIONS Tomoaki Ito, MD Department of Surgery Juntendo University Shizuoka Hospital Shizuoka, Japan USE OF ESOPHAGEAL STENTS Anee Sophia Jackson, MD Research Fellow Division of Thoracic Surgery Swedish Cancer Institute Seattle, Washington; Surgical Resident Department of General Surgery MedStar Georgetown University Hospital Washington, DC MANAGEMENT OF BARRETT’S ESOPHAGUS Brian Jacob, MD Associate Professor of Surgery Department of Surgery Icahn School of Medicine at Mount Sinai New York, New York MANAGEMENT OF RECURRENT INGUINAL HERNIA

C O N T R I B U TO R S

Lisa K. Jacobs, MD, MSPH Associate Professor of Surgery Associate Professor of Oncology Department of Surgery Johns Hopkins University Baltimore, Maryland SCREENING FOR BREAST CANCER Richard A. Jacobson, MD Resident Physician Department of Surgery Rush University Medical Center Chicago, Illinois GALLSTONE PANCREATITIS Ammar A. Javed, MD Lead Postdoctoral Research Fellow Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF LIVER HEMANGIOMAS VASCULAR RECONSTRUCTION DURING THE WHIPPLE OPERATION Fabian Johnston, MD, MHS, FACS Associate Professor of Surgery and Oncology Chief, Section of Gastrointestinal Surgical Oncology Director, Peritoneal Surface Malignancy Program Program Director, Complex General Surgical Oncology Fellowship Division of Surgical Oncology Johns Hopkins University Baltimore, Maryland MANAGEMENT OF GASTRIC ADENOCARCINOMA Douglas W. Jones, MD Division of Vascular and Endovascular Surgery Boston Medical Center; Assistant Professor of Surgery Boston University School of Medicine Boston, Massachusetts CHRONIC MESENTERIC ISCHEMIA Ronald C. Jones, MD Chief Emeritus Department of Surgery Baylor University Medical Center Dallas, Texas ROLE OF STEREOTACTIC BREAST BIOPSY IN THE MANAGEMENT OF BREAST DISEASE

Kimberly Joseph, MD Voluntary Attending Surgeon Department of Trauma and Burns John H. Stroger Hospital of Cook County Chicago, Illinois NUTRITION THERAPY IN THE CRITICALLY ILL SURGICAL PATIENT Anthony N. Kalloo, MD Moses and Helen Golden Paulson Professor of Gastroenterology Director, Division of Gastroenterology & Hepatology Johns Hopkins Hospital Baltimore, Maryland OBSTRUCTIVE JAUNDICE: ENDOSCOPIC THERAPY Andreas M. Kaiser, MD, FACS, FASCRS Professor of Clinical Surgery USC Division of Colorectal Surgery Keck School of Medicine at USC University of Southern California Los Angeles, California SURGICAL MANAGEMENT OF FECAL INCONTINENCE Pridvi Kandagatla, MD Resident Department of Surgery Henry Ford Hospital Wayne State University Detroit, Michigan INFLAMMATORY BREAST CANCER Lillian S. Kao, MD, FACS Professor Department of Surgery McGovern Medical School University of Texas Health Science Center at Houston Houston, Texas

xvii

Vikram S. Kashyap, MD Chief, Division of Vascular Surgery and Endovascular Therapy Department of Surgery University Hospitals Cleveland Medical Center Cleveland, Ohio MANAGEMENT OF ANEURYSMS OF THE EXTRACRANIAL CAROTID AND VERTEBRAL ARTERIES Khaled M. Kebaish, MD, FRCS Professor Department of Orthopedic Surgery Johns Hopkins University Baltimore, Maryland SPINE AND SPINAL CORD INJURIES Electron Kebebew, MD Professor of Surgery Chief, Division of General Surgery Harry A. Oberhelman Jr. and Mark L. Welton Professor Department of Surgery Stanford University Stanford, California MANAGEMENT OF THYROIDITIS Matthew P. Kelley, MD General Surgery Resident Department of Surgery Division of Colon and Rectal Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF LARGE BOWEL OBSTRUCTION Scott R. Kelley, MD, FACS, FASCRS Assistant Professor of Surgery Colon and Rectal Surgery Mayo Clinic Rochester, Minnesota

ENDOCRINE CHANGES IN CRITICAL ILLNESS

MANAGEMENT OF RADIATION INJURY TO THE LARGE AND SMALL BOWEL

Mark Katlic, MD, FACS Chair Department of General Surgery Sinai Hospital of Baltimore Baltimore, Maryland

Kaitlyn J. Kelly, MD Assistant Professor of Surgery Division of Surgical Oncology University of California, San Diego San Diego, California

MANAGEMENT OF DISORDERS OF ESOPHAGEAL MOTILITY

MANAGEMENT OF PERITONEAL SURFACE MALIGNANCIES Kevin Kemp, MD Assistant Professor Department of Surgery, University of Nebraska Medical Center Omaha, Nebraska CATHETER SEPSIS IN THE INTENSIVE CARE UNIT

xviii

CONTRIBUTORS

Aadil A. Khan, MD, MPH, PhD Microsurgery Fellow Department of Plastic Surgery Johns Hopkins Hospital Baltimore, Maryland BREAST RECONSTRUCTION FOLLOWING MASTECTOMY: INDICATIONS, TECHNIQUES, AND RESULTS Uzer Khan, MBBS, MD Associate Professor Department of Surgery West Virginia University Morgantown, West Virginia FLUIDS AND ELECTROLYTES Misaki Kiguchi, MD Assistant Professor and Assistant Program Director Department of Vascular Surgery MedStar Georgetown University Hospital Washington, DC TREATMENT OF VARICOSE VEINS Mehreen T. Kisat, MD, MS Instructor Department of Surgery Brigham and Women’s Hospital Boston, Massachusetts MANAGEMENT OF INTRA­ABDOMINAL INFECTIONS Andrew Klein, MD, MBA, FACS Professor and Vice-­Chairman Department of Surgery; Director Comprehensive Transplant Center Cedars-­Sinai Medical Center Los Angeles, California MANAGEMENT OF BUDD-­CHIARI SYNDROME George Kokosis, MD Resident Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland LYMPHEDEMA Spogmai Komak, MD Assistant Professor Department of Surgery McGovern Medical School University of Texas Health Science Center at Houston Houston, Texas ENDOCRINE CHANGES IN CRITICAL ILLNESS

Kambiz Kosari, MD Hepato-­Pancreato-­Biliary and Transplant Surgeon Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California MANAGEMENT OF BUDD-­CHIARI SYNDROME Mark L. Kovler, MD Resident Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland APPENDICITIS Ashley N. Krepline, MD General Surgery Resident Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin PALLIATIVE THERAPY FOR PANCREATIC CANCER Helen Krontiras, MD Professor of Surgery Department of Surgery Division of Surgical Oncology University of Alabama at Birmingham Birmingham, Alabama MOLECULAR TARGETS IN BREAST CANCER Amanda Kupstas, MD Breast Oncology Fellow Department of Surgery Mayo Clinic Rochester, Minnesota CONTRALATERAL PROPHYLACTIC MASTECTOMY Kelly J. Lafaro, MD, MPH Assistant Professor Department of Surgery Johns Hopkins University School of Medicine ABLATION OF COLORECTAL LIVER METASTASES Mitchell Ryan Ladd, MD, PhD Resident Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MULTIPLE ORGAN DYSFUNCTION AND FAILURE

Eric Lambright, MD Associate Professor Thoracic Surgery Vanderbilt University Medical Center Nashville, Tennessee MANAGEMENT OF ESOPHAGEAL PERFORATION Glenn M. LaMuraglia, MD Visiting Surgeon Division of Vascular and Endovascular Surgery Massachusetts General Hospital Professor of Surgery Harvard Medical School Boston, Massachusetts LOWER EXTREMITY AMPUTATION Rachael B. Lancaster, MD Assistant Professor of Surgery Department of Surgery Division of Surgical Oncology University of Alabama at Birmingham Birmingham, Alabama MOLECULAR TARGETS IN BREAST CANCER William P. Lancaster, MD Assistant Professor Department of Surgery Medical University of South Carolina Charleston, South Carolina MANAGEMENT OF PANCREATIC NECROSIS MANAGEMENT OF CHRONIC PANCREATITIS Ron G. Landmann, MD Chief, Colon and Rectal Surgery Baptist-­MD Anderson Cancer Center Jacksonville, Florida NEOADJUVANT AND ADJUVANT THERAPY FOR COLORECTAL CANCER Julie R. Lange, MD, ScM Associate Professor Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland BENIGN BREAST DISEASE MANAGEMENT OF CUTANEOUS MELANOMA Sean J. Langenfeld, MD Associate Professor Department of Surgery Chief, Colon and Rectal Surgery University of Nebraska Medical Center Omaha, Nebraska MANAGEMENT OF ANORECTAL ABSCESS AND FISTULA

C O N T R I B U TO R S

xix

Melissa Lazar, MD Assistant Professor of Surgery Department of Surgery Thomas Jefferson University Sidney Kimmel Medical College Philadelphia, Pennsylvania

Ira L. Leeds, MD, MBA, ScM Chief Resident Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

Robert P. Liddell, MD Assistant Professor Department of Radiology and Radiological Sciences Johns Hopkins School of Medicine Baltimore, Maryland

MANAGEMENT OF RECURRENT AND METASTATIC BREAST CANCER

MANAGEMENT OF DIVERTICULAR DISEASE OF THE COLON

TRANSJUGULAR INTRAHEPATIC PORTOSYSTEMIC SHUNT

Anna M. Ledgerwood, MD Professor Michael and Marian Ilitch Department of Surgery Wayne State University; Trauma Medical Director Department of Trauma Services Detroit Receiving Hospital Detroit, Michigan

Ji Lei, MD, MSc, MBA Director, MGH Islet Transplantation GMP Laboratory Assistant Professor of Surgery Harvard Medical School Massachusetts General Hospital Boston, Massachusetts

BLUNT CARDIAC INJURY MANAGEMENT OF EXTREMITY COMPARTMENT SYNDROME Candice Y. Lee, MD Cardiothoracic Surgery Fellow Department of Cardiovascular and Thoracic Surgery Allegheny General Hospital Pittsburgh, Pennsylvania EXTRACORPOREAL MEMBRANE OXYGENATION FOR RESPIRATORY FAILURE IN ADULTS Jong O. Lee, MD Professor Department of Surgery University of Texas Medical Branch Medical Director, Burn Intensive Care Unit Shriners Hospital for Children Galveston, Texas BURN WOUND MANAGEMENT Sang W. Lee, MD, FACS, FASCRS Charles W. and Carolyn Costello Chair in Colorectal Diseases Professor and Chief of Colon & Rectal Surgery USC Keck School of Medicine Los Angeles, California SURGICAL MANAGEMENT OF THE POLYPOSIS SYNDROMES Yann-­Leei Larry Lee, MD, MS Department of Surgery Johns Hopkins Hospital Baltimore, Maryland SPLENIC SALVAGE PROCEDURES

ISLET AUTOTRANSPLANTATION FOR CHRONIC PANCREATITIS Isabelle C. Le Leannec, MD, MEng Fellow, Colon and Rectal Surgery University of Minnesota Minneapolis, Minnesota MANAGEMENT OF CROHN’S COLITIS Rebecca Levine, MD, FACS Assistant Professor Department of Surgery Montefiore Medical Center Einstein College of Medicine New York, New York MANAGEMENT OF COLONIC VOLVULUS Eric J. Ley, MD Associate Professor Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California VENTILATOR-­ASSOCIATED PNEUMONIA Mike K. Liang, MD Associate Professor Department of Surgery McGovern Medical School Houston, Texas INCISIONAL, EPIGASTRIC, AND UMBILICAL HERNIAS Yu Liang, MD Assistant Professor General Surgery University of Connecticut Health Center Farmington, Connecticut MANAGEMENT OF MALLORY-­WEISS SYNDROME

ACUTE PERIPHERAL ARTERIAL BYPASS GRAFT OCCLUSION: THROMBOLYTIC THERAPY Anne O. Lidor, MD, MPH Professor Department of Surgery University of Wisconsin School of Medicine and Public Health Madison, Wisconsin MANAGEMENT OF MORBID OBESITY Keith D. Lillemoe, MD Chief of Surgery Department of Surgery Massachusetts General Hospital; W. Gerald Austen Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ACUTE CHOLANGITIS John C. Lipham, MD Chief, Division of Upper GI & General Surgery Professor of Surgery James & Pamela Muzzy Endowed Chair in Upper GI Cancer President, American Foregut Society Department of Surgery Keck Medical Center of USC University of Southern California Los Angeles, California MAGNETIC AUGMENTATION OF THE LOWER ESOPHAGEAL SPHINCTER Jeremy M. Lipman, MD, MHPE, FACS, FASCRS Program Director General Surgery Residency Cleveland Clinic Cleveland, Ohio SURGICAL MANAGEMENT OF CONSTIPATION

xx

CONTRIBUTORS

Pamela A. Lipsett, MD, MHPE, MCCM Warfield M. Firor Endowed Professorship Department of Surgery Assistant Dean of Assessment and Evaluation Program Director, General Surgery and Surgical Critical Care Co-­Director, Surgical Intensive Care Units Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF CLOSTRIDIUM DIFFICILE COLITIS Evan C. Lipsitz, MD, MBA, FACS Chief, Division of Vascular and Endovascular Surgery Department of Cardiothoracic and Vascular Surgery Montefiore Medical Center Albert Einstein College of Medicine New York, New York AXILLOFEMORAL BYPASS GRAFTING IN THE TWENTY-­FIRST CENTURY Anna Liveris, MD Clinical Instructor Resident Physician Department of Surgery Montefiore Medical Center Albert Einstein College of Medicine New York, New York MANAGEMENT OF CYSTS, TUMORS, AND ABSCESSES OF THE SPLEEN Satinderjit S. Locham, HBSc, MBBCh, BAO Resident Department of Surgery University of California, San Diego La Jolla, California UPPER EXTREMITY ARTERIAL OCCLUSIVE DISEASE Joseph V. Lombardi, MD Professor and Chief Department of Vascular Surgery Cooper University Hospital Camden, New Jersey ENDOVASCULAR MANAGEMENT OF VASCULAR INJURIES Bonnie E. Lonze, MD, PhD Assistant Professor Department of Surgery New York University Langone Transplant Institute New York, New York HEMODIALYSIS ACCESS SURGERY

Brian E. Louie, MD, MHA, MPH, FRCSC, FACS Director, Thoracic Surgery Research and Education Co-­Director, Minimally Invasive Thoracic Surgery Program Surgical Chair, Swedish Digestive Health Institute Division of Thoracic Surgery Swedish Medical Center and Cancer Institute Seattle, Washington MANAGEMENT OF BARRETT’S ESOPHAGUS Gregory K. Low, MD Resident Physician Department of Surgery Keck School of Medicine of the University of Southern California Los Angeles, California SURGICAL MANAGEMENT OF FECAL INCONTINENCE Andrew M. Lowy, MD Professor of Surgery Chief, Division of Surgical Oncology University of California, San Diego La Jolla, California MANAGEMENT OF PERITONEAL SURFACE MALIGNANCIES Charles E. Lucas, MD Professor Michael and Marian Ilitch Department of Surgery Wayne State University; Surgeon Department of Surgery Detroit Receiving Hospital Detroit, Michigan BLUNT CARDIAC INJURY MANAGEMENT OF EXTREMITY COMPARTMENT SYNDROME Ying Wei Lum, MD Associate Professor Division of Vascular Surgery & Endovascular Therapy Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF RECURRENT CAROTID STENOSIS

George J. Magovern Jr, MD System Chair Department of Thoracic and Cardiovascular Surgery Allegheny Health Network Pittsburgh, Pennsylvania EXTRACORPOREAL MEMBRANE OXYGENATION FOR RESPIRATORY FAILURE IN ADULTS Martin A. Makary, MD, MPH Professor of Surgery and Health Policy Management Chief, Johns Hopkins Islet Transplantation Center Director, Minimally Invasive Pancreas Surgery Johns Hopkins Hospital Baltimore, Maryland TOTAL PANCREATECTOMY WITH ISLET AUTOTRANSPLANTATION Mahmoud B. Malas, MD, MHS, FACS Professor in Resident Chief of Vascular and Endovascular Surgery Vice-­Chair of Surgery for Clinical Research University of California, San Diego Health System La Jolla, California UPPER EXTREMITY ARTERIAL OCCLUSIVE DISEASE Paul N. Manson, MD Distinguished Service Professor Department of Plastic Surgery Johns Hopkins University Baltimore, Maryland NONMELANOMA SKIN CANCERS EVALUATION AND MANAGEMENT OF FACIAL INJURIES MANAGEMENT OF FROSTBITE, HYPOTHERMIA, AND COLD INJURIES Nathalie Mantilla, MD Colon & Rectal Surgeon Department of Surgery Division of Colon & Rectal Surgery John H. Stroger Jr Hospital of Cook County; Assistant Professor of Surgery Rush University Chicago, Illinois MANAGEMENT OF PRURITUS ANI

C O N T R I B U TO R S

Peter Marcinkowski, MD Resident, General Surgery Department of Surgery UNC Hospitals University of North Carolina Chapel Hill, North Carolina MANAGEMENT OF LOWER GASTROINTESTINAL BLEEDING James F. Markmann, MD, PhD Chief Division of Transplant Surgery Claude E. Welch Professor of Surgery Harvard Medical School Massachusetts General Hospital Boston, Massachusetts ISLET AUTOTRANSPLANTATION FOR CHRONIC PANCREATITIS Stephen F. Markowiak, MD, MPH Clinical Instructor Department of Surgery University of Toledo College of Medicine Toledo, Ohio MANAGEMENT OF DUODENAL ULCERS Allison N. Martin, MD, MPH Resident Department of Surgery University of Virginia Charlottesville, Virginia MANAGEMENT OF SOLITARY NECK MASS Daniel Martin, DO Fellow Colon and Rectal Surgery Colon and Rectal Clinic of Orlando Orlando, Florida DIAGNOSIS, TREATMENT, AND SURGICAL MANAGEMENT OF FISSURES-­IN-­ANO Matthew J. Martin, MD, FACS, FASMBS Director of Trauma Research Scripps Mercy Hospital Professor of Surgery Uniformed Services University of the Health Sciences Bethesda, Maryland MANAGEMENT OF PANCREATIC PSEUDOCYST Michael Martyak, MD Assistant Professor Department of Surgery Eastern Virginia Medical School Norfolk, Virginia BLUNT ABDOMINAL TRAUMA

Roxanne L. Massoumi, MD Resident Physician Department of General Surgery University of California, Los Angeles Los Angeles, California

Anya Mezina, MD, MSc Resident Department of Medicine Johns Hopkins Hospital Baltimore, Maryland

PANCREATIC DISRUPTIONS LEADING TO PANCREATIC FISTULA, PANCREATIC ASCITES, OR PANCREATIC PLEURAL EFFUSIONS

MANAGEMENT OF HEPATIC ENCEPHALOPATHY

Douglas J. Mathisen, MD Thoracic Surgeon Department of Surgery Massachusetts General Hospital MEDIASTINAL MASSES J. Greg Mawn, MD Senior Resident Department of Orthopaedic Surgery Johns Hopkins University School of Medicine Baltimore, Maryland EARLY MANAGEMENT OF PELVIC RING DISRUPTION Laura M. Mazer, MD, MS Assistant Professor of Surgery Division of Minimally Invasive Surgery University of Michigan Ann Arbor, Michigan SURGICAL MANAGEMENT OF GASTROESOPHAGEAL REFLUX DISEASE Mary McCarthy, MD, FACS Professor Department of Surgery Wright State University School of Medicine Dayton, Ohio CHEST WALL, PNEUMOTHORAX, AND HEMOTHORAX Michael McColl, BS Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF MALE BREAST CANCER David W. McFadden, MD Murray-­Heilig Chairman and Professor Department of Surgery University of Connecticut Farmington, Connecticut MANAGEMENT OF MALLORY-­WEISS SYNDROME Ashley D. Meagher, MD, MPH Assistant Professor Department of Surgery Indiana University Health Methodist Hospital Indianapolis, Indiana ANTIFUNGAL THERAPY IN THE SURGICAL PATIENT

xxi

Fabrizio Michelassi, MD Lewis Atterbury Stimson Professor and Chairman Department of Surgery Weill Cornell Medical College Surgeon-­in-­Chief New York-­Presbyterian Hospital at Weill Cornell Medical Center New York, New York MANAGEMENT OF CROHN’S DISEASE OF THE SMALL BOWEL David J. Milia, MD Associate Professor Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin ANTIBIOTICS FOR CRITICALLY ILL PATIENTS Barbra S. Miller, MD Associate Professor Director, NW Thompson Endocrine Surgery Fellowship Department of Surgery Section of General Surgery Division of Endocrine Surgery University of Michigan Ann Arbor, Michigan MANAGEMENT OF ADRENAL CORTICAL TUMORS Emily J. Miraflor, MD Assistant Professor Department of Surgery University of California, San Francisco East Bay Oakland, California MANAGEMENT OF PHEOCHROMOCYTOMA Erica L. Mitchell, MD, MEd Medical Director Vascular and Endovascular Surgery Salem Health Medical Group Salem Health Hospitals and Clinics Salem, Oregon PSEUDOANEURYSMS AND ARTERIOVENOUS FISTULAS

xxii

CONTRIBUTORS

Daniela Molena, MD Surgical Director of Esophageal Cancer Surgery Program Thoracic Surgery Memorial Sloane Kettering Cancer Center New York, New York

Christopher R. Morse, MD Assistant Professor Thoracic Surgery Massachusetts General Hospital Harvard University Boston, Massachusetts

Katie S. Nason, MD, MPH Vice-­Chair of Surgery for Research Department of Surgery University of Massachusetts Baystate Health Center Springfield, Massachusetts

MULTIMODALITY THERAPY IN ESOPHAGEAL CANCER

MANAGEMENT OF ACQUIRED ESOPHAGEAL RESPIRATORY TRACT FISTULA

MANAGEMENT OF ZENKER’S DIVERTICULUM

George Molina, MD, MPH Fellow in Complex Surgical Oncology Department of Surgery Massachusetts General Hospital Brigham and Women’s Hospital Dana-­Farber Cancer Institute Boston, Massachusetts

Peter Muscarella II, MD Professor Department of Surgery Montefiore Medical Center New York, New York

Yosef Nasseri, MD, FACS, FASCRS Attending Colorectal Surgeon Cedars-­Sinai Medical Center Surgery Group of Los Angeles Los Angeles, California

MANAGEMENT OF CYSTS, TUMORS, AND ABSCESSES OF THE SPLEEN

ANORECTAL STRICTURE

Amol Narang, MD Assistant Professor Department of Radiation Oncology and Molecular Sciences Johns Hopkins University School of Medicine Baltimore, Maryland

Naiem Nassiri, MD Associate Professor of Surgery (Vascular) Department of Surgery Yale University School of Medicine New Haven, Connecticut; Chief, Vascular and Endovascular Surgery VA Connecticut Healthcare System West Haven, Connecticut

CONDYLOMA ACUMINATA

INTRAOPERATIVE RADIATION FOR PANCREATIC CANCER

RAYNAUD’S PHENOMENON

John R. Montgomery, MD Research Fellow Department of General Surgery University of Michigan Ann Arbor, Michigan

Rahul Narang, MD Assistant Professor of Surgery Department of Surgery Montefiore Medical Center New York, New York

PERIOPERATIVE OPTIMIZATION

MANAGEMENT OF ISCHEMIC COLITIS

Derek T. Moore, MD Department of Surgery University of Massachusetts Medical School/ Baystate Baystate Medical Center Springfield, Massachusetts

Mayur Narayan, MD, MPH, MBA, MHPE, FACS, FCCM, FICS Associate Professor Department of Surgery Division of Trauma, Burns, Critical and Acute Care Weill Cornell Medicine; Associate Attending Surgeon Associate Program Director Surgical Critical Care Fellowship New York-­Presbyterian Hospital/Weill Cornell Medical Center New York, New York

INTRADUCTAL PAPILLARY MUCINOUS NEOPLASMS OF THE PANCREAS John R.T. Monson, MD, FRCS, FACS, FASCRS Executive Director, Colorectal Surgery Center for Colon and Rectal Surgery AdventHealth Orlando, Florida

MANAGEMENT OF ZENKER’S DIVERTICULUM Robert J. Moraca, MD EXTRACORPOREAL MEMBRANE OXYGENATION FOR RESPIRATORY FAILURE IN ADULTS Benjamin Moran, MD Visiting Instructor Department of Surgery University of Maryland Medical School Baltimore, Maryland GLUCOSE CONTROL IN THE POSTOPERATIVE PERIOD

MEDICAL MANAGEMENT OF THE BURN PATIENT Nicholas A. Nash, MD Associate Professor Department of Surgery University of Louisville Louisville, Kentucky TENETS OF DAMAGE CONTROL

Victoria Needham, MD Resident Department of Surgery Montefiore Medical Center New York, New York MANAGEMENT OF COLONIC VOLVULUS Matthew A. Nehs, MD Assistant Professor Department of Surgery Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts NONTOXIC GOITER Lisa A. Newman, MD, MPH Director, Interdisciplinary Breast Program Department of Surgery Weill Cornell Medicine–New York Presbyterian Hospital Network New York, New York INFLAMMATORY BREAST CANCER Quang Nguyen, MD Interventional Radiologist Department of Radiology Medstar Good Samaritan Hospital Baltimore, Maryland TRANSARTERIAL CHEMOEMBOLIZATION FOR LIVER METASTASES

C O N T R I B U TO R S

Madhuri V. Nishtala, MD General Surgery Resident Department of Surgery University of Wisconsin Madison, Wisconsin MANAGEMENT OF TOXIC MEGACOLON Nicholas Nissen, MD Director, Liver Transplant and Hepato-­ Pancreato-­Biliary Surgery Department of Surgery Cedars-­Sinai Medical Center Los Angeles, California MANAGEMENT OF BUDD-­CHIARI SYNDROME Bolin Niu, MD Fellow Division of Gastroenterology and Hepatology Department of Medicine Thomas Jefferson University Hospital Philadelphia, Pennsylvania MANAGEMENT OF REFRACTORY ASCITES Erik R. Noren, MD, MS Resident Department of Surgery Keck School of Medicine University of Southern California Los Angeles, California SURGICAL MANAGEMENT OF THE POLYPOSIS SYNDROMES Jeffrey A. Norton, MD Professor Department of Surgery Stanford University School of Medicine Stanford, California MANAGEMENT OF PANCREATIC ISLET CELL TUMORS EXCLUDING GASTRINOMA Melanie Nukala, MD Vascular Surgery Fellow Department of Vascular Surgery Saint Louis University St. Louis, Missouri

Izi Obokhare, MD Associate Professor General Surgery, Colon and Rectal Surgery Advanced Laparoscopic/Robotic Surgery Department of Surgery Texas Tech University Health Sciences Center Amarillo, Texas MANAGEMENT OF HEMORRHOIDS Brant K. Oelschlager, MD Professor & Chief Byers Endowed Professor of Esophageal Research Division of General Surgery University of Washington Medical Center Seattle, Washington ENDOSCOPIC TREATMENT OF BARRETT’S ESOPHAGUS Terence O’Keeffe, MBChB, FACS, FCCM Professor Department of Surgery Banner University Medical Center Tucson, Arizona COAGULOPATHY IN THE CRITICALLY ILL PATIENT Charles S. O’Mara, MD, MBA Professor Associate Vice Chancellor for Clinical Affairs Department of Surgery University of Mississippi Medical Center Jackson, Mississippi BALLOON ANGIOPLASTY AND STENTS IN CAROTID ARTERY OCCLUSIVE DISEASE Patrick B. O’Neal, MD Staff Surgeon Veteran’s Administration–Boston Healthcare System Assistant Professor Department of Surgery Boston University Boston, Massachusetts

MANAGEMENT OF INFECTED GRAFTS

NECROTIZING SKIN AND SOFT TISSUE INFECTIONS

Iheoma Y. Nwaogu, MD Clinical Fellow Department of Surgery University of California, San Francisco San Francisco, California

Cecilia T. Ong, MD Surgical Resident Duke University Durham, North Carolina

ADRENAL INCIDENTALOMA

MANAGEMENT OF COMMON BILE DUCT STONES

xxiii

Mark S. Orloff, MD Professor Department of Surgery University of Rochester Rochester, New York PORTAL HYPERTENSION: ROLE OF SHUNTING PROCEDURES Tawakalitu O. Oseni, MD Instructor Department of Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF THE AXILLA IN BREAST CANCER Greg M. Osgood, MD Assistant Professor of Orthopedic Surgery Johns Hopkins University School of Medicine Baltimore, Maryland EARLY MANAGEMENT OF PELVIC RING DISRUPTION Shane E. Ottmann, MD Transplant Surgeon Department of Surgery Johns Hopkins Hospital Baltimore, Maryland PORTAL HYPERTENSION: ROLE OF SHUNTING PROCEDURES Heidi N. Overton, MD Resident Department of General Surgery Johns Hopkins University School of Medicine Baltimore, Maryland THE ABDOMEN THAT WILL NOT CLOSE Shakirat Oyetunji, MD, MPH Chief Resident Department of Cardiothoracic Surgery University of Washington Seattle, Washington MANAGEMENT OF TRACHEAL STENOSIS H. Leon Pachter, MD, FACS George David Stewart Professor Chairman, Department of Surgery New York University School of Medicine New York, New York MANAGEMENT OF DIAPHRAGMATIC INJURIES

xxiv

CONTRIBUTORS

Javier Salgado Pagacnik, MD Assistant Professor of Surgery University of Texas Southwestern Medical Center Dallas, Texas MANAGEMENT OF RECTAL CANCER Barnard J.A. Palmer, MD, Med Associate Professor Department of Surgery University of California, San Francisco–East Bay Oakland, California MANAGEMENT OF PHEOCHROMOCYTOMA Nikhil Panda, MD General Surgery Resident Department of Surgery Massachusetts General Hospital Clinical Fellow in Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ACQUIRED ESOPHAGEAL RESPIRATORY TRACT FISTULA Theodore N. Pappas, MD, FACS Distinguished Professor of Surgical Innovation Chief of Advanced Oncologic and Gastrointestinal Surgery Duke University School of Medicine Durham, North Carolina

Marco G. Patti, MD Professor of Medicine and Surgery Department of Medicine and Surgery, Center for Esophageal Diseases and Swallowing University of North Carolina at Chapel Hill Chapel Hill, North Carolina MANAGEMENT OF PARAESOPHAGEAL HIATAL HERNIA Timothy M. Pawlik, MD, MPH, PhD, FACS, FRACS(Hon) Professor and Chair Department of Surgery The Ohio State University Columbus, Ohio MANAGEMENT OF MALIGNANT LIVER TUMORS Bruce A. Perler, MD, MBA Julius H. Jacobson II Professor Vice Chair for Clinical Operations and Financial Affairs Chief Emeritus, Division of Vascular Surgery & Endovascular Therapy Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland; Vice-­President American Board of Surgery Philadelphia, Pennsylvania

Jason D. Prescott, MD, PhD Assistant Professor Department of Surgery Johns Hopkins School of Medicine Baltimore, Maryland SURGICAL MANAGEMENT OF THYROID CANCER Leigh Ann Price, MD Director, National Burn Reconstruction Center Department of Surgery MedStar Good Samaritan Hospital Assistant Professor Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland ELECTRICAL INJURIES AND LIGHTNING INJURIES Brandon W. Propper, MD, FACS, RPVI Program Director, General Surgery Department of Vascular Surgery San Antonio Military Medical Center San Antonio, Texas; Associate Professor of Surgery Department of Surgery Uniformed Services University of the Health Sciences Bethesda, Maryland

MANAGEMENT OF COMMON BILE DUCT STONES

OPEN REPAIR OF ABDOMINAL AORTIC ANEURYSMS

THE DIABETIC FOOT

Lisa Park, MD Resident Physician Department of General Surgery Cleveland Clinic Cleveland, Ohio

Erica Pettke, MD Resident Department of Surgery Mount Sinai West Hospital New York, New York

Scott G. Prushik, MD Vascular Surgeon Division of Vascular and Endovascular Surgery St. Elizabeth’s Medical Center Brighton, Massachusetts

SURGICAL MANAGEMENT OF CONSTIPATION

MANAGEMENT OF COLON POLYPS

Pankaj Jay Pasricha, MBBS, MD Professor Departments of Medicine and Neuroscience Johns Hopkins University School of Medicine Baltimore, Maryland

Henry A. Pitt, MD Professor of Surgery Lewis Katz School of Medicine at Temple University Philadelphia, Pennsylvania

MANAGEMENT OF MOTILITY DISORDERS OF THE STOMACH AND SMALL BOWEL Andrew B. Peitzman, MD Mark M. Ravitch Professor of Surgery Department of Surgery University of Pittsburgh Pittsburgh, Pennsylvania CURRENT MANAGEMENT OF RECTAL INJURY

MANAGEMENT OF HEPATIC ABSCESSES Amani D. Politano, MD, MS Assistant Professor Department of Vascular Surgery Oregon Health & Science University Portland, Oregon PSEUDOANEURYSMS AND ARTERIOVENOUS FISTULAS

ENDOVASCULAR TREATMENT OF ABDOMINAL AORTIC ANEURYSMS Thaddeus J. Puzio, MD Trauma and Surgical Critical Care Fellow University of Texas Health Science Center at Houston Houston, Texas FLUID AND ELECTROLYTE THERAPY T. Robert Qaqish, MD, MSc Chief Resident General Surgery Sinai Hospital of Baltimore Baltimore, Maryland MANAGEMENT OF DISORDERS OF ESOPHAGEAL MOTILITY

C O N T R I B U TO R S

Andre R. Ramdon, MD Department of Surgery Philadelphia VA Medical Center Philadelphia, Pennsylvania MANAGEMENT OF RUPTURED ABDOMINAL AORTIC ANEURYSMS Christine Ramirez, MD Acute Care Surgeon St. Luke’s University Health Network Bethlehem, Pennsylvania; Clinical Assistant Professor Department of Surgery Temple University Philadelphia, Pennsylvania MANAGEMENT OF TRAUMATIC BRAIN INJURY

Christopher R. Reed, MD Resident Department of Surgery Duke University Medical Center Durham, North Carolina

Sean P. Roddy, MD Professor Department of Surgery Albany Medical College Albany, New York

TRACHEOSTOMY

MANAGEMENT OF RUPTURED ABDOMINAL AORTIC ANEURYSMS

Thomas Reifsnyder, MD Associate Professor Department of Surgery Johns Hopkins Bayview Medical Center Baltimore, Maryland MANAGEMENT OF TIBIOPERONEAL ARTERIAL OCCLUSIVE DISEASE

MANAGEMENT OF MORBID OBESITY

HEMODIALYSIS ACCESS SURGERY

Yesenia Rojas-­Khalil, MD Assistant Professor of Surgery Section of Colorectal Surgery Michael E. DeBakey Department of Surgery Baylor College of Medicine Houston, Texas

Taylor S. Riall, MD, PhD Professor Department of Surgery University of Arizona Tucson, Arizona

BALLOON ANGIOPLASTY AND STENTS IN CAROTID ARTERY OCCLUSIVE DISEASE

PROPER USE OF CHOLECYSTOSTOMY TUBES

ATHLETIC PUBALGIA: THE “SPORTS HERNIA” Chandrajit P. Raut, MD, MSc Associate Surgeon Department of Surgery Brigham and Women’s Hospital; Surgery Director Center for Sarcoma and Bone Oncology Dana-­Farber Cancer Institute; Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF SOFT TISSUE SARCOMA Bradley N. Reames, MD, MS Assistant Professor Department of Surgery Section of Surgical Oncology University of Nebraska Medical Center Omaha, Nebraska MANAGEMENT OF GASTRIC ADENOCARCINOMA MANAGEMENT OF PERIAMPULLARY CANCER

Andrew P. Rogers, MD General Surgery Resident Department of Surgery University of Wisconsin School of Medicine and Public Health Madison, Wisconsin

ACUTE MESENTERIC ISCHEMIA

Daniel E. Ramirez, MD Assistant Professor Department of Vascular Surgery University of Mississippi Jackson, Mississippi

Bruce Ramshaw, MD, FACS Professor and Chair Department of Surgery University of Tennessee Graduate School of Medicine; Adjunct Professor Graduate and Executive MBA Program Haslam College of Business University of Tennessee Knoxville, Tennessee

xxv

J. David Richardson, MD Professor Department of Surgery University of Louisville School of Medicine Louisville, Kentucky

MANAGEMENT OF CHRONIC ULCERATIVE COLITIS

Charles G. Rickert, MD, PhD Surgical Resident Department of General Surgery Massachusetts General Hospital Boston, Massachusetts

John R. Romanelli, MD, FACS Professor of Surgery University of Massachusetts Medical School/ Baystate; Minimally Invasive Surgery Fellowship Director Medical Director, Metabolic and Bariatric Surgery Medical Director, Baystate General Surgery Practice Baystate Medical Center Springfield, Massachusetts

ISLET AUTOTRANSPLANTATION FOR CHRONIC PANCREATITIS

MANAGEMENT OF ZENKER’S DIVERTICULUM

Addi Z. Rizvi, MD Vascular and Endovascular Surgeon Providence Vascular Institute Sacred Heart Medical Center Spokane, Washington

Ronnie A. Rosenthal, MD, FACS Professor of Surgery Yale University School of Medicine New Haven, Connecticut; Chief, Surgical Service VA Connecticut Health Care System West Haven, Connecticut

TENETS OF DAMAGE CONTROL

PROFUNDA FEMORIS RECONSTRUCTION Thomas N. Robinson, MD, MS, FACS Chief of Surgery Rocky Mountain VA Medical Center Professor of Surgery University of Colorado Aurora, Colorado OPTIMIZING PERIOPERATIVE CARE OF THE OLDER ADULT

OPTIMIZING THE PERIOPERATIVE CARE OF THE OLDER ADULT Gedge D. Rosson, MD Associate Professor Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland BREAST RECONSTRUCTION FOLLOWING MASTECTOMY: INDICATIONS, TECHNIQUES, AND RESULTS NERVE INJURY AND REPAIR

xxvi

CONTRIBUTORS

Grace S. Rozycki, MD, MBA, RDMS Professor Department of Surgery Johns Hopkins School of Medicine Baltimore, Maryland

Bashar Safar, MBBS Assistant Professor of Surgery Department of General Surgery Johns Hopkins University Baltimore, Maryland

SURGICAL USE OF ULTRASOUND IN THE TRAUMA AND CRITICAL CARE SETTINGS

PET SCANNING IN THE MANAGEMENT OF COLORECTAL CANCER

Mario Rueda, MD, FACCS Medical Director, Trauma and Surgical Critical Care Site Director, Surgery Residency St. Mary’s Medical Center; Assistant Professor of Surgery Florida Atlantic University Charles E. Schmidt College of Medicine West Palm Beach, Florida

Zeyad T. Sahli, MD Resident Department of General Surgery University of Virginia Charlottesville, Virginia

INITIAL ASSESSMENT AND RESUSCITATION OF THE TRAUMA PATIENT Jonathon O. Russell, MD Assistant Professor Director of Endoscopic and Robotic Thyroid and Parathyroid Surgery Department of Otolaryngology–Head and Neck Surgery Johns Hopkins University Baltimore, Maryland

SURGICAL MANAGEMENT OF SECONDARY AND TERTIARY HYPERPARATHYROIDISM Joseph V. Sakran, MD, MPH, MPA Assistant Professor Department of Surgery Johns Hopkins Hospital Baltimore, Maryland

Joseph R. Scalea, MD Assistant Professor of Surgery and Immunology Director, Pancreas Transplantation Department of Surgery Division of Transplantation University of Maryland Baltimore, Maryland PANCREAS TRANSPLANTATION Thomas M. Scalea, MD, FACS, MCCM Francis X. Kelly Professor of Trauma Surgery Director, Program in Trauma Physician-­in-­Chief University of Maryland School of Medicine R. Adams Cowley Shock Trauma Center Baltimore, Maryland AIRWAY MANAGEMENT IN THE TRAUMA PATIENT

PANCREATIC AND DUODENAL INJURIES

GLUCOSE CONTROL IN THE POSTOPERATIVE PERIOD

Cesar A. Santa-­Maria, MD, MSCI Assistant Professor Department of Oncology Johns Hopkins University Baltimore, Maryland

Philip R. Schauer, MD Professor of Surgery Bariatric and Metabolic Institute Cleveland Clinic Cleveland, Ohio

MANAGEMENT OF THYROID NODULES

ADVANCES IN NEOADJUVANT AND ADJUVANT THERAPY FOR BREAST CANCER

CARDIOVASCULAR DISEASE RISK REDUCTION AFTER BARIATRIC SURGERY

Colleen M. Ryan, MD Professor Department of Surgery Massachusetts General Hospital Harvard Medical School Staff Surgeon, Shriners Hospitals for Children-­Boston Boston, Massachusetts

Samuel Sarmiento, MD, MPH, MBA Postdoctoral Fellow Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

David W. Schechtman, MD Resident Department of General Surgery Brooke Army Medical Center San Antonio, Texas; Teaching Fellow Department of Surgery Uniformed Services University of the Health Sciences Bethesda, Maryland

ELECTRICAL INJURY AND LIGHTNING INJURIES Bethany C. Sacks, MD, MEd Assistant Professor Clerkship and Curriculum Director Department of Surgery Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF INGUINAL HERNIA Justin M. Sacks, MD, MBA, FACS Assistant Professor Director, Oncological Reconstruction Department of Plastic and Reconstructive Surgery Johns Hopkins University School of Medicine Baltimore, Maryland LYMPHEDEMA

MANAGEMENT OF MALE BREAST CANCER Robert Sawyer, MD, FACS, FIDSA, FCCM Professor and Chair of Surgery Western Michigan University Homer Stryker MD School of Medicine Adjunct Professor of Engineering and Applied Sciences Western Michigan University Kalamazoo, Michigan; Adjunct Professor of Surgery University of Virginia School of Medicine Charlottesville, Virginia SURGICAL SITE INFECTIONS

THE DIABETIC FOOT Francisco Schlottmann, MD, MPH Associate Attending Surgeon Department of Surgery Hospital Alemán of Buenos Aires University of Buenos Aires Buenos Aires, Argentina MANAGEMENT OF PARAESOPHAGEAL HIATAL HERNIA

C O N T R I B U TO R S

Andrew Schulick, MD, MBA, FACS Vascular and Endovascular Surgery Department of Surgery Johns Hopkins School of Medicine Baltimore, Maryland; Chief, Vascular Surgery Suburban Hospital Bethesda, Maryland MANAGEMENT OF VASCULAR INJURIES Richard D. Schulick, MD, MBA, FACS Professor and Chair Department of Surgery Director University of Colorado Cancer Center University of Colorado School of Medicine Aurora, Colorado MANAGEMENT OF ECHINOCOCCAL CYST DISEASE OF THE LIVER Samuel I. Schwartz, MD Assistant Professor Department of Vascular Surgery Massachusetts General Hospital Boston, Massachusetts LOWER EXTREMITY AMPUTATION Rebecca Scully, MD, MPH Resident Department of Surgery Brigham and Women’s Hospital Boston, Massachusetts GANGRENE OF THE FOOT Colby J. Seegmiller, DO Resident Department of General Surgery Gundersen Health System La Crosse, Wisconsin MANAGEMENT OF PERIPHERAL ARTERIAL THROMBOEMBOLISM Naomi M. Sell, MD, MHS Resident Department of Surgery Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF PILONIDAL DISEASE Gregory Semon, DO, FACS, FACOS Assistant Professor Department of Surgery Division of Acute Care Surgery Wright State University Boonshoft School of Medicine Dayton, Ohio CHEST WALL, PNEUMOTHORAX, AND HEMOTHORAX

Katherine Senter, MD Department of Surgery Alaska Trauma and Acute Care Surgery Anchorage, Alaska PENETRATING NECK TRAUMA Samir Kaushik Shah, MD, MPH Instructor of Surgery Department of Vascular Surgery Harvard Medical School Brigham and Women’s Hospital Boston, Massachusetts GANGRENE OF THE FOOT Jay G. Shake, MD, MS, FACS Director, Cardiovascular Intensive Care Unit Department of Surgery University of Mississippi Medical Center Professor of Surgery and Anesthesiology University of Mississippi Jackson, Mississippi CARDIOVASCULAR PHARMACOLOGY Jonathan G. Sham, MD Assistant Professor Department of Surgery University of Washington School of Medicine Seattle, Washington MANAGEMENT OF PERIAMPULLARY CANCER Jessica B. Shank, MD Assistant Professor Department of Surgery Division of Surgical Oncology University of Nebraska Medical Center Omaha, Nebraska

xxvii

Shmuel Shoham, MD Associate Professor Department of Medicine Johns Hopkins University School of Medicine Baltimore, Maryland EPIDEMIOLOGY, PREVENTION, AND MANAGEMENT OF OCCUPATIONAL EXPOSURE TO BLOODBORNE INFECTIONS Jason K. Sicklick, MD, FACS Associate Professor of Surgery Division of Surgical Oncology University of California San Diego Cancer Center University of California San Diego Health System La Jolla, California MANAGEMENT OF CYSTIC DISORDERS OF THE BILE DUCTS Carrie A. Sims, MD, PhD, FACS Chief of Trauma, Critical Care, and Burns The Ohio State University Wexner Medical Center Columbus, Ohio INJURIES TO SMALL AND LARGE BOWEL Vikesh K. Singh, MD, MSc Associate Professor of Medicine Division of Gastroenterology Johns Hopkins University School of Medicine Director of Endoscopy Johns Hopkins Hospital Baltimore, Maryland ENDOSCOPIC THERAPY FOR ESOPHAGEAL VARICEAL HEMORRHAGE

SURGICAL MANAGEMENT OF THYROID CANCER

TOTAL PANCREATECTOMY WITH ISLET AUTOTRANSPLANTATION

Josef A. Shehebar, MD, FACS, FASCRS Attending Surgeon, Colon and Rectal Surgery NYU Langone Medical Center Director, Colon and Rectal Surgery NYU Langone Hospital–Brooklyn New York, New York

Matthew R. Smeds, MD, FACS Chief, Division of Vascular and Endovascular Surgery Program Director, Vascular Surgery Training Programs Associate Professor, Department of Surgery Saint Louis University St. Louis, Missouri

RECTOVAGINAL FISTULA David Shibata, MD Scheinberg Endowed Chair in Surgery Professor and Chair Department of Surgery University of Tennessee Health Science Center Memphis, Tennessee MANAGEMENT OF TUMORS OF THE ANAL REGION

MANAGEMENT OF INFECTED GRAFTS Barbara L. Smith, MD, PhD Director, Breast Program Division of Surgical Oncology Massachusetts General Hospital Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF THE AXILLA IN BREAST CANCER

xxviii

CONTRIBUTORS

Philip W. Smith, MD Associate Professor Department of Surgery University of Virginia Charlottesville, Virginia MANAGEMENT OF SOLITARY NECK MASS Mark Soliman, MD, FACS, FASCRS Program Director Minimally Invasive Colorectal Surgical Fellowship Colon and Rectal Clinic of Orlando Orlando, Florida DIAGNOSIS, TREATMENT, AND SURGICAL MANAGEMENT OF FISSURES-­IN-­ANO Ian Solsky, MD, MPH General Surgery Resident Department of Surgery Montefiore Medical Center New York, New York MANAGEMENT OF ISCHEMIC COLITIS Renganaden Sooppan, MD Vascular Surgery Fellow Department of Surgery Johns Hopkins Hospital Baltimore, Maryland MANAGEMENT OF RECURRENT CAROTID STENOSIS Rebecca Sorber, MD Resident Department of General Surgery Johns Hopkins Hospital Baltimore, Maryland BRACHIOCEPHALIC RECONSTRUCTION Julie Ann Sosa, MD, MA, FACS Leon Goldman MD Distinguished Professor and Chair Department of Surgery University of California, San Francisco San Francisco, California MANAGEMENT OF HYPERTHYROIDISM Krishnan Sriram, MD, FCCM, FRCS(C), FACS Tele-­Intensivist VISN 23 Tele-­Intensive Care System Veterans Affairs Minneapolis, Minnesota NUTRITION THERAPY IN THE CRITICALLY ILL SURGICAL PATIENT

Scott R. Steele, MD, MBA, FACS, FASCRS Chairman Department of Colorectal Surgery Cleveland Clinic Cleveland, Ohio MANAGEMENT OF TOXIC MEGACOLON Deborah Stein, MD, MPH Professor and Chair of Surgery Zuckerberg San Francisco General; Vice-­Chair of Trauma and Critical Care Surgery Department of Surgery University of California, San Francisco San Francisco, California MANAGEMENT OF TRAUMATIC BRAIN INJURY Toby B. Steinberg, MD Adult Cardiothoracic Anesthesiology Fellow Department of Anesthesia and Critical Care Hospital of the University of Philadelphia Philadelphia, Pennsylvania MANAGEMENT OF DESCENDING THORACIC AND THORACOABDOMINAL AORTIC ANEURYSMS Kent A. Stevens, MD, MPH Associate Professor Department of Surgery Johns Hopkins Medical Institutions; Director, Adult Trauma Services Johns Hopkins Hospital Baltimore, Maryland THE ABDOMEN THAT WILL NOT CLOSE Steven M. Strasberg, MD, FRCS(C), FACS, FRCS(Ed) Pruett Professor of Surgery Section of Hepato-­Pancreato-­Biliary Surgery Washington University in St. Louis St. Louis, Missouri ASYMPTOMATIC GALLSTONES Lee L. Swanström, MD, FACS, FASGE, FRCSEng Professor of Surgery Institute for Image Guided Surgery IHU-­Strasbourg Strasbourg, Alsace, France; Director Division of Minimally Invasive and GI Surgery The Oregon Clinic Portland, Oregon MANAGEMENT OF ESOPHAGEAL CANCER

Mark A.Talamini, MD, MBA Chair, Department of Surgery Chief, Surgical Services Stony Brook Medicine Stony Brook, New York USE OF STRICTUREPLASTY IN CROHN’S DISEASE Winson Jianhong Tan, MBBS(Hon), MRCS(Ed), MMed(Surg), FRCS(Ed) Fellow Division of Surgery, Colorectal Service Memorial Sloan-­Kettering Cancer Center New York, New York; Consultant Department of General Surgery Sengkang General Hospital Department of Colorectal Surgery Singapore General Hospital Singapore SURGICAL MANAGEMENT OF COLON CANCER Kenneth K.Tanabe, MD Professor of Surgery Harvard Medical School Chief of Surgical Oncology Massachusetts General Hospital Boston, Massachusetts MANAGEMENT OF BENIGN LIVER TUMORS James M.Tatum, MD General and Minimally Invasive Surgery Long Beach Memorial Medical Center Long Beach, California MAGNETIC AUGMENTATION OF THE LOWER ESOPHAGEAL SPHINCTER Ali Tavakkoli, MD, FACS, FRCS Associate Professor Department of Surgery Director, Minimally Invasive and Weight Loss Surgery Fellowship Co-­Director, Center for Weight Management and Metabolic Surgery Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts METABOLIC CHANGES FOLLOWING BARIATRIC SURGERY

C O N T R I B U TO R S

James Taylor, MBBChir, MPH General Surgery Resident Department of Surgery Johns Hopkins University Baltimore, Maryland PET SCANNING IN THE MANAGEMENT OF COLORECTAL CANCER

Alastair M.Thompson, BSc(Hon), MBChB, MD, FRCSEd(Gen) Olga Keith Weiss Chair of Surgery Division of Surgical Oncology Baylor College of Medicine Houston, Texas

Ralph P.Tufano, MD, MBA Charles W. Cummings MD Professor Departments of Otolaryngology–Head and Neck Surgery Johns Hopkins University School of Medicine Baltimore, Maryland

DUCTAL AND LOBULAR CARCINOMA IN SITU OF THE BREAST

MANAGEMENT OF THYROID NODULES

John R.Taylor III, MD Assistant Professor Department of Surgery Division of Acute Care Surgery University of Arkansas for Medical Sciences Little Rock, Arkansas

Myles Tieszen, MD, FACS Assistant Professor Department of Surgery Creighton University School of Medicine Omaha, Nebraska

COAGULATION ISSUES AND THE TRAUMA PATIENT

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

Spence M.Taylor, MD President Chair, Board of Managers GHS Health Sciences Center Prisma Health-­Upstate; Professor of Surgery University of South Carolina School of Medicine Greenville, South Carolina

Susan Tsai, MD, MHS Associate Professor of Surgical Oncology Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin

AORTOILIAC OCCLUSIVE DISEASE Christine B.Teal, MD Associate Professor Director, Breast Care Center Department of Surgery George Washington University Washington, DC A SURGEON’S PRACTICAL GUIDE TO BREAST IMAGING Maria E.Tecos, MD Resident Department of General Surgery University of Nebraska Medical Center Omaha, Nebraska ACUTE KIDNEY INJURY IN THE INJURED AND CRITICALLY ILL Ronald Tesoriero, MD, FACS Assistant Professor Department of Surgery University of Maryland Chief, Trauma Critical Care Program in Trauma R. Adams Cowley Shock Trauma Center Baltimore, Maryland GLUCOSE CONTROL IN THE POSTOPERATIVE PERIOD

xxix

PALLIATIVE THERAPY FOR PANCREATIC CANCER UNUSUAL PANCREATIC TUMORS Miriam W.Tsao, MD Fellow Department of Surgery University of Tennessee Health Science Center Memphis, Tennessee MANAGEMENT OF TUMORS OF THE ANAL REGION Theodore N.Tsangaris, MD, MBA Surgical Director, Jefferson Breast Cancer Center Sidney Kimmel Cancer Center Thomas Jefferson University; Professor of Surgery and Chief of Breast Surgery Department of Surgery Sidney Kimmel Medical College Philadelphia, Pennsylvania MANAGEMENT OF RECURRENT AND METASTATIC BREAST CANCER Anthony P.Tufaro, DDS, MD, FACS Professor and Chief Section of Plastic Surgery University of Oklahoma Health Science Center Oklahoma City, Oklahoma; Professor Department of Plastic Surgery Johns Hopkins University Baltimore, Maryland ABDOMINAL WALL RECONSTRUCTION

Lily Tung, MD Vancouver General Hospital Vancouver, BC, Canada INJURIES TO SMALL AND LARGE BOWEL Robert Udelsman, MD, MBA, FACS, FACE Director, Endocrine Surgeon Endocrine Neoplasia Institute Miami Cancer Institute; Professor of Surgery Florida International University Miami, Florida; Professor of Surgery Emeritus Yale University School of Medicine New Haven, Connecticut EVALUATION AND MANAGEMENT OF PERSISTENT OR RECURRENT PRIMARY HYPERPARATHYROIDISM Prashanth Vallabhajosyula, MD, MS Associate Professor of Surgery Department of Surgery University of Pennsylvania Philadelphia, Pennsylvania MANAGEMENT OF DESCENDING THORACIC AND THORACOABDOMINAL AORTIC ANEURYSMS Fons van den Berg, MD PhD Candidate Department of Surgery Amsterdam UMC Amsterdam, The Netherlands GALLSTONE PANCREATITIS Marissa Vane, MD, MSc Department of Surgery Maastricht UMC+ Maastricht, The Netherlands ABLATIVE TECHNIQUES IN THE TREATMENT OF BENIGN AND MALIGNANT BREAST DISEASE Arthur Jason Vaught, MD Department of Maternal-­Fetal Medicine, Gynecology, and Obstetrics Johns Hopkins University School of Medicine Baltimore, Maryland MULTIPLE ORGAN DYSFUNCTION AND FAILURE

xxx

CONTRIBUTORS

Jula Veerapong, MD Associate Professor Department of Surgery University of California, San Diego San Diego, California

Tracy S. Wang, MD, MPH, FACS Professor of Surgery Chief, Section of Endocrine Surgery Medical College of Wisconsin Milwaukee, Wisconsin

MANAGEMENT OF PERITONEAL SURFACE MALIGNANCIES

MANAGEMENT OF HYPERTHYROIDISM

Ana M.Velez-Rosborough, MD Fellow Division of Trauma and Surgical Critical Care Department of Surgery Ryder Trauma Center/Jackson Memorial Hospital University of Miami Miller School of Medicine Miami, Florida

Andrew L. Warshaw, MD, FACS, FRCSEd(Hon) Surgeon-­in-­Chief Emeritus Department of Surgery Massachusetts General Hospital; W. Gerald Austen Distinguished Professor of Surgery Department of Surgery Harvard Medical School Boston, Massachusetts

ACID-­BASE PROBLEMS

MANAGEMENT OF ACUTE NECROTIZING PANCREATITIS

George Velmahos, MD, PhD, MSEd Chief, Division of Trauma, Emergency Surgery, and Critical Care Department of Surgery Massachusetts General Hospital John F. Burke Professor of Surgery Harvard Medical School Boston, Massachusetts MANAGEMENT OF ACUTE CHOLECYSTITIS Charles M.Vollmer Jr, MD Professor Department of Surgery University of Pennsylvania Philadelphia, Pennsylvania NEOADJUVANT AND ADJUVANT THERAPY FOR PANCREATIC CANCER Carl M. Wahlgren, MD, PhD Chief and Professor Department of Vascular Surgery Karolinska Institutet/Karolinska University Hospital Stockholm, Sweden BUERGER’S DISEASE (THROMBOANGIITIS OBLITERANS) Christi Walsh, BS, MSN Department of Surgery Johns Hopkins Hospital Baltimore, Maryland TOTAL PANCREATECTOMY WITH ISLET AUTOTRANSPLANTATION

PRIMARY HYPERPARATHYROIDISM

Travis P. Webb, MD, MHPE Professor Department of Surgery Medical College of Wisconsin Milwaukee, Wisconsin PREOPERATIVE BOWEL PREPARATION: IS IT NECESSARY? IS NASOGASTRIC TUBE NECESSARY AFTER ALIMENTARY TRACT SURGERY? Sharon R. Weeks, MD General Surgery Resident Department of Surgery Walter Reed National Military Medical Center Bethesda, Maryland PORTAL HYPERTENSION: ROLE OF SHUNTING PROCEDURES Clifford R. Weiss, MD, FACS Associate Professor of Radiology, Surgery, and Biomedical Engineering Department of Radiology/Division of Interventional Radiology Johns Hopkins University School of Medicine Baltimore, Maryland TRANSHEPATIC INTERVENTIONS FOR OBSTRUCTIVE JAUNDICE Matthew J. Weiss, MD, FACS Deputy Physician-­in-­Chief Director of Surgical Oncology Northwell Health New Hyde Park, New York MANAGEMENT OF LIVER HEMANGIOMAS

Ethel D. Weld, MD Assistant Professor of Medicine Department of Infectious Diseases Department of Clinical Pharmacology Johns Hopkins University School of Medicine Baltimore, Maryland EPIDEMIOLOGY, PREVENTION, AND MANAGEMENT OF OCCUPATIONAL EXPOSURE TO BLOODBORNE INFECTIONS Katarina Wells, MD, MPH Director of Colorectal Research Department of Surgery Baylor University Medical Center MANAGEMENT OF RECTAL PROLAPSE Hunter Wessells, MD, FACS Professor and Nelson Chair, Department of Urology Adjunct Professor, Department of Surgery Affiliate Member, Harborview Injury Prevention and Research Center University of Washington Seattle, Washington UROLOGIC COMPLICATIONS OF PELVIC FRACTURE Steven D. Wexner, MD, PhD(Hon), FACS, FRCS(Eng), FRCS(Ed), FRCSI(Hon), Hon FRCS(Glasg) (Hon) Director, Digestive Diseases Center Chair, Department of Colorectal Surgery Cleveland Clinic Florida; Affiliate Professor Florida Atlantic University College of Medicine; Clinical Professor Florida International University College of Medicine Weston, Florida MANAGEMENT OF SOLITARY RECTAL ULCER SYNDROME Richard L. Whelan, MD Professor Department of Surgery Mount Sinai West Hospital New York, New York MANAGEMENT OF COLON POLYPS Elizabeth Wick, MD Associate Professor Division of General Surgery University of California, San Francisco San Francisco, California MANAGEMENT OF CROHN’S COLITIS

C O N T R I B U TO R S

Timothy K. Williams, MD Associate Professor Vascular and Endovascular Surgery Wake Forest Baptist Health Winston-­Salem, North Carolina

Cameron D. Wright, MD Professor of Surgery Massachusetts General Hospital Harvard Medical School Boston, Massachusetts

THORACIC OUTLET SYNDROME

PRIMARY TUMORS OF THE THYMUS

Douglas Wiseman, MD Endocrine Surgery Fellow Department of Surgical Oncology National Cancer Institute Bethesda, Maryland

Stephen C.Yang, MD Arthur B, and Patricia B. Modell Endowed Chair in Thoracic Surgery Professor of Surgery and Oncology Department of Surgery Division of Thoracic Surgery Johns Hopkins Medical Institutions Baltimore, Maryland

MANAGEMENT OF THYROIDITIS Meredith O. Witten, MD Breast Surgical Oncology Fellow Department of Surgery Division of Surgical Oncology University of Alabama in Birmingham Birmingham, Alabama MOLECULAR TARGETS IN BREAST CANCER Christopher L. Wolfgang, MD, PhD Chief of Surgical Oncology Professor of Surgery, Pathology, and Oncology Department of Surgery Johns Hopkins Hospital Baltimore, Maryland

ACHALASIA OF THE ESOPHAGUS Raymond Yap, MBBS, BMedSci, MSurgEd, FRACS Colorectal Surgeon Cabrini Department of Surgery Monash University Melbourne, Australia CONDYLOMA ACUMINATA Shlomo Yellinek, MD, FACS, FRCS, FRCS(Ed) Shaare Zedek Medical Center Jerusalem, Israel

VASCULAR RECONSTRUCTION DURING THE WHIPPLE OPERATION

MANAGEMENT OF SOLITARY RECTAL ULCER SYNDROME

Daniel J. Wong, MD, MHS Resident Department of Surgery Beth Israel Deaconess Medical Center Boston, Massachusetts

Heather L.Yeo, MD, MHS, MBA, MS Assistant Professor Department of Surgery Assistant Professor of Healthcare Policy and Research New York-­Presbyterian Weill Cornell Medical Center New York, New York

POPLITEAL AND FEMORAL ARTERY ANEURYSM Douglas E. Wood, MD, FRCSEd Henry N. Harkins Professor and Chair Department of Surgery University of Washington Seattle, Washington MANAGEMENT OF TRACHEAL STENOSIS Tinsay Woreta, MD, MPH Assistant Professor of Medicine Division of Gastroenterology and Hepatology Johns Hopkins University School of Medicine Baltimore, Maryland MANAGEMENT OF HEPATIC ENCEPHALOPATHY

MANAGEMENT OF SMALL BOWEL OBSTRUCTION MANAGEMENT OF CROHN’S DISEASE OF THE SMALL BOWEL Ben L. Zarzaur, MD, MPH Associate Professor Department of Surgery Indiana University School of Medicine Indianapolis, Indiana ABDOMINAL COMPARTMENT SYNDROME AND MANAGEMENT OF THE OPEN ABDOMEN

xxxi

Victor M. Zaydfudim, MD, MPH Associate Professor of Surgery Section of Hepatobiliary and Pancreatic Surgery Division of Surgical Oncology University of Virginia Charlottesville, Virginia CYSTIC DISEASE OF THE LIVER Sarah Zeb, MD Breast Imager Diagnostic Radiology Johns Hopkins Radiology Baltimore, Maryland SCREENING FOR BREAST CANCER Martha A. Zeiger, MD, FACS Director, Surgical Oncology Program National Cancer Institute National Institutes of Health Bethesda, Maryland SURGICAL MANAGEMENT OF SECONDARY AND TERTIARY HYPERPARATHYROIDISM Michael Zenilman, MD, FACS Professor, Department of Surgery Weill Cornell Medical Medicine; Chair, Department of Surgery New York Methodist Hospital New York, New York OPTIMIZING PERIOPERATIVE CARE OF THE OLDER ADULT Luke X. Zhan, MD, PhD Vascular and Endovascular Surgeon Providence Vascular Institute Providence Medical Group Sacred Heart Medical Center Spokane, Washington PROFUNDA FEMORIS RECONSTRUCTION Ruoyan Zhu, BS Research Associate Surgery Group of Los Angeles Los Angeles, California ANORECTAL STRICTURE Nicholas J. Zyromski, MD Professor Department of Surgery Indiana University School of Medicine Indianapolis, Indiana MANAGEMENT OF GALLSTONE ILEUS

Preface

The first edition of Current Surgical Therapy was published in 1984. The textbook has thus been in existence for more than 35 years, and this is the thirteenth edition. In each edition, we have updated the material to reflect the continuing evolution of the field of general surgery. The textbook continues to be perhaps the most popular surgical book in the United States, and as long as it fulfills a need we plan to continue the publication every 3 years. It has been a special privilege and honor for the two editors to be able to review contributions from surgeons around the country and, indeed, from around the world, on what they believe is the current surgical therapy for virtually all general surgical topics. It is an enjoyable task and keeps two surgeons who care for surgical patients current on all general surgical topics. The thirteenth edition contains 263 chapters. This represents a decrease of about 20 chapters from the twelfth edition. There have been comments in recent years that the text was becoming too big to easily manage. We have responded by eliminating the section on minimally invasive surgery; instead, authors have incorporated this approach in each chapter, where appropriate. In addition, some chapters have been eliminated and new ones added. As with prior editions, nearly every chapter has been written by a new author. All authors have contributed their specific and personal thoughts on the current surgical therapy of the disease about which they are experts. Therefore, to obtain a broad view of the topic, the reader may want to review the contributions of the other experts in the last two or three editions of Current Surgical Therapy. As with the past editions, disease presentation, pathophysiology, and diagnosis are discussed only briefly, with the emphasis on current surgical therapy. When an operative procedure is discussed, an effort has been made to include brief and concise descriptions with

figures and diagrams when possible. Current Surgical Therapy is written for surgical residents, fellows, and fully trained surgeons in private practice or in an academic setting. Many have told us that it is an excellent textbook to review before taking the general surgical boards or recertifying. In addition, medical students have given us feedback that they believe the text is of value to them. However, Current Surgical Therapy is not written principally for medical students. We believe a more classic surgical textbook with substantial sections on disease presentation, diagnosis, and pathophysiology is more appropriate for medical students. We remain grateful to the many surgeons throughout the country, as well as to the international surgeons, who participated in creating this textbook. Most of the potential authors whom we solicit respond enthusiastically to the opportunity to present their expert views. Their efforts obviously are what make this textbook a success. In addition, we could not have compiled this textbook without the herculean efforts of Ms. Irma Silkworth, who has been involved with virtually all of these editions. Ms. Katie DeFrancesco at Elsevier has also been a terrific help and stands out in the publishing industry. Both editors continue to enjoy and thrive in our chosen profession of general surgery. In recruiting medical students into our specialty over the last 40 years, I have used the statement, “If you pick a profession you love, you never have to work the rest of your life.” In our view, that profession is surgery. Finally, we would like to dedicate this edition, as with the others, to the surgical house staff and fellows at the Johns Hopkins Hospital, who are “the best of the best.” Andrew M. Cameron, MD John L. Cameron, MD

xxxiii

Esophagus

Esophageal Function Tests

Consensus guidelines have provided direction in using esophageal function test findings toward defining the mechanisms of esophageal symptoms. The Chicago Classification describes specific criteria in diagnosing esophageal motility disorders. The Lyon Consensus characterizes the results from esophageal testing for the diagnosis of gastroesophageal reflux disease, and further establishes a motor classification of GERD. The esophageal function tests utilized in the diagnosis of esophageal disorders are flexible fiberoptic endoscopy, high resolution esophageal manometry (Box 1), ambulatory reflux monitoring, contrast radiography of the esophagus, and occasionally, endoluminal functional lumen imaging probe (Endo-­FLIP). The specific contribution that each of these tests plays in the diagnosis and management of esophageal diseases will be discussed below.

manometry, a plot of esophageal peristalsis is generated in response to a 5-­mL water swallow and is bounded superiorly by the upper esophageal sphincter and inferiorly by the lower esophageal sphincter within the esophagogastric junction (EGJ). Esophageal motor function is then evaluated using three different metrics. The integrated relaxation pressure (IRP) is the nadir pressure over 4 seconds when EGJ relaxation is expected within a 10-­second window after upper esophageal sphincter relaxation. The distal contractile integral (DCI) is an assessment of the strength of esophageal smooth muscle contraction, incorporating length, amplitude, and duration of the contracting segments of the esophagus. The distal latency is the measure of the timing of esophageal peristalsis and is measured from upper esophageal sphincter relaxation to the contractile deceleration point, where fast esophageal body peristaltic progression transitions to slower EGJ emptying function. Within the Chicago Classification, achalasia is characterized by IRP values above the upper limit of normal (generally 15 mm Hg in many systems). However, in the setting of absent peristalsis, an IRP cutoff of 10 mm Hg may indicate type I achalasia (Figs. 2 and 3). In addition, absent peristalsis with at least 20% of swallows with panesophageal pressurization should raise suspicion for type II achalasia regardless of IRP. Type III achalasia (spastic) is associated with premature or spastic distal esophageal contractions in at least 20% of swallows (Fig. 4). There is a subset of patients with elevated IRP with preserved peristalsis that are best described as having EGJ outflow obstruction (Table 1). The standard treatment for achalasia has been a distal esophageal myotomy combined with a partial fundoplication (either Dor or Toupet), with success rates (improvement in dysphagia and the ability to regain weight) of 80% for type I achalasia and 95% for type II achalasia. Lower success rates (60%) have been reported for distal esophageal myotomy in patients with type III achalasia, most likely because the length of the myotomy is inadequate. Peroral endoscopic myotomy (POEM) increasingly has been used in the management of these patients. A submucosal tunnel is created in the esophagus approximately 10 cm proximal to the gastroesophageal junction, and a myotomy of circular muscle layers is distally extended to 2 cm into the cardia. The short-­term follow-­up results with POEM in most series have shown results that are equivalent in symptom relief to a distal esophageal myotomy. However, there is a much higher incidence of gastroesophageal reflux disease (as high as 40%) because patients do not have an associated antireflux procedure at the time of the myotomy. 

Achalasia

nn ESOPHAGOGASTRIC

Achalasia is the best-­defined primary motor disorder of the esophagus. Patients present with dysphagia to solids and liquids and associated weight loss. Although many patients may have had upper endoscopy or contrast esophagography (Fig. 1) to rule out mechanical reasons for their dysphagia, high-­resolution esophageal manometry is the best test for the diagnosis of achalasia. With high-­resolution

EGJ outflow obstruction was formerly called hypertensive lower esophageal sphincter. Using a definition of EGJ outflow obstruction based solely on the IRP with the exclusion of achalasia allows this diagnosis to be combined with another diagnosis dependent on the esophageal body motility. EGJ outflow obstruction may be

Richard J. Battafarano, MD, PhD

T

he swallowing mechanism is a dynamic process between the muscles of the esophageal wall and its neural innervation and involves the coordinated interplay between the upper esophageal sphincter and lower esophageal sphincter (LES) that allows the propagation of a bolus of food from the oropharynx to the stomach. Esophageal dysfunction can lead to a number of symptoms, including dysphagia, odynophagia, regurgitation of food or liquid, and gastroesophageal reflux. One cannot underestimate the importance of taking a detailed history directly from a patient with esophageal disease because it is often a constellation of symptoms that suggest the true source of the patient’s problem. Gastroesophageal reflux disease (GERD) is so prevalent that many patients with any esophageal symptoms are simply treated with antacid therapy for long periods of time before other sources for their esophageal symptoms are sought. This chapter provides a brief overview of the most commonly encountered esophageal function disorders and a review of the indication and interpretation of the esophageal function tests used in their diagnosis. nn DISORDERS

FUNCTION

OF ESOPHAGEAL

JUNCTION OUTFLOW OBSTRUCTION

1

2

Esophageal Function Tests

BOX 1  Indications for Esophageal Manometry Definite • Evaluation of unexplained esophageal symptoms • Accurate placement of pH, pH impedance probes • Evaluation of peristaltic function before either antireflux surgery or esophageal myotomy • Evaluation of postoperative dysphagia • Evaluation of rumination (with impedance) • Evaluation of scleroderma  Emerging • Evaluation of peristaltic function before bariatric surgery and lung transplantation • Follow-­up of achalasia after therapy From Vaezi MF, Pandolfino JE, Vela MF, et al. White Paper AGA: optimal strategies to define and diagnose gastroesophageal reflux disease. Clin Gastroenterol Hepatol. 2017;15:1162-1172.

nn DIFFUSE

ESOPHAGEAL SPASM

Diffuse esophageal spasm (DES) is an uncommon condition that accounts for less than 10% of esophageal motility abnormalities. DES is characterized by uncoordinated contractions of the esophagus that typically result in symptoms of chest pain, dysphagia, or both. The esophagogram may be abnormal, but manometry is usually necessary for the diagnosis. As with achalasia, the introduction of high-­resolution manometry changed the diagnostic criteria for DES. Initially, the Chicago Classification based both high-­resolution manometry (HRM) and conventional manometry on the same criteria of rapid or simultaneous contractions but modified this to a parameter known as distal latency, which is only apparent with HRM and is a more reliable indicator of DES. Distal latency is more likely associated with the onset of inhibitory myenteric neuron activity after contractions and seems shorter in patients with DES. This results in an increased state of contractions in the distal esophagus. Current guidelines define DES by HRM as patients who have a normal integrated relaxation pressure at the LES but who have a distal latency less than 4.5 seconds in 20% of wet swallows. Medical treatment options include nitrates, sildenafil, and tricyclic antidepressants, which often help to alleviate the noncardiac chest pain. Diazepam and lorazepam have been used if nitrates, sildenafil, or tricyclic antidepressants have failed to control symptoms. Use of proton pump inhibitors (PPIs) for treatment of concomitant GERD also may be helpful. Surgical interventions such as laparoscopic or thoracoscopic extended myotomy can be effective in well-­selected patients with refractory DES.  nn HYPERCONTRACTILE “NUTCRACKER”

ESOPHAGUS

The definition of hypercontractile, or “nutcracker” esophagus was updated in the latest version of the Chicago Classification to include only patients with a DCI greater than 8000 mm Hg • s • cm in greater than 20% of swallows, excluding patients with a single altered swallow (Figs. 5 and 6). In comparison to diffuse esophageal spasm, the peristaltic contractions propagate normally, and the LES relaxes appropriately. Diltiazem has been shown to lower distal peristaltic pressures and may reduce chest pain; however, these results are not reliably reproducible. As in DES, nitrates, sildenafil, and tricyclic antidepressants may be useful in the treatment of the noncardiac chest pain.  nn INEFFECTIVE

FIG. 1  Classic bird’s beak appearance of achalasia on a contrast esophagogram.

caused by an anatomic abnormality at the cardia (hiatal hernia, diseases of the esophageal wall, etc.) or be idiopathic with normal anatomy. In the setting of an anatomic abnormality, such as a paraesophageal hernia, surgical correction is associated with excellent durable results. In patients without an anatomic abnormality, the clinical significance of this diagnosis is still uncertain because some investigators believe that EGJ outflow obstruction may be an early or incomplete expression of a variant of achalasia. Because a significant number of patients with EGJ outflow obstruction report a spontaneous resolution of their symptoms over time without intervention, the recommendation for invasive interventions such as POEM or laparoscopic distal esophageal myotomy should be made only in those patients with persistent symptoms and associated weight loss. 

ESOPHAGEAL MOTILITY

Ineffective esophageal motility is defined in the Chicago Classification system as a DCI of less than 450 mm Hg • s • cm in 50% of the swallows. A variant of ineffective esophageal motility called fragmented peristalsis occurs when the patient’s high-­resolution manometry demonstrates normal DCI but greater than 5 cm breaks in more than 50% of the swallows. Finally, absent contractility is defined as a DCI less than 100 mm Hg • s • cm in 100% of the swallows and is primarily observed in patients with connective tissue disorders such as scleroderma. Because there are no pharmacologic agents that restore or improve peristalsis in these patients, the patients’ associated GERD is usually the focus of therapy.  nn GASTROESOPHAGEAL

REFLUX DISEASE

GERD is a common disorder that often brings patients to medical attention for acid-­suppression therapy. It is reported to affect approximately 20% of the US population and is associated with a significant cost to the US healthcare system. There is a degree of physiologic reflux that is considered normal and usually is limited and not nocturnal. The association of patient-­reported symptoms and pathologic confirmed GERD is approximately 70% in many studies. Before recommending esophageal testing for diagnosis, most patients are empirically treated with a trial of PPI therapy. For those patients who

3

E S O P H AG U S WS(1) %

WS(2)

WS(3)

WS(4)

WS(5)

WS(6)

WS(7)

Swallow

50

Achalasia: simultaneous mirror image swallow responses 0 mm Hg

Proximal 25.0 24.0

(no peristalsis)

Wave progression graph-wet swallow sequence analysis-eso body UES I II III IV V

25 30 35 40 45 0

50 0

20

00:01

00:02

5

10

15

20

Time (s)-placement relative to UES 00:03

FIG. 2  Manometry tracing of achalasia. The swallow study shows mirror-­image swallow responses in the esophageal body. UES, Upper esophageal sphincter.

develop recurrent symptoms after a trial of PPI therapy, upper endoscopy with esophageal biopsy is the next step in the diagnostic algorithm. The identification of high-­grade esophagitis (grades C or D), Barrett’s metaplasia (Fig. 7), or a peptic stricture are considered confirmatory evidence for GERD. Grade B esophagitis provides adequate evidence for initiation of medical management of GERD. However, additional testing is warranted to confirm the diagnosis before committing a patient to long-­term PPI therapy or antireflux surgery.  nn AMBULATORY ACID

MONITORING

REFLUX

Acid exposure in the esophagus is measured with an intraluminal pH probe with one of two methods: either an intraluminal tube with a nasopharyngeal catheter or a wireless Bravo pH probe (Medtronic). Both methods provide the physician with similar data, particularly relating to the amount of time the esophagus is exposed to acid reflux. When the information is correlated with a symptom log, determination is possible of whether the patient’s symptoms are related to acid exposure within the esophagus. This information is commonly expressed using six standard parameters (Box 2) to calculate a DeMeester score or a composite pH score. A score of less than 14.72 (95th percentile of normal) is considered physiologic reflux, whereas a score greater than 14.72 is considered abnormal. Acid exposure in the esophagus is recorded according to the position of the patient (supine or upright) and the relation of acid exposure with meals. Accurate placement of either the indwelling esophageal catheter or the wireless

probe is facilitated by placement during high-­resolution esophageal manometry, and so these tests are often scheduled together.  nn ESOPHAGEAL

IMPEDANCE

Impedance monitoring measures bolus transport by measuring the resistance to electrical conductivity of the esophagus and its contents. Impedance testing is an important adjunct to traditional pH testing because it can be useful for the evaluation of acid and bile (nonacid) reflux. Impedance measurement works by using low AC voltage to apply an electrical potential between two electrodes on a catheter separated by an isolator. Because air, liquid, and esophageal mucosa have unique impedance characteristics, identification of the material that is bridging the electrodes can be accurately assessed. Air is resistant to current flow and has a high impedance; liquid has a low impedance value. Esophageal tissue has an indeterminate range and is used as a baseline during monitoring. With multiple electrodes along a catheter system, identification of changes in impedance makes possible determination of the direction of bolus transport within the esophagus and identification of reflux of a bolus that has cleared the esophagus but comes back up from the stomach (Fig. 8).

GERD Classification The recently introduced classification of motor disorders in GERD characterizes morphologic and motor abnormalities in GERD, and this GERD classification is intended to be used in conjunction with the

4

Esophageal Function Tests Swallow 1 15.0

Pharynx UES 20.0

20.0

25.0

30.0

Esophagus

35.0

40.0

FIG. 3  High-­resolution manometry in a patient with achalasia. Note the simultaneous contractions and lack of swallow propagation. The lower esophageal sphincter is not hypertensive, but complete relaxation is not seen. LES, Lower esophageal sphincter; PIP, pressure inversion point; UES, upper esophageal sphincter.

A

Diaphragm 45.0

44.0 PIP 45.5

LES 45.0

48.0 Gastric 49.0 50.0 10 sec

B

0.0 Stomach 4.10

150.0 0.10

C

FIG. 4  High-­resolution manometry demonstrating achalasia subtypes. (A) Type I is characterized by minimal esophageal pressurizations. (B) Type II, with panesophageal pressurization wave (black line demarcates 30 mm Hg contour). (C) Type III, with premature spastic contractions.

E S O P H AG U S

5

TABLE 1  Interpretation of High-­Resolution Esophageal Manometry; Chicago Classification, V3.0 Step

Metrics

EVALUATION OF ESOPHAGEAL FUNCTION Evaluation of EGJ function Integrated relaxation pressure (IRP), abnormal if median value greater than upper limit of normal

Category

Characteristics

EGJ outflow obstruction

Type I: absent peristalsis or pressurization Type II: absent peristalsis, panesophageal pressurization in ≥20% Type III: premature peristalsis in ≥20% EJGOO: intact peristalsis

Evaluation for major motor disorders

Distal contractile integral (DCI): Major motor disorder hypercontractile if >8000 mm Hg/ cm/s; absent if 8000 mm Hg/cm/s in ≥20% Distal esophageal spasm: DL 10 mm

Histology

Tubular

Tubulovillous

Villous

Dysplasia

Mild

Moderate

Severe

aSpigelman

stage I, score 1–4; stage II, score 5–6; stage III, score 7–8; stage IV, score 9–12.

Other extraintestinal manifestations, although rare, include supernumerary teeth, cerebellar medulloblastoma, and cancers of the liver, biliary tree, adrenal glands, and thyroid. The risk for thyroid cancer for patients with FAP is approximately 2%, which is double that of the general population. Screening thyroid ultrasound is recommended annually, also starting at 20 years of age. Thyroid nodules larger than 1 cm warrant fine-­needle aspiration. 

Desmoid Tumors Desmoids are rare in the general population but develop in up to 30% of FAP patients. These locally invasive abdominal wall and intraabdominal/retroperitoneal myofibroblastic tumors typically develop 2

238

Surgical Management of the Polyposis Syndromes

TABLE 3  Derivation of Spigelman Stage From Scores Total Points

Spigelman Stage

Suggested Interval to Next Duodenoscopy

0

0

5y

1–4

I

3–5 y

5–6

II

3y

7–8

III

1y

9–12

IV

Consider duodenectomy vs ­endoscopy in 6 mo

to 3 years after surgery and occur around 30 years of age. Etiology and pathogenesis are not well understood. Desmoids can develop spontaneously, are the third most common cause of death associated with FAP and have been noted to be associated with trauma. Risk factors associated with the development of desmoids are mutations in the 3′ end of the APC gene, female gender, extraintestinal manifestations, and a family history of desmoid disease. Of desmoids, 10% grow rapidly, 10% resolve spontaneously, 30% vacillate between cycles of growth and regression, and 50% remain stable or grow very slowly. Extraabdominal desmoids are best treated with surgical extirpation with a 1-­cm margin, although recurrence is high with documented rates of 20% to 50%. Early excision is recommended to decrease the size of the resultant abdominal wall defect. Intraabdominal/retroperitoneal desmoids can invade the mesentery and surrounding structures resulting in obstruction, hemorrhage, ischemia, and perforation. The primary treatment is medical and includes nonsteroidal antiinflammatory agents (sulindac, celecoxib), estrogen antagonists (tamoxifen, toremifene, raloxifene), and chemotherapy (vinblastine, methotrexate, doxorubicin, dacarbazine). Surgical removal is difficult and often impossible if the root of the mesentery is involved. Resection with completely uninvolved margins (R0) will result in recurrence 50% of the time. If possible, nonresective procedures such as diversion and bypass are preferable for palliation. Ureteral obstruction is best treated with stenting. There is not a defined screening regimen for desmoid tumors, although computed tomography and magnetic resonance imaging can be used, especially for patients with an increased risk of developing desmoids. 

Attenuated FAP In contrast to classic FAP, attenuated FAP (aFAP) occurs at a later age (30s and 40s), with fewer than 100 polyps, found predominantly in the right colon. If untreated, the risk of colorectal malignancy is 70% and cancer develops later in life (58 years of age on average). Fundic gland polyps and duodenal adenomas occur frequently, but other extracolonic and intestinal manifestations, including gastric adenomas, desmoids, and CHRPE, typically are not seen in aFAP. For those with a family history or identified APC mutation suggestive of aFAP, screening colonoscopy should begin at 20 years of age and be repeated every 1 to 2 years. With the predilection for polyp development in the right colon, a formal colonoscopy is recommended. 

Mutation Y-­Homolog–Associated Polyposis Mutation Y-­homolog (MYH)-­associated polyposis (MAP) is the only polyposis syndrome with an autosomal recessive inheritance pattern. It results from a biallelic mutation in the MYH gene located on chromosome 1p34. The number of polyps associated with MAP is variable (tens to hundreds) with a median around 50. Polyps are found most

commonly in the left colon and occur at a median age of 48 years. Polyposis is not required for the diagnosis of MAP. Up to 20% of such patients present with colorectal cancer without history of synchronous polyps. If untreated, the risk of colorectal malignancy is around 75% for males and 72% for females by 70 years of age. Extraintestinal manifestations can be associated with MAP, although these are exceedingly rare. Because of the phenotypic overlap with FAP, genetic testing for the MYH mutation is typically performed when no APC mutation is detected, there are fewer than 100 adenomatous polyps, and the family history is irrelevant or does not reveal a dominant mode of inheritance. Although there are no defined endoscopic screening criteria for MAP, an initial endoscopic (colonoscopy and EGD) evaluation should be performed starting around 25 to 30 years of age. If no polyps are appreciated, endoscopy should be repeated every 3 to 5 years and more frequently if present. As in the case of FAP, patients have an increased risk for developing thyroid malignancy, and annual screening ultrasound of the thyroid is recommended.  nn CHEMOPREVENTION Although clinic trials have shown that nonsteroidal antiinflammatory drugs (sulindac, exisulind, celecoxib) and aspirin can reduce the size and number of adenomas in the colon and rectum, there was not an appreciable reduction in cancer. Chemoprevention is not recommended as a primary therapy for polyposis syndromes and is not an appropriate alternative to prophylactic surgery. Situations in which chemoprevention can be entertained include treating ileal pouch anal anastomosis polyps, a high family risk of desmoid tumors, delayed surgery, and unwillingness or inability to tolerate polypectomy or completion proctectomy.  nn SURGERY The goal of prophylactic surgery for polyposis syndromes is to prevent colorectal cancer. The timing and type of surgery offered depends on a multitude of factors, including clinical presentation, family history, and, if known, the site of the chromosomal mutation. Severe polyposis (more than 1000 colonic or 20 rectal polyps) and APC mutations between codons 1250 and 1464 carry a higher risk of cancer, and surgery should be offered as early as possible. Surgery also should be pursued early for symptomatic disease. For those with a high risk of desmoid disease (family history, mutation in the 3′ end of the APC gene, female gender, extracolonic manifestations), surgery should be delayed as long as possible to decrease the chance of desmoid tumors developing. Young patients should have surgery delayed, if possible, to allow for adequate physical, social, and intellectual maturity. For patients with classic FAP, surgery typically occurs around 16 to 20 years of age. Surgical options include open or minimally invasive total proctocolectomy with creation of an end or continent ileostomy, total abdominal colectomy with creation of an ileorectal anastomosis (IRA), and a total proctocolectomy with creation of an ileal pouch anal anastomosis (IPAA). A thorough preoperative discussion of the advantages and drawbacks to each approach is necessary to achieve the best patient outcome.  nn PROCTOCOLECTOMY WITH

END-­ILEOSTOMY

A proctocolectomy with end Brooke ileostomy has a low rate of complications. Most patients do not elect for creation of an incontinent stoma if they are candidates for other continence-­restoring approaches. Indications for this approach are patient preference, low rectal cancer requiring an abdominoperineal resection, rectal cancer requiring postoperative pelvic radiation, inability to create an IPAA (inadequate mesenteric length), and poor sphincter function. In

L A R G E B OW E L

some cases, patients with chronic pouch dysfunction may elect for conversion to an end-­ileostomy. The procedure is carried out in an oncologic approach secondary to the risk of a preoperatively unrecognized cancer. A perineal intersphincteric dissection is carried out preserving the external sphincter and levator ani muscles. The perineum is closed in layers and the greater omentum, if present, is mobilized and placed in the pelvis to prevent future bowel obstructions. After closure of the abdomen the ileostomy is matured in a standard evaginated Brooke fashion.

Proctocolectomy With Continent Ileostomy The continent ileostomy was initially described by Nils Kock (Kock’s pouch) in 1969; it allows patients an option for an ileostomy without need for a cumbersome bag. Modifications and revisions to the original Kock’s continent ileostomy have been described (Barnett continent intestinal reservoir and T-pouch). Each features an ileal pouch reservoir and valve mechanism to prevent leakage of accumulating stool. Although the technique lost favor when IPAA was established as the first-­choice technique, continent ileostomy remains an option for select patients undergoing proctocolectomy for FAP, including those with inadequate sphincter function, insufficient small bowel mesentery, a history of low rectal or anal cancer, failed existing IPAA, and/or in whom traditional Brooke ileostomy is undesirable. Contraindications to construction of a continent ileostomy include Crohn’s disease, morbid obesity, marginal small bowel length, and a psychologic or physical disability that would preclude understanding or being able to perform daily stoma intubation. 

Total Abdominal Colectomy With IRA Colectomy with IRA should be considered only in cases of attenuated or mild polyposis (fewer than 20 rectal or 1000 colonic adenomas), rectal polyps smaller than 3 cm, no colorectal dysplasia or cancer, a distensible and compliant rectum, and in patients with an intact sphincter mechanism that are willing to adhere to strict follow up. This rectal-­sparing approach is an appealing alternative in younger patients of reproductive age to decrease the risk of impotence and reduced fecundity. Strict rectal surveillance (every 6–12 months) must be adhered to because of the increased risk of future neoplastic changes. The risk of rectal carcinoma can reach up to 40% by 30 years, although this is based on literature from the pre-­IPAA era. In patients who require a completion proctectomy, an end-­ileostomy, restorative IPAA, or continent ileostomy remain viable options. 

239

artery cephalad to the head of the pancreas, proximal division of the ileocolic artery, and relaxing incisions of the mesentery over tension points along the superior mesenteric artery. Rectal dissection is completed in the total mesorectal excision plane, and transection of the rectum with a 30-­mm transverse stapler should occur 2 to 3 cm above the dentate line in the anal transition zone. After reach has been verified, a J configuration is fashioned with each limb measuring between 12 and 15 cm in length. The limbs are paired in an antimesenteric fashion and are held in orientation with interrupted stay sutures. For those without evidence of adenomas in the anal transition zone, or dysplasia in the lower rectum, a double-­stapled IPAA can be fashioned; otherwise, an anal mucosectomy and handsewn IPAA is recommended. After creation of the IPAA, an air insufflation leak test is performed and, if necessary, a protective loop ileostomy fashioned, which should be created as close to the pouch as possible to decrease issues with high output. In selected patients, the operation can be completed with good results without the creation of a diverting loop ileostomy.

Double-­Stapled Technique An enterotomy is made in the antimesenteric apex of the pouch, and a linear cutting stapler is used to divide the walls of the two limbs, creating a common channel (Fig. 1). A pursestring suture then is fashioned around the enterotomy, and the anvil from a circular stapler is placed inside the pouch, where it is held in place by tightening the purse string (Fig. 2). The circular stapler then is placed transanally. After appropriate orientation, the circular stapler cartridge spike is advanced either above or below the transverse rectal staple line and attached to the anvil. The stapler then is closed, approximating the pouch and anus (Fig. 3). 

Handsewn Technique An anal canal mucosectomy is performed starting at the dentate line. Raising the mucosa with a submucosal injection (Fig. 4) of dilute saline and epinephrine (1:200,000) facilitates dissection of the mucosa away from the internal sphincter muscle (Fig. 5), which can be completed sharply or with electrocautery. After the mucosa and

Restorative Proctocolectomy With IPAA Initially described in 1978 by Parks and Nicholls, the restorative proctocolectomy has become the most common continence-­preserving procedure performed in patients who are appropriate candidates. Indications include severe polyposis (>20 rectal or >1000 colon adenomas), rectal polyps larger than 3 cm in size, colonic dysplasia or cancer, dysplastic rectal polyps, and in patients with an intact sphincter mechanism willing to adhere to strict follow up. The restorative pouch can be fashioned in two limbs (J), three limbs, four limbs, or isoperistaltic configurations. The J pouch, because of its ease of construction and excellent functional outcomes, has become the most common choice for surgeons. A total colectomy is performed in an oncologic fashion, and the ileum is transected flush with the cecum. To provide adequate perfusion to the pouch, it is imperative to preserve the ileal branches of the ileocolic and distal mesenteric arteries. Evaluation for adequacy of reach of the small bowel to the deep pelvis should be undertaken before creation of the pouch. If the proposed apex of the pouch anal anastomosis can be advanced 3 to 4 cm below the inferior edge of the pubis, one can feel confident of successful reach for anastomosis. Strategies to decrease tension at the anastomosis include complete mobilization of the small bowel mesentery to the root of the superior mesenteric

FIG. 1  J pouch creation. (Courtesy Mayo Foundation for Medical Education and Research.)

240

Surgical Management of the Polyposis Syndromes

FIG. 4  Submucosal injection. (Courtesy Mayo Foundation for Medical Education and Research.)

FIG. 2 Anvil in J pouch. (Courtesy Mayo Foundation for Medical Education and Research.)

Levator ani FIG. 5 Anal mucosectomy. (From Kelley SR, Dozois EJ. Ulcerative colitis. In: Clark S, ed. Colorectal Surgery: A Companion to Specialist Surgical Practice, 5th ed. Edinburgh: Elsevier; 2014:134.)

Dentate line

FIG. 3  Stapled ileal pouch anal anastomosis. (Courtesy Mayo Foundation for Medical Education and Research.)

proximal rectum have been removed circumferentially, the pouch is brought down gently to the level of the dentate line. An enterotomy is made in the apex of the pouch, if not already created, and anchored in position by placing a suture in each of the four quadrants, incorporating a full-­thickness bite of the pouch, internal sphincter muscle, and mucosa. Sutures are placed between the anchoring stitches to complete the anastomosis (Fig. 6). 

Postoperative Surveillance After creation of an end-­ileostomy, IPAA, or IRA, annual surveillance with flexible endoscopy facilitates early detection and removal of adenomas, dysplasia, and carcinomas. The patient must understand the need for a lifelong surveillance regimen. Histologic evaluation of random biopsies and polyps should be performed to exclude dysplasia and cancer, particularly in any areas of chronic scarring. More frequent surveillance is performed for increased numbers or size of polyps. Severe dysplasia and villous adenomas more than 1 cm in size should prompt a completion proctectomy in those with an IRA.  nn HAMARTOMATOUS

SYNDROMES

POLYPOSIS

Peutz-­Jeghers Syndrome Peutz-­Jeghers syndrome is an autosomal dominant inherited disease resulting most commonly from a mutation in the LKB1 (STK11) tumor suppressor gene located on chromosome 19p13. Approximately 30% to 40% will occur de novo. Hamartomatous polyps are found throughout the gastrointestinal tract, although most

L A R G E B OW E L

FIG. 6  Hand-­sewn ileal pouch anal anastomosis. (Courtesy Mayo Foundation for Medical Education and Research.)

commonly in the small intestine. Polyps vary in size and tend to become pedunculated as they grow larger. Extraintestinal manifestations are common, with the hallmark phenotypic feature in adolescence being mucocutaneous hyperpigmentation that can affect the perioral and buccal region, eyes, nostrils, perianal region, fingers and toes, and hands and feet. Hyperpigmentation dissipates as a patient ages. Hamartomatous polyps and mucocutaneous pigmentation confirms a diagnosis of Peutz-­Jeghers syndrome. Although the overall number of polyps tend to be small (52%, accuracy 100%) and SUVmax (decreases >42%, accuracy 94%). Jansen and colleagues also confirmed these findings, showing a significant reduction in SUVmax was detectable after the first week of therapy. A meta-­analysis performed by Maffione and colleagues looked at 10 studies, with a total of 203 patients, to evaluate the value of FDG-­ PET/CT to detect early response of patients with LARC receiving nCRT. FDG-­PET/CT was found to have a sensitivity and specificity of 79% and 78%, respectively. Patients who achieve a pCR after nCRT have a better disease-­ free survival (DFS) and OS compared with partial responders and nonresponders. As such, a watch-­and-­wait approach can be adopted, with robust surveillance and early detection of regrowths allowing for a high rate of successful salvage surgery. FDG-­PET/CT has been suggested to form part of the surveillance algorithm. Perez and colleagues prospectively monitored 99 patients who underwent FDG-­ PET/CT at baseline and after 6 weeks and 12 weeks after completion of neoadjuvant treatment before clinical assessment. Sixteen patients (16%) had a pCR and were managed without surgery. FDG-­PET/CT was used during a strict follow-­up program; however, the authors found that at 6 weeks the scan was able to detect less than 50% of true complete responders. The use of FDG-­PET/CT in the management of LARC continues to evolve; however, as of yet there is no definitive consensus. It has the potential to detect early responders, predict survival, and aid in the determination of pCR to guide nonoperative intervention.  nn DETECTION

OF RECURRENCE

Local and systemic recurrence after CRC surgery occurs in up to 30% of patients in the first 2 years. Early detection allows for higher resectability and better survival, with 5-­year survival rates of 30% to 40% in selected patients with single organ metastatic disease. The most common sites for recurrence are liver locally—especially for rectal cancer—and lung. Postoperative surveillance protocols based on the site and stage of the original cancer include clinic visits with history and physical

examination, CEA levels, endoscopy (sigmoidoscopy or colonoscopy), and CT scanning for 5 years after surgical resection. Intensive follow-­up to ensure the detection of recurrences does not always diminish mortality rates; thus the optimal monitoring strategy remains in doubt. Several studies have shown that FDG-­PET is sensitive and specific in detecting recurrence in CRC patients, thus affecting management. This remains the most common use of FDG-­PET to date, particularly in patients who have had potentially curative resection of liver or lung metastases. However, the clinical value and overall efficacy of FDG-­PET/CT in surveillance are not yet established. A recent open-­ label multicenter trial conducted by Sobhani and colleagues enrolled patients in remission of CRC (stage II perforated, stage III, or stage IV) after curative surgery. One hundred twenty patients were in the intervention arm of 6-­monthly FDG-­PET/CT, for 3 years, whereas 119 patients in the control arm underwent usual monitoring alone (3-­ monthly physical and tumor marker assays, 6-­ monthly liver ultrasonography and chest radiography, and 6-­monthly whole-­body ceCT). The trial found that the failure rate (unresectable recurrence or death) was not significantly different between the intervention and control arms (29.2% and 23.7%). Multivariate analysis also showed no difference. The median time to diagnosis of unresectable recurrence (months) was significantly shorter in the intervention group (7 [range, 3–20] vs 14.3 [range, 7.3–27], P = .016); however, the mean cost per patient was significantly higher in the intervention group. The authors concluded that, when FDG-­PET/CT is added every 6 months, increased costs are seen without decreasing treatment failure rates in patients in remission of CRC. Also, neither OS nor DFS was better in the intervention arm. Several other studies have found improved survival with intensified follow-­up; however, the components of monitoring vary considerably, so that no definitive conclusions can be drawn. A Cochrane review of 15 studies including 5403 participants with stage II or III CRC—despite variability in settings and follow-­up intensity—shows more salvage surgery with curative intent in patients in the group undergoing intensified follow-­up. However, a meta-­analysis from Vargas and colleagues of 16 randomized controlled trials, including 11 with survival data, revealed that intensified monitoring was not associated with better survival. So, although FDG-­PET/CT appears to detect recurrence at an earlier time point, doing so does not appear to impact survival but does incur a financial burden.

Detection of Recurrence in Patients with Elevated Carcinoembryonic Antigen CEA is produced by the columnar and goblet cells of the colon, as well as colonic cancer cells, and has a half-­life of 3 to 11 days. Serial determination of plasma CEA concentration is widely used in the postoperative surveillance of CRC; however, clinicians face a major challenge when the CEA is elevated but no evident relapse can be localized. CEA can also be elevated in smokers, patients with inflammatory bowel disease, or other epithelial tumors, resulting in 60% to 70% sensitivity and 80% specificity in the diagnosis of recurrent CRC. Studies have demonstrated a median lead time of 9 months between serum CEA elevation and detection of recurrent disease. Furthermore, normal CEA levels do not exclude tumor recurrence, and an increased CEA does not provide information of the location of the recurrence. Thus imaging is required to confirm and localize recurrence, and in current clinical practice ceCT is the modality of choice. However, it is often difficult to differentiate between pelvic recurrence and postoperative fibrosis. To discover the value of FDG-­ PET or FDG-­ PET/CT in the detection of recurrent CRC in patients with elevated CEA, Lu and colleagues conducted a systematic review and meta-­analysis of the literature in 2013. A total of 510 patients from 11 studies (10 retrospective) were included. The authors showed that both FDG-­PET and FDG-­PET/CT performed well with a high sensitivity (90.3% and 94.1%, respectively for FDG-­PET and FDG-­PET/CT) and specificity

L A R G E B OW E L

(80% and 77.2%, respectively), with equally impressive accuracy (89% and 92.3%, respectively). In addition, FDG-­PET or FDG-­PET/ CT detected 20% of patients with CEA elevation resulting from other causes. Vallam and colleagues demonstrated a correlation between the degree of CEA elevation and the likelihood of recurrence of CRC. They retrospectively analyzed all PET/CT scans performed for elevated CEA during surveillance after complete resection of primary tumor followed by adjuvant therapy. In their sample of 104 patients with elevated CEA, 62 patients (59.6%) were found to have recurrent disease. At CEA levels less than 5, 5.1 to 10, 10.1 to 15, 15.1 to 50, and greater than 50 ng/mL, disease recurred in 10%, 45%, 70%, 94%, and 100% of patients, respectively. Sensitivity and specificity of PET/CT to were 92.7% and 95.2%, respectively, thus indicating that PET/CT is a valuable tool to detect recurrence. When comparing standard ceCT imaging with FDG-­PET/CT, the latter appears to be superior in detecting recurrence in patients with elevated CEA. This was confirmed by Ozkan and colleagues in a retrospective study that included 69 patients, showing a sensitivity and specificity of 97% and 61% for FDG-­PET/CT, compared with 51% and 61% for ceCT. 

Detection of Recurrence in Patients With Normal Carcinoembryonic Antigen Several studies have looked at the utility of FDG-­PET/CT to detect recurrence in patients with normal CEA levels. Sanli and colleagues retrospectively reviewed 235 patients with recurrence, of which 118 had a normal CEA. The sensitivity and specificity of detecting recurrence were 100% and 84%, respectively, whereas in those with elevated CEA the sensitivity was 97.1%, and specificity was 84.6%. The investigators therefore concluded that, regardless of the CEA levels, FDG-­PET/CT can successfully detect recurrence. These results have been supported by several other studies, including the work of Zhang and colleagues, who compared ceCT to FDG-­ PET/CT in patients with different CEA concentrations. FDG-­PET/ CT had a sensitivity, specificity, and accuracy of 95.2%, 82.6%, and 92.5%, respectively, whereas ceCT showed values of 80.7%, 73.9%,

Neoadjuvant and Adjuvant Therapy for Colorectal Cancer Ron G. Landmann, MD, and Alexandra W. Elias, MD

C

olon and rectal cancer, or colorectal cancer (CRC), is a common disease in the United States with an estimated 97,229 new cases of colon cancer and 43,030 new cases of rectal cancer in 2018. The lifetime risk of developing CRC is approximately 1 in 22 (4.49%) for men and 1 in 24 (4.15%) for women. Although the mortality and incidence rates have been declining over the last few decades, CRC remains the third leading cause of cancer-­related deaths among both men and women in the United States, with an estimated 50,630 cancer-­related deaths in 2018. The decline in both incidence and mortality is likely due to several factors, such as cancer prevention, improved screening, and potentially curative therapies. There are currently more than 1 million CRC survivors in the United States. For early stage CRC (stages I through III), surgical resection remains the basis of curative treatment. Recurrence, particularly distant recurrence, despite appropriate surgical resection is common

265

and 93%. The sensitivity and accuracy were significantly higher for FDG-­PET/CT compared with ceCT, and no statistical difference was found between patients with normal and increased CEA levels. Thus, even when CEA levels are not elevated, clinical, endoscopic, or conventional suspicion of recurrence should be evaluated with FGD-­PET/CT.  nn CONCLUSIONS The use of FDG-­PET and FDG-­PET/CT in the evaluation of CRC continues to evolve. Although the use of PET for initial staging appears to be limited and not cost-­effective, it is evident from the data presented in this chapter that PET can play an important role at several key stages in a patient’s journey through CRC treatment. In patients with suspected recurrent disease and in patients with liver metastases who might not be amenable to surgery, PET has been shown to have a distinct advantage compared with conventional imaging. In addition, PET can help to determine response to neoadjuvant therapy, especially in the case of rectal cancer, and can help to guide prognosis. Despite these findings, the databank for PET remains largely retrospective, with few prospective studies or clinical trials contributing to the field. More randomized controlled trials would add greater weight to the evidence presented in this chapter.

Suggested Readings Rymer B, Curtis NJ, Siddiqui MR, Chand M. FDG PET/CT Can assess the response of locally advanced rectal cancer to neoadjuvant chemoradiotherapy: evidence from meta-­analysis and systematic review. Clin Nucl Med. 2016;41:371–375. Serrano PE, Gafni A, Gu CS, et al. Positron emission tomography-­computed tomography (PET-­CT) versus no PET-CT in the management of potentially resectable colorectal cancer liver metastases: cost implications of a randomized controlled trial. J Oncol Pract. 2016;12:e765–e774. Sobhani I, Itti E, Luciani A, et  al. Colorectal cancer (CRC) monitoring by 6-­monthly 18FDG-­PET/CT: an open-­label multicentre randomised trial. Ann Oncol. 2018;29:931–937. Ye Y, Liu T, Lu L, et al. Pre-­operative TNM staging of primary colorectal cancer by (18)F-­FDG PET-­CT or PET: a meta-­analysis including 2283 patients. Int J Clin Exp Med. 2015;8:21773–22185.

and thought to be due to micrometastatic disease that is not readily detectable by current methods. For this reason, neoadjuvant and adjuvant therapies are often utilized to target micrometastases with the goal of complete eradication and prolonged survival. The challenge lies in determining which patients will benefit from additional therapy beyond surgical resection. The decision to add neoadjuvant or adjuvant therapy is based in large part on risk calculators, which take into consideration not only clinical tumor stage, but also patient-­specific data, clinicopathologic features, such as lymphovascular invasion, and molecular profiling, such as microsatellite instability (MSI) status, KRAS, NRAS, and BRAF mutations. Surgical, adjuvant, and neoadjuvant therapeutic approaches to CRC are different for colon versus rectal cancer and are discussed separately in this chapter (Fig. 1). nn STAGING Formal CRC staging is essential to risk stratification for guiding clinicians’ treatment decisions for individual patients. Guidelines from the National Comprehensive Cancer Network (NCCN) for preoperative workup of newly diagnosed CRC should include the following: colonoscopy (plus consideration of rigid proctoscopy for rectal cancer) with biopsy and pathology review; complete blood counts, chemistry profile, and carcinoembryonic antigen (CEA) serum level; computed tomographic (CT) scan of chest, abdomen, and pelvis for colon

L A R G E B OW E L

(80% and 77.2%, respectively), with equally impressive accuracy (89% and 92.3%, respectively). In addition, FDG-­PET or FDG-­PET/ CT detected 20% of patients with CEA elevation resulting from other causes. Vallam and colleagues demonstrated a correlation between the degree of CEA elevation and the likelihood of recurrence of CRC. They retrospectively analyzed all PET/CT scans performed for elevated CEA during surveillance after complete resection of primary tumor followed by adjuvant therapy. In their sample of 104 patients with elevated CEA, 62 patients (59.6%) were found to have recurrent disease. At CEA levels less than 5, 5.1 to 10, 10.1 to 15, 15.1 to 50, and greater than 50 ng/mL, disease recurred in 10%, 45%, 70%, 94%, and 100% of patients, respectively. Sensitivity and specificity of PET/CT to were 92.7% and 95.2%, respectively, thus indicating that PET/CT is a valuable tool to detect recurrence. When comparing standard ceCT imaging with FDG-­PET/CT, the latter appears to be superior in detecting recurrence in patients with elevated CEA. This was confirmed by Ozkan and colleagues in a retrospective study that included 69 patients, showing a sensitivity and specificity of 97% and 61% for FDG-­PET/CT, compared with 51% and 61% for ceCT. 

Detection of Recurrence in Patients With Normal Carcinoembryonic Antigen Several studies have looked at the utility of FDG-­PET/CT to detect recurrence in patients with normal CEA levels. Sanli and colleagues retrospectively reviewed 235 patients with recurrence, of which 118 had a normal CEA. The sensitivity and specificity of detecting recurrence were 100% and 84%, respectively, whereas in those with elevated CEA the sensitivity was 97.1%, and specificity was 84.6%. The investigators therefore concluded that, regardless of the CEA levels, FDG-­PET/CT can successfully detect recurrence. These results have been supported by several other studies, including the work of Zhang and colleagues, who compared ceCT to FDG-­ PET/CT in patients with different CEA concentrations. FDG-­PET/ CT had a sensitivity, specificity, and accuracy of 95.2%, 82.6%, and 92.5%, respectively, whereas ceCT showed values of 80.7%, 73.9%,

Neoadjuvant and Adjuvant Therapy for Colorectal Cancer Ron G. Landmann, MD, and Alexandra W. Elias, MD

C

olon and rectal cancer, or colorectal cancer (CRC), is a common disease in the United States with an estimated 97,229 new cases of colon cancer and 43,030 new cases of rectal cancer in 2018. The lifetime risk of developing CRC is approximately 1 in 22 (4.49%) for men and 1 in 24 (4.15%) for women. Although the mortality and incidence rates have been declining over the last few decades, CRC remains the third leading cause of cancer-­related deaths among both men and women in the United States, with an estimated 50,630 cancer-­related deaths in 2018. The decline in both incidence and mortality is likely due to several factors, such as cancer prevention, improved screening, and potentially curative therapies. There are currently more than 1 million CRC survivors in the United States. For early stage CRC (stages I through III), surgical resection remains the basis of curative treatment. Recurrence, particularly distant recurrence, despite appropriate surgical resection is common

265

and 93%. The sensitivity and accuracy were significantly higher for FDG-­PET/CT compared with ceCT, and no statistical difference was found between patients with normal and increased CEA levels. Thus, even when CEA levels are not elevated, clinical, endoscopic, or conventional suspicion of recurrence should be evaluated with FGD-­PET/CT.  nn CONCLUSIONS The use of FDG-­PET and FDG-­PET/CT in the evaluation of CRC continues to evolve. Although the use of PET for initial staging appears to be limited and not cost-­effective, it is evident from the data presented in this chapter that PET can play an important role at several key stages in a patient’s journey through CRC treatment. In patients with suspected recurrent disease and in patients with liver metastases who might not be amenable to surgery, PET has been shown to have a distinct advantage compared with conventional imaging. In addition, PET can help to determine response to neoadjuvant therapy, especially in the case of rectal cancer, and can help to guide prognosis. Despite these findings, the databank for PET remains largely retrospective, with few prospective studies or clinical trials contributing to the field. More randomized controlled trials would add greater weight to the evidence presented in this chapter.

Suggested Readings Rymer B, Curtis NJ, Siddiqui MR, Chand M. FDG PET/CT Can assess the response of locally advanced rectal cancer to neoadjuvant chemoradiotherapy: evidence from meta-­analysis and systematic review. Clin Nucl Med. 2016;41:371–375. Serrano PE, Gafni A, Gu CS, et al. Positron emission tomography-­computed tomography (PET-­CT) versus no PET-CT in the management of potentially resectable colorectal cancer liver metastases: cost implications of a randomized controlled trial. J Oncol Pract. 2016;12:e765–e774. Sobhani I, Itti E, Luciani A, et  al. Colorectal cancer (CRC) monitoring by 6-­monthly 18FDG-­PET/CT: an open-­label multicentre randomised trial. Ann Oncol. 2018;29:931–937. Ye Y, Liu T, Lu L, et al. Pre-­operative TNM staging of primary colorectal cancer by (18)F-­FDG PET-­CT or PET: a meta-­analysis including 2283 patients. Int J Clin Exp Med. 2015;8:21773–22185.

and thought to be due to micrometastatic disease that is not readily detectable by current methods. For this reason, neoadjuvant and adjuvant therapies are often utilized to target micrometastases with the goal of complete eradication and prolonged survival. The challenge lies in determining which patients will benefit from additional therapy beyond surgical resection. The decision to add neoadjuvant or adjuvant therapy is based in large part on risk calculators, which take into consideration not only clinical tumor stage, but also patient-­specific data, clinicopathologic features, such as lymphovascular invasion, and molecular profiling, such as microsatellite instability (MSI) status, KRAS, NRAS, and BRAF mutations. Surgical, adjuvant, and neoadjuvant therapeutic approaches to CRC are different for colon versus rectal cancer and are discussed separately in this chapter (Fig. 1). nn STAGING Formal CRC staging is essential to risk stratification for guiding clinicians’ treatment decisions for individual patients. Guidelines from the National Comprehensive Cancer Network (NCCN) for preoperative workup of newly diagnosed CRC should include the following: colonoscopy (plus consideration of rigid proctoscopy for rectal cancer) with biopsy and pathology review; complete blood counts, chemistry profile, and carcinoembryonic antigen (CEA) serum level; computed tomographic (CT) scan of chest, abdomen, and pelvis for colon

266

Neoadjuvant and Adjuvant Therapy for Colorectal Cancer

Resected colon cancer

Stage II Yes T4, 70, Borderline PS

5–FU/LV or capecitabine

• Clinical factors • Molecular markers

FIG. 1 Algorithm for use of adjuvant treatment for colorectal cancer. 5-­FU/LV, 5-­fluorouracil and leucovorin; FOLFOX, folinic acid, 5-­fluorouracil, oxaliplatin; LN, lymph nodes; dMMR, defective mismatch repair; MSI-­H, microsatellite instability-­high; PS, performance status; XELOX, capecitabine and oxaliplatin.

cancer versus CT chest with magnetic resonance imaging (MRI) of the abdomen and pelvis with contrast for rectal cancer. Endorectal ultrasound for rectal cancer is no longer indicated unless a patient has contraindications to MRI, such as an incompatible pacemaker. One of the reasons preoperative staging is particularly important in rectal cancer is because it determines whether patients receive neoadjuvant radiation therapy (RT), with the goal of down-­staging. Positron emission tomography is not recommended (NCCN Guidelines, 2018). CRC is staged based on the TNM staging system (T, primary tumor; N, regional lymph nodes; M, distant metastasis) adopted by the American Joint Committee on Cancer (AJCC; Table 1). In the most recent version of the AJCC Cancer Staging Manual (8th edition, 2016), several modifications were made to assist clinicians with prognostication, including the additional subdivision of M1c, which details peritoneal carcinomatosis as a poor prognostic factor; clarification to how tumor deposits in lymph nodes are defined, reintroduction of the “L” and “V” elements to better identify lymphatic and vessel invasion; and the identification of MSI status, KRAS, NRAS, and BRAF mutations as additional prognostic and predictive factors. This is discussed further in later sections. The importance of accurate staging is reflected in Fig. 2, which outlines the predicted 5-­year survival rates.  nn RESECTED

COLON CANCER

Adjuvant Therapy for Colon Cancer Stage II Colon Cancer

For stage III colon cancer, the addition of adjuvant fluorouracil-­ containing chemotherapy regimens has demonstrated a significant survival benefit and has been established as the standard of care since a systematic review in 1988; however, the use of adjuvant chemotherapy in stage II colon cancer is controversial. Trials for adjuvant chemotherapy in stage II disease have not shown a clear benefit, although some studies have shown small increases in disease-­free survival (DFS), particularly in patients with high-­risk disease. Because most patients with R0 resections of stage II disease have a good predicted 5-­year survival rate (see Fig. 2), an excellent benefit from adjuvant therapy would be needed to demonstrate improved overall outcomes. Additionally, stage IIC patients have a poorer overall survival (OS) at 5 years compared with stage IIIA patients; thus there may be a subset

of patients who would derive improved long-­term outcomes with adjuvant chemotherapy. Although randomized trials have failed to demonstrate a statistically significant improvement in OS with adjuvant chemotherapy for patients with stage II cancer, the number of patients with high-­risk stage II disease in these studies is likely inadequate to demonstrate benefit. For this reason, the American College of Clinical Oncology states the risks and benefits of adjuvant chemotherapy should be discussed with patients, and individualized treatment plans should be constructed. There are three AJCC-­ approved web-­ based prognostic tools, which can be used to help counsel patients. To predict recurrence and survival, these tools take into consideration such factors as age, sex, race, body mass index, performance status, T category, tumor differentiation/grade, number of regional lymph nodes evaluated, number of positive regional lymph nodes, and tumor location. In addition, gene assays such as Oncotype DX Colon and Coloprint, as well as circulating DNA profiles, can help predict recurrence and response to therapy; however, they should be used with caution, as prospective trials have not demonstrated that these tests can determine which patients benefit from adjuvant chemotherapy. High-­risk features for which a clinician may be more inclined to offer adjuvant therapy to a patient with stage II colon cancer include having inadequate lymph node sampling (15 mm, impacted, difficult cannulation, altered anatomy), sphincterotomy and endoscopic papillary large balloon dilation can result in high success rates of complete clearance. When this combined approach is used, a large endoscopic sphincterotomy is not required. Patients with

G A L L B L A D D E R A N D B I L I A RY T R E E

497

prior biliary sphincterotomy do not need extension of the sphincterotomy before endoscopic papillary large balloon dilation. The diameter of the dilating balloon ranges from 12 to 20 mm (short 5–5.5 cm esophageal, pyloric, or colonic dilating balloons), and is chosen based on the diameter of the stone or stones and the maximal diameter of the bile duct. There are other techniques that can be used for stone extraction of difficult bile duct stones such as mechanical lithotripsy using a basket and the digital single operator cholangioscopy system with electrohydraulic lithotripsy or laser lithotripsy, the latter with more than 95% rate of stone clearance with low rate of adverse events. If stone removal is unsuccessful, biliary decompression should be accomplished by placement of a biliary stent or nasobiliary drain. Endoscopic sphincterotomy and stone extraction without subsequent cholecystectomy may be appropriate in some patients with comorbidities and high surgical risk. Nevertheless, biliary symptoms recur twice as commonly in patients whose gallbladder remains in situ with a 5-­year risk of significant biliary adverse events leading to cholecystectomy as high as 15%. 

T   ABLE 2 ASGE Proposed Strategy to Assign Risk of Choledocholithiasis in Patients With Symptomatic Cholelithiasis Based on Clinical Predictors Clinical Predictors of Choledocholithiasis

Very strong   CBD stone on transabdominal US   Clinical ascending cholangitis   Bilirubin >4 mg/dL Strong   Dilated CBD on US (>6 mm with gallbladder in situ)   Bilirubin level 1.8–4 mg/dL Moderate   Abnormal liver biochemical test other than bilirubin   Age older than 55 years   Dilated CBD on US/cross-sectional imaging

Benign Biliary Strictures

Likelihood of Choledocholithiasis Based On Clinical Predictors

Probability

Presence of any very strong predictor

High (>50%)

Presence of both strong predictors

High (>50%)

No predictors present

Low (< 10%)

All other patients

Intermediate (10%–50%)

ASGE, American Society for Gastrointestinal Endoscopy; CBD, common bile duct; US, ultrasound.

ERCP is indicated in the evaluation and treatment of benign biliary strictures, congenital bile duct abnormalities, and postoperative adverse events such as anastomotic strictures and biliary leaks. Biopsies and brushings can help define the etiology of benign biliary strictures and diagnostic yield may increase with cholangioscopy directed biopsies. Biliary injury is rarely recognized during surgery; symptoms appear most often in the early postoperative course or months or years after surgery. Benign biliary strictures may be dilated with hydrostatic balloons (maximum biliary dilator diameter 10 mm) or graduated catheters passed over a guidewire. Benign biliary strictures amenable to endoscopic dilation include those secondary to chronic pancreatitis, dominant strictures in PSC, postoperative

Symptomatic Patient with Cholelithiasis

Likelihood of CBD Stone Based on Clinical Predictors (Table 2)

Low

Intermediate

High

ORa

Preoperative ERCP

Laparoscopic Cholecystectomy No Cholangiography

Negative Laparoscopic Cholecystectomy

Laparoscopic IOC or Laparoscopic Ultrasound

If Positive, or If Unavailable

Positive

ORa Laparoscopic Common Bile Duct Exploration

Preoperative EUS or MRCP

Postoperative ERCP

FIG. 1 American Society for Gastrointestinal Endoscopy suggested management algorithm for patients with symptomatic cholelithiasis based on the risk of probability for choledocholithiasis. aDepending on costs and local expertise. (Modified from Tse F, Barkun JS, Barkun AN. The elective evaluation of patients with suspected choledocholithiasis undergoing laparoscopic cholecystectomy. Gastrointest Endosc. 2004;60:437-­448.)

498

Obstructive Jaundice: Endoscopic Therapy

T   ABLE 3  Bismuth Classification for Benign Biliary Strictures Type

Criteria

I

Low common hepatic duct stricture with a length of common hepatic duct stump greater than 2 cm

II

Proximal common hepatic duct stricture with a common hepatic duct stump less than 2 cm

III

Hilar stricture, no residual common hepatic duct, but the hepatic ductal confluence is preserved

IV

Hilar stricture with involvement of hepatic ductal confluence and loss of communication between right and left hepatic duct

V

Involvement of an aberrant right sectorial duct alone or with concomitant stricture of the common hepatic duct

strictures, and strictures caused by stone disease. Single or multiple plastic stents may be used to maintain patency after initial dilation. Serial endoscopic dilation and maximal caliber stent placement can be used to achieve prolonged ductal patency in most benign postoperative strictures. Despite high success rates, this approach is technically demanding and requires an average of five ERCPs. Fully covered self-­expandable metal stents (cSEMS) are an alternative approach in benign biliary strictures. The cSEMS expand to lumen sizes larger than those of plastic stents, without the risk of tissue ingrowth and embedding such as in uncovered and partially covered metal stents, allowing an indwell time up to 12 months if necessary, and without compromising the ease of removal. Chronic pancreatitis accounts for 10% of all common bile duct strictures. Treatment is indicated for patients with jaundice and/or cholangitis, those with significant biliary dilation (12–15 mm) proximal to the stricture, and with abnormal liver function tests (AP >3 times the normal value for 6 months or longer). In patients with bile duct strictures resulting from chronic pancreatitis the use of multiple plastic stents over a long period (14 months) compared with a single stent increases the chance of long-­term overall success of 65.2% with a high risk of restenosis (17%). In patients with PSC, ERCP is no longer indicated for diagnosis in a routine basis. An MRCP is a noninvasive imaging test with comparable diagnostic accuracy to ERCP in the diagnosis of dominant strictures. A dominant stricture is defined as a lumen diameter of 1.5 mm or less in the common bile duct and 1 mm or less in the common hepatic duct and it is seen in 45% to 58% of patients with PSC. In case of unclear results of an MRCP or clinical deterioration of a patient with prior diagnosis of PSC (worsening cholestasis, jaundice, or cholangitis), ERCP can aid in the diagnosis of a dominant stricture. Benign strictures in PSC patients respond well to endoscopic therapy with balloon dilation with or without stent placement. Limited data suggest that balloon dilation is often sufficient and that the use of stents may be associated with an increased risk of adverse events and cholangitis. Postsurgical strictures occur after orthotopic liver transplantation and laparoscopic cholecystectomy in 3% to 13% and 0.2% to 0.7%, respectively. Also, strictures at bile-­enteric anastomoses are seen more often every day in clinical practice. Strictures recognized early in the postoperative/operative period are often associated with a bile leak caused by direct trauma, whereas delayed presentation is commonly associated with ischemic injury and resultant fibrosis. A commonly used classification for postoperative strictures is the Bismuth classification, based on location of the stricture within the bile tree (Table 3). Clinical presentation (elevation of liver function tests, pain, jaundice) depends on the degree of bile duct obstruction.

Endoscopic treatment usually involves serial placement of multiple large bore plastic stents over a 1-­year period, with exchange every 3 to 4 months. Balloon dilation of anastomotic biliary strictures within the first 4 weeks of surgery carries an increased risk of anastomotic dehiscence and thus a less aggressive approach is suggested in this setting. Success rates for this approach range from 74% to 90% with recurrence rates as high as 30% within 2 years of stent removal. Distal postoperative biliary strictures (Bismuth I and II) are associated with better success rates compared to proximal hilar strictures (Bismuth III). A 2016 randomized control trial comparing plastic stents versus cSEMS in benign biliary strictures concluded that patients with benign biliary strictures and a bile duct diameter 6 mm or more in whom the cSEMS would not overlap the cystic duct, cSEMS should be considered an appropriate option. The main potential benefit of cSEMS in the management of refractory benign biliary strictures is their large caliber and longer duration of patency allowing them to be left in place longer, resulting in fewer procedures for serial dilations and placement of multiple plastic stents. 

Malignant Biliary Strictures The most widely recognized cause of malignant distal biliary obstruction is pancreatic head cancer. Other causes of malignant obstructive jaundice are cholangiocarcinomas (CCA), gallbladder cancer, and metastatic tumors compressing the biliary system. ERCP is the most accepted and widely used method for biliary decompression in these scenarios. In those patients with unresectable malignant biliary obstruction, effective biliary decompression improves symptoms and enables patients to undergo palliative therapies. On the other hand, in those patients who are surgical candidates, routine preoperative biliary intervention may worsen outcomes. In a randomized control trial comparing preoperative biliary drainage using plastic stents versus early surgery without drainage in resectable pancreatic cancer showed that endoscopic preoperative biliary drainage with placement of a plastic stent did not have a beneficial effect on the surgical outcome. In a recent published meta-­analysis including 32 studies showed that the available evidence argues against preoperative biliary drainage in patient with resectable pancreatic head adenocarcinoma. With the advent of neoadjuvant chemotherapy used to downstage potentially unresectable tumors in the hope of improving the outcome, preoperative biliary drainage is clinically relevant. Preoperative biliary drainage during the period of neoadjuvant treatment might be best achieved with metal stents, which have a higher patency rate than plastic stents. In patients with unresectable pancreatic head adenocarcinoma and obstructive jaundice, ERCP with SEMS are superior to plastic stents in the palliative setting. A recent meta-­analysis of 19 studies involving 1989 patients, cSEMS were associated with significantly lower occlusion rates, less therapeutic failure (7% vs 13%), less need for reintervention and lower rates of cholangitis (8% vs 21%) than plastic stents. In distal malignant biliary obstruction, there is the option of cSEMS vs. uncovered SEMS, each of which have its advantages and disadvantages. A meta-­analysis involving eight studies concluded that the group of patients with cSEMS had lower incidence of adverse events, with no significant difference in dysfunctions; however, cSEMS trends to be better, with no difference in stent patency, patient survival, and complications. Malignant hilar strictures are categorized according to the Bismuth-­Corlette classification (Table 4). At the time of presentation, only a minority (48 hr)

Moderately Severe

Mild

Low-fat diet as tolerated

Prophylactic antibiotics are not indicated in patients with no signs of infection. Historically, prophylactic antibiotics were given on the premise of preventing the progression of sterile pancreatic and peripancreatic necrosis to infected necrosis. However, studies have shown that prophylactic antibiotics confer no benefits, predispose to multidrug-­resistant and fungal infections, and should not be administered unless there is clear evidence of infection (cholangitis, bacteremia, proven infected pancreatic necrosis). However, it may be difficult to be sure if infection is present because patients with acute pancreatitis often manifest signs and symptoms of SIRS such as fever, tachycardia, and leukocytosis, all of which can mimic infection. When infection is suspected, CT imaging is indicated. Infected necrosis can be diagnosed by imaging demonstrating gas within the areas of pancreatic necrosis. Although this finding is highly specific, it is not sensitive. Until recently, percutaneous sampling by needle aspiration of the necrotic collection for culture was routinely advocated to assess

Succeeds

Trial of low-fat oral diet for 3-4 days

Severe

Fails

Nasogastric feeds (within 2 days) Fails Nasojejunal feeds Fails Total parenteral nutrition

FIG. 1  Nutritional treatment algorithm for patients with acute pancreatitis based on severity of disease.

PA N C R E A S

for infected necrosis when there was clinical suspicion for infection but no definitive evidence on imaging. However, it has a relatively high false-­negative rate of approximately 20% and should be used selectively. It should be performed if there is no clear-­cut evidence of infection (i.e., gas bubbles on CT) and prolonged unwellness, and a positive aspirate culture will prompt initiation of antibiotics. If the patient has ongoing inflammatory signs and a negative needle aspiration will not change the treatment plan, a decision to treat empirically based on the clinical diagnosis is justified. Patients with necrosis and “persistent unwellness” (failure to thrive) over a period of weeks often harbor occult infection. Our studies have shown that 40% of these patients have infected necrosis when cultures are obtained, many times after negative needle aspirates. If infection is suspected or confirmed, antimicrobial therapy should be initiated. The majority of infections are caused by enteric organisms. Empiric antibiotic treatment should consist of a regimen with an appropriate spectrum of antimicrobial activity and known to penetrate pancreatic necrosis such as a carbapenem or gram-­ negative agent (quinolone or third-­or fourth-­generation cephalosporin) plus metronidazole. If drainage or debridement is performed, a sample should be taken for culture to guide treatment. If patients are started on antibiotics for infected necrosis in the absence of microbiologic data and continue to clinically deteriorate, the regimen should be empirically broadened to include gram positive, fungal, and multidrug-­resistant organisms in addition to performing other maneuvers to achieve source control (e.g., drainage, debridement) and considering extrapancreatic sources of the clinical deterioration.  nn KEY

COMPLICATIONS

Choledocholithiasis and Cholangitis In patients presenting with pancreatitis caused by gallstones, routine ERCP should not be performed. The majority of inciting stones will pass into the duodenum spontaneously. However, retained stones or cholangitis may occur in parallel with the ongoing pancreatitis. In cases of suspected persistent choledocholithiasis without any concern for cholangitis, a period of watchful waiting of 24 to 48 hours with serial liver function tests and clinical monitoring is reasonable as spontaneous passage of gallstones is common and results in resolution of the obstruction without the need for intervention. These patients should be managed like any other patients with possible choledocholithiasis, which is discussed elsewhere in this text. In cases of suspected cholangitis, blood cultures should be drawn, followed by the initiation of broad-­spectrum antibiotics and urgent ERCP to relieve the biliary obstruction. 

Abdominal Compartment Syndrome Abdominal compartment syndrome occurs when organ dysfunction, most prominently oliguric acute kidney injury and respiratory failure with high peak inspiratory pressures, is caused by intra­abdominal hypertension. Although intraabdominal hypertension is defined as intra­abdominal pressures exceeding 12 mm Hg, abdominal compartment syndrome rarely occurs with intraabdominal pressures less than 20 mm Hg. The intra­abdominal pressure is measured by instilling a small volume of fluid into the bladder via a urinary catheter and transducing the catheter. Three main factors can contribute to the development of abdominal compartment syndrome in acute pancreatitis: (1) visceral edema from large-­volume crystalloid resuscitation, (2) ascites similarly related to fluid resuscitation and occasionally to portal system thrombosis, and (3) space occupying retroperitoneal and peripancreatic necrosis and fluid collections. This complication typically occurs early in the course of disease when the inflammatory response is at its height amid active fluid resuscitation. Initial treatment is the cessation of unnecessary volume infusion, volume removal if possible via diuretic administration or ultrafiltration, interventions to decrease hollow-­viscera volume (nasogastric drainage, rectal drainage, prokinetics), and sedation and neuromuscular

503

blockade to increase abdominal wall compliance. In patients with a significant volume of ascites, percutaneous drainage can sometimes result in a dramatic decrease in intraabdominal pressure and should be attempted before decompressive laparotomy is considered. When all other strategies fail, decompressive laparotomy is the definitive treatment for abdominal compartment syndrome. It is critical to note that debridement of pancreatic necrosis should not be performed at the time of abdominal decompression as very early debridement increases mortality. There are a variety of techniques for managing the open abdomen. We prefer to use a vacuum-­assisted abdominal closure, with closure beginning 24 to 72 hours after the initial decompression and proceeding in a staged fashion every 24 to 48 hours thereafter if primary closure is not possible. 

Pseudoaneurysm Pseudoaneurysms occur when enzyme-­rich pancreatic fluid autodigests and weakens the walls of pancreatic and visceral arteries, leading to wall defects. The most commonly involved arteries are the splenic, branches of the gastroduodenal arcade, and the hepatic artery (Fig. 2A). These pseudoaneurysms are life threatening because of their potential for major hemorrhage. Pseudoaneurysms are typically asymptomatic until they bleed, and are often incidentally identified during CT or magnetic resonance imaging in the course of serial evaluations. We recommend aggressive angioembolization of virtually all pseudoaneurysms found in pancreatitis fields because there is no reliable means of clinical monitoring and the consequences of hemorrhage are potentially grave. When pseudoaneurysms do rupture, they may present initially with a small volume “sentinel” bleed that, if recognized, can provide a window for intervention before massive hemorrhage ensues. In patients with external drainage catheters, any bleeding from the drain must be taken seriously and considered a possible pseudoaneurysm until proven otherwise. Similarly, pseudoaneurysm should be in the differential diagnosis when upper gastrointestinal bleeding occurs in a patient with transgastric stents or a gastric anastomosis to the necrosis cavity (discussed later). In stable patients, CT angiography (CTA) can effectively assess for pseudoaneurysm, provide a picture of any surrounding local complications (such as necrosis or pseudocyst), and provide a road map for angioembolization. In hemodynamically unstable patients with clinical suspicion for a bleeding pseudoaneurysm, there may not be time for CTA; immediate angiography may be required. Surgical control of a bleeding pseudoaneurysm within a bed of pancreatic necrosis is extremely challenging, and operative intervention should be reserved for patients with severe hemodynamic instability when angiography is not immediately available or where angiographic interventions have failed. 

Pseudocyst Acute pancreatic and peripancreatic collections are common local complications of acute pancreatitis. These collections usually do not require intervention, with a minority of them progressing to walled-­off necrosis or pseudocysts when they are not reabsorbed. Pseudocysts are organized collections of pancreatic fluid that have a connection to the pancreatic duct system, and persist more than 4 weeks after the index pancreatitis episode. Most pancreatic pseudocysts resolve spontaneously, or remain asymptomatic without intervention. In symptomatic cases, symptoms typically develop as a result of pressure symptoms on adjacent structures (usually stomach or duodenum), resulting in pain or early satiety. Less commonly, infection, rupture, or hemorrhage, may occur (Fig. 2B). In symptomatic patients, standard management consists of internal drainage into the gastrointestinal tract. This is most commonly performed endoscopically. Surgical drainage, which may have a higher rate of resolving the pseudocyst in a single procedure, should be considered when patients have failed endoscopic drainage, when there is a disconnected pancreatic duct (requiring internal drainage or resection), for larger pseudocysts in dependent locations (i.e.,

504

Management of Acute Necrotizing Pancreatitis

paracolic gutters), or when other intraabdominal procedures are indicated (i.e., laparoscopic cyst gastrostomy at the time of laparoscopic cholecystectomy to prevent recurrent biliary pancreatitis). For patients with hemorrhage into the pseudocyst, a CTA should be performed to determine the appropriate next steps. Bleeding may be from a major adjacent visceral artery branch or from a small unnamed vessel in the pseudocyst wall. If CTA demonstrates a significant arterial bleed, the bleeding should be treated akin to a pseudoaneurysm and managed as detailed in the previous section. If CTA demonstrates no clear evidence of an arterial bleed, the bleed seen on initial imaging is likely a sidewall bleed within the pseudocyst, and the pseudocyst can be managed without directly addressing the culprit vessel. Be cognizant of the fact that cystic neoplasms can masquerade as a pseudocyst, especially in cases in which a prior history of pancreatitis is unclear. 

Portal System Thrombosis Portal system thrombosis is often asymptomatic and is typically found incidentally on imaging (Fig. 2C). Long-­term sequelae are the

result of chronic portal hypertension, the most obvious of which is the formation of varices although these are rarely complicated by bleeding. Complete superior mesenteric or portal vein thromboses can occasionally have more dramatic presentation however, including massive ascites, propagation into the mesenteric venous tributaries with bowel infarction, and encephalopathy. Good evidence for the management of portal system thromboses in pancreatitis is lacking. We typically anticoagulate with intravenous unfractionated heparin or subcutaneous low-­molecular-­weight heparin, with a bridge to oral anticoagulation for 3 to 6 months when no further invasive procedures planned. Anticoagulation seems to be effective, with partial recanalization of the portal venous system occurring in 63% to 93% of patients, and complete recanalization in 34% to 45%. If recanalization occurs, it typically happens within the first 6 months, and anticoagulation beyond that point is unlikely to be effective. In case of acute symptomatic thrombosis, endovascular thrombectomy or thrombolysis can be considered, and in the very rare case with associated bowel infarction requiring laparotomy, surgical thrombectomy is an option. 

LEFT

A

B

C

D

FIG. 2  Potential complications of acute pancreatitis. (A) Splenic artery pseudoaneurysm with active hemorrhage (red arrow). (B) Pancreatic pseudocyst directly posterior to the stomach (red arrow). (C) Superior mesenteric vein thrombosis (red arrow). (D) Disconnected duct syndrome. The red arrow indicates a viable segment of pancreatic tail disconnected from the head, with a central area of nonviable pancreatic tissue.

PA N C R E A S

Disconnected Duct Syndrome The disconnected duct syndrome refers to complete necrosis of a central portion of the pancreas with associated disruption of the main pancreatic duct such that a viable tail segment is “disconnected” from any drainage pathway to the ampulla of Vater (Fig. 2D). After resolution of the necrosis, this can present a number of problems including external pancreatic fistula if external drainage is in place, recurrent pseudocyst, or chronic recurrent pain resulting from obstructive pancreatitis because the disconnected segment produces exocrine secretions that have nowhere to go. If a disconnected duct is identified early in the course of pancreatitis, then the management strategy for the necrosis should incorporate some form of internal drainage (see the sections on endoscopic and surgical transgastric necrosectomy and dual-­modality therapy later in this chapter) to avoid prolonged external drainage. If endoscopic therapy is chosen, then soft silastic transgastric stents should be left in place indefinitely. In case of recurrent obstructive pancreatitis in the disconnected tail, endoscopic pancreatic duct stenting may be able to reestablish drainage to the ampulla. Surgical options include internal drainage, usually into a Roux limb of jejunum, or resection of the disconnected distal segment. Distal pancreatectomy has a slightly higher rate of successful symptomatic management (95% vs 80% in one recent series) but at the cost of more intraoperative blood loss, worse long-­term endocrine function, and usually splenectomy. One of the most important aspects of this syndrome is to recognize it. In our experience, many patients referred with this problem have been told “this is life after a bad episode of acute pancreatitis,” sometimes for years, when in fact therapeutic options may be available.  nn INTERVENTIONAL

MANAGEMENT

The primary indication for operative intervention in necrotizing pancreatitis is known or suspected infected pancreatic necrosis. Patients with sterile necrosis with gastric outlet or intestinal obstruction or persistent failure to thrive beyond 8 weeks may also benefit from necrosectomy and, as mentioned previously, a significant minority of

505

them may also have occult infection. When intervention is required, it should ideally be delayed for a minimum of 4 weeks whenever possible to allow for encapsulation and demarcation of the necrosis. This waiting period applies to all of the techniques described later. The strategy for intervention for infected necrosis typically involves a step­up approach. Patients undergo percutaneous or endoscopic drainage of the infected cavity as the first intervention, with necrosectomy reserved for those who fail to improve after the initial drainage. The step-­up approach is associated with less major morbidity and in some studies lower mortality rates compared with primary open necrosectomy. In 30% to 40% of cases, percutaneous or endoscopic drainage alone may obviate the need for necrosectomy. The question of when to step up should be tailored to the individual patients depending on the extent and location of the necrosis and clinical progress after initial drainage, but there are no hard criteria established for when to step up. It is worth emphasizing that the majority of patients with infected necrosis require some form of necrosectomy, not just drainage and antibiotic treatment alone. Because the step-­up approach has been popularized and the number of facilities capable of percutaneous and endoscopic intervention has proliferated, it has been our impression that the misconception that drainage suffices for the management of infected necrosis has grown in parallel. This frequently results in cases in which an appropriate initial intervention (such as percutaneous drainage) is undertaken, but patients are inappropriately allowed to languish because of a failure to recognize the need to step up and debride. A complete strategic plan should be established before the initial intervention because the route and method of percutaneous drainage may dictate or limit subsequent options. Fig. 3 depicts our algorithm for selecting the optimal method of intervention; Table 3 depicts their relative strengths and weaknesses.

Video-­Assisted Retroperitoneal Debridement Video-­ assisted retroperitoneal debridement (VARD) is the technique most closely associated with minimally invasive necrosectomy because it was used almost exclusively in the landmark 2010 study

Infected WOPN

Transgastric access? Yes Surgical debridement Consider if: - Disconnected duct - Large volume dense necrosis - Needs cholecystectomy - Physiologically stable

No

Endoscopic debridement

Retroperitoneal access?

- Default technique

Yes

No

Video-assisted retroperitoneal debridement

Sinus tract endoscopy

Enteric or pancreatic fistula

Slow to resolve

Consider dual modality therapy

FIG. 3  Decision algorithm for selecting the optimal intervention approach in patients with infected walled-­ off pancreatic necrosis. Surgical debridement refers to transgastric debridement, which can be performed open or laparoscopically. Dual-­modality therapy refers to a combination of endoscopic transgastric drainage and external necrosectomy via video-­assisted retroperitoneal debridement or sinus tract endoscopy. WOPN, walled-­off pancreatic necrosis.

506

Management of Acute Necrotizing Pancreatitis

TABLE 3 Advantages and Disadvantages of Different Pancreatic Necrosectomy Approaches Open Transperitoneal Necrosectomy

Video-­Assisted ­Retroperitoneal Debridement

Transgastric Necrosectomy

Advantages

Flexibility in access

Disadvantages

High rates of morbid- Requires a clear anatomic window ity and mortality with posterior stomach Endoscopic approach usually requires multiple reinterventions

Sinus Tract Endoscopic ­Debridement

Lacks external drainage system Does not require a trans- Flexibility in access (transperitoneEndoscopic approach has decreased gastric window al, retroperitoneal, intercostal) morbidity when compared to Decreased morbidity com- Decreased wound morbidity open necrosectomy pared with open transcompared with open necroSurgical approach allows rapid peritoneal necrosectomy sectomy or video-­assisted debridement and simultaneous Uses equipment familiar to retroperitoneal debridement cholecystectomy surgeons

that established the step-­up approach as the usual standard of care. VARD relies on a retroperitoneal drain placement as the initial intervention. This is most commonly placed via the left flank in the window between the left kidney and descending colon. If drainage alone does not resolve the infected necrosis, the drain is used as a guide for the VARD procedure. Thus, careful attention to the exact route of placement is critical and should be discussed in detail with the radiologist performing the drainage procedure. When the decision is made to proceed with VARD, the patient is placed in a partial lateral decubitus position at a 30-­to 45-­degree angle, with the percutaneous drain prepped into the field. A 4-­to 5-­cm lateral flank incision is made over the drain, and electrocautery is used for the dissection along its tract through the retroperitoneum into the necrotic cavity. It is critical for the surgeon to have diligently studied the preoperative CT scan to understand the relationship of the surrounding structures (colon, kidney, splenic vessels) to the course of the drain tract to avoid injury to them (Fig. 4). Retractors and a laparoscope are used to provide visualization. On entering the cavity, suction and irrigation are used to clear liquid debris. If the cavity is superficial enough, ring forceps can be used to excavate the necrotic material. For deeper cavities, longer rigid retractors are used to create a working space, and laparoscopic graspers can be used for debridement. Once debridement has been completed, drains are placed. We typically bring the drain(s) out through a separate stab incision(s) and close the initial wound in multiple layers to reduce the risk of leakage through the wound. A video of the procedure can be found at https://youtu.be/9ErPBAnrOAU. VARD uses equipment familiar to surgeons, such as a laparoscope for visualization, as well as ring or laparoscopic forceps for debridement. Similar to open necrosectomies, VARD allows rapid and large-­volume debridement of necrosis, and complete debridement is usually achievable in a single procedure. VARD has a number of limitations, however. First, there is a relatively high rate of external pancreatic fistula. Second, wound complications, although generally minor, are fairly common. Finally, it requires a retroperitoneal window that is wide and safe enough for a cutdown along the drain tract into the cavity. 

Sinus Tract Endoscopic Debridement Sinus tract endoscopic debridement (STE) is another step-­up option when percutaneous drainage fails to resolve pancreatic necrosis. STE is relatively free of anatomic requirements because any necrotic collection that can be accessed percutaneously can be accessed by STE.

Increased wound compli- Requires familiarization with cations compared with equipment (rigid nephroscope, sinus tract endoscopic intraoperative fluoroscopy) debridement Often requires multiple Requires a retroperitoneal ­interventions drainage route Pancreatic fistula Pancreatic fistula

Thus, STE is an especially good option when the optimal drainage routes into the necrosis are transperitoneal, intercostal, or through very small windows between vital structures. As with VARD, planning for STE should begin at the time of percutaneous drainage in collaboration with interventional radiology. The path of the percutaneous drain should be chosen in a manner that allows access to the entire necrotic cavity, typically by entering at one end. Positioning of the patient on the table can be supine or in partial decubitus position based on the location of the drain. The drain is cut at the skin level, and prepped and draped into the surgical field (Fig. 5). Under fluoroscopic guidance, a guidewire is threaded through the percutaneous drain into the necrotic cavity, and the drain removed over the wire. A nephrostomy balloon dilator catheter is then introduced over the wire until the tip of the balloon is within the cavity, and the balloon inflated with contrast to a pressure of 20 psi. The balloon is kept inflated for 1 minute to allow for tract dilation. Subsequently, a 30Fr working sheath is inserted over the balloon dilator catheter, and a rigid nephroscope then inserted with continuous irrigation of the cavity to facilitate visualization. A grasper is then introduced through the nephroscope and debridement performed in a systematic manner to develop a working space within the cavity. The goal is to debride until a healthy circumference of granulation tissue is identified. However, it is unnecessary and dangerous to debride necrotic tissue that is tightly adherent to the walls of the cavity because of the risk of bleeding. During sinus tract endoscopy, even minor bleeding can be problematic because it impairs visualization in the relatively small working space, especially when swirling in the irrigant. In this circumstance, the bleeding areas can be managed by packing the cavity with a half-­inch gauze packing-­strip through the sheath for a few minutes. This is often enough to allow the procedure to proceed. Advancing the sheath past the point of bleeding to “jail” and tamponade the bleeding tissue can also be effective and allow debridement to continue at a different point in the cavity. An alternative is to stop the continuous irrigation and continue the debridement in a dry fashion, so that the blood can pool on the floor of the cavity instead of swirling in the irrigant (portrayed in the video). If complete debridement cannot be achieved at the time of the first STE; it is always safer to irrigate the cavity and return in 2 to 3 days to repeat the debridement, at which time the necrosis is typically much less adherent. Once the debridement procedure is completed, a new drainage catheter is placed over the wire before it being removed. If a repeat procedure is planned, a separate 7Fr catheter is placed adjacent to the drainage catheter to allow for continuous irrigation, which loosens tightly adherent necrotic tissue within the cavity. The drain is

PA N C R E A S

A

B

C

D

507

E FIG. 4  Operative steps for video-­assisted retroperitoneal debridement of pancreatic necrosis. (A) Axial view of the percutaneous catheter that is used as a guide as the dissection is carried out into the retroperitoneum. (B) The cut down on the percutaneous drain carried out through the fascia with electrocautery. (C) Retractors and a laparoscope are inserted through the incision into the necrosis cavity. (D) A laparoscope is used for better cavity visualization, and laparoscopic graspers are used to manually debride and remove necrotic material. (E) Necrotic specimen at the end of the procedure.

508

Management of Acute Necrotizing Pancreatitis

A

B

C

D

E FIG. 5  Operative steps for sinus tract endoscopic debridement of pancreatic necrosis. (A) An Amplatz wire is inserted through the cut end of the drain into the necrosis cavity. (B) A 30Fr nephrostomy balloon dilator catheter is inserted over the wire until the tip of the balloon is within the cavity of the necrotic tissue. (C) A balloon dilator is inflated with contrast for dilation of the tract. (D) A 30Fr working sheath is inserted over this balloon into the cavity. (E) A rigid nephroscope is advanced into the cavity through the sheath and a grasper is passed through the nephroscope for debridement of the necrosis cavity.

then secured to the skin, concluding the procedure. A full video of the procedure can be found at https://youtu.be/e05-­SVI-­7rA. Aside from versatility in allowing access to almost any intraabdominal collection, the other main advantage of STE is that wound complications are uncommon because the drain entry site is the only incision. However, STE requires familiarization with equipment not frequently used by surgeons, such as fluoroscopically guided dilators and the rigid nephroscope, and the relatively small graspers mean multiple procedures are usually required for full debridement of larger necrotic collections. 

Endoscopic Transluminal Necrosectomy Endoscopic transluminal necrosectomy is typically performed through the gastric wall but can also be done through the duodenal

wall. Endoscopic transgastric necrosectomy is our preferred approach for the majority of cases when there is a clear transgastric window into the necrotic cavity, and when the burden of pancreatic necrosis is in continuity with the posterior stomach. A lumen-­apposing metal stent can be placed to facilitate transluminal drainage after the initial debridement and to simplify reintervention. The primary benefits of this approach are the resultant low rate of new-­onset organ failure, minimal pain, and absence of an external drain, thus obviating any pancreatic fistula. The primary drawbacks are that it can only be applied to necrotic collections that are accessible through the gastrointestinal tract, and that it usually requires multiple re-­interventions (median of three interventions in the Pancreatitis, Endoscopic Transgastric Versus Primary Necrosectomy in Patients With Infected Necrosis trial) and occasionally requires other modalities to complete debridement (20% in the same trial). 

PA N C R E A S

Surgical Transgastric Necrosectomy Operative transgastric necrosectomy can be performed in an open or laparoscopic fashion. It accrues the main benefit of endoscopic transgastric necrosectomy—namely, the lack of external drains or pancreatic fistula—but typically does not require multiple procedures as surgical instrumentation generally allows for a more complete debridement at the initial operation. Additionally, it allows for a simultaneous cholecystectomy to be performed when required to reduce the risk of recurrence in patients with gallstone pancreatitis. By combining transgastric debridement with what essentially is a cystgastrostomy, it also allows for more durable control of high-­grade pancreatic fistulas such as that with a disconnected distal remnant. As with the endoscopic approach, surgical transgastric necrosectomy should be reserved for patients whose main necrotic burden lies in the lesser sac posterior to the stomach. 

Dual-­Modality Drainage Dual-­modality drainage consists of combining endoscopic transgastric drainage with percutaneous drainage. In the initial description, this was performed as drainage alone without necrosectomy by either approach. The result was a low mortality and low rate of external fistula but a long duration of external drainage. Combining this approach with either transgastric or minimally invasive surgical necrosectomy may reduce the overall duration of drainage. We use this combined approach in two scenarios. (1) When planned transgastric endoscopic necrosectomy reveals a large volume of dense necrosis which will require multiple endoscopic necrosectomies, but which could be efficiently and effectively managed with the larger instrumentation available surgically. (2) When a large volume of necrosis tracks away from the stomach inferiorly through the retroperitoneum or down the retrocolic gutters, which can be more efficiently addressed via a surgical retroperitoneal approach, but we wish to incorporate internal drainage to minimize the duration of external drainage and the risk of prolonged pancreatic fistula. Incorporation of internal drainage minimizes the duration of external drainage and the risk of prolonged pancreatic fistula. Essentially, dual-­modality drainage accrues the main benefit of endoscopic drainage (its low rate of external fistula formation and the main benefits of surgical necrosectomy), larger instrumentation that allows more rapid debridement and the ability to access portions of the necrosis that are difficult to reach endoscopically. 

Transperitoneal Necrosectomy Although the benefits of minimally invasive approaches to necrosectomy have been largely adopted by centers managing patients with necrotizing pancreatitis, it is important to remember that open operative debridement still represents a useful tool in appropriately selected patients. Most trials supporting the use of the step-­up approach only enrolled patients with necrosis anatomically amenable to those interventions, and also report an unusually high mortality rate with patients undergoing open debridement (∼40%), which limits the generalizability of the results in the operative group. In these studies, surgical necrosectomy was not used as part of a step-­up approach (there was no preoperative drainage), but using it only if percutaneous drainage failed would likely improve outcomes. A contemporary series from our institution analyzing 68 patients who underwent open necrosectomy demonstrated an in-­hospital mortality rate of 8.8% despite severe disease by prognostic indices on admission. This suggests that, in part owing to advances in critical care and better understanding in the timing and indications for operative intervention, surgical transperitoneal necrosectomy remains a useful tool, especially if the expertise and resources for minimally invasive necrosectomy are not available. We still occasionally use it for patients with widespread necrosis that is not completely accessible endoscopically and percutaneously. Because the necrosis in this scenario is typically multifocal

509

and/or enmeshed between vital structures, it is typically difficult to access laparoscopically, and so most of our surgical necrosectomies are done open.  nn RECURRENCE

PREVENTION

The recurrence prevention strategy for pancreatitis is dependent on its etiology. Almost one-­half of all causes of acute pancreatitis are caused by gallstones, with the true prevalence likely even higher because some pancreatitis with undefined etiology may be due to unrecognized biliary sludge or microlithiasis. The risk of recurrent biliary complications from gallstones after an episode of biliary pancreatitis nears 35%; the most effective mitigation strategy is cholecystectomy, which reduces recurrence rates to 0% to 4%. In mild biliary pancreatitis, same admission cholecystectomy is the standard of care. In cases in which there are peripancreatic fluid collections, current guidelines recommend waiting until peripancreatic collections resolve, or 6 weeks from the onset of pancreatitis to allow the collections to organize. Cholecystectomy can generally be performed safely at the time of laparoscopic or open necrosectomy unless the biliary tree is obscured by the inflammatory mass. When using a minimally invasive step-­up approach to necrosis, we typically wait until the necrosis is fully treated to perform cholecystectomy; however, this delay may expose the patient to the risk of further biliary tract complications. If necessary, choledocholithiasis or cholangitis can be treated endoscopically, and acute cholecystitis can be treated with endoscopic or percutaneous decompression as a bridge to eventual cholecystectomy. In patients who are poor surgical candidates, an endoscopic sphincterotomy may be adequate prophylaxis as it reduces biliary complication rates significantly. An unknown proportion of apparently idiopathic pancreatitis cases may be caused by undetected gallstones or biliary microlithiasis. One randomized trial demonstrated a reduced incidence of recurrent pancreatitis after cholecystectomy in patients who had been evaluated for typical causes of pancreatitis, did not have an etiology identified, and were randomized to cholecystectomy or watchful waiting. The odds ratio for recurrence was 5 in the nonoperative group compared with the cholecystectomy group, and the number needed-­to-­treat was 5 to prevent an episode of recurrent pancreatitis. In patients who initially present with idiopathic pancreatitis, the decision to perform a cholecystectomy should be individualized. We currently offer cholecystectomy to idiopathic pancreatitis patients who are good surgical candidates and avoid it in poor surgical candidates when we believe the risks outweigh the potential benefits. For patients presenting with alcoholic pancreatitis, the root cause needs to be addressed: alcohol dependence. Alcoholic acute pancreatitis is usually a result of 5 to 10 years of overconsumption of alcohol and is rarely from an isolated binge. As such, it is prudent for surgeons to recognize that it may be beyond their means to successfully achieve behavioral change in a deep-­seated habit cultivated over years. Randomized controlled trials have shown that behavioral interventions that include counseling and longitudinal assistance for associated psychosocial issues significantly decrease alcohol consumption and dependency scores. Such patients should be enrolled in institutional programs that provide the necessary support to minimize the risk of future recurrences. Hypertriglyceridemia can typically be successfully managed with diet modification and fibrate therapy. As noted in the section on etiology, hyperparathyroidism is the most common cause of hypercalcemia-­induced pancreatitis and is most commonly treated with parathyroidectomy. Finally, surgeons should be wary of the possibility of a pancreatic neoplasm as a possible etiology, especially in elderly patients or patients with “idiopathic” pancreatitis. Early-­stage pancreatic adenocarcinoma or mucinous neoplasm can obstruct the pancreatic duct and cause pancreatitis. Detecting them in follow-­up imaging may provide an opportunity to intervene and treat at an early stage. 

510

Gallstone Pancreatitis

nn SUMMARY

Suggested Readings

Acute pancreatitis is a common but potentially life-­threatening disease. Imaging is not always necessary to establish a diagnosis and is more helpful in delineating the demarcation of necrosis from healthy tissue later in the disease process. Patients should be expeditiously fluid-­ resuscitated early in the course of disease while remaining conscious of the potential complications of over-­resuscitation. Institution of an oral diet early in the course of mild pancreatitis and enteral tube feeding in severe disease form the foundation of nutritional support, parenteral nutrition being reserved for patients who cannot tolerate enteral feeding. Antibiotics should not be administered prophylactically in the absence of clinical suspicion or evidence for infection. Infected necrosis is the primary indication for mechanical intervention in acute pancreatitis. Infection is usually diagnosed by a combination of radiographic and clinical factors. Intervention, when indicated, should be initiated with a step-­up approach; decisions about the best mode and route for intervention should optimally be made by a multidisciplinary group of surgeons, interventional radiologists experienced with percutaneous drainage, and interventional endoscopists capable of endoscopic transluminal intervention. A variety of techniques are now available for interventional management of necrotizing pancreatitis and its sequelae with the different strengths and weaknesses of the alternatives as described previously. An individualized approach based on patient anatomy, physiology, and preference should be used to choose the best technique in each case to obtain optimal results.

Bakker OJ, van Santvoort HC, van Brunschot S, et al. Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial. JAMA. 2012;307:1053–1061. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg. 2000;232:175–180. Fagenholz PJ. Sinus tract endoscopic debridement of pancreatic necrosis; 2018. https://youtu.be/e05-­SVI-­7rA. Fagenholz PJ. Video assisted retroperitoneal debridement of pancreatic necrosis; 2018. https://youtu.be/9ErPBAnrOAU. Fagenholz PJ, Thabet A, Mueller PR, et al. Combined endoscopic transgastric drainage and video assisted retroperitoneal pancreatic debridement: The best of both worlds for extensive pancreatic necrosis with enteric fistulae. Pancreatology. 2016;16(5):788–790. Madenci AL, Michailidou M, Chiou G, et  al. A contemporary series of patients undergoing open debridement for necrotizing pancreatitis. Am J Surg. 2014;208(3):324–331. Räty S, Pulkkinen J, Nordback I, et al. Can laparoscopic cholecystectomy prevent recurrent idiopathic acute pancreatitis?: a prospective randomized multicenter trial. Ann Surg. 2015;262(5):736–741. van Brunschot S, Hollemans RA, Bakker OJ, et al. Minimally invasive and endoscopic versus open necrosectomy for necrotising pancreatitis: a pooled analysis of individual data for 1980 patients. Gut. 2018;67(4):697–706. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-­up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med. 2010;362:1491–1502. Working Group IAP/APA Acute Pancreatitis Guidelines. IAP/APA evidence-­ based guidelines for the management of acute pancreatitis. Pancreatology. 2013;13(4 suppl 2):e1–e15. Zyromski NJ, Nakeeb A, House MG, Jester AL. Transgastric pancreatic necrosectomy: how I do it. J Gastrointest Surg. 2016;20(2):445–449.

Gallstone Pancreatitis

nn PATHOPHYSIOLOGY

John C. Alverdy, MD, FACS, Richard A. Jacobson, MD, and Fons van den Berg, MD

A

cute pancreatitis (AP) remains a leading cause of emergency department visits, hospital admissions, and healthcare costs worldwide. In the United States, the incidence has risen concordantly with an aging population and the increased prevalence of obesity. AP resulting from gallstones is the most common etiology in the developed world, followed by alcohol ingestion, although locoregional rates vary based on population characteristics. In patients with mild disease, cholecystectomy should be performed during the same admission to prevent complications of recurrent disease. Twenty percent of all patients admitted with a diagnosis of AP develop severe disease for which intensive care therapy and interventional procedures other than cholecystectomy may be appropriate. In this latter population, the surgical community has been slow to fully adopt more minimally invasive techniques that are associated with lower morbidity and improved survival. nn EPIDEMIOLOGY AP accounts for more than 330,000 emergency department visits and contributes to roughly 5400 deaths per year in the United States. More than US$2.5 billion are spent annually on the treatment of AP and its complications. Gallstone pancreatitis (GSP) is the most common etiology, accounting for up to 60% of all cases, followed by alcoholic AP. Risk factors for GSP are patient related, stone related, or anatomic. Factors such as age, female gender, gallstones smaller than 5 mm, biliary sludge, 20 or more gallstones, and a large cystic duct diameter have been demonstrated to influence the incidence, course, and outcome of the disease. 

The molecular pathophysiology of GSP is defined by inappropriate and excessive activation of intrapancreatic proteases that lead to autodigestion of the pancreatic tissue. The subsequent inflammatory response results in local and systemic complications of AP as outlined in the following section. Anatomically, it is recognized that obstructive biliary events incite this cascade of pancreatic inflammation and autodigestion; however, the full process is incompletely understood. Three leading theories include: (1) obstruction of the sphincter of Oddi leading to back-­pressure and stasis in the pancreatic duct; (2) bile reflux into the pancreatic duct; and (3) duodenal contents refluxing into the pancreatic duct. Irrespective of these mechanical events, the burst of protease activation that follows is what damages the pancreatic acini, which in some cases leads to ductal disruption, local inflammation, secondary infection, and all subsequent complications. Infection of a peripancreatic fluid collection or necrotic pancreatic parenchyma is the leading cause of morbidity and mortality in AP. Importantly, these are not primary necrotizing infections that destroy pancreatic tissue, but rather secondary infections by translocated bacteria presumed to originate from the gastrointestinal tract. The morbidity and mortality associated with infected pancreatic necrosis has resulted in several attempts to prevent the translocation of intestinal bacteria with prophylactic intravenous antibiotics, selective digestive decontamination with oral antibiotics, or probiotic regimens. The timing, dose, route, and utility of microbiome-­altering regimens remains controversial in GSP.  nn CLINICAL

PRESENTATION

Patients presenting with AP appear constitutionally similar to those in septic shock. They often appear unwell and complain of acute onset, constant visceral-­ type pain in the epigastrium generally

510

Gallstone Pancreatitis

nn SUMMARY

Suggested Readings

Acute pancreatitis is a common but potentially life-­threatening disease. Imaging is not always necessary to establish a diagnosis and is more helpful in delineating the demarcation of necrosis from healthy tissue later in the disease process. Patients should be expeditiously fluid-­ resuscitated early in the course of disease while remaining conscious of the potential complications of over-­resuscitation. Institution of an oral diet early in the course of mild pancreatitis and enteral tube feeding in severe disease form the foundation of nutritional support, parenteral nutrition being reserved for patients who cannot tolerate enteral feeding. Antibiotics should not be administered prophylactically in the absence of clinical suspicion or evidence for infection. Infected necrosis is the primary indication for mechanical intervention in acute pancreatitis. Infection is usually diagnosed by a combination of radiographic and clinical factors. Intervention, when indicated, should be initiated with a step-­up approach; decisions about the best mode and route for intervention should optimally be made by a multidisciplinary group of surgeons, interventional radiologists experienced with percutaneous drainage, and interventional endoscopists capable of endoscopic transluminal intervention. A variety of techniques are now available for interventional management of necrotizing pancreatitis and its sequelae with the different strengths and weaknesses of the alternatives as described previously. An individualized approach based on patient anatomy, physiology, and preference should be used to choose the best technique in each case to obtain optimal results.

Bakker OJ, van Santvoort HC, van Brunschot S, et al. Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial. JAMA. 2012;307:1053–1061. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg. 2000;232:175–180. Fagenholz PJ. Sinus tract endoscopic debridement of pancreatic necrosis; 2018. https://youtu.be/e05-­SVI-­7rA. Fagenholz PJ. Video assisted retroperitoneal debridement of pancreatic necrosis; 2018. https://youtu.be/9ErPBAnrOAU. Fagenholz PJ, Thabet A, Mueller PR, et al. Combined endoscopic transgastric drainage and video assisted retroperitoneal pancreatic debridement: The best of both worlds for extensive pancreatic necrosis with enteric fistulae. Pancreatology. 2016;16(5):788–790. Madenci AL, Michailidou M, Chiou G, et  al. A contemporary series of patients undergoing open debridement for necrotizing pancreatitis. Am J Surg. 2014;208(3):324–331. Räty S, Pulkkinen J, Nordback I, et al. Can laparoscopic cholecystectomy prevent recurrent idiopathic acute pancreatitis?: a prospective randomized multicenter trial. Ann Surg. 2015;262(5):736–741. van Brunschot S, Hollemans RA, Bakker OJ, et al. Minimally invasive and endoscopic versus open necrosectomy for necrotising pancreatitis: a pooled analysis of individual data for 1980 patients. Gut. 2018;67(4):697–706. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-­up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med. 2010;362:1491–1502. Working Group IAP/APA Acute Pancreatitis Guidelines. IAP/APA evidence-­ based guidelines for the management of acute pancreatitis. Pancreatology. 2013;13(4 suppl 2):e1–e15. Zyromski NJ, Nakeeb A, House MG, Jester AL. Transgastric pancreatic necrosectomy: how I do it. J Gastrointest Surg. 2016;20(2):445–449.

Gallstone Pancreatitis

nn PATHOPHYSIOLOGY

John C. Alverdy, MD, FACS, Richard A. Jacobson, MD, and Fons van den Berg, MD

A

cute pancreatitis (AP) remains a leading cause of emergency department visits, hospital admissions, and healthcare costs worldwide. In the United States, the incidence has risen concordantly with an aging population and the increased prevalence of obesity. AP resulting from gallstones is the most common etiology in the developed world, followed by alcohol ingestion, although locoregional rates vary based on population characteristics. In patients with mild disease, cholecystectomy should be performed during the same admission to prevent complications of recurrent disease. Twenty percent of all patients admitted with a diagnosis of AP develop severe disease for which intensive care therapy and interventional procedures other than cholecystectomy may be appropriate. In this latter population, the surgical community has been slow to fully adopt more minimally invasive techniques that are associated with lower morbidity and improved survival. nn EPIDEMIOLOGY AP accounts for more than 330,000 emergency department visits and contributes to roughly 5400 deaths per year in the United States. More than US$2.5 billion are spent annually on the treatment of AP and its complications. Gallstone pancreatitis (GSP) is the most common etiology, accounting for up to 60% of all cases, followed by alcoholic AP. Risk factors for GSP are patient related, stone related, or anatomic. Factors such as age, female gender, gallstones smaller than 5 mm, biliary sludge, 20 or more gallstones, and a large cystic duct diameter have been demonstrated to influence the incidence, course, and outcome of the disease. 

The molecular pathophysiology of GSP is defined by inappropriate and excessive activation of intrapancreatic proteases that lead to autodigestion of the pancreatic tissue. The subsequent inflammatory response results in local and systemic complications of AP as outlined in the following section. Anatomically, it is recognized that obstructive biliary events incite this cascade of pancreatic inflammation and autodigestion; however, the full process is incompletely understood. Three leading theories include: (1) obstruction of the sphincter of Oddi leading to back-­pressure and stasis in the pancreatic duct; (2) bile reflux into the pancreatic duct; and (3) duodenal contents refluxing into the pancreatic duct. Irrespective of these mechanical events, the burst of protease activation that follows is what damages the pancreatic acini, which in some cases leads to ductal disruption, local inflammation, secondary infection, and all subsequent complications. Infection of a peripancreatic fluid collection or necrotic pancreatic parenchyma is the leading cause of morbidity and mortality in AP. Importantly, these are not primary necrotizing infections that destroy pancreatic tissue, but rather secondary infections by translocated bacteria presumed to originate from the gastrointestinal tract. The morbidity and mortality associated with infected pancreatic necrosis has resulted in several attempts to prevent the translocation of intestinal bacteria with prophylactic intravenous antibiotics, selective digestive decontamination with oral antibiotics, or probiotic regimens. The timing, dose, route, and utility of microbiome-­altering regimens remains controversial in GSP.  nn CLINICAL

PRESENTATION

Patients presenting with AP appear constitutionally similar to those in septic shock. They often appear unwell and complain of acute onset, constant visceral-­ type pain in the epigastrium generally

PA N C R E A S

radiating to the back. Associated symptoms include nausea and vomiting. The pain is exacerbated by oral intake and may be relieved when the patient leans forward. Other diagnoses to be considered are acute cholecystitis, choledocholithiasis, and peptic ulcer disease. On examination, patients can present with tachycardia driven, in part, by the pain associated with AP in mild cases or by severe hypovolemia in more moderate to severe cases. Low-­grade fevers are common and not necessarily indicative of an acute infectious process. Abdominal tenderness can be a prominent finding on physical examination. The presence of bruising in the flank areas, umbilicus, or inguinal regions indicates hemorrhagic pancreatitis that has resulted in dissection of blood along the retroperitoneal planes. The presence of fever, jaundice, or acholic stools may indicate obstructive choledocholithiasis or cholangitis.  nn DIAGNOSIS The diagnosis of GSP is generally made in the emergency setting and is based on clinical, radiographic, and biochemical factors. Blood chemistries suggestive of AP in the setting of gallstones documented by ultrasound usually confirm the diagnosis. The pancreatic enzymes amylase and lipase are usually elevated in patients with AP. Sensitivity and specificity of these assays depends on the specific enzyme measured and the threshold used to define a positive test. Amylase is considered nonspecific, whereas lipase is considered specific for AP, but neither delineates GSP from other etiologies. Most centers use a threshold of positivity defined as three to four times the normal value; however, the duration of symptoms should be considered when interpreting levels of these enzymes. Other circulating factors such as C-­reactive protein and interleukin 6 have been validated as useful markers of the severity of AP, but they do not discriminate AP from other inflammatory or infectious processes. Elevated alanine aminotransferase or lymphocyte/neutrophil ratio may distinguish GSP from other etiologies of AP; however, this distinction is generally based on the absence of a history of alcohol abuse and evidence of obstructing biliary stones or sludge on imaging. Other markers of the acute phase response induced by AP such as interleukin 8 have been validated to rise in the setting of AP but are uncommonly used in clinical practice. Ultrasound is the initial imaging test of choice as it is a noninvasive and highly sensitive test to detect the presence of gallstones. Because patients with GSP may present with a significant ileus, the sensitivity of ultrasound may be decreased by overshadowing of bowel gas. A limitation of ultrasound in this setting, however, is that it cannot assess the severity of pancreatic inflammation and therefore contrast-­enhanced computed tomography (CT) is the test of choice to determine the extent of the AP. Ideally, a pancreas protocol CT with arterial and portal phases is most useful to delineate the degree of inflammation and disruption of the pancreatic parenchyma and main duct. Most authors recommend a CT scan at least 72 hours after the onset of symptoms because radiographic signs of local complications may not be evident before this and therefore are not actionable. On CT, peripancreatic edema (in the anterior pararenal space, transverse mesocolon, or small bowel mesentery), fat stranding, or nonenhancement of the pancreatic parenchyma (indicating necrosis) are indicative of AP.

Severity The severity of GSP is a function of the degree of acute cholecystitis and the degree of pancreatic necrosis and inflammation that is present. Most often, the gallbladder is not inflamed because the disease is more of a consequence of the migration of gallstones into the common bile duction than acute gallbladder inflammation. However, it is important to recognize the severity of both as an important aspect of the treatment strategy. Stratification by severity is of prognostic value and predicts which patients will require intensive care and invasive drainage procedures. The severity of systemic inflammation generally dictates the need for intensive care monitoring and can usually be assessed on admission. Although most patients present with mild

511

TABLE 1  Severity of AP According to 2012 Revision of the Atlanta Criteria Severity

Characteristics

Mild

No organ failure No local or systemic complications

Moderately severe

Transient organ failure ( 4 weeks

< 4 weeks

> 4 weeks

Acute peripancreatic fluid collection

Pancreatic pseudocyst

Acute necrotic collection

Walled off necrosis

PA N C R E A S

nn PATHOPHYSIOLOGY Pancreatic pseudocyst is a well-­formed fluid-­filled cystic mass that typically abuts the pancreas, although it less commonly can be partially or wholly intrapancreatic. The primary etiology of the cyst is disruption of the main pancreatic duct or associated ductal branches, with contained leakage of pancreatic ductal fluid and parenchymal secretions. Therefore, the cyst is typically filled with bicarbonate-­rich and high amylase content fluid with inactivated pancreatic enzymes. Although there is certainly a moderate amount of microscopic debris and inflammatory mediators within the typical PS, there should be little to no macroscopic debris or solid component present. The fluid typically collects in the lesser sac and, over time, develops a well-­formed capsule that can be distinguished from a true cyst by the lack of an endothelial lining. Interestingly, the incidence of PS is higher with alcoholic pancreatitis versus gallstone-­related and other causes. Whether this is due to a true pathophysiologic difference between these types of pancreatitis or to confounding factors including higher severity and recurrence risks with alcohol-­induced pancreatitis is unknown. Unlike necrotizing pancreatitis and WON where the risk of infection is extremely high (30%–70%), infection of a PS is rare. In most cases where infection does occur, it is secondary to percutaneous/ endoscopic instrumentation of the PS or adjacent pancreas, or to the presence of associated pancreatic necrosis that was initially not recognized or that represents progression of disease from IEP to NP. Unlike a noninfected PS, infected pseudocysts are rarely asymptomatic and will typically manifest local signs of inflammation, a relatively rapid increase in the size of the PS, and systemic signs of sepsis if not promptly addressed. Finally, there are relatively rare scenarios where PS may be seen in association with necrotizing pancreatitis, and it is important to be able to clearly differentiate this from walled off necrosis. This usually happens as a delayed complication following the necrosis of a segment of pancreas or following endoscopic/surgical debridement of a segment of the pancreatic body or tail, and with a persistent pancreatic ductal leak leading to PS formation. In these cases, the diagnosis of PS is made based on the absence of any solid or necrotic components of the fluid collection and is purely related to the ductal leak and not active pancreatic necrosis. Among the most challenging of these cases is the patient with a “disconnected duct” syndrome, caused by necrosis of the mid-­body segment of the pancreas resulting in an essentially free floating distal pancreatic body/tail segment and open ductal system that is no longer in continuity.  nn INITIAL

PRESENTATION AND EVALUATION

Clinical Evaluation As described in the preceding sections, the underlying etiology and presenting symptoms or complaints of patients with PS are highly variable. There are essentially no symptoms or examination findings that are specific to PS, and the majority of patients will have either minimal or no symptoms. If symptoms are present, they most commonly will feature vague upper abdominal and/or back pain, abdominal fullness or bloating, early satiety, or pain shortly after meals, or less commonly symptoms of gastric outlet obstruction. More acute presentations featuring concerning abdominal exam findings and abnormal hemodynamics are uncommon, and usually only seen in cases of infection of the PS (pancreatic abscess) or in PS-­related hemorrhage. In many cases in current practice, the diagnosis of a peripancreatic fluid collection is made based on either screening or surveillance radiographic imaging studies. These are often done as part of the workup for an identified episode of severe acute pancreatitis or for persistent symptomatology concerning for a possible PS or other local complication associated with acute or CP, or pancreatic trauma. Elevated serum amylase levels are seen in approximately 50% of patients with PS and are often mistakenly attributed to “recurrent pancreatitis” or to failure of resolution of the index episode.

527

Similar to the physical examination, there are no patient history questions that are high yield or specific for the diagnosis of PS. However, the history can be critical in identifying the likely etiology and in narrowing down the differential of a pancreatic/peripancreatic cystic mass identified on imaging. These questions should focus on the timing of any current or prior episodes of pancreatitis or acute abdominal pain episodes, symptomatology concerning for CP, any prior major abdominal trauma, and then common etiologic factors for pancreatitis including alcohol use, biliary disease, and medications. Patients with a possible PS who have no history of acute/CP episodes or symptoms (or trauma) should prompt consideration of an alternative diagnosis of cystic neoplasm of the pancreas. However, up to 15% of those with PS have no clearly identified antecedent pancreatitis episode or trauma, and a significant percent of pancreatic cystic neoplasms may initially present as an episode of acute pancreatitis. The correct diagnosis can usually be made using a combination of patient history and findings on imaging studies (as outlined in the following section), and less commonly may require percutaneous or endoscopic evaluation with fluid sampling and/or biopsy. Signs and symptoms of pancreatic insufficiency should also be elicited, as this may significantly impact surgical management decisions involving pancreatic resection. 

Diagnostic Imaging for PS Abdominal radiologic imaging with CT and/or MRI has become the standard for evaluating the patient with a known or suspected PS, and for surveilling patients to determine resolution versus persistence of the PS (Fig. 2). In addition to providing critical information about the size, location, and characteristics of the PS, these studies allow for evaluation of the local anatomy to guide any planned intervention, help to differentiate PS from other cystic neoplasms, and can identify any major associated pancreatic abnormalities that may drastically alter the treatment plan. In particular, CT scan is highly accurate for identifying pancreatic necrosis and for differentiating a true PS from necrotizing pancreatitis with walled-­off necrosis. CT scan is also the modality of choice for surveillance of an identified early peripancreatic fluid collection to evaluate for resolution versus progression to a PS, or to assess resolution versus progression in a known PS. It is particularly important for the surgeon to integrate the patient history and carefully review the CT scan characteristics in differentiating true PS from WON or other diagnoses as we continue to see the term pseudocyst erroneously applied as a catch-­all diagnosis in radiologic reports. Although transabdominal ultrasonography has little role in the initial diagnostic imaging for PS, it can be useful for characterization and serial surveillance imaging in patients with a known PS and to avoid the need for multiple repeat CT or MRI scans. The role of MRI with cholangiopancreatography (MRCP) as an alternative or complementary study has also increased over time because of the improved ability to evaluate details of the pancreas and biliopancreatic ductal system compared with CT. This can provide critical information that will significantly alter the management, including signs of CP with ductal strictures and dilation, ductal communication with the PS, pancreatic atrophy or calcification, disconnected duct anatomy, and signs suggestive of cystic neoplasm or malignancy rather than PS. In the past, many surgeons advocated for routine endoscopic retrograde cholangiopancreatography (ERCP) in these patients to assess the ductal anatomy and to demonstrate any patent communication between the main pancreatic duct and the PS cavity. We have found that the detail and reliability of MRCP evaluation of the ductal system has supplanted the need for routine ERCP, and that an MRCP will often be requested by most gastroenterologists before proceeding to ERCP. In cases in which MRCP is not available or indeterminate, or has identified a ductal abnormality requiring further delineation or intervention, the patient should usually proceed to ERCP for definitive delineation of pancreatic ductal anatomy and any communication with the PS. In addition to being diagnostic, ERCP can provide adjunctive or even definite

528

MANAGEMENT OF PANCREATIC PSEUDOCYST

PS

A

B

FIG. 2  (A) Computed tomography scan demonstrates a small pseudocyst in evolution in the mid-­body of the pancreas. (B) Magnetic resonance imaging with T2-­weighted imaging shows round and homogeneously enhancing lesion (arrow) in the distal pancreatic body consistent with a pseudocyst. PS, pseudocyst.

TABLE 1  Comparison of Characteristics of PS, WON, and Cystic Neoplasms of the Pancreas PS

WON

Cystic Neoplasm

History

Acute IEP episode or trauma

Acute NP episode

No prior pancreatitis or acute mild episode

Timing

>4 weeks from IEP event

>4 weeks from NP event

Mature cyst present at initial imaging

Symptoms

Minimal to asymptomatic

Systemically ill/sepsis

Minimal to asymptomatic

CT appearance

Round/oval, well-­formed wall, no debris/septations

Irregular, thick wall, debris, loculations, associated pancreatic necrosis

Irregular, multiple septations, dilated main pancreatic duct, duct wall nodules (IPMN)

Pancreas involvement

Extrapancreatic

Extrapancreatic

Intrapancreatic

Usual location

Body and tail

Body and tail

Pancreatic head (IPMN), body/tail (MC and SC)

Cyst wall

Thick, no epithelialization

Very thick, no epithelialization

Epithelialized cyst wall

Cyst fluid

High amylase, low CEA, no mucin, no epithelial cells

High amylase, low CEA, no mucin, no epithelial cells

Low amylase, high CEA, mucin (IPMN or MC), epithelial cells

Malignant potential

No

No

Yes, highest for IPMN

CEA, carcinoembryonic antigen; IEP, interstitial edematous pancreatitis; IPMN, intraductal pancreatic mucinous neoplasm; MC, mucinous cystic neoplasm; NP, necrotizing pancreatitis; PS, pseudocyst; SC, serous cystic neoplasm; WON, walled-­off necrosis.

therapy via balloon dilation of ductal strictures, stenting across pancreatic ductal defects, or providing internal drainage of the PS by transpapillary stent placement. Endoscopic ultrasound (EUS), either alone or in conjunction with ERCP, is an increasingly used modality that can provide additional detailed pancreatic and ductal anatomy information and ultrasound-­guided cyst fluid sampling or tissue biopsy. This is most commonly indicated when there is concern that the PS may in fact be a pancreatic cystic neoplasm, and in particular with concern for a possible intraductal papillary mucinous neoplasm (IPMN). In this setting, EUS has a reported sensitivity of up to 90% and specificity approaching 100% and can also reliably distinguish benign from malignant lesions. Findings consistent with an IPMN or other cystic neoplasm on ERCP or EUS include mucin extruding from the ampulla, hyperechoic nodules in the duct wall, cyst septations, cyst fluid with mucin or elevated CEA, and epithelial cells on cytology or needle biopsy. 

Differential Diagnosis As should be obvious from the preceding material, one of the most critical components in assessing these lesions and determining the optimal management strategy is to narrow down the differential diagnosis to the correct conclusion. Of utmost importance is not taking a radiologic report stating “pancreatic pseudocyst” at face value and assuming that is the correct diagnosis. Fortunately, the differential is usually limited to only several likely possibilities, and these can readily be discerned by careful review of the history, imaging findings, and select use of additional diagnostic modalities such as ERCP or EUS. The potential diagnoses in these cases usually consists of PS versus walled-­off necrosis or a cystic neoplasm of the pancreas, and Table 1 lists some of the key factors and diagnostic criteria that can be helpful in making the correct diagnosis. However, it must be emphasized that very few of these criteria are absolute, and individual factors should not be considered in isolation but rather in sum total to come to the

PA N C R E A S

529

correct diagnosis. One example of this is the common gestalt that a cystic mass in a patient with a current or recent history of acute pancreatitis is likely to be a PS, but epidemiologic studies have demonstrated that 20% to 40% of cystic neoplasms may initially present as an episode of acute pancreatitis. However, unlike PS, which takes weeks to mature, cystic neoplasms are fully formed with a mature cyst wall on the initial imaging studies done at the time of admission or within the first several days. This should prompt additional diagnostic evaluation to evaluate for an IPMN or other cystic neoplasm, and to ensure that a malignant or premalignant lesion is not written off as a PS. Fig. 3 shows side-­by-­side examples of characteristic cross-­ sectional imaging appearance and features of PS versus WON versus a cystic neoplasm (IPMN). 

Ductal Evaluation A key component of the assessment and deciding on the optimal management strategy for PS is an understanding of the interplay between the pancreatic ductal anatomy or associated anomalies and the likely response of the PS to specific interventions. In discussing this in teaching conferences, a standard assertion is that after a PS is identified on cross-­sectional imaging, an ERCP should be performed to determine “if it communicates with the pancreatic duct” (Fig. 4). However, actual practice is not that straightforward, and has also evolved significantly as noninvasive imaging of the ductal anatomy has improved drastically. By definition, if the lesion is a true PS, then it either has a communication with the pancreatic ductal system, or it had one that has subsequently sealed. In addition to the issue of an active communication with the duct, other ductal abnormalities associated with pancreatitis that can be seen with PS are critical to delineate. These include the disconnected duct syndrome described previously, ductal stricture with obstruction, and irregular dilation or the string of beads ductal appearance consistent with CP. A useful system for categorizing the duct anatomy and anomalies associated with a PS has been described by Nealon and colleagues (Fig. 5). This ductal characterization is not only important for taxonomy, but it also has major implications in selecting the appropriate intervention and in the risk of failure or complications following interventions. The ductal anatomy can hereby be categorized as type I to type IV, and then within each type can be subcategorized (using “a” or “b”) by whether a demonstrable connection with the PS is present or absent.  nn MANAGEMENT AND

A

B

INTERVENTIONS

Arguably, the most important distinction in selecting the optimal management strategy and intervention for PS is the presence of absence of associated symptomatology. We generally concur with the recommendation from the 2013 evidence-­based guideline on acute pancreatitis by the American College of Gastroenterology, stating: “Asymptomatic pseudocysts and pancreatic and/or extrapancreatic necrosis do not warrant intervention regardless of size, location, and/ or extension (moderate recommendation, high quality of evidence).” However, there are exceptions to this rule such as the inability rule out a cystic neoplasm, rapidly enlarging PS, and patients in high-­risk occupations or locations where immediate medical attention will not be readily available. In addition to the lower likelihood of spontaneous resolution, the risk of subsequent complications such as rupture or infection are higher in larger lesions (≥10 cm) and therefore the decision for intervention versus observation should be individualized to each patient. The oft-­quoted “rule of sixes” that intervention is indicated for PS larger than 6 cm and persisting longer than 6 weeks should largely be abandoned in favor of a more individualized and nuanced approach as outlined here. For symptomatic PS or those associated with pancreatic or pancreatic-­ ductal pathology that requires intervention, there are now a wide variety of options that range from minimally invasive endoscopic or percutaneous interventions to major open surgery. Key factors in selecting the optimal intervention include not only

C FIG. 3  Characteristic computed tomography imaging findings for (A) large cystic mass displacing stomach (arrow) with round, regular shape and no solid component consistent with pseudocyst. (B) Irregular cystic fluid collection with debris and air consistent with walled-­off necrosis. (C) Irregular septated cystic mass in head of pancreas consistent with a cystic neoplasm.

530

MANAGEMENT OF PANCREATIC PSEUDOCYST

Type Ia

Type Ib

Type IIa

Type IIb

Type IIIa

Type IIIb

Type IVa

Type IVb

FIG. 4  Endoscopic retrograde cholangiopancreatography in a patient with a pseudocyst demonstrates normal main pancreatic duct (white arrow) with distal communication into pseudocyst cavity (black arrow).

the individual procedure’s efficacy and complication profile, but also the likelihood of recurrence or need for multiple interventions. In addition, many of these options, particularly some of the advanced endoscopic interventions, are technically difficult and require a significant level of advanced skills and comfort by the gastroenterologist or surgeon. We have found that in many review papers or chapters about PS, these options are all laid out as if they are all routinely and equally available, whereas in reality they may be limited to select settings (urban, high-­volume, academic), and either patient transfer or proceeding with an alternative intervention must be decided.

Percutaneous Drainage Image-­guided percutaneous drainage (PD) is among the least invasive of the available options for PS, but it should be used in only very specific scenarios as the primary intervention. Although PD can usually achieve complete cyst drainage/decompression, it is associated with a higher rate of treatment failure and PS recurrence, as well as the complication of a persistent pancreatic fistula if used in the setting of a patent communication between the PS and the pancreatic ductal system. In a large national study of more than 14,000 patients comparing PD with open surgery for PS, PD was independently associated with increased complications, PS recurrence, and mortality. However, this (and most) study suffers from significant selection bias, and it did not have an endoscopic intervention group for comparison. Another concern with performing PD, particularly for minimally symptomatic PS, is bacterial seeding and infection which then requires urgent drainage. Although there is a risk of converting a PS to a pancreatic abscess with any instrumentation or surgical intervention, the risk appears to be higher for percutaneous drainage versus most other endoscopic or surgical drainage procedures. Percutaneous drainage of the PS also does nothing to address any proximal pancreatic duct pathology if present, and therefore we recommend using PD as a primary therapeutic option only in the setting of a PS with a normal pancreatic duct and no communication (type Ia). When used in the setting of type II to IV pancreatic ductal anatomy, the reported failure rates of PD are high (50%–90%). However, PD may be indicated even in the setting of known ductal pathology as a temporizing and/or initial diagnostic maneuver. These scenarios would mainly include patients with severe physiologic disturbance or comorbid disease that are prohibitive for other more invasive interventions, or in patients with an acutely infected PS (pancreatic abscess). PD has also been proposed by some experts to help with the

FIG. 5  Classification of associated pancreatic duct anatomy and anomalies: type I, normal duct; type II, ductal stricture/obstruction; type III, disconnected distal duct; and type IV, irregular duct dilation, consistent with chronic pancreatitis. Subgroups are identified by “a” for no ductal communication with pseudocyst and “b” ductal communication present. (From Nealon WH, et al., A unifying concept: pancreatic ductal anatomy both predicts and determines the major complications resulting from pancreatitis. J Am Coll Surgeons. 2009;208:790-­799.)

diagnosis and delineation of the distal pancreatic duct segment when imaging via standard modalities is not possible (i.e., the disconnected duct). In these cases, imaging of the distal pancreatic duct segment can be done with injection of contrast via the percutaneous drain under fluoroscopy, which can aid with characterization of the disconnected ductal segment and planning for the optimal endoscopic or operative intervention. 

Endoscopic Interventions Endoscopic techniques, equipment, and experience with both diagnostic and therapeutic interventions for PS have advanced dramatically over the past decade, and are now arguably the mainstay for management of most uncomplicated PS. In addition to being less invasive than surgical drainage procedures, endoscopic techniques can be used to simultaneously address both drainage of the PS and many of the associated pancreatic ductal lesions or injuries. Endoscopic drainage procedures for PS can be broadly categorized as either internal (via the pancreatic duct) or transmural (via the abutting stomach or

PA N C R E A S

*

*

*

*

A

531

*

*

B

FIG. 6  (A) Endoscopic cystogastrostomy with transgastric needle aspiration to confirm pseudocyst location, which is visually identified by the bulging of the posterior gastric wall. (B) A covered self-­expanding metal stent is deployed to create and maintain the cystogastrostomy tract.

duodenum). In addition, these procedures can either be performed via standard endoscopy or with the addition of EUS guidance. Unlike percutaneous drainage, the presence of a known ductal communication with the PS (Fig. 4) is not a relative contraindication to internal or transmural endoscopic drainage procedures and can often be simultaneously addressed during the endoscopic drainage procedure. An internal endoscopic PS drainage procedure refers to accessing and transpapillary stenting of the main pancreatic duct to allow internal drainage of the PS fluid back into the duct and duodenum. This obviously is only applicable to PS with a clearly identified and patent communication with the pancreatic ductal system and without proximal obstruction or discontinuity of the duct and is most suitable for the type Ib or IIb duct (Fig. 5). Although this can also be effective for the type IVb duct, there is a higher recurrence rate and likely continued severe symptomatology related to the accompanying CP that should be addressed as outlined below. For the type IIIb duct (disconnected duct), this is usually not a viable alternative because the duct is in discontinuity. There are published descriptions of attempting blind or image guided stent placement across the defect, or alternatively accessing the distal duct segment through the gastric wall under EUS guidance and placing a transgastric stent. However, these advanced endoscopic procedural skillsets are relatively uncommon and it does not address the long-­term problem of the disconnected duct and the isolated distal pancreatic segment. Endoscopic transmural drainage procedures have become increasingly common, and in many centers have largely replaced surgical drainage for uncomplicated PS or in patients who have prohibitive risk factors for surgical intervention. These are done by creating a wide communication between the PS and the lumen of either the stomach (most common) or duodenum (less common). This is typically done in sequence by: (1) endoscopic inspection of the gastric lumen to identify the site of “bulging” of the posterior wall due to extrinsic compression from the PS, (2) needle aspiration to confirm the location (Fig. 6A), (3) creation of a small gastrotomy into the PS via needle-­knife, (4) balloon dilation of the tract, and (5) placement of either multiple small double pigtail stents or a single larger self-­ expanding covered metal stent (Fig. 6B). Tissue from the cyst wall should be collected for pathologic evaluation to definitively rule out a cystic neoplasm or malignancy. In select cases of larger PS or with significant debris present, a nasocystic drain may be left behind for continued cavity irrigation and drainage. Prior or simultaneous ERCP is performed if there is a patent ductal communication or a

proximal pancreatic duct stricture requiring transpapillary stenting or other endoscopic intervention to optimize internal drainage and prevent PS recurrence. These stents or pigtail catheters should be left in place for at least 6 weeks because recurrence has been found to be significantly higher with stent/drain removal before 6 weeks. Repeat imaging should be performed to document successful drainage and resolution of the PS prior to stent or drain removal (Fig. 7). A small randomized trial and multiple nonrandomized series have demonstrated improved outcomes with endoscopic versus surgical PS drainage, although the incidence of requiring additional interventions is higher. In addition, these studies used open surgery as the comparison group, and it is unclear if different results will be seen with the increased utilization of minimally invasive surgical approaches to PS. EUS is an important adjunct to these procedures as it can help exactly localize the PS and guide the drainage procedure, assess for complete evacuation and collapse of the PS, and assess for any additional anatomic abnormalities or signs of an alternative diagnosis. There have now been two randomized trials demonstrating the superiority of EUS-­guided versus standard endoscopic PS drainage, and we recommend routine use of this adjunct if available. This is particularly critical if endoscopic cyst-­duodenostomy is being performed because there will often not be a visible bulge to locate the optimal puncture site and EUS can readily identify critical structures including the pancreatic duct and common bile duct to avoid iatrogenic injury. It is imperative to both the safety and success of endoscopic transmural drainage procedures that enough time has lapsed to allow the PS to mature and have a well-­formed capsule that is adherent to the stomach or duodenum. In addition to simply estimating the time interval, it is critical to review the preoperative cross-­sectional imaging for signs of an adequately thick and well-­formed PS wall, and to assess the location and anatomic relationship of the PS to the adjacent stomach or duodenum. A PS that is not directly adjacent to and abutting the gastric or duodenal wall is generally not a candidate for safe transmural drainage and usually will require a surgical drainage procedure. Optimal endoscopic candidates should also have a PS wall thickness of between 3 and 10 mm, with a common recommendation for surgical drainage in PS with wall thickness greater than 10 mm. 

Surgical PS Drainage Procedures With the marked improvements in endoscopic equipment, techniques, and experience over the past several decades, the need for operative

532

MANAGEMENT OF PANCREATIC PSEUDOCYST H

Spin: -90 Tilt: 0

PS

Spin: -90 Tilt: 0

A

B

FIG. 7  (A) Computed tomography (CT) scan after endoscopic cystogastrostomy shows stent in good position between the stomach and pseudocyst. (B) CT scan 4 weeks later shows adequate drainage with collapse of pseudocyst (PS).

intervention for PS has decreased significantly. However, surgical interventions remain a key and important aspect of PS management and are still frequently required in cases of PS with complex associated ductal or pancreatic pathology. In addition, operative drainage or resectional procedures are still required in cases of failure of percutaneous or endoscopic treatments and may also be used more frequently in settings where there is less access to advanced therapeutic endoscopists. It is not uncommon to see patients who have undergone multiple percutaneous and/or endoscopic attempts at drainage and associated complications prior to any surgical referral. In cases where there has been clear failure or recurrence after an adequate endoscopic drainage procedure, we encourage an initial consultation with an experienced surgeon for input on the pros and cons of additional endoscopic or percutaneous attempts versus prompt surgical intervention. Surgical treatment for PS typically involves creation of an anastomosis between the cyst wall and the gastrointestinal tract to facilitate drainage and decompression of the fluid, any debris, and obliteration of the cavity. The two most common procedures for accomplishing this are cystogastrostomy and cystojejunostomy, and less commonly cystoduodenostomy, which is typically reserved for smaller PS in the head of the pancreas. Although these operations have traditionally been performed as an open procedure, they can now be readily accomplished using minimally invasive techniques such as laparoscopy and more recently robotic-­assisted surgery. Although this discussion is often presented as either endoscopic or surgical therapy, the optimal results are obtained with a combined multimodality approach (if needed) to address both the PS and the underlying etiology, particularly for PS with an identified ductal communication. Most commonly this means combining surgical drainage with endoscopic interventions such as ERCP and pancreatic duct stenting and/or stricture dilation to optimize transpapillary drainage. In highly select cases of large PS that are very well encapsulated and not adherent to surrounding structures, the cyst can be excised en bloc with concomitant ligation or obliteration of any patient fistula with the pancreatic duct. Selection of the optimal drainage procedure should be based on careful review of the anatomic relationships identified on preoperative cross-­sectional imaging, and with consideration of any factors, such as gastric varices or prior gastrointestinal surgery, that would dictate a different surgical plan. Because the majority of PS is limited to the lesser sac and directly abut the posterior wall of the stomach, cystogastrostomy is the simplest and most commonly performed transmural drainage procedure. Advantages of this approach include

minimal manipulation of the gastrointestinal tract, the avoidance of any direct dissection into the lesser sac and pancreas, and shorter operative times. Disadvantages include the usual need for an anterior gastrotomy to access the site and a possibly increased risk of infectious and bleeding complications versus cystojejunostomy, although this continues to be debated in the literature. We have found surgical cystogastrostomy to be a safe and effective intervention, and with a complication profile similar to cystojejunostomy. Cystogastrostomy is most commonly performed using a transgastric approach through the anterior gastric wall (Fig. 8). For open cystogastrostomy, a longitudinal gastrotomy is made in the anterior wall of the mid-­stomach and centered as much as possible over the PS. This incision will be more forgiving if extension proximally and/or distally is needed and can be easily closed in longitudinal or transverse fashion to avoid narrowing the gastric lumen. Stay sutures on each edge of the gastrotomy are helpful to retract the edges and widely expose the posterior wall of the stomach. Inspection typically reveals an obvious bulge at the location of the PS, but if unclear then needle aspiration or intraoperative ultrasound can be performed to confirm the location. Entry into the PS through the posterior stomach wall is then made and the cyst fluid and any debris are suctioned. An adequate anastomosis is then created between the PS wall and the posterior stomach which can either be handsewn or stapled. We prefer a stapled anastomosis for both speed and simplicity. A fragment of the cyst wall should be excised for pathologic analysis, and any remaining fluid or debris is evacuated through the wide anastomosis. The anterior gastrotomy can then be closed with either a running suture or stapled. Less commonly, cystogastrostomy can be performed via an exogastric approach by direct exposure of the PS in the lesser sac and then direct creation of the anastomosis to the posterior wall of the stomach. This is typically reserved for smaller pseudocysts that make transgastric localization more difficult, and that can be readily exposed without risk of cyst rupture or damage to the stomach or pancreas. Minimally invasive cystogastrostomy typically uses the same basic sequence to approach the pseudocyst and create the anastomosis (Fig. 9A). Although either a handsewn or stapled anastomosis (Fig. 9B) can be performed laparoscopically, the stapled approach is much simpler versus laparoscopic intracorporeal suturing for both the anastomosis and the anterior gastrotomy closure. An alternative technique uses a transgastric laparoscopic approach in which the trocars are inserted into the gastric lumen, which is then insufflated to allow a working space to perform the cystogastrostomy. Either balloon-­tipped trocars

PA N C R E A S

533

Stomach Posterior wall gastrotomy

Biopsy of cyst wall

Anterior and posterior wall of stomach Pseudocyst cavity

Pseudocyst cavity Closure of gastrotomy Cyst wall Posterior wall of stomach

FIG. 8  Surgical cystogastrostomy procedure showing anterior and posterior gastrotomies to expose and then enter the pseudocyst (upper left and inset), creation of handsewn cystogastrostomy anastomosis (middle), and closure of anterior gastrotomy (bottom). (From Doane SM, Yeo CJ. In: Yeo J, editor. Shackelford’s Surgery of the Alimentary Tract. Philadelphia: Elsevier; 2019.)

A

B

FIG. 9  Laparoscopic or robotic-­assisted stapled cystogastrostomy is performed by (A) making an anterior vertical gastrotomy to expose the posterior gastric wall and (B) insertion of linear stapler with one end into pseudocyst cavity and the other end in the stomach to create the cystogastrostomy anastomosis.

534

MANAGEMENT OF PANCREATIC PSEUDOCYST

Transverse colon

Pseudocyst Middle colic artery Cystotomy

Biopsy

Outer layer Roux-en-Y jejunal loop

Inner layer of posterior row Inner layer of anterior row Enterotomy

FIG. 10  Surgical cystojejunostomy between a jejunal Roux limb and a large pseudocyst bulging into the transverse colon mesentery (top) with completion of handsewn posterior layers (middle two panels) and then completion of anterior layer (bottom). (From Doane SM, Yeo CJ. In: Yeo J, editor. Shackelford’s Surgery of the Alimentary Tract. Philadelphia: Elsevier; 2019.)

or anterior gastric stay sutures at each trocar site are used to prevent the trocars from dislodging and to maintain gastric insufflation. More recently, these procedures have been described using a robotic surgery platform, which are nearly identical to the laparoscopic approach but with much better camera optics and visualization, and with greater ease of intracorporeal robotic-­assisted suturing. Cystojejunostomy is the second most commonly performed drainage procedure for PS and offers similar efficacy and success rates as cystogastrostomy. This is typically performed using a proximal segment of jejunum that reaches easily to the area of the PS, and then constructed as a Roux-­en-­Y with the cystojejunostomy to the Roux limb and then a jejunojejunostomy to restore continuity (Fig. 10). Cystojejunostomy is clearly indicated for pseudocysts that are not abutting the stomach or duodenum or are located outside of the lesser sac, but can be used for transmural drainage of almost any PS. Advantages over cystogastrostomy include the ability to drain PS in almost any abdominal location and the ability to drain multiple PS, whereas disadvantages include the need for two anastomoses and the more technically demanding nature of the anastomosis. Unlike the cystogastrostomy, it does usually require exposure and visualization of the PS wall for at least enough length to create an anastomosis. Technically, the anastomosis is similar to the cystogastrostomy, and can be stapled (circular or linear) or handsewn. There are some data suggesting that a stapled cystojejunostomy is associated with increased complications (bleeding and infection) compared with a handsewn method, although the series are relatively small and underpowered. The most important factor for technical success with cystojejunostomy is waiting until a thick and well-­formed wall is present that can readily hold sutures and support an anastomosis. If a linear stapled anastomosis is performed, then particular attention should be paid to closure of the common enterotomy to avoid narrowing of the intestinal lumen which can lead to failure to resolve the PS or to an anastomotic leak. 

Surgery for PS With Major Pancreatic or Ductal Pathology A cornerstone of successful management and long-­term resolution for PS is that the evaluation and management strategy must also focus on identifying critical, related pancreatic or ductal disease that requires concomitant intervention. Several series have shown that failure and recurrence rates are high when a drainage procedure alone is performed in patients with major ductal strictures or obstruction, and that continued symptoms are the rule in patients with PS associated with CP. In these cases, we recommend approaching these scenarios with the mindset that the pancreatic/ductal pathology is the primary issue that requires intervention, and the PS is a secondary effect or manifestation that can only be resolved by treating the underlying cause. For patients with disconnected duct syndrome (type III in Fig. 5) associated with mid-­body necrosis and loss of the central segment of the main pancreatic duct, surgical intervention should focus on either restoring drainage to the distal disconnected pancreatic segment or performing operative resection of that segment. Characterization of the distal pancreatic duct is important for preoperative planning, and, often, an MRCP can be adequate. If not, then percutaneous drainage of the PS can be performed as an initial measure to decompress the PS and then to perform a fistulogram via injection of contrast through the drain. Surgical options then would include a PS drainage procedure alone (cystogastrostomy or cyst jejunostomy), restoration of pancreatic duct drainage via a pancreatojejunostomy to the distal segment, or resection via distal pancreatectomy/splenectomy. Intraoperative decompression or excision of the PS is performed, but no formal PS drainage procedure needs to be performed. Deciding among these options is a complex and highly individualized process, and referral or consultation with an experienced pancreatic surgeon is strongly recommended. In addition to the anatomic considerations, a key consideration is the likelihood of postoperative pancreatic insufficiency or

PA N C R E A S

diabetes with resection of the disconnected segment versus preservation and drainage. It is also important to clearly distinguish this patient population from the less common cohort who has a disconnected duct and distal segment resulting from a traumatic injury to the pancreas. These patients usually have an otherwise normal pancreas outside of the zone of injury, may have multiple associated injuries that impact the treatment plan, and will have a different risk/benefit ratio for operative drainage versus resection of the distal pancreatic segment. These operations can be particularly difficult because the frequent presence of significant scarring, inflammation, and distorted anatomy, and may require opening the PS cavity to access the distal pancreatic segment. If drainage via pancreatojejunostomy is selected, this should be done with a Roux-­en-­Y reconstruction and passage of the roux limb through a window in the transverse mesocolon. The type of anastomosis (duct-­to-­mucosa, invaginating, lateral) will be based on the size of the duct, the presence of any strictures, the character of the pancreatic parenchyma (hard vs soft) and individual surgeon preference and experience. An alternative option that has been described is drainage into the stomach via a pancreatogastrostomy. This option has the advantage of obviating the need for any small bowel manipulation and requires only one anastomosis instead of two. However, many surgeons have less familiarity with this type of anastomosis, and the inflammation of the posterior gastric wall from the pancreatitis and the PS often makes this option less feasible. Resection of the disconnected pancreatic segment via distal pancreatectomy is the other viable option. Although at first glance this would seem to be the simpler and superior option versus a complicated drainage procedure and high-­risk pancreatic anastomosis, resection may have significant longer term adverse effects that must be considered. Although distal pancreatectomy is routinely performed for other indications (neoplasm, trauma) without significant concern for a functional impact, this should not be extrapolated to these scenarios. Among patients who have had a significant episode of necrotizing pancreatitis with loss of some or all of the mid-­pancreas (and likely dysfunction of the remaining pancreas), the impact of removing the disconnected distal pancreatic segment must be carefully considered. This can result in significant and lifelong problems, including pancreatic exocrine insufficiency and a high risk of new-­onset diabetes or conversion of non-­insulin-­dependent diabetes to permanent insulin dependence. An additional concern with resection is the need for concomitant splenectomy, and the long-­term risk of postsplenectomy infectious complications. Although spleen-­preserving distal pancreatectomy is routinely performed in other patient populations, in these scenarios, it is often not practical or possible because of the dense inflammatory and fibrotic changes that make it difficult and risky to attempt to separate the pancreas from the splenic artery and particularly the splenic vein. We advocate for resection in patients who have preservation of a significant portion of the proximal pancreas and no existing pancreatic exocrine or endocrine deficiencies, and where intraoperative inspection confirms that a safe mobilization and resection can be performed. Although this option is less ideal for most patients with pancreatitis-­associated ductal loss, we consider distal pancreatectomy with or without splenectomy to be the better option for most patients with disconnected duct anatomy resulting from trauma. The final cohort that must be clearly distinguished and approached differently are those with PS and symptomatic CP with associated irregular ductal dilation (type IV in Fig. 5). Longitudinal studies have demonstrated that spontaneous resolution of a PS in this patient population is significantly lower (4 weeks) AND well-formed PS wall on imaging studies

• CT/MRI with necrosis suggests WON • Systemic signs of sepsis or imaging signs of infection (air, inflammation) • Priority is antibiotics and urgent drainage

• Categorize duct anatomy (type I to IV) • Identify any ductal communication (a or b) • Options: MRCP, ERCP, EUS, & fistulogram

5. Patient status & associated issues

• Systemically ill = infection and/or necrosis • Comorbidities and prior interventions • Varices, splenic vein thrombosis, chronic pancreatitis, pancreatic ascites

6. Optimal intervention

• Percutaneous for duct type Ia or to temporize • Transpapillary drainage if duct communication • Endoscopic vs surgical transmural drainage • Pancreas resect/reconstruction: type III/ IV ducts

7. Monitoring & follow-up

• Watch for postprocedure infection/bleeding • Repeat CT or MRI to assess resolution • Endoscopic: leave stent/pigtails >6 weeks

FIG. 11  Critical information and decision points for the management of pancreatic pseudocyst from initial diagnostic workup through postoperative care and follow-­up. APFC, Acute peripancreatic fluid collection; CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography; EUS, endoscopic ultrasound; MRCP, magnetic resonance cholangiopancreatography; MRI, magnetic resonance imaging; PS, pancreatic pseudocyst; WON, walled-­off necrosis. Teoh AY, Dhir V, Jin ZD, Kida M, Seo DW, Ho KY. Systematic review comparing endoscopic, percutaneous and surgical pancreatic pseudocyst drainage. World J Gastrointest Endosc. 2016;8:310–318. Varadarajulu S, Christein JD, Tamhane A, Drelichman ER, Wilcox CM. Prospective randomized trial comparing EUS and EGD for transmural drainage of pancreatic pseudocysts (with videos). Gastrointest Endosc. 2008;68:1102–1111. Zhao X, Feng T, Ji W. Endoscopic versus surgical treatment for pancreatic pseudocyst. Dig Endosc. 2016;28:83–91.

PA N C R E A S

Pancreatic Ductal Disruptions Leading to Pancreatic Fistula, Pancreatic Ascites, or Pancreatic Pleural Effusions Roxanne L. Massoumi, MD, and O. Joe Hines, MD

nn DEFINITIONS Following is a list of terms commonly associated with pancreatic ductal disruption:   

nn Acute peripancreatic fluid collection: A fluid collection near

nn

nn

nn

nn

nn

nn

nn

the pancreas that occurs within the first 4 weeks after the onset of pancreatitis. This is to be differentiated from a pseudocyst that occurs later. Disconnected pancreatic duct system: A disruption in the main pancreatic duct resulting in a disturbance of the flow of pancreatic juice from the upstream (body/tail) to the downstream (head/ uncinate) portions of the pancreas. This often occurs as a result of pancreatic necrosis and is commonly found in the watershed neck portion of the pancreas where the splenic and pancreaticoduodenal arcades meet (Fig. 1A). When the ductal disruption is complete and at its most extreme, it will be evident on abdominal imaging, such as a computed tomography (CT) scan. Pancreatic ascites: A pancreatic fistula that communicates with the peritoneal cavity, usually beginning as a leak into the lesser sac. When a pseudocyst is also present, it is presumed that the leak is from the pseudocyst. This is typically well-­tolerated unless the pancreatic enzymes are activated, which can happen if the collection is secondarily infected with microorganisms. Pancreatic ductal disruption: Disturbance in any of the pancreatic ducts along the length of the main pancreatic duct or pancreatic duct branches. Disruptions may be demonstrated by imaging or is assumed to be present in the case of amylase rich fluid sampled from a drain or collection. Pancreatic fistula: Loss of pancreatic fluid through an area of ductal disruption out of the pancreas. A pancreatic fistula can develop to the intraperitoneal, retroperitoneal, or thoracic cavities and also to other organs or the skin. Pancreatic necrosis: An area of the pancreas that has lost the blood supply; this is diagnosed by contrast-­enhanced CT scan as a lack of enhancement of a portion of the pancreas. When the necrosis is in the tissue outside of the parenchyma, it is termed peripancreatic necrosis. Clinically, the development of necrosis is important because this indicates a more severe condition with both local and systemic complications. Pancreatic pleural effusion: A leak of pancreatic fluid from a disrupted duct into either one of the pleural spaces. The leak will first communicate with the retroperitoneal cavity and then would connect into the thoracic cavity via the esophageal hiatus or directly through the diaphragm. A ductal leak near the portal vein would likely communicate with the right chest and a leak in the distal duct to the left chest. Pancreatic pseudocyst: A well-­circumscribed intra-­or peripancreatic fluid collection that forms from pancreatitis. A fluid collection is characterized as pseudocyst 4 to 6 weeks after an episode of acute pancreatitis.

537

nn Walled-­ off pancreatic necrosis: A parenchymal or peri-­

pancreatic collection of fluid and necrotic tissue surrounded by a well-­defined inflammatory wall occurring 4 to 8 weeks following pancreatitis. 

nn CLINICAL

PRESENTATION AND DIAGNOSIS

When to Suspect Ductal Disruption Inflammation of the pancreas, whether from pancreatic pathology, pancreatic surgery, or trauma, can lead to a pancreatic ductal disruption. Although a certain amount of hyperamylasemia and/ or abdominal pain can be normal and transient with pancreatitis or after surgery, persistent symptoms, or suspicious laboratory results after 1 week are concerning for a ductal leak. Often, the surgeon will be suspicious of this event after reviewing initial imaging. A certain level of ductal disruption can be expected when a patient is noted to have a peripancreatic fluid collection, abscess, pseudocyst, or pancreatic necrosis. After pancreatic trauma, grades III through V can be expected to develop a pancreatic ductal disruption and leak. High suspicion must also be maintained after a major pancreatic surgery and is a main reason for leaving a drain to control the leak. 

Diagnosing a Ductal Disruption Because pancreatic enzymes can be particularly caustic to the retroperitoneum and peritoneal cavity, prompt diagnosis and treatment of a pancreatic duct disruption is critical. The first step for diagnosis is a CT scan with intravenous contrast; this will often show a peripancreatic fluid collection or, in the case of a chronic condition, a pancreatic duct dilation. The location of a fluid collection may help discern the area of duct disruption. For example, ventral leaks often gather in the lesser sac, dorsal tail leaks into the retroperitoneum near the left pararenal space, and dorsal head leaks into the retroperitoneum near the right pararenal space. Given the enclosed space, these fluid collections are unlikely to cause peritonitis and instead often form pseudocysts. Other useful imaging modalities include endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopancreatography. In the postoperative setting, drains are typically left in the vicinity of the surgery and comparing a drain with serum amylase level can aid in diagnosis. If no drain is present, fluid can be drained percutaneously under imaging guidance. When a fistula is present, the drain amylase is often more than or equal to three times the serum amylase. A disrupted pancreatic duct can also present as a pleural effusion if pancreatic fluid leaks into the thoracic cavity. This can occur in the left chest from a disruption of the pancreatic body or tail and travel through the pleuroperitoneal foramina or hiatus in to the chest. Pancreatic fluid can also enter the right thorax with a disruption in the pancreatic neck or the dorsal pancreatic head and travel behind the hepatoduodenal ligament, also ultimately tracking up through the pleuroperitoneal foramina into the chest. When a leak is near the esophagus, pancreatic fluid can track up into the mediastinum. Studies have shown that patients who undergo early operative intervention tend to have earlier resolution of their pancreaticopleural fistula. Persistent pancreatic fluid leakage requires drainage. Inadequate drainage can lead to pancreatic pseudocyst formation, erosion into surrounding structures such as blood vessels or organs, pancreatic necrosis, or communication with another body compartment such as the mediastinum, pleural cavity, or abdomen. Patients who are left untreated for prolonged periods may develop further pancreatic damage secondary to unresolving inflammation around the pancreas. This could begin a vicious cycle ultimately leading to sepsis and multisystem organ failure. 

538

PANCREATIC DUCTAL DISRUPTIONS

A

P

C

B

D

FIG. 1  (A) This patient, with severe acute necrotizing pancreatitis, developed necrosis and an acute peripancreatic fluid collection resulting in a disconnected pancreatic duct. (B) Initial management was percutaneous drainage, which resulted in near-­complete resolution of the collection but a continued high-­output pancreatic fistula. Endoscopic retrograde cholangiopancreatography revealed a disrupted duct in the neck of the pancreas (C) that was stented, resulting in resolution of the remaining collection and healing of the disruption (D).

nn TREATMENT

Percutaneous and Transgastric Catheter Drainage Options After maintaining a high index of suspicion in at-­risk patients and locating a fistula-­derived fluid collection on imaging, early drainage is imperative. Because the peripancreatic area is often very inflamed, these patients are a high surgical risk, and treatment should be as minimally invasive as possible in the stable patient. Percutaneous drainage is the gold standard for the treatment of fluid collections from pancreatic ductal disruption. Percutaneous drainage is typically performed transabdominally via CT guidance through a window clear of major vessels, intestines, or organs. Drains are usually about 12Fr in diameter and can be upsized as needed; sometimes, multiple catheters are required to control the leakage. The drains are attached to a bag with minimal

suction (such as an accordion bag or bulb) and should be flushed with 10 to 20 mL of normal saline multiple times per day to maintain patency. The size of the drained fluid collection will need to at first be frequently monitored to confirm correct placement and functioning. Repeat CT scans are performed on postprocedure days 3 to 5. Once there is confirmation that the drain is functioning appropriately, this interval can be increased to weekly or as needed for an acute change in the patient’s clinical status. The need to upsize the drain if it is not functioning well or add additional drains is common. If possible, drains should be positioned near the site of the ductal disruption and drainage (and not necessarily in the center of the collection) for maximal efficacy. Drain fluid should be sent for amylase level, white blood cell count, bacterial culture, and gram stain. Once the patient is clinically improved, the drainage is minimal, and/or there is confirmation of fluid collection collapse on imaging, the drain can be removed. Once adequate drainage is

PA N C R E A S

achieved, pancreatic necrosis is typically halted. During this process, secondary fluid collections, pancreatic ascites, or pleural effusions may appear. These secondary collections should be drained as well if clinically significant; however, priority should remain adequate drainage at the leak source. In general, after operating on the pancreas, particularly after a pancreaticoduodenectomy, a drain is left in the operated vicinity in the case of a pancreatic fistula. Recent studies, however, have shown that in patients at a low risk for leak, routine drain placement may be unnecessary. Moreover, most patients who have undergone pancreaticoduodenectomy and develop a pancreatic fistula, can be best managed nonoperatively with percutaneous drainage, as described here. As mentioned previously, to perform a percutaneous drainage, a clear window for access into the cavity must be available. In some instances, this window is not available and if the collection is in close proximity to the stomach, a transgastric approach is preferred. Further advantages of this approach include an internal system of drainage and avoidance of the formation of future pancreatic-­cutaneous fistulas. The drainage process could take between days to months but is preferred to open surgical necrosectomy for most patients. Drainage is minimally invasive and can be completed in the stable patient prophylactically, whereas surgery is typically reserved for patients in which drainage has been inadequate or there is concern for other pathology such as injury or necrosis of other organs. Patients treated with drainage have also shown superior long-­term outcomes and tend to have lower rates of morbidity and mortality. 

Endoscopic Drainage Options Endoscopically, pancreatic fluid collections can be drained via an ERCP-­placed stent into the main pancreatic duct (Fig. 1). Placing this stent will allow pancreatic juice to flow down the path of least resistance and through the proper anterograde direction into the intestines, rather than leaking into the retroperitoneum. When ductal disruption is mid-­main pancreatic duct, the stent can be used to traverse the defect and connect the proximal and distal ends of the severed duct. One possible drawback to this endoscopic approach is the risk of post-­ERCP pancreatitis. In the setting of a pancreatic pseudocyst, endoscope ultrasound-­ guided stenting can provide a route for internal drainage when the area of walled off necrosis is adjacent to either the stomach or the duodenum. This can be performed as either a cyst-­gastrostomy or cyst-­duodenostomy. Necrosectomies can also be performed endoscopically, rather than surgically. Although endoscopic drainage has the advantage of avoiding painful external catheters, it is not without its own set of risks. Internal drainage may introduce foreign intestinal flora into the pancreatic space and transgastric stents can erode into surrounding vessels and organs. There is also the disadvantage of not being able to further characterize the drain fluid by assessing its appearance or sending it for laboratory tests. Procedures combining percutaneous and endoscopic drainage have been described. 

Surgical Management Although the traditional approach to pancreatic necrosis was early surgical necrosectomy, the standard of care is now the minimally invasive step-­up approach. The step-­up approach advocates for repeated percutaneous drainage, followed by video-­assisted retroperitoneal dissection, and with the last option being open necrosectomy. When following the step-­up approach, it was found that approximately 35% of patients were able to avoid open surgery, but mortality was similar between the two treatment methods.

539

Surgical treatment for a pancreatic leak is dependent on the location of the duct disruption. Distal leaks can be controlled with distal pancreatectomy. Repeated drainage before an attempt at surgery, however, is warranted and the distal pancreas may become atrophic and the leak volume may decrease. Although this may be a long process and involve many drainage procedures, it will spare the patient a surgery and preserve some pancreatic endocrine/exocrine function. Proximal duct leaks refractory to drainage would require operative drainage with a Roux-­en-­Y jejunal limb anastomosed to the most proximal aspect of the leak. Often, this involves tracing a drainage catheter, which by definition is in the fistula tract, toward the pancreas. However, failure of both endoscopic and percutaneous drainage in this area is fortunately rare. Retroperitoneal dissection in this setting can be difficult secondary to the dense adhesions in the area and the left-­sided portal hypertension caused by a thrombosed or scarred splenic vein. 

Pancreatic Ascites Pancreatic ascites occurs when pancreatic fluid drains into the abdominal cavity and traditionally was felt to be a surgical emergency requiring wide drainage. This is no longer the case, and it is now largely treated medically with bowel rest, total parenteral nutrition, and anti-­secretagogues such as octreotide. Other anti-­secretagogues that been used include atropine, glucagon, acetazolamide, gabexate mesylate, and ulinastatin. The ductal anatomy is interrogated with imaging to identify the duct leak location, and interventions to control this include percutaneous and endoscopic with drainage and ductal stenting. In most cases of acute pancreatitis, pancreatic ascites resolves with these interventions and over time as the pancreatitis resolves. For patients with chronic pancreatitis, this is less likely because of the scarred nature of the duct and ductal stones obstructing pancreatic juice flow, and may require a surgical drainage procedure Roux-­en-­Y. 

Pancreatic Pleural Effusions Although pancreatic pleural effusions can result from acute pancreatitis, it is more often a complication of chronic pancreatitis with ductal disruption (Fig. 2). Treatment begins with bowel rest, total parenteral nutrition, repeated thoracentesis, tube thoracostomy, anti-­ secretagogues such as octreotide; interventional techniques using ERCP-­aided stenting or nasopancreatic drainage; or surgery typically consisting of excision of the damaged portion of the pancreatic duct and drainage of fluid collections or remaining pancreatic tissue with a Roux-­en-­Y limb of jejunum. Conservative treatment has an historical efficacy of 30% to 60% with a recurrence rate of 15%. Surgical therapy has been typically reserved for patients in whom medical therapy has failed and has a success rate as high as 90%, but with up to 20% rate of recurrence. The operative strategy is highly variable and depends on the ductal anatomy. After reviewing the literature, we found that in patients with complications from pancreatic pleural effusions surgery was successful more often than medical therapy (94% vs 31%). 

Multidisciplinary Team Approach The treatment of pancreatic ductal disruptions can be highly complex and a multidisciplinary team of surgeons, gastroenterologists, interventional radiologists, dieticians, nursing staff, hospitalists, pharmacists, and intensive care staff is critical to the success for the patient. The management is best guided by a weekly conference attended by members of the team where the patient’s case is presented, imaging is reviewed, and a plan formulated. Pancreatic ductal disruption is an uncommon and complex condition requiring particular expertise in order to resolve the complications that accompany this diagnosis.

540

PANCREATIC DUCTAL DISRUPTIONS

A

B

C FIG. 2  Patient with a pancreatic pleural effusion and fistula. (A) Initial chest radiographs demonstrates opacification of the left middle and lower lung zones consisted with moderate sized pleural effusion. (B) After interval drainage of the left pleural effusion, there was a left lower hydropneumothorax with a dense surrounding pleural lining consistent with a trapped left lung. In addition, a curvilinear lucency was seen under the right hemidiaphragm representing intraperitoneal air following a laparotomy to address the fistula. (C) Preoperative computed tomography scans demonstrate evidence of chronic pancreatitis with peripancreatic cystic collections and pseudocysts. One such pseudocyst extends superoposteriorly into the left upper quadrant, where it communicates with the pleural space. A large left pleural effusion is associated with this finding. Other findings include portal vein thrombosis, splenic vein thrombosis, and a 15-­mm splenic artery aneurysm.

Suggested Readings Gluck M, Ross A, Irani S, et  al. Dual modality drainage for symptomatic walled-­off pancreatic necrosis reduces length of hospitalization, radiological procedures, and number of endoscopies compared to standard percutaneous drainage. J Gastrointest Surg. 2012;16(2):248–256. Kazanjian KK, Hines OJ, Eibl G, et al. Management of pancreatic fistulas after pancreaticoduodenectomy: results in 437 consecutive patients. Arch Surg. 2005;140(9):849–854.

King JC, Reber HA, Shiraga S, et al. Pancreatic-­pleural fistula is best managed by early operative intervention. Surgery. 2010;147(1):154–159. McMillan MT, Malleo G, Bassi C, et al. Drain management after pancreatoduodenectomy: reappraisal of a prospective randomized trial using risk stratification. J Am Coll Surg. 2015;221(4):798–809. van Santvoort HC, Besselink MG, Bakker OJ, et  al. A step-­up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med. 2010;362(16):1491–1502.

PA N C R E A S

Management of Chronic Pancreatitis William P. Lancaster, MD, and David B. Adams, MD

C

hronic pancreatitis (CP) is a chronic, progressive, inflammatory disease of the pancreas that is characterized by irreversible destruction of functional pancreatic parenchyma and subsequent fibrosis. In the United States, the estimated incidence is 14 per 100,000 with annual healthcare costs estimated to be greater than $2.6 billion. The predominant clinical feature of CP is abdominal pain, most frequently epigastric often with radiation to the back. The pancreas is the most highly innervated visceral organ and the pain associated with CP is unmatched by other abdominal viscera. Characteristic of the CP pain syndrome is allodynia, hyperalgesia, and centralization of pain. Other common clinical features include signs related to loss of pancreatic parenchyma. Symptoms of exocrine pancreatic insufficiency include bloating, steatorrhea, weight loss or inability to gain weight. Progressive islet cell loss is manifested clinically as diabetes mellitus (type 3c). The exocrine, endocrine, and neurologic (chronic pain) dysfunction associated with CP do not follow a parallel course and each may progress at a different rate. CP is a multifactorial disease with a strong genetic predisposition. Several risk factors known to contribute to the development of CP are summarized conveniently by the TIGAR-­O classification in Box 2. Alcohol remains the most common risk factor for the development of CP in Western countries. However, the North American Pancreatitis Study group has recently reported that excessive alcohol consumption is responsible for only 44.5% of CP cases, with the remaining cases attributable to other causes. It is likely that alcohol plays a lesser causative role in the development of pancreatitis with the actual mechanism being alcohol as an exacerbating factor of pre­existing physiologic or genetic circumstances. Mutations in several genes, including cationic trypsin (PRSS1), pancreatic secretory trypsin inhibitor (SPINK1), and cystic fibrosis transmembrane conductance regulator (CFTR) contribute to the development and irreversible progression of CP. Although the genetic underpinnings of CP are continually and currently being elucidated, hereditary pancreatitis is an entity distinct from genetic pancreatitis with a more aggressive natural history. This disease is responsible for 2% to 3% of cases of CP in the United States and is caused by a germline mutation resulting in gain of function of the PRSS1 gene resulting in constitutively active cationic trypsin and subsequent pancreatic autodigestion and inflammation. Patients experience early-­onset CP, developing symptoms in many cases before the age of 20 years. Autoimmune pancreatitis is a rare cause of CP seen in less than 5% of patients. It is characterized by distinct radiologic and histologic features, most often an inflammatory pancreatic head mass and biopsy showing lymphocytic infiltration without malignant cells. It is treated with corticosteroids with complete resolution in most cases. nn DIAGNOSTIC

EVALUATION

The evaluation of patient with suspected CP begins with a careful history and physical examination. The presenting complaint is most often abdominal pain that is often life-­limiting and without clear etiology. Attention should be paid to the location and character of the pain. Most commonly, patients describe epigastric burning pain that radiates to the middle of the back. The pain is often constant with few identifiable aggravating or alleviating factors. The patient should be questioned regarding dietary and bowel habits. Especially in the later stages of disease, patients will restrict their diet to carbohydrates alone and consume very little protein and fat because these cause

541

their pain to worsen, and, in many cases, patients are malnourished though they are of normal or even overweight. Greasy, foamy, floating, and frequent stools suggest steatorrhea and associated exocrine pancreas insufficiency. Imaging studies are helpful in the diagnosis of CP. The most common imaging findings in the setting of CP were described recently in the NAPS2 study and include pancreatic duct dilation (68%), atrophy (57%), calcifications (55%), pancreatic duct irregularity (51%), and pancreatic pseudocysts (32%). Historically, plain abdominal x-­ray was the initial imaging study of choice for the diagnosis of CP, with the finding of upper abdominal calcifications considered pathognomonic, but this is only demonstrated in a subset of cases and in the setting of advanced disease. Computed tomography (CT) is now the most common initial imaging study and has a sensitivity of 47% to 80% and specificity of 90%. However, pancreatic calcifications seen on CT must be interpreted with caution as approximately 40% of patients will have other pancreatic pathology. CT is also useful for the identification of the complications associated with CP, such as pancreatic pseudocysts, infection, hemorrhage, pseudoaneurysm formation, pancreatic fistula, and biliary or gastrointestinal obstruction. Magnetic resonance cholangiopancreatography produces detailed images of the hepatobiliary and pancreatic systems; in the setting of advanced disease, it has a sensitivity of 75%, but it is of low yield in the early stages of disease, in which the sensitivity is only 25%. Endoscopic imaging techniques have the highest sensitivity and specificity for the diagnosis of CP. There are criteria for establishing the diagnosis for both endoscopic retrograde cholangiopancreatography (Cambridge classification) and endoscopic ultrasonography (Endoscopic Ultrasound-­Rosemont criteria). Both diagnostic systems are based on imaging findings that support the diagnosis of CP such as ductal dilation and calcifications. Endoscopic retrograde cholangiopancreatography (ERCP) is arguably the most sensitive and specific test (ranging from 70% to 100%) for the diagnosis of CP. Endoscopic ultrasound can be a valuable aid in the diagnosis of CP in early-­stage disease before the development of overt anatomic abnormalities. These modalities are of limited utility for the surgeon because surgical intervention is primarily aimed at alleviating the consequences of longstanding disease.  nn MEDICAL

MANAGEMENT

The goals of medical management of CP are to control pain, treat pancreatic exocrine insufficiency, and maintain glucose homeostasis. Alleviation of severe abdominal pain can be treated with non­narcotic analgesics, although this is often unsuccessful and thus opioids are often required to achieve adequate pain control. Helpful adjuncts to opioid therapy are pregabalin and gabapentin and frequently can reduce the total dosage of opioids required. In patients with steatorrhea or inability to maintain weight, pancreatic enzyme replacement therapy should be initiated.

Endoscopic Treatment Endoscopic therapies commonly are used as first-­line interventions in the setting of obstructive CP. The goal is to relieve pressure in the pancreatic duct and to facilitate drainage of pancreatic secretions into the duodenum. ERCP is the mainstay of these treatment modalities because it allows dilation of pancreatic duct strictures, division of the sphincter of Oddi, stent placement, and stone removal. ERCP-­based therapies are most effective in the setting of main pancreatic duct obstruction caused either by strictures or stones. Pancreatic duct strictures can be dilated endoscopically and often require stent placement followed by serial stent exchanges usually every 3 months for a period of 2 years. Pancreatic stent removal is associated with a 30% to 40% risk of restenosis. Endoscopic

542

Management of Chronic Pancreatitis

BOX 1  Indications for Surgery in Chronic Pancreatitis • Pain • Relapsing pancreatitis (inflammatory mass in the head, ­pancreatic ductal strictures) • Complicated pancreatic pseudocyst • Biliary obstruction • Duodenal obstruction • Bleeding pseudoaneurysm • Sinistral portal hypertension with recurrent bleeding • Concern for malignancy   

BOX 2 TIGAR-­O Classification of Etiology of Chronic Pancreatitis Toxic-­Metabolic Alcohol Tobacco Hypercalcemia Chronic renal failure Other toxins  Idiopathic Early onset Late onset Tropical  Genetic PRSS1 (hereditary pancreatitis) CFTR SPINK1 Alpha-­1 antitrypsin deficiency  Autoimmune Recurrent and Severe Acute Pancreatitis Postnecrotic Recurrent acute pancreatitis Ischemic/vascular  Obstructive Pancreas divisum Tumor (intraductal papillary mucinous neoplasm, adenocarcinoma, etc.)   

sphincterotomy is performed commonly during ERCP in the setting of CP because it facilitates pancreatic duct decompression, stent placement and stone extraction. Restenosis occurs in 14% of cases after sphincterotomy. Pancreatic duct stones can be removed endoscopically, but stones larger than 5 mm in diameter often require mechanical or extracorporeal shock wave lithotripsy before extraction. Complications associated with ERCP include bleeding, perforation, and post-­ERCP pancreatitis.  nn SURGICAL

MANAGEMENT

Surgical management of CP is primarily reserved for patients who have failed medical and endoscopic therapy and for patients that have developed adjacent organ anatomic complications related to peripancreatic fibrosis such as biliary obstruction, duodenal obstruction, and splenic vein occlusion with gastric variceal hemorrhage. The goals of surgical therapy are palliation of abdominal pain, improvement in quality of life, and relief of secondary complications (Box 1). To that end, the selection of which operation is appropriate for a given

FIG. 1  Computed tomography scan showing extensive proximal pancreatic duct calcification with a dilated distal duct.

patient is based on the pattern of disease evident on cross-­sectional imaging. Broadly, the different operations for CP can be categorized as drainage procedures, resection procedures, or combination resection and drainage procedures.

Drainage Procedures Longitudinal Pancreatojejunostomy The goal of a drainage procedure is to provide surgical relief of pancreatic ductal obstruction. The mainstay of the management of chronic fibrocalcific pancreatitis is the Puestow procedure, originally described by Puestow and later modified by Partington and Rochelle. The pancreatic duct is opened longitudinally along the anterior aspect of the pancreas and a side-­to-­side pancreatojejunostomy is created to Roux-­en-­Y limb of jejunum. The anastomosis can be created in one layer or two, typically with permanent suture. Lateral pancreatojejunostomy is reserved for patients with a pancreatic duct size of 6 mm or greater and without significant enlargement of the pancreatic head. It is important to clear the duct of all stones and debris to optimize drainage. Pancreatoscopy and electrohydraulic lithotripsy can be a useful adjunct in this regard (Fig. 1). The calcific pancreas with a dilated duct also lends itself to a laparoscopic approach. 

Lateral Pancreatojejunostomy Lateral pancreatojejunostomy results in partial or complete pain relief in up to 90% of patients and is accomplished with acceptable morbidity and mortality. At least 25% of patients will develop insulin dependent diabetes during long-­term follow-­up, and this likely reflects the progressive and irreversible course of the disease. Persistent tobacco smoking and alcohol consumption are significant risk factors for recurrent abdominal pain and surgical failure. 

Combined Resection and Drainage Procedures Frey’s Procedure Combined resection and drainage procedures were developed to enhance drainage of the dominant duct in the head of the pancreas and the ducts to the uncinate process while preserving pancreatic parenchyma and the duodenum. This operation is best suited for patients with pancreatic duct dilation secondary to obstruction with an enlarged pancreatic head. It combines the lateral pancreatojejunostomy with local pancreatic head resection. The advantages of the Frey procedure are that it is not necessary to divide the pancreatic neck, which can be particularly treacherous in the setting of severe

PA N C R E A S

543

chronic inflammation. There are presumed benefits with regard to long-­term metabolic function with parenchymal and duodenal preservation. The gastroduodenal artery is the right-­sided boundary of the pancreatic head resection and often must be ligated when encountered. Local pancreatic head resection is limited to the portion of pancreas that is superficial to the intrapancreatic portion of the common bile duct. If the bile duct is entered, it can be effectively drained into the Roux limb. The bile duct edges can be sutured to the surrounding pancreatic head to prevent late stricture. If stricture does occur, a choledochoduodenostomy can be performed. Care must be taken to avoid injuring the underlying portal vein as substantial hemorrhage can occur. 

Beger’s Procedure The Beger procedure is an operation that combines pancreatic head resection with drainage of the distal pancreatic duct while preserving the duodenum. Along with the Frey procedure, the Beger procedure is referred to as duodenal preserving pancreatic head resection. It is best applied in circumstances where the pancreatic head is enlarged without significant distal pancreatic duct obstruction. In this operation, the pancreatic neck is divided and a subtotal resection of the pancreatic head is performed. A Roux limb is then used for two pancreatic anastomoses, one in a side-­to-­side fashion to the pancreatic head and one in an end-­to-­side fashion to the pancreatic body and tail. The Beger operation affords a greater degree of local pancreatic head resection but has the disadvantages of requiring division of the pancreatic neck and the construction of two pancreatic anastomoses. The Berne modification of the Beger procedure preserves pancreatic parenchyma overlying the SMV and portal vein, and, like the Frey procedure, affords of degree of safety when CP presents as a large inflammatory mass in the head of the pancreas. The downside of the Berne, Beger, and Frey procedures is reoperation for biliary stenosis and pancreatic cancer. Several studies have compared the results of the Frey and Beger operations and the rates of pain relief and endocrine dysfunction are similar. Pain relief can be achieved in up to 94% of patients with either operation and the rate of endocrine dysfunction ranges from 10% to 30%. Rates of exocrine dysfunction are similar. Most of the randomized prospective studies comparing resection and drainage procedures originate from Eastern European experiences where genetic and environmental factors appear to be associated with markedly enlarged pancreatic head masses associated with obstructive complications. In the United States, experience with this scenario is less common and the predominant indication for surgery in the United States is pain, less frequently associated with biliary, duodenal, and splenic vein stenosis. 

Resection Procedures Pancreatoduodenectomy Pancreatoduodenectomy (Whipple operation) involves resection of the pancreatic head, duodenum, and distal bile duct. It is most commonly applied in patients with an enlarged pancreatic head but is also useful in cases of head-­dominant disease with biliary or duodenal obstruction or in situations where malignancy is suspected (Fig. 2). It can be applied for both large and small pancreatic ducts. Classically, the operation involves resection of the duodenum with antrectomy but the pylorus can be preserved in some cases by division of the first portion of the duodenum and creation of a duodenojejunostomy (pylorus-­preserving pancreatoduodenectomy). It is our experience that, in cases where it is technically possible, preservation of the pylorus results in a high rate of delayed gastric emptying and intolerance of oral intake and thus it is our standard practice to divide the stomach proximal to the pylorus (pylorus-­ablating pancreatoduodenectomy). Patients with CP have chronically elevated levels of CCK that leads to impaired gastric emptying, which may be improved with pyloric resection. Opponents of pancreatoduodenectomy argue that it is an overly aggressive operation performed for benign disease and that efforts should be made to preserve pancreatic parenchyma

FIG. 2  Computed tomography scan showing an enlarged pancreatic head with calcification.

to allay future risk of endocrine and exocrine dysfunction. In cases where malignancy is suspected or cannot be excluded, pancreatoduodenectomy should be undertaken. In appropriately selected patients, pancreatoduodenectomy provides improvement in abdominal pain and quality of life. It is theorized that the radical nature of pancreatoduodenectomy would lead to high rates of endocrine and exocrine dysfunction, with some studies reporting rates of 50%. However, in a recent randomized controlled trial comparing pancreatoduodenectomy to duodenal preserving pancreatic head resection, rates of endocrine dysfunction were quite low (0.5 cm and 1 cm but no more than 2 cm

T2

Tumor >2 cm but no more than 4 cm

T3

Tumor >4 cm in greatest dimension

T4

Tumor involves celiac axis, superior mesenteric artery, and/or common hepatic artery

N1

Metastases in 1 to 3 nodes

N2

Metastases in 4 or more nodes

M Category Unchanged STAGE Stage IA

T1

N0

M0

Stage IB

T2

N0

M0

Stage IIA

T3

N0

M0

Stage IIB

T1, T2, T3

N1

M0

Stage III

T1, T2, T3

N2

M0

T4

Any N

M0

Any T

Any N

M1

Stage IV

From Amin MB, Edge S, Greene F, et al., eds. AJCC Cancer Staging Manual. Vol. 8. New York: Springer International; 2017. Courtesy the American College of Surgeons.

trial) or FOLFIRINOX (from the PRODIGE-­24 trial) should be considered first-­line therapy in the adjuvant setting. Recent data from the PREOPANC-­1 trial support a total neoadjuvant therapy approach to systemic therapy, to treat micrometastatic disease, maximize completion of therapy, and test tumor biology. In the metastatic setting, both FOLFIRINOX (from the PRODIGE trial) and gemcitabine with nabpaclitaxel (from the MPACT trial) have been shown to improve survival compared to gemcitabine alone and are considered standard of care. Unlike PDAC, few trials exist to support the multimodality treatment of distal cholangiocarcinoma. Although retrospective studies of adjuvant radiotherapy are heterogeneous and show mixed results, the regimen of gemcitabine and capecitabine chemotherapy followed by capecitabine chemoradiation has recently gained favor based on phase II data suggesting a median survival of 35 months and 2-­year survival of 65% in patients with extrahepatic cholangiocarcinoma and gallbladder carcinoma (from the SWOG S0809 trial). However, current guidelines suggest observation, gemcitabine-­or fluoropyrimidine-­based chemotherapy, or fluoropyrimidine chemoradiation may be considered in the adjuvant setting. In patients with advanced (unresectable and metastatic) biliary tract cancer, the largest phase III trial to date (the ABC-­02 trial) established gemcitabine and cisplatin combination chemotherapy as the standard of care. Additional chemotherapies considered in this setting include oxaliplatin, capecitabine, and 5-­fluorouracil, and radiotherapy (external beam or brachytherapy) may be considered in select cases. Like distal cholangiocarcinoma, randomized data to support management of ampullary and duodenal adenocarcinoma are sparse, aside from the early trials evaluating mixed periampullary cancers discussed earlier. Currently, patients with ampullary adenocarcinoma are treated in accordance with the ESPAC-­4 trial. Although supported by multiple retrospective series, the use

of chemoradiation in ampullary adenocarcinoma is controversial, and little data exist to guide the management of advanced disease. Similarly, data evaluating multimodality therapy in duodenal adenocarcinoma are scant. As a result, patients are often treated by extrapolating the literature of colon cancer, as node-­positive patients and those with advanced disease are offered oxaliplatin and fluoropyrimidine-­based chemotherapies as first-­line systemic therapy.  nn SURGICAL

MANAGEMENT

Pancreaticoduodenectomy PD can be divided broadly into three phases: (1) abdominal exploration to confirm the absence of metastases; (2) resection of the malignancy; and (3) pancreatobiliary and enteric reconstruction. Multiple variations on the procedure exist including standard PD (which includes en bloc distal gastrectomy) versus pylorus-­preserving PD (PPPD) (Fig. 1), antecolic versus retrocolic gastrojejunostomy (GJ), and numerous variants of pancreaticojejunostomy (PJ). However, none of these technical details have been shown to significantly augment postoperative outcomes and largely vary by a surgeon’s institution and training history. What follows here is a generalized description based on commonly performed techniques at The Johns Hopkins Hospital. 

Abdominal Exploration The abdomen is explored through either a limited midline laparotomy or via diagnostic staging laparoscopy depending on the surgeon’s preference. An initial exploration is emphasized. The peritoneal cavity is closely inspected for evidence of tumor spread, including all visceral and parietal peritoneal surfaces, the omentum, and entire bowel. Intraoperative ultrasound enables better

PA N C R E A S

A

FIG. 1  Standard pancreaticoduodenectomy with hemigastrectomy and gastrojejunostomy (A) versus pylorus-­preserving pancreaticoduodenectomy with duodenojejunostomy (B). (From Cameron JL, Sandone C. Atlas of Gastrointestinal Surgery, vol 1, 2nd ed. Shelton, CT: People’s Medical Publishing House; 2007.)

B

characterization of indeterminate hepatic lesions found on preoperative imaging, and biopsy with frozen sectioning allows for definitive evaluation of suspicious tissue prior to committing the patient to PD. 

Open Resection After it is deemed appropriate to proceed with PD, the incision may be extended or if a diagnostic staging laparoscopy was utilized, a laparotomy may then be performed. Many surgeons mobilize the hepatic flexure of the colon to aid the performance of a generous Kocher’s maneuver. Extensive Kocher’s maneuver to the level of the left renal vein allows better exposure posterior to the portal vein (PV) and superior mesenteric vein (SMV). After removing lymphatic tissue around the left renal vein and intraabdominal aorta (station 16), the root of the superior mesenteric artery (SMA) can be exposed. The lesser sac is entered and the transverse mesocolon is separated from the bare area of the duodenum and pancreatic head. The middle colic vein is traced down to its union with the right gastroepiploic vein, which classically forms a common trunk (i.e., gastrocolic trunk or trunk of Henle) draining into the SMV. The gastroepiploic vein is ligated and anterior exposure of the SMV is obtained. Portal lymphadenectomy is performed after dissecting the gallbladder down from the hepatic cystic plate with the goal to skeletonize the three portal tubular structures. The common hepatic duct (CHD) is divided near the cystic duct junction. The PV is skeletonized and followed down to the pancreatic neck. Care is taken to identify and preserve any variant hepatic arterial anatomy, most commonly a replaced right hepatic artery, which can be found in up to 15% of patients. After identifying the origin of the gastroduodenal artery (GDA) from the common hepatic artery (CHA), the GDA is test-­clamped to ensure the patient has adequate flow in the proper hepatic artery (PHA). Some patients with celiac stenosis become reliant on retrograde flow from the GDA to supply the liver, therefore ligating the GDA in these patients can be catastrophic. Celiac stenosis identified from preoperative imaging should be addressed prior to proceeding with the PD. After GDA test-­clamping, the GDA is doubly ligated with silk tie

547

and suture ligated with Prolene and divided. We add a surgical clip to the GDA stump to aid in angiographic identification should a GDA pseudoaneurysm develop postoperatively. In a standard PD, the stomach is divided proximal to the antrum with a GI or Endo-­ GIA stapler. The jejunum is divided with another firing of the stapler 20 cm distal to the ligament of Treitz (Treitz). Dissection of the small bowel mesentery is carried proximally until the Treitz is divided and the distal duodenum and proximal jejunum can be passed underneath the superior mesenteric vessels into the previous plane of dissection on the patient’s right side. At this point, a tunnel between the SMV­PV confluence and the pancreatic neck is carefully made from above and below. The pancreatic neck is divided with attention to protect the underlying PV and SMV. The pancreatic neck and head can now be carefully dissected off of the PV-­SMV junction, taking care to identify and control major venous tributaries that include the vein of Belcher (superior pancreaticoduodenal vein) and the first jejunal branch of the SMV. The PV and SMV are retracted to the patient’s left to expose the uncinate process and the right side of the SMA. The inferior pancreaticoduodenal artery should be ligated during the separation of uncinate from the SMA. The surgical specimen should now be free and sent to pathology for evaluation. We routinely check intraoperative frozen sections of the pancreatic neck and uncinate margins, and the CBD margin if applicable. 

Open Reconstruction Sequential reconstruction includes PJ, hepaticojejunostomy (HJ), and GJ for standard PD or duodenojejunostomy for PPPD. For construction of a retrocolic PJ/HJ, the jejunal limb is passed through the transverse mesocolon to the right of the middle colic vessels. The GJ is typically created in an antecolic fashion via a Hofmeister or Reichel-­Polya anastomosis. As various elements of the PD have evolved and improved over the past several decades, the proclivity for the PJ to leak remains the operation’s Achilles’ heel. A variety of techniques have been developed in an attempt reduce the incidence of PJ leak, none of which have been shown to be superior. These include the broad categories

548

Management of Periampullary Cancer

of invagination versus duct-­to-­mucosa technique. We commonly perform an end-­to-­side two-­layer duct-­to-­mucosa PJ consisting of an interrupted, nonabsorbable outer layer and interrupted absorbable inner layer. We typically use a 3-­8 Fr pediatric feeding tube as a stent across the PJ. The HJ is created 5-­10 cm distal on the antimesenteric jejunal limb in a position that prevents undue tension or twisting on either anastomosis. This is performed in a single-­layer interrupted fashion with absorbable suture. Finally, the GJ is created in an antecolic fashion using either a two-­layer handsewn technique with absorbable suture, or with a stapler near the greater curve of the stomach with the common enterotomy closed with two layers of absorbable suture. Two soft, closed-­suction drains are placed anterior and posterior to the HJ and PJ in most cases to monitor for and drain a postoperative pancreatic fistula (POPF). Recent multi­institutional data suggest that drains can safely be precluded altogether in patients at low risk for POPF as determined by their Fistula Risk Score. If a drain is placed, early removal based on drain amylase levels has been associated with lower clinically significant POPF rates as well as shorter hospital stays. 

Major Vascular Resections Venous Resections Resection of the portomesenteric confluence is performed when it is not technically feasible to dissect the pancreatic tumor off the underlying veins without leaving gross disease behind or if dissection cannot be performed safely. If venous resection is to be undertaken, vascular control proximal and distal to the PV-­SMV confluence must be obtained. The type of venous reconstruction is based on the longitudinal and circumferential degree of vein involvement. When a limited aspect (less than one-­third circumference) of the vein requires resection, tangential resection with primary closure or patch venoplasty is sufficient. More extensive involvement generally requires segmental resection with either a primary anastomosis or interposition graft repair. Defects of up to 5 cm can be anastomosed primarily with appropriate mobilization (Fig. 2). If the vein ends cannot be reapproximated without undue tension, we prefer using an autologous vein graft (e.g., left renal, saphenous, jugular veins) over synthetic (e.g., polytetrafluoroethylene, Gore-­Tex) grafts owing to the potential for infection and reduced long-­term patency rates. Finally, in selected patients with preoperative venous occlusion in whom adequate portomesenteric collateralization has developed, SMV resection without reconstruction can be performed safely. However, this approach should be reserved for high volume centers with experience in this technique. 

Arterial Resections Major arterial resections for PDAC remain controversial. Poor overall survival combined with significant morbidity associated with arterial reconstructions limit the potential benefit of these procedures. Isolated CHA resection can be amenable to primary anastomosis. SMA resection with autologous vein or arterial transposition can be performed in highly selected cases. We do not recommend routine use of SMA resection, as the aggressive biology of PDAC often supersedes the survival benefit of these radical resections. 

Minimally Invasive Approaches PD via minimally invasive surgical (MIS) approaches has gained popularity over the last decade. Included among these are laparoscopic PD, robotic PD (RPD), and combined modalities. Multiple reports have demonstrated oncologic equivalence when comparing MIS and open PD, and associate the MIS approach with decreased pain, shorter length of stay, and fewer wound

FIG. 2  Resectable tumor involvement at the portal vein–superior mesenteric vein (PV-­SMV) confluence. This is considered resectable because invasion is limited to a single target vessel above and below the necessary region of resection (black lines) and does not extend too high on the PV. The vein segment and mass are resected en bloc, followed by a primary end-­to-­end reconstruction (inset). Ligation and division of the splenic vein, even absent direct involvement, is often necessary to mobilize the PV and SMV sufficiently for primary anastomosis. (Courtesy Corrine Sandone, copyright Johns Hopkins University.)

complications. Although no study has clearly demonstrated an oncologic benefit with MIS PD, many proponents believe it has the potential to decrease postoperative complications that may delay or even prevent delivery of adjuvant therapy. A key for successful execution of MIS is careful selection of anatomically and pathologically favorable candidate patients, especially for surgeons new to the approach. Although multiple techniques have been reported, what follows is a description of the approach to RPD at The Johns Hopkins Hospital.

Robotic Pancreaticoduodenectomy Setup and Resection The supine split-­leg position is convenient for the bedside assistant. After establishing pneumoperitoneum, we insert a 12-­mm trocar in the umbilical position and utilize the robotic camera for a diagnostic staging laparoscopy to evaluate for both metastatic disease, and anatomic features (e.g., adhesions) that may prohibit successful RPD. If deemed appropriate to proceed, four robotic trocars (8 mm) are placed transversely across the abdomen at the midpoint of the anterior superior iliac spine and the 12th rib, and the da Vinci Xi system (Intuitive) is docked from the patient’s left side (Fig. 3). The ligamentum teres and falciform ligament are divided and secured with an endoloop that is then passed transabdominally to retract the liver anteriorly and cranially. The lesser sac is entered by dividing the gastrocolic ligament and the right gastroepiploic artery is controlled. After controlling the right gastric artery, an Endo-­GIA stapler is used to divide the stomach approximately 5 cm proximal to the pylorus. The portal triad is skeletonized to expose the CHD, CBD, PHA, and PV. The CHD is transected with scissors and controlled with a clipped vessel loop to limit bile leakage for the remainder of the resection. The PHA is traced proximally to identify the right gastric

PA N C R E A S

549

aorta are exposed from the patient’s right side. However, in contrast to open PD, the Treitz can often be divided from the right side of the superior mesenteric vessels and the first portion of the jejunum is pulled through the native duodenal tunnel prior to division with the Endo-­GIA stapler. Transection of the uncinate along the right side of the SMA is the most challenging aspect of RPD. A vessel sealer can be used to separate the uncinate from the nearby SMA. However, in the setting of tumor abutting SMA, we will often use a combination of hook cautery, harmonic scalpel, and a bipolar Maryland clamp to more finely dissect the tissue off the SMA. Finally, the gallbladder is taken down from the cystic plate. The Whipple specimen and the gallbladder are removed from a Pfannenstiel’s incision using a large Endo-­Catch (Medtronic) bag. 

Robotic Pancreaticoduodenectomy Reconstruction

FIG. 3 Trochar placement for robotic pancreaticoduodenectomy. (Modified from Galvez D, et al. Technical considerations for the fully robotic pancreaticoduodenectomy. J Vis Surg. 2017;3:81.)

Neck of pancreas

After hemostasis is confirmed, the anastomoses are performed in the same order as open PD: PJ, followed by HJ, and finally the GJ. We favor a retrocolic end-­to-­side two-­layer PJ, with the posterior row comprised of a 3-­0 running V-­Loc suture (Covidien), 5-­0 interrupted PDS for the inner duct-­to-­mucosa layer, and the anterior row comprised of an additional 3-­0 running V-­Loc suture. As in open PD, we typically stent the PJ with a pediatric feeding tube. The HJ is created 5 to 10 cm distal to the PJ using 5-­0 PDS suture in an interrupted fashion. The anastomosis typically requires 10-­12 interrupted sutures depending on the size of the bile duct. Finally, we create an antecolic side-­to-­side isoperistaltic GJ with a blue-­load Endo-­GIA stapler, closing the enterotomy with a 3-­0 V-­Loc suture in two layers. We routinely use the existing lateral robotic port sites to place two 19Fr closed-­suction Blake drains anterior and posterior to the PJ.  nn POSTOPERATIVE

SMV

Head of pancreas

Duodenum FIG. 4  Intraoperative view of the trunk of Henle (arrow). SMV, superior mesenteric vein.

artery, GDA, and CHA. The right gastric artery is controlled with a clip and the GDA is test-­clamped in a similar fashion to the open technique. The GDA is triply ligated with a 2-­0 silk tie and clips and divided with the robotic scissors. The PV is identified at the superior border of the pancreas, taking care not to divide the superior pancreaticoduodenal vein on the lateral side of the PV. The SMV is identified at the inferior border of the pancreas, taking care to identify the gastrocolic trunk of Henle. The trunk of Henle is dissected out sharply and the right gastroepiploic vein is ligated with clips and divided preserving the middle colic vein if technically feasible (Fig. 4). A retropancreatic tunnel on top of the PV is carefully created with blunt dissection. The neck of pancreas is divided with electrocautery or harmonic scalpel. A Kocher maneuver is performed using a combination of electrocautery and the surgical energy. The inferior vena cava and abdominal

CARE

Over the past decade, enhanced recovery after surgery (ERAS) protocols have introduced evidence-­based pathways to standardize and improve the postoperative care of patients undergoing PD. Although specific ERAS guidelines vary by institution, the ERAS Society has published PD-­specific recommendations based on available evidence. These include preoperative smoking cessation, avoidance of hyperglycemia, an attempt at near-­zero fluid balance, early drain removal, early feeding, and early/scheduled postoperative mobilization (Table 2). ERAS programs have repeatedly demonstrated an ability to decrease the length of hospital stay, however, high-quality studies showing their effects on other important endpoints are limited. These serve as general guidelines to help protocolize postoperative care systems, and should not take precedence over a clinician’s clinical judgment while caring for individual patients.  nn POSTOPERATIVE

COMPLICATIONS

As PD entered the second decade of the twenty-­first century, mortality continues to decrease at high volume centers. However, morbidity associated with the procedure remains high even in experienced hands. The most common postoperative complications are delayed gastric emptying (DGE, 14% to 45%), POPF (7% to 25%), and wound infection (10% to 40%). Postpancreatectomy hemorrhage (PPH) is another important complication that occurs less frequently, however, has potentially devastating and life-­threatening consequences if not managed appropriately.

Delayed Gastric Emptying DGE is a functional gastroparesis commonly occurring after PD. The International Study Group of Pancreatic Surgery (ISGPS) definition subcategorizes DGE into three classifications based on the severity

550

Management of Periampullary Cancer

TABLE 2  Enhanced Recovery After Surgery Society Guidelines for Perioperative Care for Pancreaticoduodenectomy Item

Summary and Recommendations

Evidence Level

Recommendation Grade

Preoperative counselling

Patients should receive dedicated preoperative counselling routinely.

Low

Strong

Perioperative biliary drainage

Preoperative endoscopic biliary drainage should not be undertaken routinely in patients with a serum bilirubin concentration POD 7

14

+

+

C

>14 days or reinsertion >POD 14

21

+

+

Definition: Functional gastroparesis after surgery without mechanical obstruction as determined by upper gastrointestinal contrast series or endoscopic evaluation. Modified from Wente MN, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142:761-­768. POD, postoperative day.

BOX 1  2016 International Study Group of Pancreatic Surgery updated criteria for ­postoperative pancreatic fistula Amylase >3 times upper limit institutional normal serum amylase value ↓ Biochemical leak ↓ Persistent drainage >3 weeks*; clinically relevant change in management of postoperative pancreatic fistula; percutaneous or endoscopic drainage*; angiographic procedures for bleeding*; signs of infection without organ failure* ↓ Grade B pancreatic fistula ↓ Reoperation,a organ failure,a deatha ↓ Grade C pancreatic fistula  

Modified from Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;161:584-­591. aTreatment/event postoperative pancreatic fistula related.

of symptoms and the intensity of required interventions (Table 3). The pathophysiology of DGE is not fully understood but is likely multifactorial and modulated by both technical factors and concurrent patient morbidity. The management is primarily supportive and involves ruling out other causes of oral intolerance with fluoroscopic and cross-­sectional imaging. A 2016 Cochrane review found no difference in DGE rates when comparing standard PD versus PPPD, and antecolic versus retrocolic GJ. When compared to end-­to-­side GJ, several studies have demonstrated lower rates of DGE when performing side-­to-­side GJ, however, both are commonly performed at high volume centers. 

Pancreatic Fistula POPF remains one of the most recalcitrant complications after all types of pancreatic surgery. Its severity can range from an asymptomatic laboratory finding to a life-­threatening systemic condition (Box 1). Multiple technical adaptations to the PJ have been attempted over the years, including invagination, duct-­to-­mucosa anastomoses, and multiple layer closures, none of which have demonstrated a difference in rates of POPF. Stenting across the PJ has also failed to show any effect on POPF rates in multiple studies, but is still commonly performed. Over time, the most reliable predictors of POPF development have been patient-­and procedural-­related factors including gland texture, duct diameter, intraoperative blood loss, and underlying pancreatic pathology. These factors comprise a commonly used Fistula Risk Score calculator, which some surgeons use to determine the need for postoperative perianastomotic external drainage. Conservative management results in spontaneous fistula closure in up to 90% of cases, usually within 4 weeks. Parenteral nutrition and octreotide analogues may also play a role in reducing fistula output. On rare occasions, patients with severe clinical instability or signs of sepsis and organ dysfunction may require surgical reexploration and repair or revision of the PJ anastomosis. 

Postpancreatecotomy Hemorrhage PPH is potentially one of the most lethal complications following PD, particularly if not diagnosed and treated expeditiously (Table 4). Early PPH is most commonly the result of inadequate surgical hemostasis and is best treated with a return trip to the operating room. Late PPH often results from inflammatory process (e.g., intestinal ulceration, POPF) leading to arterial pseudoaneurysm formation, rupture, and hemorrhage. GDA pseudoaneurysm rupture is classically seen 5 or more days postoperatively and can be proceeded by a so-­called herald bleed. Interestingly, the presenting location of the hemorrhage can be either an intraluminal or intra­abdominal, often obfuscating the diagnosis. In either case, pseudoaneurysmal bleeding is best treated with endovascular coil embolization or covered stenting, depending on the precise location and anatomy. 

552

Management of Periampullary Cancer

TABLE 4  International Study Group of Pancreatic Surgery Consensus Definitions of Postpancreatectomy Hemorrhage A. DEFINITIONS Time of Onset Early hemorrhage (≤ 24 hours after the end of the index operation) Late hemorrhage (> 24 hours after the end of the index operation) Location Intraluminal (anastomotic suture lines, cut surface of the pancreas, stress ulceration, pseudoaneurysm) Extraluminal (arterial or venous vessels, diffuse bleeding from resection area, anastomotic suture lines, pseudoaneurysm) Severity of Hemorrhage Mild Decrease in hemoglobin concentration 3 units packed cells Need for invasive treatment (interventional angiographic embolization or relaparotomy) B.  GRADING SCALE Grade

Onset, Severity, and Location

Clinical Condition

A

Early, mild, intraluminal or extraluminal bleeding

Good

B

Early, severe, intraluminal or extraluminal bleeding Late, mild, intraluminal or extraluminal bleeding

Good to moderately impaired

C

Late, severe, intraluminal or extraluminal bleeding

Severely impaired, life-­threatening

Modified from Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al. Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery. 2007;142:20-­25.

nn SUMMARY Periampullary cancer is a heterogenous group of malignancies presenting unique challenges in its diagnosis and management. Modern therapies require a multidisciplinary treatment team and multimodality therapy. PD is the surgical treatment for these cancers and prognosis is overwhelmingly determined by a patient’s disease biology. Advances in MIS techniques and new multimodal therapies offer the prospect for improved outcomes in patients with these difficult diseases.

Suggested Readings Bassi C, et al. The 2016 update of the international study group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;3:584–591. Cameron J, Sandone C. Pancreaticoduodenectomy (pylorus-­ preserving Whipple procedure). Atlas of Gastrointestinal Surgery. 2nd ed. Vol 1. Hamilton, ON: BC Decker Inc; 2007:284–305.

Giuliano K, et  al. Technical aspects of pancreaticoduodenectomy and their outcomes. Chin Clin Oncol. 2017;6:64. Griffin JF, Poruk KE, Wolfgang CL. Pancreatic cancer surgery: past, present, and future. Chin J Cancer Res. 2015;27:332–348. He J, Ahuja N, Makary MA, Cameron JL, Eckhauser FE, Choti MA, et al. 2564 resected periampullary adenocarcinomas at a single institution: trends over three decades. HPB (Oxford). 2014;16:83–90. Torphy RJ, et al. Comparing short-­term and oncologic outcomes of minimally invasive versus open pancreaticoduodenectomy across low and high volume centers. Annals of Surgery. 2018 (epub ahead of print). Wente MN, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142:761–768. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al. Post-­ pancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery. 2007;142:20–25.

PA N C R E A S

Vascular Reconstruction During the Whipple Operation Ammar A. Javed, MD, and Christopher L. Wolfgang, MD, PhD

T

he development of multidrug regimens such as gemcitabine/ nab-­paclitaxel or the FOLFIRINOX regimen for use in patients undergoing oncologic resection of localized pancreatic cancer (pancreatic ductal adenocarcinoma [PDAC]) has resulted in improved systemic control. Over this same time, improvements in radiation delivery in the form of stereotactic body radiation have improved the ability to achieve an R0 resection in patients with major vessel involvement (stage III). Taken together, the improved systemic control and ability to achieve clear margins had led to more aggressive surgical approaches to PDAC. The cancer-­specific outcomes with these more complex operations are superior to what was reported historically with less advanced tumors. Moreover, these operations can be performed with low mortality, acceptable morbidity, and a high likelihood of returning to a good quality of life. Resection and reconstruction of the portovenous axis with a Whipple operation has become commonplace. The resection and reconstruction of major arteries such as common hepatic artery and superior mesenteric artery (SMA) remains controversial and is performed infrequently currently. The exception to this is the resection of cancers of the body of the pancreas that involve the celiac axis and are resected through a distal pancreatectomy/splenectomy with en bloc resection of the celiac trunk. This chapter discusses what determines resectability in patients with stage III (major vessel involvement) PDAC of the head of the pancreas, as well as the pertinent surgical anatomy and surgical technique. nn PREOPERATIVE

EVALUATION AND DETERMINING RESECTABILITY

A high-­quality pancreas protocol computed tomography (CT) is the single most important test in determining resectability as it is very accurate in assessing the local tumor relationships with the major vessels. A pancreas protocol CT consists of a dual-­phase CT with arterial and venous phase imaging. Arterial phase images are acquired at 25 to 30 seconds after intravenous (IV) contrast injection, and venous phase images are acquired 60 to 70 seconds postinjection. The contrast is infused through a peripheral IV at 4 to 5 mL/s and water (1000 mL) is used as an oral contrast agent. All images are reviewed in axial, multiplanar (coronal and sagittal planes), and three-­dimensional volume rendering and maximum intensity projection. Recent improvement in imaging and the introduction of cinematic rendering (Fig. 1), has significantly improved the ability to appreciate the extent of vascular involvement, and its use is recommended for the evaluation of these patients at all centers where this modality is available. This should not be accomplished by routine exploration of stage III disease. Fig. 1 shows examples of each. Localized pancreatic cancer that involves the main arteries of the upper abdomen (SMA, common hepatic artery, or celiac trunk) or the portovenous axis is classified as stage III. The degree of vascular involvement is determined and can be classified according to the classification system proposed by Ishikawa (Fig. 2). Two joint consensus statement divided stage III pancreatic cancer into either borderline resectable or locally advanced based on the extent of vessel involvement. The details of this staging system are beyond the scope of this

553

chapter; however, controversy exists among high-­volume surgeons in regard to what defines resectable and on whom should be explored. In general, borderline resectable consists of patients whom have portovenous involvement that is technically reconstructible or no more than abutment of a major artery, while locally advanced patients have portovenous involvement that cannot be reconstructed or complete encasement of a major artery. Patients with borderline resectable tumors have a high probability of an oncologic resection and should undergo exploration. Certainly, a significant percentage of patients classified as locally advanced and “unresectable” can actually undergo an oncologic resection. Determination of which patients is this category should undergo exploration is much more nuanced than borderline resectable and beyond the scope of this chapter. PDAC is a highly systemic disease marked by early dissemination and most often presenting as metastatic disease. Even in the minority of patients who are diagnosed with localized and potential curable cancers, the predominant pattern of failure following resection is systemic. Thus, to cure localized pancreatic cancer two battles need to be won, a local battle fought with surgical resection and in some cases with the addition of radiation, and a systemic battle fought with chemotherapy and/or biological therapy. All stage III patients should undergo preoperative therapy (neoadjuvant in borderline resectable and induction in locally advanced) before exploration.  nn SURGICAL ANATOMY

OF THE PORTAL VEIN AND SUPERIOR MESENTERIC VEIN

The confluence of the superior mesenteric, splenic, and portal veins is located posterior to the pancreatic neck. The main trunk of the superior mesenteric vein (SMV) is constituted from the jejunal and ileal branches that merge caudal to the pancreas. This main trunk drains the gastroepiploic and middle colic veins. The inferior mesenteric vessel can either drain into the splenic vein before its confluence with the portal vein (PV)/SMV or may directly drain into the SMV (Fig. 3). The first jejunal branch of the SMV is frequently located posterior to the SMA as it enters the posteromedial ileal branch. This is the junction where these two structures merge to form the main truck of the SMV. The jejunal branch also receives a branch from the uncinate process as it passes posterior to the SMA.  nn SURGICAL TECHNIQUE Tumors of the head of the pancreas are resected through a Whipple operation that removes the distal stomach, the duodenum, and a small portion of the proximal jejunum, the gallbladder, distal bile duct, and the head of the pancreas. The enteric reconstruction consists of three anastomoses: pancreatojejunostomy (or less commonly pancreaticogastrostomy), hepaticojejunostomy, and a gastrojejunostomy. The details of the standard Whipple operation are described elsewhere in this text. The modifications of the procedure for involvement of the portal vein, splenic vein, and SMV confluence are described within this chapter. The general principle of approaching tumors of the confluence is to perform the entire dissection leaving the specimen attached only to the PV-­SMV and then clamping and resecting the vein to remove the specimen. Specifically, the colon is mobilized off of the head of the pancreas, the duodenum is Kockerized, the common hepatic duct and stomach are divided. The jejunum is divided beyond the ligament of Trietz, the ligament is opened and the third and fourth portions of the duodenum and proximal jejunum are passed beneath the superior mesenteric vessels into the right upper quadrant. Vascular control is then obtained with vessel loops on the portal vein superiorly and the SMV inferiorly. If the inferior mesenteric vein (IMV) aberrantly enters the SMV and not the splenic vein, a common variant, control of this vessel can also be obtained. In addition,

554

Vascular Reconstruction During the Whipple Operation

A

B

C

D

FIG. 1  Pancreatic adenocarcinoma with encasement of the portal vein–superior mesenteric vein confluence. (A) Axial portal venous phase computed tomography image demonstrating an ill-­defined hypoenhancing mass in the pancreatic head (arrow) with abrupt caliber change of dilated main pancreatic duct (arrowhead). (B) Coronal portal venous phase computed tomography image shows tumor encasement and severe narrowing of the portal vein–superior mesenteric vein confluence (arrow). (C) Cinematic rendering of the pancreatic head mass (arrow) with dilated main pancreatic duct (arrowhead). (D) Cinematic rendering better illustrates the degree of portal vein-­superior mesenteric vein narrowing (arrow) compared with (B) two-­dimensional coronal reconstruction. (From Chu LC, Johnson PT, Fishman EK. Cinematic rendering of pancreatic neoplasms: preliminary observations and opportunities. Abdom Radiol. 2018.)

Collateral vessel

SV

Tumor

SV

SV

SV

SV

Tumor

Tumor

Tumor

PV

PV

PV

PV

PV

Tumor Fat plane

SMV

Ileal branch of SMV

A

SMV

SMV

SMV

IMV Jejunal branch of SMV Ileal branch of SMV

B

Jejunal branch of SMV

SMV IMV

IMV

IMV

Ileal branch of SMV

C

Jejunal branch of SMV

Ileal branch of SMV

D

Jejunal branch of SMV

IMV

Ileal branch of SMV

Jejunal branch of SMV

E

FIG. 2  Ishikawa classification of portal vein (PV)–superior mesenteric vein (SMV) involvement. (A) Normal vessel with fat plane between tumor and vessel. (B) Smooth shift of vessel with normal caliber. (C) Unilateral narrowing of vessel. (D) Bilateral narrowing of vessel with patent lumen. (E) Bilateral narrowing of vessel with closed lumen and formation of collateral veins. IMV, inferior mesenteric vein; SV, splenic vein.

555

PA N C R E A S

the coronary vein entering the portal vein from the left should be ligated and divided. At this point, the tunnel can be created behind the neck of the pancreas realizing that the only major tributary that is not controlled is the splenic vein. Controlling the splenic vein is difficult prior to the division of the neck. Once the neck is divided vessel loop control of the splenic vein should be obtained. The final step prior to en bloc resection of the PV-­SMV is to divide the uncinate process in a manner that clears the right side of the SMA for 270 degrees in axial plane. The easiest way to accomplish this is to divide the splenic vein thus exposing the anterior wall of the SMA. This will allow retraction of the specimen to the right and the uncinate process can be divided to the left of the vein (Fig. 4). Alternatively, if the

Portal vein Splenic vein Superior mesenteric vein

Inferior mesenteric vein

Gastroepiploic vein Ileal branch

Jejunal branch

FIG. 3  Peripancreatic vascular anatomic structures. Volume-­rendered computed tomography images show normal peripancreatic venous anatomic structures.

tumor involves the SMV only below the insertion of the splenic vein, the splenic vein can be preserved and the uncinate can be divided by retracting the PV-­SMV anteriorly. Either way, at this time, the only structure attaching the Whipple specimen to the abdomen is the region of vascular involvement to the PV-­SMV. Based on the extent of vascular involvement, appropriate resection and reconstruction is performed as detailed later in this chapter.  nn VENOUS

RESECTION AND RECONSTRUCTION

Several techniques have been described to resect and reconstruct venous involvement of the PV-­SMV. Four types have been proposed by the ISGPS (Fig. 5). The selection of the appropriate technique depends on the extent of vascular involvement, caliber of the vessel, extent of narrowing of vessel, and the surgeon’s preference. When there is invasion of a small circumference of the vessel, a vascular clamp can be placed to partially occlude the vessel, and a small tangential resection of the involved vessel can be performed (Fig. 6). This can be achieved by longitudinal dissection along the lateral aspect of the clamp. The use of a longitudinal incision is debatable, however, and some surgeons prefer a transverse incision because it minimizes the extent of narrowing of the vessel. Depending the size of the opening in the vessel and its caliber, the resulting defect can be closed primarily using a running 5-­0 Prolene suture (lateral venorrhaphy or transverse venorrhaphy) (Fig. 7). If significant narrowing of the vessel is anticipated, the closure can be performed using a venous patch (patch venoplasty), which can be fashioned from the saphenous vein, peritoneal patch, or a bovine pericardial patch (Fig. 7). In the event of a more extensive encasement (circumferential involvement) of the confluence, a segmental resection may be required (Fig. 8). This can be reconstructed using multiple techniques depending on the length of the resected segment as well as the tension on the remaining distal and proximal ends (Fig. 9). For all cases requiring a segmental resection, the inflow and outflow control is established both proximal and distal to the tumor involvement by

Portal vein

Portal vein

Vein graft

SMA SMV SMA

A

SMV

B

FIG. 4 Technique of venous resection and reconstruction. (A) Segmental resection of the superior mesenteric vein (SMV) is performed while the specimen is still attached to the superior mesenteric artery (SMA) before completion of the retroperitoneal dissection. (B) An internal jugular vein interposition graft enables medial retraction of the reconstructed superior mesenteric–portal vein confluence, allowing access to the retroperitoneum for standard dissection of the tumor from the lateral wall of the SMA. (From Leach SD, Lee JE, Charnsangavej C, et al. Survival following pancreaticoduodenectomy with resection of the superior mesenteric-­portal vein confluence for adenocarcinoma of the pancreatic head. Br J Surg. 1998;85[5]:611-­617.)

556

Vascular Reconstruction During the Whipple Operation

PV

PV

PV

SMV

SMV SMV

1

2

SV

SV

SV

SV

3

PV

PV

IJV

SV

IJV

SMV

SMV

4a

4b

FIG. 5 Techniques of venous reconstruction according to International Study Group of Pancreatic Surgery classification. (1) Transverse venous closure. (2) Venous patch angioplasty. (3) Segmental venous resection and primary anastomosis. (4a) Interposition of the internal jugular vein (IJV) graft. (4b) Interposition of the IJV graft with reanastomosis of the splenic vein (SV) to the IJV graft. PV, portal vein.

Rarely, the resected segment of vein is long and a primary anastomosis cannot be performed. In these cases, a venous interposition graft may be needed (Fig. 9 C–D). The most suitable grafts are the internal jugular vein or the left renal vein between the IVC and the gonadal vein as their caliber is similar to that of the portal and superior mesenteric veins. The use of polytetrafluoroethylene as a graft is associated with high incidence of postoperative thrombosis and should not be performed. It is unclear whether the splenic vein needs to be reimplanted when divided to resect and reconstruct the confluence. The rationale to do so is to avoid left-­sided portal hypertension and gastric varices. In our practice, we implant the splenic vein if it receives flow from the IMV. If the IMV enters directly into the SMV, we do not reimplant the vessel. We do not routinely systemically heparinize in reconstruction of the PV-­SMV. However, it is performed by some groups using the intravenous administration of 2500 to 3000 units of heparin 2 to 3 minutes before the occlusion. Recently, our institution has reported on a few cases of vascular resection without reconstruction in cancers with complete occlusion of the SMV with left-­sided collateralization of venous return. In these cases, intraoperatively, the superior mesenteric vessel is ligated at the level of the ileal and jejunal branches, and no reconstruction is performed. Upon follow-­up, these patients did not develop any complication of obstructed venous flow. Although this is not common practice, it might be a suitable option for select patients who present with localized disease with isolated involvement of the distal superior mesenteric vein and the presence of collateral vessels.  nn POSTOPERATIVE FIG. 6  Pancreatic tumor adherent to a small segment of the portal vein-­ superior mesenteric vein complex requiring tangential resection. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

placing vascular clamps. The vessels are then divided both proximal and distal to the location of the tumor. Typically for a resected segment that is 3 to 4 cm in length a primary anastomosis can be performed by approximating the vessels. Once approximated, both ends are sutured together using a 5-­0 Prolene suture is used to perform the anastomosis in a running manner (Fig. 9 A–B).

CARE

Although perioperative mortality of patients undergoing the Whipple operation has declined significantly, postoperative morbidity remains high. The most frequent complications include delayed gastric emptying, postoperative pancreatic fistula, postpancreatectomy hemorrhage, and surgical site infections. The routine postoperative management of patients undergoing a Whipple operation without vascular reconstruction is detailed in another chapter. The postoperative management of patients who undergo the Whipple operation with vascular resection is similar to that of patients not requiring a vascular resection. In particular for these patients, although systemic anticoagulation is not required, 300 mg of rectal aspirin is administered daily until the patient is tolerating a regular diet at which time it is switched to 81 mg daily of oral aspirin.

PA N C R E A S

PV

Splenic vein PV

PV

Splenic vein

Splenic vein

Patch graft

SMV SMV

A

557

SMV

B

C

FIG. 7 Tangential resection of tumor and repair using. (A) Primary venorrhaphy–longitudinal venorrhaphy. (B) Primary venorrhaphy–transverse venorrhaphy. (C) Patch venoplasty. PV, portal vein; SMV, superior mesenteric vein. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

graft occlusion or thrombus, a duplex should be performed. In the event that the duplex is inconclusive because of presence of bowel gas or the dressing, a CT angiogram is warranted. For partial thrombosis and with no concern for a technical problem, systemic anticoagulation should be initiated. If complete occlusion exists and/or there is concern for a technical problem, the patient should return to the operating room for surgical thrombectomy and revision of the reconstruction.  nn

FIG. 8  Pancreatic tumor encasing the portal vein-­superior mesenteric vein complex requiring segmental resection. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

Furthermore, the patient should be monitored for signs of thrombosis and graft occlusion. The most common signs of this are acute onset of ascites, oliguria, increased fluid requirement, and hypotension. Of note, the liver function tests may frequently be normal even with complete occlusion of the portal vein. If there is a suspicion for

POSTOPERATIVE APPEARANCE OF THE RECONSTRUCTED VENOUS VASCULATURE

One of the less frequently discussed aspects of vascular resection and reconstruction during a Whipple operation is the postoperative appearance of the reconstructed vasculature. A vast majority of patients who require these procedures are those that have malignant tumors of the pancreas. These tumors are at a high risk of recurrence; therefore, postoperative imaging is performed at regular intervals to rule out recurrence of disease. Interestingly, in patients undergoing a Whipple procedure with a concurrent vascular resection, aberrant or irregular appearance of the venous structures is common because of the new postoperative anatomy of the vessels. Postoperative appearance of the venous vasculature was recently studied and defined by our group, and two distinct features of the vasculature were described that included changes in the appearance of the vasculature itself and identification of perivenous changes. Changes in the vasculature can be observed in approximately one-­ half of all patients undergoing venous resections. These can be further classified into concentric smooth narrowing, eccentric irregular narrowing, or the presence of a thrombus (Fig. 10). The most frequent form is concentric smooth narrowing, whereas development of thrombus is a rare event. On the other hand, perivenous changes comprise perivenous fluid-­like collections or perivenous soft-­tissue thickening that has a mass-­like effect (Fig. 11). Perivenous changes are also frequently encountered in these patients, with perivenous fluid collections being more common of the two. These changes observed on postoperative imaging overlap considerably with findings suspicious for the recurrence of disease. Interestingly, however, in the previously mentioned study on long-­term follow-­up, a majority of patients with these changes did not develop local recurrence, suggesting that these changes in fact are related to the surgically altered anatomy of the vasculature. We believe that a

558

Vascular Reconstruction During the Whipple Operation

Splenic vein

Splenic vein

PV Splenic vein

PV

PV

Interposition Graft

SMV

SMV

PV

Interposition Graft

SMV

A

B

Splenic vein

SMV

C

D

FIG. 9  Segmental resection of tumor and reconstruction with primary end-­to-­end anastomosis with (A) and without (B) preservation of the splenic vein. In cases of long segment resection, reconstruction can be performed with interposition graft, with (C) and without (D) preservation of the splenic vein. PV, portal vein; SMV, superior mesenteric vein. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

A

B

C

FIG. 10  Categories of postoperative portal vein–superior mesenteric vein complex changes. (A) Concentric smooth narrowing. (B) Eccentric irregular narrowing. (C) Venous thrombosis. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

FIG. 11  Categories of perivenous findings. (A) Perivenous soft tissue thickening, which can have a mass-­like appearance. (B) Perivenous fluid collection. (From Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol [NY]. 2018;43[5]:1193-­1203.)

A

B

PA N C R E A S

thorough knowledge of the postoperative appearance of the venous vasculature is important for all physicians involved in the care of these patients. If the radiologists, surgeons, or medical oncologists are not well versed in this area, there is a high risk of misdiagnosis of recurrence of disease, which in turn triggers a significant change in the management of these patients.  nn SUMMARY With an increased utilization of effective systemic and locoregional therapies in the neoadjuvant setting, a significant proportion of patients with vascular involvement are now undergoing resection. Vascular resection and reconstruction add to the complexity of an already challenging surgical procedure; however, with appropriate preoperative imaging, thorough surgical planning, detailed knowledge of the available techniques, and surgical experience, these procedures can be performed safely. When performed by experienced surgeons at high-­volume centers, outcomes similar to that of the standard Whipple operation can be achieved in patients undergoing these complex resections. The oncologic benefits of these procedures have been well-­established, and as these procedures become more common, patients who would have otherwise succumb to their disease will have a shot at undergoing curative resection. Last, changes in the postoperative appearance of the vasculature associated with altered vascular

Palliative Therapy for Pancreatic Cancer Ashley N. Krepline, MD, and Susan Tsai, MD, MHS

P

atients with advanced pancreatic cancer often face myriad physical ailments. The overarching goal of palliative care is to improve in quality of life and alleviate pain and suffering for patients. Approaching patients about palliative care can be challenging because patients often have a misperception that palliative care is mutually exclusive from active cancer therapy. In reality, palliative care can be administered simultaneously with active cancer treatment and at times allows patients to return to a condition to be able to tolerate oncologic treatments. In other instances, focus on palliation of symptoms allows patients to achieve a better quality of life with their loved ones toward the end of life. Effective palliation of symptoms often requires consideration of the severity of symptoms while balancing the expected life expectancy and goals of care for the patient. Often, patients may present with multiple interrelated symptoms, which requires a multidisciplinary approach with the input of surgeons, medical oncologists, radiation oncologists, gastroenterologists, and radiologists to effectively address all symptoms. Guidance from palliative care specialists may further augment the development of a multidisciplinary, multimodal plan to comprehensively address symptoms while minimizing hospitalizations and ease the transition to end of life planning. Because of the anatomic location, pancreatic cancers can affect adjacent structures through invasive growth and obstruction. Patients commonly experience gastric outlet obstruction from duodenal obstruction, distal biliary obstruction causing jaundice, and debilitating back or epigastric pain related to the tumor infiltration into the celiac plexus. In addition, debilitating ascites may occur either secondary to carcinomatosis or occlusion of the superior mesenteric vein/portal vein. Finally, patients with advanced pancreatic cancer have the highest rates of depression and suicide among all cancers. Effective counseling and medication can help to alleviate emotional pain and suffering. This chapter will provide an overview of palliative approaches for these problems.

559

anatomy should be considered when evaluating these patients postoperatively to reduce the risk of misdiagnosis of recurrent disease.

Suggested Readings Beane JD, House MG, Pitt SC, et  al. Pancreatoduodenectomy with venous or arterial resection: a NSQIP propensity score analysis. HPB (Oxford). 2017;19(3):254–263. Croome KP, Farnell MB, Que FG, et al. Pancreaticoduodenectomy with major vascular resection: a comparison of laparoscopic versus open approaches. J Gastrointest Surg. 2015;19(1):189–194; discussion 194. Gage MM, Reames BN, Ejaz A, et al. Pancreaticoduodenectomy with en bloc vein resection for locally advanced pancreatic cancer: a case series without venous reconstruction. Chin Clin Oncol. 2018;7(1):7. Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol (NY). 2018;43(5):1193–1203. Katz MH, Fleming JB, Pisters PW, et  al. Anatomy of the superior mesenteric vein with special reference to the surgical management of first-­ order branch involvement at pancreaticoduodenectomy. Ann Surg. 2008;248(6):1098–1102. Miyazaki M, Yoshitomi H, Takano S, et al. Combined hepatic arterial resection in pancreatic resections for locally advanced pancreatic cancer. Langenbecks Arch Surg. 2017;402(3):447–456. Peters NA, Javed AA, Cameron JL, et  al. Modified Appleby procedure for pancreatic adenocarcinoma: does improved neoadjuvant therapy warrant such an aggressive approach? Ann Surg Oncol. 2016;23(11):3757–3764.

nn INITIAL

EVALUATION

Patients with advanced pancreatic cancer face difficult treatment decisions and open communication about goals of care will help patients to make sound, informed decisions. Fundamental questions which should be addressed with the patient and his/her family are summarized in Box 1. Understanding personal goals and preferences will help shape conversations about treatment recommendations and engage patients in shared decision making. It is also important to elicit patient preferences for how information is communicated and what level of detail is suitable. Along with a foundational understanding, treatment goals should be revisited as necessary at future encounters. Longitudinal assessment of symptoms at the time of diagnosis and throughout the course of the treatment will help to identify and proactively address problems. Patients with advanced pancreatic cancer are frequently monitored with imaging, such as a computed tomography (CT) scan of the chest, abdomen, and pelvis, and basic laboratory tests. Although these studies are obtained to assess treatment response, they should also be closely examined for impending complications. These may manifest as signs of current or impending biliary obstruction, bowel obstruction, ascites, tumor infiltration into the abdominal neural plexus (celiac axis), or significant portal venous narrowing/obstruction. In addition, routine assessment of pain, nausea, vomiting, abdominal distension, and cachexia can be helpful. Referral to palliative care services can facilitate addressing multiple symptoms. Notably, patients who are comanaged by palliative care specialists have been found to have decreased rates of intensive care admissions, emergency room visits, and repeated hospitalizations.  nn BILIARY

OBSTRUCTION

More than 70% of patients with pancreatic cancer will present with biliary obstruction leading to jaundice and malabsorption. Jaundice can cause debilitating pruritus, abdominal pain, fatigue, weight loss, nausea, vomiting, and anorexia, which can be reversed with biliary decompression. Durable relief of biliary obstruction is necessary to improve quality of life and normalize liver function tests to allow initiation of chemotherapy. Several series have demonstrated that endoscopic retrograde cholangiopancreatography (ERCP) can

PA N C R E A S

thorough knowledge of the postoperative appearance of the venous vasculature is important for all physicians involved in the care of these patients. If the radiologists, surgeons, or medical oncologists are not well versed in this area, there is a high risk of misdiagnosis of recurrence of disease, which in turn triggers a significant change in the management of these patients.  nn SUMMARY With an increased utilization of effective systemic and locoregional therapies in the neoadjuvant setting, a significant proportion of patients with vascular involvement are now undergoing resection. Vascular resection and reconstruction add to the complexity of an already challenging surgical procedure; however, with appropriate preoperative imaging, thorough surgical planning, detailed knowledge of the available techniques, and surgical experience, these procedures can be performed safely. When performed by experienced surgeons at high-­volume centers, outcomes similar to that of the standard Whipple operation can be achieved in patients undergoing these complex resections. The oncologic benefits of these procedures have been well-­established, and as these procedures become more common, patients who would have otherwise succumb to their disease will have a shot at undergoing curative resection. Last, changes in the postoperative appearance of the vasculature associated with altered vascular

Palliative Therapy for Pancreatic Cancer Ashley N. Krepline, MD, and Susan Tsai, MD, MHS

P

atients with advanced pancreatic cancer often face myriad physical ailments. The overarching goal of palliative care is to improve in quality of life and alleviate pain and suffering for patients. Approaching patients about palliative care can be challenging because patients often have a misperception that palliative care is mutually exclusive from active cancer therapy. In reality, palliative care can be administered simultaneously with active cancer treatment and at times allows patients to return to a condition to be able to tolerate oncologic treatments. In other instances, focus on palliation of symptoms allows patients to achieve a better quality of life with their loved ones toward the end of life. Effective palliation of symptoms often requires consideration of the severity of symptoms while balancing the expected life expectancy and goals of care for the patient. Often, patients may present with multiple interrelated symptoms, which requires a multidisciplinary approach with the input of surgeons, medical oncologists, radiation oncologists, gastroenterologists, and radiologists to effectively address all symptoms. Guidance from palliative care specialists may further augment the development of a multidisciplinary, multimodal plan to comprehensively address symptoms while minimizing hospitalizations and ease the transition to end of life planning. Because of the anatomic location, pancreatic cancers can affect adjacent structures through invasive growth and obstruction. Patients commonly experience gastric outlet obstruction from duodenal obstruction, distal biliary obstruction causing jaundice, and debilitating back or epigastric pain related to the tumor infiltration into the celiac plexus. In addition, debilitating ascites may occur either secondary to carcinomatosis or occlusion of the superior mesenteric vein/portal vein. Finally, patients with advanced pancreatic cancer have the highest rates of depression and suicide among all cancers. Effective counseling and medication can help to alleviate emotional pain and suffering. This chapter will provide an overview of palliative approaches for these problems.

559

anatomy should be considered when evaluating these patients postoperatively to reduce the risk of misdiagnosis of recurrent disease.

Suggested Readings Beane JD, House MG, Pitt SC, et  al. Pancreatoduodenectomy with venous or arterial resection: a NSQIP propensity score analysis. HPB (Oxford). 2017;19(3):254–263. Croome KP, Farnell MB, Que FG, et al. Pancreaticoduodenectomy with major vascular resection: a comparison of laparoscopic versus open approaches. J Gastrointest Surg. 2015;19(1):189–194; discussion 194. Gage MM, Reames BN, Ejaz A, et al. Pancreaticoduodenectomy with en bloc vein resection for locally advanced pancreatic cancer: a case series without venous reconstruction. Chin Clin Oncol. 2018;7(1):7. Javed AA, Bleich K, Bagante F, et al. Pancreaticoduodenectomy with venous resection and reconstruction: current surgical techniques and associated postoperative imaging findings. Abdom Radiol (NY). 2018;43(5):1193–1203. Katz MH, Fleming JB, Pisters PW, et  al. Anatomy of the superior mesenteric vein with special reference to the surgical management of first-­ order branch involvement at pancreaticoduodenectomy. Ann Surg. 2008;248(6):1098–1102. Miyazaki M, Yoshitomi H, Takano S, et al. Combined hepatic arterial resection in pancreatic resections for locally advanced pancreatic cancer. Langenbecks Arch Surg. 2017;402(3):447–456. Peters NA, Javed AA, Cameron JL, et  al. Modified Appleby procedure for pancreatic adenocarcinoma: does improved neoadjuvant therapy warrant such an aggressive approach? Ann Surg Oncol. 2016;23(11):3757–3764.

nn INITIAL

EVALUATION

Patients with advanced pancreatic cancer face difficult treatment decisions and open communication about goals of care will help patients to make sound, informed decisions. Fundamental questions which should be addressed with the patient and his/her family are summarized in Box 1. Understanding personal goals and preferences will help shape conversations about treatment recommendations and engage patients in shared decision making. It is also important to elicit patient preferences for how information is communicated and what level of detail is suitable. Along with a foundational understanding, treatment goals should be revisited as necessary at future encounters. Longitudinal assessment of symptoms at the time of diagnosis and throughout the course of the treatment will help to identify and proactively address problems. Patients with advanced pancreatic cancer are frequently monitored with imaging, such as a computed tomography (CT) scan of the chest, abdomen, and pelvis, and basic laboratory tests. Although these studies are obtained to assess treatment response, they should also be closely examined for impending complications. These may manifest as signs of current or impending biliary obstruction, bowel obstruction, ascites, tumor infiltration into the abdominal neural plexus (celiac axis), or significant portal venous narrowing/obstruction. In addition, routine assessment of pain, nausea, vomiting, abdominal distension, and cachexia can be helpful. Referral to palliative care services can facilitate addressing multiple symptoms. Notably, patients who are comanaged by palliative care specialists have been found to have decreased rates of intensive care admissions, emergency room visits, and repeated hospitalizations.  nn BILIARY

OBSTRUCTION

More than 70% of patients with pancreatic cancer will present with biliary obstruction leading to jaundice and malabsorption. Jaundice can cause debilitating pruritus, abdominal pain, fatigue, weight loss, nausea, vomiting, and anorexia, which can be reversed with biliary decompression. Durable relief of biliary obstruction is necessary to improve quality of life and normalize liver function tests to allow initiation of chemotherapy. Several series have demonstrated that endoscopic retrograde cholangiopancreatography (ERCP) can

560

Palliative Therapy for Pancreatic Cancer

be safely and effectively performed in the vast majority of patients with malignant obstructive jaundice. Early complications after ERCP occur in 5% of patients and most commonly include ERCP-­induced pancreatitis, infection, or bleeding. The most common delayed complication following biliary stenting is recurrent jaundice secondary to stent occlusion. Self-­expanding metal stents (SEMS) are clearly superior to plastic stents with regard to stent patency and stent-­related complications. Because of the smaller lumen, plastic biliary stents are more likely to occlude from biliary sludge or bacterial overgrowth requiring frequent stent exchanges, whereas SEMS are more likely to be occluded with tumor ingrowth of the stent. In general, the patency of a plastic stent is proportional to the size of the stent, for example, an 8Fr stent will be patent for 8 weeks. In contrast, a 10-­mm diameter SEMS will provide an equivalent drainage as nine 10Fr plastic stents, and therefore provide much more durable biliary drainage. Covered SEMS have become increasingly used and have demonstrated decreased rates of occlusion and therefore longer stent patency (Table 1). In addition to stent occlusion, another common complication of biliary stents is stent migration, which is more frequently seen with plastic and covered metal stents as opposed to uncovered SEMS. Operative biliary bypass is an alternative strategy to allow for drainage of an obstructed biliary tree. There are several techniques that may be utilized for a biliary bypass. The most common and durable technique used is a retrocolic choledochojejunostomy with a Roux-­en-­Y reconstruction. Alternatively, a loop of jejunum may be used rather than a Roux-­en-­Y. This is a quicker operation and does not require a jejunojejunostomy but is associated with a higher incidence of cholangitis and is not preferred to the Roux-­en-­Y. If a patient has not had a prior cholecystectomy, a cholecystectomy should be performed in addition to the biliary drainage procedure, even in the absence of cholelithiasis or symptoms. Resolution of biliary obstruction after endoscopic stent placement or surgical biliary bypass have been shown to be equivalent in several retrospective studies. In a series of 98 patients, 52 underwent surgical biliary bypass and 46 underwent placement of SEMS. Of the 52 patients undergoing surgical bypass, 24% developed perioperative

morbidity and 7% with SEMS developed periprocedural complication. Recurrent biliary obstruction as a result of SEMS occlusion occurred in 40% of patients, whereas recurrent biliary obstruction was not identified in patients who underwent surgical bypass. However, the increased durability of a hepaticojejunostomy must be weighed against the potential for perioperative morbidity and an increased length of hospital stay (13 vs 3 days) when compared with SEMS. Biliary bypass has been shown to be more durable than SEMS placement; however, even when including procedure related readmissions and repeat endoscopy for stent exchange, biliary stent placement has also been associated with fewer costs.  nn GASTRIC

OUTLET OBSTRUCTION

Nausea and vomiting in patients with advanced pancreatic cancer can be due to mechanical and/or nonmechanical problems. An initial evaluation should include a thorough history and physical and radiographic studies to identify obstruction related either to the primary tumor (gastric outlet obstruction) or a metastatic deposit (malignant bowel obstruction). In the absence of a mechanical etiology, the symptoms may be related to malignant gastroparesis or to chemotherapy and may be best alleviated with pharmacologic agents. Patients with gastric outlet obstruction suffer from nausea, vomiting, anorexia, weight loss, and malnutrition. Clinical suspicion can be supported by the presence of a dilated stomach on CT scan, and may be confirmed by an upper gastrointestinal fluoroscopy or endoscopy to confirm the diagnosis (Fig. 1A). Traditionally, a gastrojejunostomy is performed to relieve the area of obstruction and palliate symptoms. Palliative gastrojejunostomy for malignant gastric outlet obstruction has been associated with a 30% rate of morbidity, but with advances in surgical treatment, the morbidity has decreased to approximately 10%. Gastrojejunostomy can be performed via an open or laparoscopic approach. Classically, a retrocolic gastrojejunostomy is performed by sewing the posterior wall of the stomach to a loop of jejunum in a stapled or handsewn side-­to-­side fashion. This approach is favored over a Roux limb to improve the emptying of the jejunal limb. Studies comparing efficacy of laparoscopic compared with open gastrojejunostomy are limited; however, current studies demonstrate a quicker return to oral intake with a laparoscopic approach and a trend towards decreased length of hospital stay. More recently with the advent of endoscopic techniques, duodenal stenting has become an alternative strategy to manage malignant gastric outlet obstruction. Duodenal stenting has become a well-­ established technique to treat malignant gastric outlet obstructions (Table 2). A SEMS is endoscopically placed across a malignant stricture or area of narrowing. Duodenal stenting is typically associated with oral intake within 24 hours after placement, significantly faster than after an open bypass procedure. In a systematic review of 606

BOX 1  Questions Providers May Use to Engage Patients in Communicating Goals of Care What is your understanding of your illness? What are your hopes and fears? What are your goals and priorities? What outcomes are unacceptable to you? What would a good day look like?   

TABLE 1  Rates of Clinical Success and Complications Among Patients With Malignant Biliary Obstruction With Self-­Expanding Metal Biliary Stents No. of Patients

Rate of Stent Occlusion (%)

Median Time to Uncovered Metal Stent Occlusion (mo)

Yokota et al, BMC Gastroenterol, 2017

44

45

6.6

6.7

Sampaziotis et al, Eur J Gastroenterol ­Hepatol, 2015

99

44

3.7

4.1

Kitano et al, Am J Gastroenterol, 2013

120

37

5.6

7.8

Eum et al, Dig Endosc, 2013

107

34

Lee et al, Gastrointest Endosc, 2013

500

38

26.3

11.8

Maire et al, Am J Gastroenterol, 2006

59

31

7.0

—

Elwir et al, Dig Dis Sci, 2013

44

16

4.3

—

Study

Survival After Stent Placement (mo)

4.4

561

PA N C R E A S S

R

160 mm

A

B

FIG. 1  Diagnosis and management of gastric outlet obstruction. (A) Upper gastrointestinal series demonstrating obstruction in the first portion of the duodenum. (B) Palliation with duodenal and biliary stenting.

TABLE 2  Rates of Clinical Success and Complications Among Patients With Malignant Gastric Outlet Obstruction Undergoing Duodenal Stenting No. of Patients

Median Survival (mo)

Clinical Success Rate (Tolerate Oral Intake) (%)

Complication Rate (%)

Ding et al, J Oncol, 2013

94

2

90

5

Nassif et al, Endoscopy, 2003

63

1.75

95

30

Dormann et al, Endoscopy, 2003

606

4

89

27

Maire et al, Am J Gastroenterol, 2006

24

11

92

0

Graber et al, Endoscopy, 2007

51

2.5

84

11

Oh et al, Gastrointest Endosc, 2015

292

—

83

29

Study

patients with malignant gastric outlet obstruction, duodenal stents were successfully placed in 97% of patients. Of the patients who underwent successful placement of a stent, 89% had a clinical success, defined as ability to tolerate soft or regular diet. Reintervention for recurrent symptoms, most commonly resulting from stent occlusion secondary to tumor infiltration, occurred in 18% of patients at a median of 125 days. This can often be managed with the placement of an additional covered SEMS within the existing stent. Other complications included stent migration which occurred in 5% of patients, and a major bleeding or perforation event which occurred in 1% of patients. If there is associated jaundice, biliary stenting should be performed before duodenal stent placement, as placement of a biliary stent after placement of a duodenal stent is technically difficult (Fig. 1B). The decision to perform duodenal stenting or surgical gastrojejunostomy depends on expected survival and goals of care. In patients who undergo duodenal stenting, rapid return of oral intake is possible, but the long-­term risk of stent occlusion secondary to tumor in growth is high, especially after 4 months. Comparatively, rates of recurrent gastric outlet obstruction after undergoing

gastrojejunostomy are quite low, estimated around 2%, but carry an added risk of perioperative morbidity and mortality.  nn PROPHYLACTIC

BILIARY AND GASTRIC BYPASS

Palliation therapy implies that the patient is suffering from active symptoms. But, at times, palliative care decisions are made to prevent future complications in asymptomatic patients. The classic example of this is when patients who were thought to have operable disease are discovered at the time of surgery to have an unresectable pancreatic cancer. Among asymptomatic patients, future complications from local tumor extension results in duodenal obstruction in up to 20% of cases and biliary obstruction in 65% of cases. In a study of 87 patients undergoing exploratory laparotomy for intended pancreatic cancer resection, patients that were found to be unresectable at time of operation were randomized into prophylactic hepaticojejunostomy or prophylactic hepaticojejunostomy with gastrojejunostomy. Hepaticojejunostomy with gastrojejunostomy was performed in 44 patients, and no patient developed subsequent symptoms of gastric outlet obstruction. In

562

Palliative Therapy for Pancreatic Cancer

contrast, 8 (19%) of the 43 patients undergoing hepaticojejunostomy alone developed symptoms of gastric outlet obstruction requiring intervention. Similarly, in a second smaller randomized study of 65 patients who had a nontherapeutic laparotomy, 29 patients underwent hepaticojejunostomy and 36 patients underwent hepaticojejunostomy with gastrojejunostomy. Gastric outlet obstruction occurred in 2 (6%) of the 36 patients who underwent double bypass compared with 12 (41%) of the 29 patients who underwent hepaticojejunostomy alone. Neither study identified increased morbidity, mortality, or difference in overall survival between the two surgical intervention arms; however, it is important to note that the overall perioperative morbidity and mortality from these two trials were 29% to 32% and 0% to 2%, respectively. This early experience supported the routine use of double bypass at the time of a nontherapeutic laparotomy for pancreatic cancer. The largest series of palliative double bypass included 583 patients from a single institution. The series reported a 14% major complication rate and a 1.6% overall mortality rate. In addition, 18% of patients required readmission within 30 days and recurrent biliary obstruction or gastric outlet obstruction occurred in 9% of patients. Although concomitant prophylactic gastric and biliary drainage at the time of nontherapeutic laparotomy has been the standard practice for many years, more recent minimally invasive approaches of endoscopic stenting has led to a declining use of routine double bypass. Critics of routine surgical bypass at the time of a nontherapeutic laparotomy have reported that 98% of patients who do not receive a surgical bypass can be effectively palliated without an operation. In a study of 155 patients with advanced pancreatic cancer, only 4 patients required a future surgical intervention, and the median overall survival was approximately 6 months. Therefore, the authors argued that a prophylactic palliative operation has a detrimental impact on a patient’s quality of life when less invasive treatments are available. Ultimately, the decision regarding endoscopic versus surgical palliation requires knowledge about the (1) severity of symptoms, (2) estimation of patient’s performance status, and (3) understanding of the patient’s goals of care to best balance the durability of the intervention with respect to anticipated overall survival (Fig. 2). In the current era of multimodality chemotherapy, the median overall

survival of patients with metastatic pancreatic cancer is 11 months with the multidrug regimen 5-­FU, leucovorin, irinotecan, and oxaliplatin. Increasingly, patients are undergoing a diagnostic laparoscopy to rule out radiographically occult metastatic disease prior to resection. If patients are found to have radiographically occult metastatic disease and have symptoms of gastric obstruction, which have been inadequately managed with endoscopic therapy, then an open or laparoscopic gastrojejunostomy is indicated. Endoscopic management is usually successful in alleviating biliary obstruction, and therefore hepaticojejunostomy may be reserved for select patients who have been inadequately palliated with endoscopic stenting or for whom, gastroduodenal obstruction prevents endoscopic access. In patients who are asymptomatic, endoluminal approaches will likely be successful in palliating future symptoms and surgical bypass can be avoided. With expectant nonsurgical management, patients can avoid perioperative complications and receive systemic therapy for their metastatic disease with minimal delay.

Palliative Pancreaticoduodenectomy A small volume of literature exists describing the benefits of palliative resection for pancreatic cancer. In such series, improved overall survivals were noted among patients who undergo palliative microscopic (R1) or macroscopic (R2) resections compared with patients who undergo palliative surgical bypass. It should be noted that many series which report a survival benefit for palliative pancreaticoduodenectomy include patients who had a R1 resection from an operation performed with curative intent. The survival of these patients, in whom a curative intent surgery was performed, is not surprisingly superior to patients with obvious metastatic disease or residual gross tumor. When curative intent patients are eliminated from the analysis, the survival benefit is lost. In a systematic review of four cohort studies which included 138 patients with pancreaticoduodenectomy performed for palliation has been associated increased risk of morbidity and mortality 1.75 (95% CI, 1.35–2.26), P < .0001) and 2.98 (95% CI, 1.31–6.75, P < .009), respectively. Currently, there are no data to support palliative pancreaticoduodenectomy (R2 resection) for pancreatic cancer. 

Pancreatic Cancer

Non-therapeutic laparotomy

Previous biliary stent

Yes

Gastric outlet obstruction

No

Biliary stent ineffective

Biliary stent effective

Biliary obstruction

Biliary bypass

No further intervention

Biliary stent

Previous duodenal stent

Yes

No

Gastrojejunostomy

Symptomatic

Gastrojejunostomy FIG. 2 Algorithm for palliation in patients with unresectable pancreatic cancer.

Asymptomatic

No further intervention

PA N C R E A S

nn MALIGNANT

BOWEL OBSTRUCTION

Malignant bowel obstruction resulting from peritoneal carcinomatosis is an ominous complication. These patients often suffer from severe, debilitating malnutrition and cancer cachexia which complicates their operative candidacy. In addition, most patients have concurrent malignant ascites, adding to the risk for postoperative complications, including wound dehiscence and seeding of the surgical incision causing local pain. Depending on the location of the obstruction, the success of endoscopic approaches may be limited, especially if the obstruction occurs within the mid to distal small bowel. Patients may find some alleviation of symptoms when abstaining from oral intake. Relief of nausea and vomiting may require gastric decompression with a nasogastric tube in the acute setting. In very select patients who have no ascites and very low volume disease, an operative exploration and surgical bypass may be considered. If patients are not candidates for surgical palliation, after a short interval (5–7 days), they may benefit from transitioning the nasogastric tube to a venting gastrostomy tube for long term palliation. If surgical or endoscopic intervention is not possible, pharmacologic therapy may provide some relief. Adequate pain control with opioids is critical either by intravenous or subcutaneous route in the acute setting and can be transitioned to equivalent transdermal doses when the patient has a stable opioid requirement. Antisecretory agents may reduce splanchnic blood flow, intestinal secretions, and cramping abdominal pain. Octreotide is considered the standard of care for malignant bowel obstruction and has been effective for palliation in prospective trials, with total daily doses ranging from 300 to 1200 mcg. In addition, haloperidol is effective in relieving nausea and vomiting in patients with malignant bowel obstruction. In our opinion, it is imperative to engage a palliative care specialist for these patients to aid in the transition to comfort care.  nn MALIGNANT ASCITES Malignant ascites can lead to debilitating abdominal distension and early satiety. Ascites can develop as a result of portal hypertension or carcinomatosis. The etiology of the ascites can often be discriminated by the serum-­ascites albumin gradient (SAAG), with SAAG greater than 1.1 consistent with portal hypertension and nonperitoneal cause of ascites, while SAAG less than 1.1 is associated with carcinomatosis. Dietary sodium restriction and diuretics may be helpful in patients with ascites secondary to portal hypertension, but such measures are often ineffective in patients with carcinomatosis, with only 50% of patients noting a benefit. Therapeutic paracentesis relieves symptoms, but repeated attempts are associated with increased complications, such as infection or bowel perforation. The ideal rate of fluid withdrawal is unknown, but large volume paracentesis of up to 5 L is usually safe. Significant improvement in abdominal pressure is experienced with a removal of 5 L of fluid, but this may be associated with hypotension, renal failure, hypoalbuminemia, and pulmonary embolism. Plasma expanders, such as albumin, have been effective in preventing circulatory collapse and should be considered in select patients based on symptoms. In patients with an anticipated survival of greater than 2 months, the placement of an indwelling pigtail or tunneled catheter to control ascites with drainage at home or in the clinic may be useful. In a meta-­analysis, tunneled catheters have a lower risk of infectious complications, than nontunneled (pigtail) catheters, which are associated with a 30% incidence of infection.  nn VASCULAR THOMBOSIS The incidence of thromboembolic disease in pancreatic cancer is higher than other metastatic cancers and ranges from 10% to 20% in clinical trials and as high as 50% in autopsy series. Thrombosis can occur in the mesenteric veins resulting from tumor compression or in the peripheral veins because of the overall hypercoagulable state. Significant portal venous narrowing or thrombus should be treated with

563

anticoagulation therapy when possible to prevent propagation of the thrombus to complete occlusion. Once mesenteric venous occlusion occurs, refractory ascites may develop, as does venous hypertension, resulting in small bowel edema, ileus, and in rare cases intestinal ischemia. Low-­molecular-­weight heparin has been preferred over warfarin for anticoagulation for ease of management, but patients often struggle with long-­term repeated subcutaneous injections. With the advent of direct oral anticoagulants, patients with advanced pancreatic cancer have access to highly effective oral agents which require less monitoring.  nn TUMOR-­ASSOCIATED

MANAGEMENT

PAIN

Patients with unresectable pancreatic cancer often suffer from tumor-­ associated abdominal and back pain resulting from direct tumor infiltration into the celiac ganglion. Effective pain management often requires opioid analgesia, which can be complemented with additional adjunct analgesics for neuropathic pain as needed. Opioids are effective for neuropathic pain, but somewhat less effective for somatic or visceral pain. In addition, opioids have debilitating side effects including dizziness and sedation. More recently, the trend in pain management has been to optimize pain control with the lowest dose of opioids in combination with other adjunct analgesics to provide effective and consistent pain control. Celiac plexus block has become increasingly used in treatment of pancreatic cancer-­related pain symptoms. A celiac plexus block involves injecting a 50% alcohol solution or other neurolytic agents into the celiac ganglion or splanchnic nerves under CT, endoscopic ultrasound, or fluoroscopic guidance. In the event a patient undergoes nontherapeutic laparotomy, celiac plexus block may also be performed at that time. More than 80% of patients experience improvement in pain control after celiac plexus block in blinded or sham studies, as measured by reduction of opioid requirements. Recently, a meta-­analysis identifying 358 patients undergoing celiac plexus block found improved pain scores at 4 and 8 weeks in patients undergoing celiac plexus block with narcotic therapy compared to patients receiving narcotic therapy alone. A meta-­analysis of 145 patients demonstrated relief in 90% of patients undergoing celiac plexus block at 3 months and relief persisted in 70% to 90% of patients long-­term. Short-­term effects related to celiac plexus block include diarrhea, hypotension, and pain related to the procedure, which typically resolve with time celiac plexus block should be considered in the initial pain management plan for patients with pancreatic cancer and may be repeated in patients who demonstrated a prior benefit from neurolysis.  nn DEPRESSION The rates of depression and suicide among patients with pancreatic cancer are the highest in any cancer population. Elevated levels of circulating cytokines such as interleukin-­6 and tumor necrosis factor-­α are thought to alter neurohormonal pathways in the brain causing depressive symptoms even before the diagnosis of cancer. Studies report that depression occurs in 33% to 76% of patients and has a significant impact on quality of life. Many symptoms of depression mimic symptoms associated with pancreatic cancer (fatigue, anorexia, weight loss), complicating the diagnosis of depression. However, asking a patient whether he or she has “felt depressed most of the time” is a validated tool with good sensitivity and specificity for identifying depression even among patients who are terminally ill. Antidepressants are effective in patients with advanced cancers and supportive counseling may help patients to strengthen coping strategies and help with anticipatory grief. Depression and anxiety often occur because of unaddressed fears of death or the symptoms that may arise in the process of dying. An early referral to a palliative care specialist can be beneficial to provide address concerns and focus on quality of life for these patients. 

564

NEOADJUVANT AND ADJUVANT THERAPY FOR PANCREATIC CANCER

nn CONCLUSION It is important for surgeons to understand how to improve the quality of life of patients with advanced pancreatic cancer. A multidisciplinary approach to symptom management in coordination with a palliative care specialist will prepare patients for the myriad physical and emotional challenges that occur. Frequent assessment of available imaging and proactive solicitation of patient concerns will help providers to identify and manage symptoms more effectively. Clear communication about goals of care within the context of anticipated survival help to engage patients in shared decision making and allow providers to select the appropriate palliative treatment for each individual patient.

Suggested Readings Ding NS, Alexander S, Swan MP, et al. Gastroduodenal outlet obstruction and palliative self-­expandable metal stenting: a dual centre experience. J Oncol. 2013;2013: epub.

Neoadjuvant and Adjuvant Therapy for Pancreatic Cancer Brett L. Ecker, MD, and Charles M.Vollmer Jr, MD

P

ancreatic ductal adenocarcinoma is an aggressive malignancy with low rates of long-­term cure even after complete (R0) resection. The frequent and rapid development of systemic disease underscores the importance of multimodality treatment strategies. As such, chemotherapy is used for all stages of disease, including resectable, borderline resectable and locally advanced disease. nn DEFINING THE

OF DISEASE

CLINICAL STAGE

The American Joint Commission on Cancer staging of pancreatic adenocarcinoma follows the Tumor, Node, Metastasis framework. Although this staging paradigm is prognostic of long-­term survival, its dependence on postsurgical pathologic evaluation and its lack of definitional precision between resectable versus unresectable disease (i.e., T3 disease may classified as either) together limits its applicability in the preoperative setting. Rather, the National Comprehensive Cancer Network (NCCN) has endorsed the Intergroup clinical staging based on the tumor relationships to the key vascular anatomy and the presence of extrapancreatic disease. Following a dedicated contrast-­enhanced computed tomography (CT) scan (pancreatic protocol), tumors are classified according to the probability of achieving an R0 resection, and can be: (1) resectable; (2) borderline resectable; (3) locally advanced, or unresectable disease despite the absence of distant metastasis; and (4) disseminated. The definitions for each are presented in Table 1. The NCCN guidelines forgo potentially vague terms such as “abutment” and “encasement” in favor of degrees of contact with the vessel wall, although abutment can be generally understood to represent involvement less than 180 degrees of vessel circumference without contour irregularity or thrombosis, whereas encasement denotes more than 180 degrees involvement. Resectable pancreatic cancer (Fig. 1A) is defined by a primary tumor that does not contact any arterial vessel (celiac, hepatic, or superior mesenteric artery [SMA]) and does not contact the superior mesenteric vein/ portal vein (SMV/PV), or contacts the SMV/PV less than 180 degrees without contour irregularity. Borderline resectable disease (Fig. 1B)

Gray PJ, Wang J, Pawlik TM, et  al. Factors influencing survival in patients undergoing palliative bypass for pancreatic adenocarcinoma. J Surg Oncol. 2012;106:66–71. Kneuertz PJ, Cunningham SC, Cameron JL, et al. Palliative surgical management of patients with unresectable pancreatic adenocarcinoma: trends and lessons learned from a large, single institution experience. J Gastrointest Surg. 2011;15:1917–1927. Smith TJ, Temin S, Alesi ER, et al. American society of clinical oncology provisional clinical opinion: the integration of palliative care into standard oncology care. J Clin Oncol. 2012;30(8):880–887. Spanheimer PM, Cyr AR, Liao J, et al. Complications and survival associated with operative procedures in patients with unresectable pancreatic head adenocarcinoma. J of Surg Oncol. 2014;109:697–701. Wyse JM, Carone M, Paquin SC, et al. Randomized, double-­blind, controlled trial of early endoscopic ultrasound-­guided celiac plexus neurolysis to prevent pain progression in patients with newly diagnosed, painful, inoperable pancreatic cancer. J Clin Oncol. 29:3541–3546.

is defined by limited tumor contact less than 180 degrees with the SMA or celiac artery, any contact (1–360 degrees) with the hepatic artery, or contact with the SMV/PV more than 180 degrees, SMV contour irregularity, or SMV thrombosis. The locally advanced category (Fig. 1C) includes more than 180-­degree involvement of the SMA or celiac artery, or extensive involvement of the SMV/PV without proximal and distal targets for vascular reconstruction. Extensive SMA involvement is considered unresectable because of the poor survival and higher morbidity and mortality expected following SMA resection. Acknowledging the limitations of cross-­sectional imaging in identifying micrometastases, this clinical framework is crucial to decisions regarding the sequencing of care.

Resectable Pancreatic Cancer The clinical utility of multimodality therapy is well-­defined in the setting of resectable disease; yet, the timing of surgery relative to systemic therapy is an area of dispute and active research. Systemic chemotherapy has traditionally been administered after upfront surgery (i.e., in the adjuvant setting), although several institutions have championed the use of neoadjuvant therapy prior to surgical resection. Because complete surgical extirpation is the only opportunity for long-­term cure, upfront surgery ensures that surgical resection can occur before the disease progresses to an unresectable state. Neoadjuvant therapy also requires a tissue biopsy, subjecting the patient to potential morbidity from unnecessary pretreatment endoscopic procedures. Jaundiced patients, as is common with pancreatic head lesions, will often require additional procedures for biliary drainage, with a small but attendant risk of complications. Last, neoadjuvant therapy may increase the burden of postoperative complications. Pancreatic fistula is one of the most common and clinically relevant morbidities after pancreatic resection, and accounts for one-­third of all mortalities after pancreatoduodenectomy. Although the rate of pancreatic fistula may be less following neoadjuvant therapy, there is some evidence that fistula is associated with increased clinical burden in neoadjuvant cohorts. These data, which need to be confirmed more broadly, at least suggest that deconditioning from chemotherapy may lead patients to fail to promptly recover once a complication has occurred. In agreement, NCCN guidelines do not recommend the routine use of neoadjuvant therapy for those with clearly resectable disease without high-­risk features (e.g., elevated CA 19-­9, large tumor burden in the pancreas or surrounding lymph nodes, poor performance status) outside of a clinical trial. In contrast, the primary benefits of neoadjuvant therapy include: (1) the early treatment of micrometastatic disease, which is common and can lead to early postoperative recurrences; (2)

564

NEOADJUVANT AND ADJUVANT THERAPY FOR PANCREATIC CANCER

nn CONCLUSION It is important for surgeons to understand how to improve the quality of life of patients with advanced pancreatic cancer. A multidisciplinary approach to symptom management in coordination with a palliative care specialist will prepare patients for the myriad physical and emotional challenges that occur. Frequent assessment of available imaging and proactive solicitation of patient concerns will help providers to identify and manage symptoms more effectively. Clear communication about goals of care within the context of anticipated survival help to engage patients in shared decision making and allow providers to select the appropriate palliative treatment for each individual patient.

Suggested Readings Ding NS, Alexander S, Swan MP, et al. Gastroduodenal outlet obstruction and palliative self-­expandable metal stenting: a dual centre experience. J Oncol. 2013;2013: epub.

Neoadjuvant and Adjuvant Therapy for Pancreatic Cancer Brett L. Ecker, MD, and Charles M.Vollmer Jr, MD

P

ancreatic ductal adenocarcinoma is an aggressive malignancy with low rates of long-­term cure even after complete (R0) resection. The frequent and rapid development of systemic disease underscores the importance of multimodality treatment strategies. As such, chemotherapy is used for all stages of disease, including resectable, borderline resectable and locally advanced disease. nn DEFINING THE

OF DISEASE

CLINICAL STAGE

The American Joint Commission on Cancer staging of pancreatic adenocarcinoma follows the Tumor, Node, Metastasis framework. Although this staging paradigm is prognostic of long-­term survival, its dependence on postsurgical pathologic evaluation and its lack of definitional precision between resectable versus unresectable disease (i.e., T3 disease may classified as either) together limits its applicability in the preoperative setting. Rather, the National Comprehensive Cancer Network (NCCN) has endorsed the Intergroup clinical staging based on the tumor relationships to the key vascular anatomy and the presence of extrapancreatic disease. Following a dedicated contrast-­enhanced computed tomography (CT) scan (pancreatic protocol), tumors are classified according to the probability of achieving an R0 resection, and can be: (1) resectable; (2) borderline resectable; (3) locally advanced, or unresectable disease despite the absence of distant metastasis; and (4) disseminated. The definitions for each are presented in Table 1. The NCCN guidelines forgo potentially vague terms such as “abutment” and “encasement” in favor of degrees of contact with the vessel wall, although abutment can be generally understood to represent involvement less than 180 degrees of vessel circumference without contour irregularity or thrombosis, whereas encasement denotes more than 180 degrees involvement. Resectable pancreatic cancer (Fig. 1A) is defined by a primary tumor that does not contact any arterial vessel (celiac, hepatic, or superior mesenteric artery [SMA]) and does not contact the superior mesenteric vein/ portal vein (SMV/PV), or contacts the SMV/PV less than 180 degrees without contour irregularity. Borderline resectable disease (Fig. 1B)

Gray PJ, Wang J, Pawlik TM, et  al. Factors influencing survival in patients undergoing palliative bypass for pancreatic adenocarcinoma. J Surg Oncol. 2012;106:66–71. Kneuertz PJ, Cunningham SC, Cameron JL, et al. Palliative surgical management of patients with unresectable pancreatic adenocarcinoma: trends and lessons learned from a large, single institution experience. J Gastrointest Surg. 2011;15:1917–1927. Smith TJ, Temin S, Alesi ER, et al. American society of clinical oncology provisional clinical opinion: the integration of palliative care into standard oncology care. J Clin Oncol. 2012;30(8):880–887. Spanheimer PM, Cyr AR, Liao J, et al. Complications and survival associated with operative procedures in patients with unresectable pancreatic head adenocarcinoma. J of Surg Oncol. 2014;109:697–701. Wyse JM, Carone M, Paquin SC, et al. Randomized, double-­blind, controlled trial of early endoscopic ultrasound-­guided celiac plexus neurolysis to prevent pain progression in patients with newly diagnosed, painful, inoperable pancreatic cancer. J Clin Oncol. 29:3541–3546.

is defined by limited tumor contact less than 180 degrees with the SMA or celiac artery, any contact (1–360 degrees) with the hepatic artery, or contact with the SMV/PV more than 180 degrees, SMV contour irregularity, or SMV thrombosis. The locally advanced category (Fig. 1C) includes more than 180-­degree involvement of the SMA or celiac artery, or extensive involvement of the SMV/PV without proximal and distal targets for vascular reconstruction. Extensive SMA involvement is considered unresectable because of the poor survival and higher morbidity and mortality expected following SMA resection. Acknowledging the limitations of cross-­sectional imaging in identifying micrometastases, this clinical framework is crucial to decisions regarding the sequencing of care.

Resectable Pancreatic Cancer The clinical utility of multimodality therapy is well-­defined in the setting of resectable disease; yet, the timing of surgery relative to systemic therapy is an area of dispute and active research. Systemic chemotherapy has traditionally been administered after upfront surgery (i.e., in the adjuvant setting), although several institutions have championed the use of neoadjuvant therapy prior to surgical resection. Because complete surgical extirpation is the only opportunity for long-­term cure, upfront surgery ensures that surgical resection can occur before the disease progresses to an unresectable state. Neoadjuvant therapy also requires a tissue biopsy, subjecting the patient to potential morbidity from unnecessary pretreatment endoscopic procedures. Jaundiced patients, as is common with pancreatic head lesions, will often require additional procedures for biliary drainage, with a small but attendant risk of complications. Last, neoadjuvant therapy may increase the burden of postoperative complications. Pancreatic fistula is one of the most common and clinically relevant morbidities after pancreatic resection, and accounts for one-­third of all mortalities after pancreatoduodenectomy. Although the rate of pancreatic fistula may be less following neoadjuvant therapy, there is some evidence that fistula is associated with increased clinical burden in neoadjuvant cohorts. These data, which need to be confirmed more broadly, at least suggest that deconditioning from chemotherapy may lead patients to fail to promptly recover once a complication has occurred. In agreement, NCCN guidelines do not recommend the routine use of neoadjuvant therapy for those with clearly resectable disease without high-­risk features (e.g., elevated CA 19-­9, large tumor burden in the pancreas or surrounding lymph nodes, poor performance status) outside of a clinical trial. In contrast, the primary benefits of neoadjuvant therapy include: (1) the early treatment of micrometastatic disease, which is common and can lead to early postoperative recurrences; (2)

PA N C R E A S

565

TABLE 1  National Comprehensive Cancer Network Definitions of Resectability Tumor-­Artery Relationship

Tumor-­Vein Relationship

Resectable

No radiographic contact between tumor and celiac, hepatic or superior mesenteric artery

No tumor contact with the SMV/PV, or 180 degrees with the SMV/ PV, or contour irregularity, or short segment thrombosis (amenable to resection)

Locally advanced

Tumor contact >180 degrees with celiac or superior mesenteric artery or aorta

Occlusion of the SMV/PV without suitable targets proximally and distally for vascular reconstruction

Disseminated

Evidence of peritoneal or distant metastasis

SMV/PV, superior mesenteric vein/portal vein.

A

B

C FIG. 1  Clinical staging is based on tumor-­vessel contact. Contrast-­enhanced, axial computed tomographic images show the tumor (yellow arrow), superior mesenteric vein (SMV; blue arrowhead), and superior mesenteric artery (SMA; red arrow). (A) Resectable pancreatic cancer. Hypodense tumor is present in the pancreatic head with preservation of the fat plane between the pancreas and the SMV and without tumor abutment of the SMA. (B) Borderline resectable pancreatic cancer. Note the hypodense tumor that abuts both the SMV and the SMA. (C) Locally advanced pancreatic cancer. The hypodense tumor encases (>180 degrees) the SMA.

enhanced patient selection; (3) improved delivery and completion rates; and (4) enhanced efficacy of chemoradiotherapy (given the well-­oxygenated environment), with corresponding higher negative margin resection and lymph node-­negative rates. As many as 25% of patients will have metastatic disease discovered at operative exploration or during the postoperative recovery, thus negating any potential benefit derived from such morbid surgery. Instead, neoadjuvant therapy enriches the surgical population for those most likely to achieve a long-­term cure. Additionally, postoperative morbidity

influences the use and timing of adjuvant therapies, which are crucial to achieve the goal of long-­term survival. More than half of all pancreatoduodenectomy patients will suffer some postoperative morbidity, and prolonged postoperative recovery prevents the delivery of adjuvant therapy in nearly one-­quarter of patients. Neoadjuvant therapy ensures the delivery of some systemic therapy. Last, chemoradiotherapy has been associated with improved R0 rates in retrospective series. The benefits of each strategy (neoadjuvant vs. adjuvant) are summarized in Box 1.

566

NEOADJUVANT AND ADJUVANT THERAPY FOR PANCREATIC CANCER

At present, there is level I evidence to support the use of adjuvant therapy but no head-­to-­head randomized prospective data comparing adjuvant and neoadjuvant approaches. Established adjuvant chemotherapy regimens include monthly bolus 5-­fluorouracil, which demonstrated superiority to observation alone in the European Study for Pancreatic Cancer (ESPAC)-­1 trial, gemcitabine monotherapy following the results of the European Charité Onkologie-­001 trial, and combination gemcitabine/capecitabine, which was superior to gemcitabine monotherapy in the ESPAC-­4 trial. At present, there is level I evidence to support the use of adjuvant therapy but no headto-head randomized prospective data comparing adjuvant and neoadjuvant approaches. Established adjuvant chemotherapy regimens include monthly bolus 5-fluorouracil, which demonstrated superiority to observation alone in the European Study for Pancreatic Cancer (ESPAC)-1 trial, gemcitabine monotherapy following the results of the European Charité Onkologie-001 trial, and combination gemcitabine/capecitabine, which was superior to gemcitabine monotherapy in the ESPAC-4 trial. Additionally, more aggressive regimens are the focus of currently enrolling and recently completed trials. The PRODIGE 24-CCTG PA 6 trial evaluating gemcitabine versus FOLFIRINOX in the adjuvant setting found significant improvements in overall survival with the multidrug regimen. Nab-paclitaxel/gemcitabine is superior to gemcitabine alone in the metastatic setting and is currently under evaluation in the adjuvant setting (APACT trial), and a trial has completed enrollment. The benefits of multidrug regimens will have to be balanced against increased treatment-related toxicities in a deconditioned postoperative patient, but we can expect increasing use of FOLFIRINOX when patient fitness allows. The relative efficacy of FOLFIRINOX versus nab-paclitaxel/gemcitabine has yet to be understood. The role of adjuvant radiotherapy (RT) is less clearly defined. Both the Gastrointestinal Tumor Study Group 9173 trial and the European Organization for Research and Treatment of Cancer trial observed

BOX 1  Potential Advantages and Disadvantages of Neoadjuvant Therapy Advantages • Ability to deliver systemic therapy to all patients • Identification of patients with aggressive tumor biology (manifested as disease progression) at the time of posttreatment and preoperative restaging who thereby avoid the toxicity of surgery • Increased efficacy of radiation therapy; free radical production in a well-­oxygenated environment • Decreased radiation-­induced toxicity to adjacent normal tissue because the radiated field is resected at the time of pancreatectomy • Decreased rate of positive resection margins; superior mesenteric artery margin in particular • Decreased rate of pancreatic fistula formation • Potential for the tumor size to decrease, especially in borderline resectable tumors, which may facilitate surgical resection  Disadvantages • Potential for complications from pretreatment endoscopic procedures (endoscopic ultrasound scan and fine-­needle aspiration, and endoscopic retrograde cholangiopancreatography) • Biliary stent–related morbidity; stent occlusion during neoadjuvant therapy • Disease progression obviating resectability; loss of a window of resectability may occur (rarely) in the borderline resectable patient • Coordination of multiple physicians during the preoperative phase; discrete handoff from surgeon to medical oncologist to radiation oncologist (as occurs with adjuvant therapy) is not possible in the neoadjuvant setting   

benefits for postoperative concurrent 5-­flurouracil plus external beam RT relative to observation. However, these results were not verified in the previously mentioned ESPAC-­1 trial, which was a four-­arm study that demonstrated a trend toward worse survival for the group receiving chemoradiotherapy. In these historical trials, split-­course RT at lower doses (40 Gy) may have limited efficacy. The contemporary phase II European Organization for Research and Treatment of Cancer 40013 trial evaluated gemcitabine-­based chemoradiotherapy, and observed that the addition of RT to gemcitabine alone improved local control, although without impacting disease-­free or overall survival. In the neoadjuvant setting, several phase I and II trials have demonstrated the safety and efficacy of many of the previously mentioned therapy regimens. Currently, there is no level I evidence defining the optimal neoadjuvant regimen, or how such therapies compare to their use in the adjuvant setting. PACT-­15 (NCT01150630) was a three-­arm, randomized, phase II/III trial comparing neoadjuvant gemcitabine, cisplatin, epirubicin, and capecitabine (PEXG) to adjuvant gemcitabine monotherapy and adjuvant PEXG. As of this year, the study authors have decided not to continue with the phase III aspect of this trial, given ever-­evolving adjuvant therapy standards. Still, in the small cohort of treated patients (n = 30/arm), neoadjuvant therapy demonstrated promising efficacy (66% event-­free at 1 year vs 50% event-­free in the adjuvant PEXG group vs 23% event-­free in the adjuvant gemcitabine group). Likewise, an interim analysis published in abstract form of the phase II/III ESPAC-­5F trial comparing neoadjuvant gemcitabine-­based chemoradiation with adjuvant gemcitabine-­ based chemoradiation demonstrated a survival benefit of upfront chemoradiotherapy (median survival, 23 months vs 11 months). Several remaining trials comparing neoadjuvant with adjuvant therapy are ongoing (e.g., NEONAX, NCT02047513, neoadjuvant nab-­paclitaxel/ gemcitabine vs adjuvant nab-­paclitaxel/gemcitabine; PANACHE01-­ PRODIGE48, NCT02959879, neoadjuvant FOLFIRINOX or FOLFOX vs standard adjuvant chemotherapy; NorPACT, NCT02919787, neoadjuvant FOLFIRINOX and adjuvant gemcitabine/capecitabine vs adjuvant gemcitabine/capecitabine). In general, neoadjuvant therapy for resectable disease involves (1) the early initiation of chemotherapy (with or without RT); (2) restaging before surgical exploration, which occurs between 2 and 6 weeks after chemotherapy, and 4 and 6 weeks after chemoradiotherapy; and (3) frequent use of additional chemotherapy (with or without RT) following postoperative recovery. In summary, NCCN guidelines recommended the use of systemic therapies for all patients with resectable disease. Although the timing of such therapy can be justified in either the neoadjuvant or adjuvant setting, at present, neoadjuvant therapy is not yet considered a standard approach for patients with potentially resectable pancreatic cancer outside of the context of a clinical trial. 

Borderline Resectable Patients In contrast to resectable disease, less controversy exists regarding the use of neoadjuvant therapy before attempted resection of borderline resectable disease. This category is defined by more advanced locoregional disease at higher risk for incomplete resection, obviating the oncologic benefit of surgical resection because patients who undergo a resection with a macroscopically positive margin demonstrate long-­term survival comparable to patients who do not undergo any operation. Moreover, although borderline resectable patients are apparently radiographically localized, they have a greater risk for subclinical distant metastases. The current body of literature is limited by heterogeneity in regimens used and definitions of clinical resectability; nevertheless, many institutions have embraced neoadjuvant combination chemotherapy (with or without RT) for this scenario. The Alliance for Clinical Trials in Oncology Trial A021101 (NCT01821612) recently published the results of a prospective pilot trial using neoadjuvant FOLFIRINOX followed by capecitabine-­based chemoradiation in borderline resectable patients, defined in strict accordance with current NCCN guidelines. Among the 23 patients enrolled, 15 (68%) completed neoadjuvant

PA N C R E A S

therapy and proceeded to surgical resection, where 93% of patients underwent an R0 resection. Although not powered for a survival analysis, its notable that median survival for these patients was 21.7 months, which compares favorably to the 24-­month survival observed in ESPAC-­4 for gemcitabine monotherapy following upfront resection (for less advanced, resectable cancer). This year, the Korean multicenter, randomized controlled phase II/III trial (NCT01458717), which compared gemcitabine-­based chemoradiation in the neoadjuvant versus adjuvant setting for 50 borderline resectable patients was published. Among patients who underwent surgical resection, R0 resection rates were significantly improved with neoadjuvant therapy (82% vs 33%), which may explain the significant improvements observed in the intent-­to-­treat survival analysis (median, 21 vs 12 months). These encouraging results led to early termination of the trial. Beyond enhanced resectability, neoadjuvant therapy can lead to pathologic complete response in a rare number of patients, which is independently predictive of long-­term survival. In the Korean trial described above, a complete pathologic response was observed in 12% of patients. In a contemporary cohort of borderline resectable and locally advanced patients treated at Johns Hopkins treated with chemoradiotherapy using either FORFIRINOX or multiagent gemcitabine-­based chemotherapy, a complete pathologic response was more common following FORFIRINOX (13% vs 7%), suggesting that even greater gains may be expected than those observed with gemcitabine monotherapy in the Korean trial. In general, choice of systemic agents for neoadjuvant treatment has evolved from single-­ agent gemcitabine to combination therapies, such as FOLFIRINOX or gemcitabine/nab-­paclitaxel, given their success in the metastatic and adjuvant setting, and ongoing randomized trials will provide much anticipated data on their efficacy. 

Locally Advanced Patients Locally advanced pancreatic cancer is nonmetastatic but technically unresectable disease, usually because of interplay of the tumor with the critical local visceral vasculature. Similar to borderline resectable patients, initial management consists of chemotherapy with or without chemoradiotherapy. Although rare (5 years) in rare patients with ACC. In such situations, resection of a unifocal metastasis may be considered in an otherwise healthy patient (Fig. 1C). This also allows for histologic confirmation of the diagnosis and molecular profiling as a guide to further systemic therapy. 

A

nn SOLID

PSEUDOPAPILLARY TUMORS

SPTs of the pancreas are also quite rare but, in contrast to ACC, they have low malignant potential. SPT have been associated with several other names, including Frantz tumors, Hamoudi tumors, and papillary cystic neoplasm. SPTs are well known for their high prevalence among women, most commonly occurring in the third decade of life and even earlier (mean age, 22 years; range, 2–85 years). The most common presenting symptoms and signs (if present) include abdominal pain and/or the presence of an abdominal mass on physical examination. In the asymptomatic patient, tumors may be discovered as a palpable mass on routine physical examination or as an incidental finding on imaging for an unrelated complaint. Serologic tests are often of little value with CA19-­9 being rarely elevated. On CT imaging, SPTs can range from being completely cystic to completely solid; in our experience, a pure cystic SPT is uncommon because the cystic portion is secondary to necrotic degeneration of the primary tumor. They frequently demonstrate peripheral enhancement and central calcification and are characteristically large (often much larger than seen in Fig. 2), heterogeneously enhancing lesions with solid and cystic components. On MRI, SPTs have a low signal intensity on T1-­weighted images and a high intensity on T2-­weighted images. Although SPTs can occur throughout the pancreas, they are perhaps slightly more common in the pancreatic tail. When they occur in the pancreatic head, they can attain large size in the absence of bile duct obstruction, similar to ACC and pancreatic neuroendocrine tumors. However, SPT can grow into the wall of the superior mesenteric or portal vein; it would be a mistake to operate on a large SPT of the pancreatic head and assume that a narrowed superior mesenteric or portal vein could be successfully separated from the tumor; venous resection/reconstruction may be required. The radiographic differential diagnosis of a SPT should include other cystic neoplasms including mucinous neoplasms or serous cystadenomas, and intraductal papillary mucinous neoplasms, as well as a pancreatic neuroendocrine tumor. However, age is important; in a young woman younger than age 30, SPT and pancreatic neuroendocrine tumor would be most likely. In the absence of an inherited endocrinopathy (such as MEN1 or VHL), an SPT would be most likely. In a young woman younger than age 20, SPT would clearly be the most likely diagnosis as even in MEN1, a large pancreatic neuroendocrine tumor would be uncommon under the age of 20. FNA biopsy may be useful when routine imaging is inconclusive and diagnostic uncertainty exists; however, because of the tumor’s largely necrotic composition, FNA biopsy may often be nondiagnostic.

B

FIG. 2  (A) Axial and (B) coronal images of a contrast-­enhanced computed tomography scan from a patient with a solid pseudopapillary tumor (arrows) of the neck of the pancreas with solid and cystic characteristics.

PA N C R E A S

Some defining histologic features of SPTs include the presence of solid cellular hypervascular regions without gland formation, and the presence of branching papillary fronds with sheets and degenerative pseudopapillae. Cells stain positively for neuron-­specific enolase, CD10, and keratins; chromogranin, synaptophysin, and endocrine pancreatic enzymes are generally not expressed. SPTs often stain positive for progesterone receptors, whereas estrogen receptor positivity is more variable. There are no histologic characteristics that appear prognostic for patients with SPTs. The genetic profile associated with SPT is different from adenocarcinoma, most notably for an absence of KRAS, GNAS, and SMAD4 mutations. Almost all SPTs harbor alterations in the APC/β-­catenin pathway resulting from a mutation involving CTNNB1 (exon 3). Nuclear accumulation of β-­catenin has been described in 95% of SPTs and 74% of tumors overexpress cyclin D1, a downstream effector of β-­catenin. Interestingly, BCL9L, a β-­catenin stabilizing gene, is significantly decreased in SPT, which may help attenuate the protumorigenic effects of overactivation of the Wnt/β-­catenin pathway. In addition, genes involved in the Hedgehog and androgen receptor signaling pathways, as well as genes involved in epithelial mesenchymal transition have been shown to be activated in SPT.

Treatment Surgical resection is recommended for all patients with localized SPT. Although these tumors may be extremely large and can invade critical vasculature, most tumors are usually amenable to complete resection if the operating team is comfortable with resection and reconstruction of the superior mesenteric and/or portal veins; arterial resection/ reconstruction is required much less often. Pancreaticoduodenectomy or distal pancreatectomy can be performed with en bloc resection of involved adjacent organs when indicated. Recurrence is very uncommon; these authors have seen only one patient who had metastatic disease and, in that patient, it was present at the time of diagnosis. We have not seen a patient with SPT who developed a metachronous recurrence following a potentially curative operation. Although recurrence rates are low, long-­term surveillance is felt to be important because of the young age of most all patients at the time of diagnosis. Given the excellent survival rates following surgical resection alone, adjuvant systemic therapy is not routinely used. If metastatic disease occurs (very rare), the most common sites include liver, mesentery, and peritoneum. The management of such very patients is anecdotal and for those with single site recurrence, surgery may be a reasonable approach. 

A

nn AUTOIMMUNE

571

PANCREATITIS

Autoimmune pancreatitis (AIP) is a form of pancreatitis characterized by obstructive jaundice with or without a pancreatic mass, lymphoplasmacytic infiltration, and fibrosis of the pancreas, and a therapeutic response to corticosteroids. The incidence of AIP is unknown, but it has become less frequently found on pathologic examination of surgical specimens following operations for presumed cancer; the diagnosis of AIP is clearly being made more often without surgical resection of the pancreas. Patients with AIP often present with painless jaundice that can mimic pancreatic ductal adenocarcinoma and is due to inflammation and narrowing/stricture of the distal common bile duct. In addition, other common symptoms of AIP include weight loss and abdominal pain but usually in the absence of cachexia and pain requiring narcotic medication. Similar to patients with adenocarcinoma, many of those with AIP are diabetic with impaired glucose tolerance. AIP is currently classified into two subtypes. Type 1 AIP is associated with an elevation in serum immunoglobulin G4 (IgG4) levels and radiographic evidence of extrapancreatic involvement such as Sjögren’s syndrome, rheumatoid arthritis, primary sclerosing cholangitis, orbital pseudotumor, and inflammatory bowel disease. Extrapancreatic organ involvement can occur before, synchronous with, or after the diagnosis of AIP and type 1 is more common in older men. Biopsy of extrapancreatic sites can be helpful in making the diagnosis because the affected organs often demonstrate the characteristic lymphoplasmacytic infiltrate rich in IgG4-­positive cells. In contrast, type 2 AIP is seen in the absence of elevated IgG4 levels and associated autoimmune disease is limited to inflammatory bowel disease, which is found in approximately 30% of patients. There is also no age predominance in type 2 AIP. Serum IgG4 is the single best serologic marker of AIP with a sensitivity of 80% in patients with type 1 AIP, but only 17% in those with type 2 AIP. IgG4 elevation above twice the upper limit of normal is strongly suggestive of AIP in the setting of obstructive jaundice. The classic features of AIP on CT or MRI include a diffusely enlarged, sausage-­shaped pancreas with homogeneous attenuation and no visible pancreatic duct. However, when the predominant area of involvement is the pancreatic head and porta hepatis, the imaging characteristics can be more challenging (Fig. 3). In contrast to alcohol-­induced pancreatitis, AIP is not associated with ductal dilation, calculi, and pseudocyst formation. Importantly, although AIP may involve a stricture of the pancreatic duct, the upstream dilation characteristic of pancreatic ductal adenocarcinoma is rarely observed.

B

FIG. 3 Axial images of a contrast-­enhanced computed tomography scan from a 55-­year-­old man proven to have type 1 autoimmune pancreatitis before (A) and after (B) steroid treatment. Images taken at the level of the superior mesenteric artery (arrows). Biopsy showed lymphoplasmacytic inflammation and a positive immunohistochemical stain for immunoglobulin G4.

572

Unusual Pancreatic Tumors

However, occasionally, AIP may present as a focal mass-­forming lesion in the pancreas that can be easily confused with pancreatic ductal adenocarcinoma; in such cases, the diagnosis is usually made by the pathologist after the involved pancreas has been surgically excised. Recently, international consensus diagnostic criteria were developed for type 1 and type 2 AIP, which incorporate the findings from radiographic imaging (including ductal imaging with magnetic resonance cholangiography or endoscopic retrograde cholangiopancreatography) as well as serologic and histopathologic data. Type 1 AIP may not require a histologic diagnosis when the typical clinical, radiographic, and laboratory criteria are present, but since type 2 AIP is often seronegative and lacks other organ involvement, definitive diagnosis requires a pathologic biopsy specimen. Type 1 AIP has three essential features: (1) lymphoplasmacytic infiltrate surrounding small-­sized interlobular pancreatic ducts; (2) fibrosis centered around the ducts and veins affecting predominantly the peripancreatic adipose tissue; and (3) obliterative phlebitis affecting the pancreatic veins. Immunostaining often demonstrates abundant (>10 cells/high-­power field) IgG4-­positive cells. Type 2 AIP differs from type 1 by less prominent fibrosis, phlebitis, and the lack of IgG4 positivity. In type 2 AIP, lymphoplasmacytic infiltrates may result in obliteration of the pancreatic duct lumen, in contrast to type 1 AIP, in which the ductal epithelium is generally spared. The diagnosis of AIP (especially type 2) can be difficult, short of removing part of the pancreas as no single diagnostic test is sufficient. A correct diagnosis, made short of removing the pancreas, relies on a combination of cytology/histology of biopsy specimens, cross-­sectional and endoscopic imaging, serologic findings, and a detailed clinical history. In general, the diagnosis of AIP requires a multidisciplinary team consisting of a radiologist, pathologist, surgeon, and gastroenterologist with expertise in the disease.

Treatment AIP is highly responsive to corticosteroid therapy and when this does not occur quickly (in weeks), an alternate diagnosis (especially adenocarcinoma) should be considered. Although AIP can resolve spontaneously, treatment with corticosteroids has been associated with rapid reversal of jaundice, diabetes, and exocrine dysfunction usually within 2 to 4 weeks of starting treatment. The resolution of bile duct obstruction shortly after starting corticosteroids confirms the diagnosis. The International Consensus Diagnostic Criteria for Autoimmune Pancreatitis recommends a trial of 0.6 to 1 mg/kg per day of prednisone for a period of 2 weeks followed by reimaging and interval assessment of CA 19-­9 levels. If the diagnosis of AIP is correct, the bile duct stricture and gland enlargement should improve with steroid therapy. On clinical and radiologic improvement, the prednisone can be tapered by 5 to 10 mg/day every 1 to 2 weeks until a daily dosage of 20 mg, then decreased by 5 mg every 2 weeks. Because clinical relapse can occur in up to 30% of patients, some have advocated a more prolonged taper or the administration of low-­dose maintenance prednisone. In Japan, prednisone (2.5–7.5 mg/day) is administered for up to 3 years, which has demonstrated a lower relapse rate in type 1 AIP. Patients with type 1 AIP who experience a rapid decrease in serum IgG4 levels have a low probability of relapse. If corticosteroids are ineffective in disease relapse, other immunologic therapies, including rituximab and azathioprine, have been used. Surgery is reserved for situations where diagnostic uncertainty exists. If the diagnosis of AIP cannot be confirmed and the pancreas has been biopsied more than once, surgical resection of the pancreas (involved segment, Whipple’s or distal pancreatectomy) is the next logical approach. If the pancreas (and the patient) are operable/resectable and a trial of steroids is preferred despite diagnostic uncertainty, we would reimage in 3 to 4 weeks and in the absence of improvement, proceed to surgery. If the involved segment or entire pancreas is deemed not resectable because of local anatomy, and AIP is suspected but not confirmed, then a trial of corticosteroids is quite reasonable. If there is no response to corticosteroids, repeat biopsy should be performed. One of the most difficult clinical situations involves a patient with an

inoperable pancreas tumor in which the differential diagnosis includes both AIP and adenocarcinoma and the diagnosis cannot be established despite multiple attempts at endoscopic ultrasound (EUS)-­guided or percutaneous biopsy and a trial of corticosteroids. In this situation we would not start empiric chemotherapy (patient may not have cancer), but rather, perform a comprehensive reassessment in 4 to 6 weeks and consider rebiopsy at that time. If the correct diagnosis is, in fact, inoperable pancreatic cancer, such a treatment delay will be of little clinical significance and with time, the diagnosis will become apparent.  nn PRIMARY

PANCREATIC LYMPHOMA

Lymphomas involving predominantly the pancreas are extremely rare and can occur exclusively in the pancreas (primary pancreatic lymphoma [PPL]), via direct extension from adjacent peripancreatic lymphadenopathy (secondary pancreatic lymphoma), or originate from lymph nodes distant from the pancreas. PPL is defined by the World Health Organization as “an extranodal lymphoma arising in the pancreas with the bulk of the disease localized to this site; contiguous lymph node involvement and distant spread may be seen but the primary clinical presentation is in the pancreas with treatment directed to this site.” PPL is predominantly non-­Hodgkin’s lymphoma of B-­cell phenotype and diffuse large B-­cell lymphoma is the most common histological subtype. PPL accounts for less than 2% of extranodal lymphomas. Currently, no specific biochemical markers aid in the diagnosis of PPL. Elevated serum lactate dehydrogenase and β2-­ microglobluin levels in the setting of a normal CA 19-­9 may provide a clue to the diagnosis of PPL. PPLs predominantly occur in men (7:1) and usually present in the fifth to sixth decade of life. Common presenting symptoms include abdominal pain, and symptoms such as fever, night sweats, chills, weight loss, jaundice, and gastric or duodenal outlet obstruction resulting from the bulk of disease. In patients with PPL, CT or MRI demonstrates the presence of a large mass that focally involves the head of the pancreas (Fig. 4) or occasionally a more diffuse form that is infiltrative and can mimic the appearance of acute pancreatitis. As one would expect, patients with PPL often present with significant lymphadenopathy involving the peripancreatic lymph nodes and most notably, the retroperitoneal lymph nodes below the renal vein in the paraaortic location. The involved lymph nodes lack central necrosis or calcifications and the pancreatic duct is rarely dilated despite what appears to be a large

FIG. 4 Axial image of a primary pancreatic lymphoma involving the head of pancreas. Note the hypodense appearance of the mass (arrowheads) and the adjacent abutment of the superior mesenteric vein (long arrow) and proximity to the superior mesenteric artery (short arrow).

PA N C R E A S

pancreatic tumor. Also, narrowing or occlusion of the superior mesenteric or portal veins is rarely present in most patients despite the bulky tumor size. PPLs are avid on fluorodeoxyglucose positron emission tomography scans with uptake patterns that may be focal nodular, diffuse, or segmental. Such findings prompt biopsy, either with either EUS-­FNA (small or core biopsy needle) or percutaneous core biopsy technique. Cytopathologic features include large malignant lymphocytic nuclei, prominent nucleoli, and a background of necrosis. Immunohistochemical stains that are positive in pancreatic endocrine neoplasms, such as synaptophysin, are generally negative in PPL. The use of flow cytometry may be limited by the cellularity of some fine-­needle aspirate specimens; therefore, core needle biopsy is usually preferred when the diagnosis of PPL is suspected.

Treatment The standard of care in the management of PPL is chemotherapy alone, which provides excellent control of symptoms, including jaundice, as well as long-­term remission. PPL is most commonly treated with a multidrug regimen such as cyclophosphamide, doxorubicin, vincristine, and prednisone. Complete remission can be expected with multidrug therapy in the majority of patients with large B-­cell lymphoma. However, recurrence is common in patients older than age 60. The use of an anti-­CD20 antibody, rituximab, and cyclophosphamide, doxorubicin, vincristine, and prednisone has been associated with improved response rates of up to 85% in diffuse large B-­cell lymphoma. Other regimens include cyclophosphamide, vincristine, and prednisone and methotrexate, Adriamycin, cyclophosphamide, vincristine, prednisone, and bleomycin. Laparotomy should be reserved for patients in whom the diagnosis is uncertain despite percutaneous or endoscopic biopsy often in the setting of biliary obstruction/endoscopic stenting or for therapeutic purposes (palliative surgery) in the setting of gastrointestinal hemorrhage or gastric outlet obstruction.  nn METASTATIC

RENAL CELL CANCER

Metastatic lesions to the pancreas are also very rare and the vast majority of patients thought to have a metastasis to the pancreas actually have metastases to peripancreatic lymph nodes. The one exception is renal cell cancer (RCC), which metastasizes to the pancreatic parenchyma. Synchronous metastases can occur in up to 25% of patients with RCC and metachronous metastases may occur in up to 40% of all patients with a history of RCC. RCC metastases to the pancreas may present after an extended disease-­free interval from nephrectomy, and we have seen an anecdotal patient with a disease-­ free interval of 20 years. This emphasizes the importance of long-­term follow-­up for patients with RCC after initial nephrectomy. There are no differences in the frequency of pancreatic metastases based on the laterality of the primary tumor and metastases from RCC can occur anywhere within the pancreas; there is not a preferred location (head vs body or tail). Interestingly, many patients have solitary metastases (based on CT/MRI imaging), which are usually asymptomatic (>50%) and identified during follow-­up surveillance. Occasionally, a metastasis from RCC can erode into the duodenum and cause gastrointestinal hemorrhage (because they are highly vascular tumors) or obstruct the splenic vein resulting in gastroesophageal varices and anemia, melena, or hematemesis. In patients with more widespread disease, abdominal pain, weight loss, or jaundice may be the presenting complaint. Many patients also have extrapancreatic metastases; therefore, a thorough staging evaluation should be performed in patients with suspected or biopsy-­proven metastatic RCC to the pancreas (to include an MRI of the brain). CT is the best test for the evaluation of presumed RCC metastases to the pancreas and is often diagnostic in the absence of a biopsy. The classic hypervascular tumor may demonstrate a central area of low attenuation on the arterial phase (Fig. 5) in a patient with a history of prior surgery for a large RCC is diagnostic. The imaging characteristics can be similar to pancreatic neuroendocrine tumors but

573

this is a clinical dilemma only in patients with VHL who underwent prior nephrectomy for RCC. In contrast to the hypervascularity of RCC metastasis to the pancreas, pancreatic ductal adenocarcinoma is hypodense on the arterial phase of CT imaging. In the majority of patients with a history of RCC, the CT findings are diagnostic and there is no need for a pancreatic biopsy. However, a tissue biopsy may be helpful if there is a concern over the diagnosis on imaging. Positive immunohistochemical staining for CD10 and PAX8 can be used to distinguish between metastatic RCC and other tumors, such as clear cell carcinoma of the pancreas, clear cell pancreatic endocrine tumor, and the solid variant of serous cystadenoma.

Treatment In general, surgery is applied only to patients with isolated pancreatic metastases from RCC (uni-­or multifocal); those with synchronous extrapancreatic metastases are usually not considered for surgical treatment in the absence of pancreas-­associated complications such as bleeding or biliary/gastric outlet obstruction. Patients with isolated RCC metastases to the pancreas who undergo surgical resection may experience a long disease-­free survival. In a systematic review that identified 321 patients with resected RCC metastases to the pancreas, the 5-­year disease-­free survival and overall survival were 57% and 73%, respectively. However, with improvements in systemic therapies (targeted agents, immunotherapy) to include innovative clinical trials, patients with metastatic RCC may experience prolonged survival without surgery. In addition, antiangiogenic agents such as bevacizumab, sunitinib, and sorafenib, have shown promising results in metastatic RCC, necessitating the need for a multidisciplinary approach to the management of these patients.  nn CONCLUSIONS The major diagnostic concern in patients with a pancreatic neoplasm, either suggested or clearly demonstrated on cross sectional imaging, is pancreatic ductal adenocarcinoma. Remember that most patients with adenocarcinoma of the pancreas will also have an elevation in serum levels of CA19-­9 and/or an elevation in hemoglobin A1c (often in the setting of weight loss). For patients with a symptomatic or asymptomatic pancreatic mass, accurate staging with CT or MRI is usually followed by EUS with FNA biopsy. Although EUS-­FNA biopsy has become very safe in experienced hands, it should not be used to replace a detailed patient history, physical examination, focused laboratory evaluation and a thorough review of all imaging studies. Rather, pancreatic biopsy should be used in conjunction with all available clinical and laboratory information to arrive at an accurate diagnosis and stage of disease. If the diagnosis is not adenocarcinoma, a thorough understanding of rare and unusual pancreatic neoplasms is important to develop the correct treatment plan. Surgery is the cornerstone of therapy for SPT, ACC, and isolated RCC metastases (especially those with a long disease-­free interval) if the tumor is operable. For those patients with tumors that are inseparable from the superior mesenteric or portal vein (on preoperative imaging), surgery should only be undertaken if the operative team has experience with venous resection and reconstruction at the time of pancreatectomy. Similarly, if the tumor appears to encase the celiac artery or common hepatic artery, arterial resection and reconstruction may be necessary; such tumor-­vascular relationships are accurately delineated on preoperative imaging and should not be an unexpected finding at the time of laparotomy. Most important, patients with diagnoses as discussed in this chapter will often live many years even without surgery, mandating that surgery-­associated mortality be close to zero and morbidity be managed safely. This can be achieved by referral of patients to specialty centers (often referred to as high-­volume centers) when indicated. Finally, it is important to remember that medical therapy is the obvious choice for patients with AIP and PPL when the diagnosis can be accurately established. Usually, this is possible without an operation (open or laparoscopic).

574

Unusual Pancreatic Tumors

A

B

C FIG. 5  (A, C) Axial and (B) coronal images of a contrast-­enhanced computed tomography scan from a patient who underwent a left nephrectomy for renal cell carcinoma 15 years prior and then developed a metachronous metastasis to the pancreatic body. Note the characteristic bright enhancement of the metastatic lesion (arrow) on arterial phase imaging as well as the posterior displacement of the pancreas (arrowheads) because of the absence of the left kidney.

Suggested Readings Abraham SC, Klimstra DS, Wilentz RE, et al. Solid-­pseudopapillary tumors of the pancreas are genetically distinct from pancreatic ductal adenocarcinomas and almost always harbor beta-­catenin mutations. Am J Pathol. 2002;160(4):1361–1369. Chari ST, Longnecker DS, Kloppel G. The diagnosis of autoimmune pancreatitis: a western perspective. Pancreas. 2009;38(8):846–848. Grimison PS, Chin MT, Harrison ML, Goldstein D. Primary pancreatic lymphoma-­pancreatic tumours that are potentially curable without resection, a retrospective review of four cases. BMC cancer. 2006;6:117. Kamisawa T, Shimosegawa T, Okazaki K, et al. Standard steroid treatment for autoimmune pancreatitis. Gut. 2009;58(11):1504–1507. Lowery MA, Klimstra DS, Shia J, et al. Acinar cell carcinoma of the pancreas: new genetic and treatment insights into a rare malignancy. The oncologist. 2011;16(12):1714–1720. Papavramidis T, Papavramidis S. Solid pseudopapillary tumors of the pancreas: review of 718 patients reported in English literature. J Am Coll Surg. 2005;200(6):965–972. Reddy S, Cameron JL, Scudiere J, et  al. Surgical management of solid-­ pseudopapillary neoplasms of the pancreas (Franz or Hamoudi tumors): a large single-­institutional series. J Am Coll Surg. 2009;208(5):950–957; discussion 957-­959.

Sadot E, Yahalom J, Do RK, et al. Clinical features and outcome of primary pancreatic lymphoma. Ann Surg Oncol. 2015;22(4):1176–1184. Schmidt CM, Matos JM, Bentrem DJ, Talamonti MS, Lillemoe KD, Bilimoria KY. Acinar cell carcinoma of the pancreas in the United States: prognostic factors and comparison to ductal adenocarcinoma. J Gastrointest Surg. 2008;12(12):2078–2086. Shimosegawa T, Chari ST, Frulloni L, et al. International consensus diagnostic criteria for autoimmune pancreatitis: guidelines of the International Association of Pancreatology. Pancreas. 2011;40(3):352–358. Tanis PJ, van der Gaag NA, Busch OR, van Gulik TM, Gouma DJ. Systematic review of pancreatic surgery for metastatic renal cell carcinoma. Br J Surg. 2009;96(6):579–592. Tosoian JJ, Cameron JL, Allaf ME, et al. Resection of isolated renal cell carcinoma metastases of the pancreas: outcomes from the Johns Hopkins Hospital. J Gastrointest Surg. 2014;18(3):542–548. Webb TH, Lillemoe KD, Pitt HA, Jones RJ, Cameron JL. Pancreatic lymphoma. Is surgery mandatory for diagnosis or treatment? Ann Surg. 1989;209(1):25–30.

PA N C R E A S

Intraductal Papillary Mucinous Neoplasms of the Pancreas Carlos Fernandez-­del Castillo, MD, and George Molina, MD, MPH

I

ntraductal papillary mucinous neoplasms (IPMNs) of the pancreas were first described in Japan in 1982 and have been increasingly identified over the past 30 years because of the advances of cross-­ sectional abdominal imaging and its more frequent use in the evaluation of various abdominal complaints. At the Massachusetts General Hospital (MGH) approximately 25% of pancreatic resections are for IPMNs. Based on autopsies, it is suspected that IPMN lesions are found in up to 2% to 3% of the general population. Furthermore, IPMNs are more prevalent in older individuals, with about 10% of individuals older than 70 years being identified with an IPMN. Historically, IPMNs were thought to occur more frequently in men; however, more recent studies and our experience at MGH has been that there is most likely no sex difference in IPMN prevalence. This chapter reviews IPMN characteristics, risk of malignancy, and management recommendations. nn CLASSIFICATION

LESIONS

OF THE TYPES OF IPMN

There are two distinct types of IPMN lesions and these include main duct (MD-­IPMN) and branch duct (BD-­IPMN). Mixed-­type IPMN lesions are a third type that include features of MD-­IPMN and BD-­ IPMN (Fig. 1). IPMNs are categorized as MD-­IPMN if the main pancreatic duct has segmental or diffuse dilation that is greater than 5 mm without other causes of obstruction. Conversely, BD-­IPMNs are defined as having a pancreatic cyst that is more than 5 mm in size that communicates with a nondilated main pancreatic duct. Mixed-­ type IPMN lesions are characterized as having a pancreatic cyst that communicates with a dilated main pancreatic duct. Most IPMNs are identified incidentally when patients are undergoing cross-­sectional imaging for abdominal complaints, most often abdominal pain. Although most IPMN lesions are incidentally found using a multidetector computed tomography (MDCT) or a magnetic resonance imaging (MRI), the best imaging modality in the initial workup of a suspected IPMN is a pancreas-­protocol MDCT scan of the abdomen and pelvis (Fig. 2). This includes an arterial contrast phase and a portal-­venous contrast phase that favors imaging of the pancreas. After a lesion has been identified as being suspicious for an IPMN, the decision must be made to either proceed with observation, diagnostic intervention, or resection. The decision on how to proceed depends on the type of lesion and other characteristics that have been classified as high-­risk stigmata and worrisome features. This classification scheme and management recommendations are based on the International Association of Pancreatology (IAP) 2006 Sendai consensus guidelines that were subsequently updated to the IAP 2012 Fukuoka consensus guidelines. These guidelines were most recently revised in 2017, and they are now known as the revised IAP 2012 Fukuoka consensus guidelines. A distinct feature of IPMN lesions are papillary projections in the pancreatic ductal system. Based on the morphology of these projections, IPMNs are categorized into epithelial subtypes or phenotypes. There are four distinct epithelial phenotypes: intestinal, gastric, pancreatobiliary, and oncocytic. This categorization is important because these phenotypes are associated with tubular, colloid, and oncocytic invasive malignancy with varying degrees of survival. 

nn CLINICAL

575

PRESENTATION

At our institution, the majority of patients (57%) were asymptomatic at the initial time of presentation. Of those patients who were managed with surveillance, 83% were asymptomatic at the initial time of presentation and only 10% of patients progressed to have symptoms at the 6-­month mark. Among patients who underwent surgical resection, about 50% had symptoms at presentation that included abdominal pain (41%), weight loss greater than 10 pounds (29%), acute pancreatitis (22%), and jaundice (9%). Additionally, about 34% of patients who underwent resection at MGH had a diagnosis of diabetes mellitus, and these patients had an adjusted twofold elevated risk of having high-­grade dysplasia and invasive carcinoma.  nn FAMILY

HISTORY

Although there are currently no clear genetic disorders that are associated with IPMN lesions, at MGH, 13.9% of all patients with an IPMN lesion had a family history of pancreatic cancer. Despite this finding, a family history of pancreatic cancer was not associated with type (MD-­IPMN, BD-­IPMN, or mixed-­type IPMN), epithelial subtype, or presence of malignancy. Among these same patients, however, there was an associated increase in the incidence of concurrently occurring pancreatic ductal adenocarcinoma (PDAC) (11.1% vs 2.9%, P = .02) and extrapancreatic malignancies (35.6% vs 20.1%, P = .03). The Johns Hopkins experience has reported that a family history of pancreatic cancer was an independent risk factor for recurrence of IPMN after initial resection, and that it carried a fourfold increased risk of developing a recurrence.  nn MD-­IPMN MD-­IPMN lesions usually occur in the sixth decade of life, and they are more often associated with symptoms, with more than 50% presenting with abdominal pain, weight loss, jaundice, or pancreatitis. About 50% of all MD-­IPMN lesions are intestinal epithelial phenotype. These express MUC-­2, which produces a thick mucin that may lead to obstruction of the main pancreatic duct and subsequent pancreatitis. Intestinal epithelial phenotype IPMN lesions are associated with high-­grade dysplasia and invasive carcinoma and most often progress to colloid carcinoma. Colloid carcinomas have a median survival of 95 months after resection and are more indolent than tubular carcinomas, which have a reported postresection survival of about 35 months. Additionally, colloid carcinomas are associated with GNAS mutations. Oncocytic epithelial phenotype is associated with MD-­ IPMN, but it is only seen in 5% of all IPMN lesions. These lesions are indolent and if they transform into invasive malignancy their median survival has been reported to be approximately 132 months, which is much more favorable than colloid or tubular carcinomas. According to the revised IAP 2012 Fukuoka consensus guidelines, all MD-­IPMNs with main duct dilation of more than 10 mm, jaundice, or mural nodules should be considered for surgical resection if the patient is an appropriate surgical candidate. The reasoning behind this recommendation is the increased risk for high-­grade dysplasia and invasive carcinoma. The risk of identifying high-­grade dysplasia in a resected MD-­IPMN specimen at MGH is about 32%, and the reported risk of invasive carcinoma ranges from 44% to 70%.  nn BD-­IPMN The majority of the pancreatic cysts identified in cross-­ sectional abdominal imaging are suspected to be BD-­IPMN lesions. The most common epithelial phenotype found among BD-­ IPMN lesions is gastric epithelial subtype (83%), which does not commonly undergo malignant transformation. When malignant transformation does occur, however, BD-­IPMN lesions transform into tubular carcinomas,

576

Intraductal Papillary Mucinous Neoplasms of the Pancreas

A

B

C

FIG. 1  (A) Sagittal multidetector computed tomography image showing main duct intraductal papillary mucinous neoplasm with notable dilation (arrow) of the main pancreatic duct. (B) Axial multidetector computed tomography image showing a small cystic lesion in the pancreatic tail (arrow) consistent with branch duct intraductal papillary mucinous neoplasm. No dilation is observed in the main pancreatic duct. (C) Mixed-­type intraductal papillary mucinous neoplasm on computed tomography scan. There is notable diffuse dilation of the pancreatic duct in the head and body of the gland (arrowheads) and multiple small branch duct lesions throughout the body and tail (arrows).

A

B

C

FIG. 2  Radiographic appearance of (A) branch duct, (B) main duct, and (C) mixed-­type intraductal papillary mucinous neoplasms. (From Tanaka M. International consensus on the management of intraductal papillary mucinous neoplasm of the pancreas. Ann Transl Med. 2015;3[19]:286.)

which are more aggressive tumors. Despite being a more aggressive form of invasive cancer, the 5-­year survival of invasive BD-­IPMN is still more favorable than pancreatic ductal adenocarcinoma (55% vs 37%). Pancreatobiliary-­ type epithelial phenotype IPMNs are more often seen in BD-­IPMN lesions than in MD-­IPMN lesions and are also associated with tubular carcinoma (Fig. 3); however, pancreatobiliary type is difficult to distinguish from gastric type. Overall, the risk of invasive carcinoma associated with BD-­IPMN ranges from 9% to 17%, whereas the risk of high-­grade dysplasia is about 15% in resected BD-­IPMNs.  nn MIXED-­TYPE

IPMN

By definition, mixed-­type IPMN lesions include dilation of the main pancreatic duct and cystic lesions arising from branches of the main pancreatic duct. At our institution, the risk of high-­grade dysplasia and invasive carcinoma for mixed-­type IPMN tumors is about 31% and 28%, respectively. This is notably higher than for BD-­IPMN, but lower than what has been reported for MD-­IPMN. Although recent data have shown that the degree of main duct dilation may play a prognostic role in predicting how mixed-­type IPMN lesions behave, the revised IAP 2012 Fukuoka consensus guidelines recommend consideration for surgical resection of mixed-­type IPMN with main duct dilation larger than 10 mm, jaundice, or mural nodules and if the patient is an appropriate surgical candidate.

IAP Guidelines for BD-­IPMN Lesions The revised IAP 2012 Fukuoka consensus guidelines provide recommendations on how to proceed when a BD-­IPMN lesion is identified.

These guidelines are based on the presence or absence of high-­risk stigmata and worrisome features. According to the previous IAP 2006 Sendai consensus guidelines, surgical resection should be considered for a BD-­IPMN lesion if the patient is evaluated to be an appropriate surgical candidate and if any of the following criteria were met: any symptomatic cyst, asymptomatic cyst that was greater than 3 cm in size, main pancreatic duct dilation that was greater than 6 mm in diameter, or presence of mural nodules within the cyst. The revised IAP 2012 Fukuoka consensus guidelines were brought forth because many patients were undergoing pancreatic resections for benign IPMN lesions (Table 1 and Fig. 4). The revised IAP 2012 Fukuoka consensus guidelines categorized IPMN characteristics into high-­risk stigmata and worrisome features. Surgical resection should be considered for a BD-­IPMN lesion if the patient is identified as being an appropriate surgical candidate and if any of the following high-­risk stigmata are present: obstructive jaundice in a patient with a cystic lesion of the head of the pancreas, enhancing mural nodule greater than or equal to 5 mm, or main pancreatic duct greater than or equal to 10 mm in diameter. If high-­risk stigmata are not present, then management recommendations depend on the presence of worrisome features. Worrisome features include pancreatitis, a BD-­IPMN cyst greater than or equal to 3 cm in size, enhancing mural nodule smaller than 5 mm, thickened and/or enhancing cyst wall, main pancreatic duct measuring 5 to 9 mm in diameter, abrupt change in caliber of pancreatic duct with distal pancreatic atrophy, lymphadenopathy, increased serum level of CA19-­9, and cyst growth rate greater than or equal to 5 mm over 2 years. If worrisome features are present then the recommendation is to perform an endoscopic ultrasound (EUS) and fine-­needle aspiration (FNA) biopsy to further investigate the BD-­IPMN. Surgical resection should be considered if

577

PA N C R E A S

gastric

A

intestinal

B oncocytic

pancreatobiliary

C

D

FIG. 3 The four histologic classifications of intraductal papillary mucinous neoplasm are (A) gastric, (B) intestinal, (C) pancreatobiliary, and (D) oncocytic. (From Tanaka M, Fernandez-­del Castillo C, Adsay V, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology. 2012;12:183-­187.)

the patient is an appropriate surgical candidate and if any of the following are present during the EUS with FNA: mural nodule, suspicion for involvement of the main pancreatic duct, or cytology from the FNA biopsy that is suspicious or positive for malignancy.  nn FIELD-­DEFECT

CONCEPT

The field-­defect concept refers to the phenomenon that a patient may have an IPMN in one location in the pancreas and concomitantly have PDAC at a different site of the pancreas. The reported incidence of concomitant PDAC has ranged from 4% to 11%. In addition to concomitant PDAC, patients may have synchronous and multifocal lesions. Independent predictors of concomitantly occurring PDAC are worsening diabetes mellitus and abnormal serum CA19-­9. Furthermore, it is important to note that IPMNs can histopathologically range from low-­grade dysplasia, which can be considered to be similar to adenomas, to high-­grade dysplasia and invasive carcinoma. Because of the field defect concept, multiple IPMN lesions with varying degrees of dysplasia and presence of invasive carcinoma can be present concurrently. As such, each suspicious lesion should be risk-­ stratified individually, and cyst-­specific pancreatectomy should be performed rather than total pancreatectomy.  nn ELEVATED

SERUM CA19-­9

The revised IAP 2012 Fukuoka consensus guidelines now includes elevated serum CA19-­9 as a worrisome feature. This is based on

recent findings from a meta-­analysis of 15 studies that included 1629 patients and found that elevated serum CA19-­9 was significantly predictive of detecting invasive carcinoma in IPMN lesions (pooled sensitivity of 52% and specificity of 88%, respectively). We have found that among resected BD-­IPMN at MGH, an elevated serum CA19-­9 of 100 units/mL or higher had the highest predictive accuracy for detecting invasive carcinoma (93%), when compared with the standard cutoff of 37 units/mL (83%). However, there is no evidence to suggest a diagnostic cutoff of serum CA19-­9 for detecting high-­grade dysplasia.  nn IMAGING

MODALITIES

Many patients will be diagnosed with an IPMN based on CT imaging performed for the workup of abdominal pain or as part of the workup for another abdominal reason or pathology. If contrast was used, the portal-­venous phase was most likely not captured, and if it was captured, it was most likely suboptimal. All patients with a known or suspected IPMN should undergo a pancreas-­protocol MDCT scan or a gadolinium-­enhanced with magnetic resonance cholangiopancreatography (MRCP). A pancreas-­protocol MDCT captures the arterial and portal-­venous phase and provides the best images of the pancreas. At MGH, MDCT includes negative oral contrast with water and intravenous contrast administration. The amount of IV contrast is weight dependent and it is given via bolus at about 3 to 4 mL/s. There is a 50-­second delay from when the contrast injection starts to when the arterial phase is captured. The portal-­venous phase, also referred

578

Intraductal Papillary Mucinous Neoplasms of the Pancreas

TABLE 1  International Association of Pancreatology Consensus Guidelines for Management of BD-IPMNs Recommendations

2006 SENDAI GUIDELINES Symptomatic cyst 3 cm Main pancreatic duct >6 mm in diameter

Surgical resection should be considered if any are present in a patient who is an appropriate surgical candidate

REVISED 2012 FUKUOKA GUIDELINES High-Risk Stigmata Obstructive jaundice in a patient with cystic lesion of pancreatic head Enhancing mural nodule ≥5 mm Main pancreatic duct ≥10 mm

Surgical resection should be considered if any are present in a patient who is an appropriate surgical candidate

Worrisome Features Pancreatitis Cyst ≥3 cm Enhancing mural nodule 3 years unless patient can successfully wean off and initiate centrally acting drugs (e.g., gabapentinoids) before surgery 8. Poorly controlled psychiatric comorbidity 9. Medical noncompliance   

ARP, acute recurrent pancreatitis; CP, chronic pancreatitis; HGP, hereditary genetic pancreatitis; QOL, quality of life.

Suggested Readings Bellin MD, Freeman ML, Gelrud A, et al. Total pancreatectomy and islet autotransplantation in chronic pancreatitis: recommendations from pancreasFest. Pancreatology. 2014;14:27–35. Conwell DL, Banks PA, Sandhu BS, et al. Validation of demographics, etiology, and risk factors for chronic pancreatitis in the USA: a report of the North American Pancreas Study (NAPS) Group. Dig Dis Sci. 2017;62:2133–2140. Conwell DL, Lee LS, Yadav D, et al. American Pancreatic Association practice guidelines in chronic pancreatitis: evidence-­based report on diagnostic guidelines. Pancreas. 2014;43:1143–1162. Dunderdale J, McAuliffe JC, McNeal SF, et  al. Should pancreatectomy with islet cell autotransplantation in patients with chronic alcoholic pancreatitis be abandoned? J Am Coll Surg. 2013;216:591–596; discussion 596-­598. Fan CJ, Hirose K, Walsh CM, et al. Laparoscopic total pancreatectomy with islet autotransplantation and intraoperative islet separation as a treatment for patients with chronic pancreatitis. JAMA surgery. 2017;152:550–556. Guda NM, Muddana V, Whitcomb DC, et  al. Recurrent acute pancreatitis: international state-­ of-­ the-­ science conference with recommendations. Pancreas. 2018;47:653–666. Jalaly NY, Moran RA, Fargahi F, et al. An evaluation of factors associated with pathogenic PRSS1, SPINK1, CTFR, and/or CTRC genetic variants in patients with idiopathic pancreatitis. Am J Gastroenterol. 2017;112:1320–1329. John GK, Singh VK, Moran RA, et  al. Chronic gastrointestinal dysmotility and pain following total pancreatectomy with islet autotransplantation for chronic pancreatitis. J Gastrointest Surg. 2017;21:622–627. John GK, Singh VK, Pasricha PJ, et al. Delayed gastric emptying (DGE) following total pancreatectomy with islet auto transplantation in patients with chronic pancreatitis. J Gastrointest Surg. 2015;19:1256–1261. Lennon AM, Manos LL, Hruban RH, et al. Role of a multidisciplinary clinic in the management of patients with pancreatic cysts: a single-­center cohort study. Ann Surg Oncol. 2014;21:3668–3674. Moran RA, Klapheke R, John GK, et al. Prevalence and predictors of pain and opioid analgesic use following total pancreatectomy with islet autotransplantation for pancreatitis. Pancreatology. 2017;17:732–737. Olesen SS, Krauss T, Demir IE, et al. Towards a neurobiological understanding of pain in chronic pancreatitis: mechanisms. Quartuccio M, Hall E, Singh V, et  al. Glycemic predictors of insulin independence after total pancreatectomy with islet autotransplantation. J Clin Endocrinol Metab. 2016;102:801–809. Whitcomb DC, Shimosegawa T, Chari ST, et al. International consensus statements on early chronic pancreatitis. recommendations from the working group for the international consensus guidelines for chronic pancreatitis in collaboration with the International Association of Pancreatology, American Pancreatic Association, Japan Pancreas Society, pancreasFest working group and European Pancreatic Club. Pancreatology. 2018.

Spleen

Splenectomy for Hematologic Disorders Ciro Andolfi, MD, and Marco P. Fisichella, MD

T

he spleen is a lymphoid organ that has a dual function: hemopoietic, especially during the in utero period, and immune. The following hematologic disorders and indications are the main reasons for elective splenectomy: hypersplenism, symptoms related to splenomegaly, and decreased blood counts related to sequestration or autoimmune damage. Hypersplenism is a common complication and is defined as splenomegaly with cytopenia. Symptoms associated with splenomegaly include abdominal pain, early satiety, weight loss, and abdominal distension. In addition, splenectomy should be considered for patients with unexplained splenomegaly. Splenectomy for hematologic conditions rarely leads to a cure of the underlying hematologic disorder, but it may be beneficial for the resolution of hematologic abnormalities and improving symptoms. Hematologic diseases that require splenectomy may be broadly classified as red blood cell (RBC), platelet, lymphoproliferative, and myeloproliferative disorders (Box 1). nn RED

CELL MEMBRANE AND HEMOLYTIC DISORDERS

Autoimmune Hemolytic Anemia Autoimmune hemolytic anemia (AIHA) is secondary to antibodies directed against one or several components of the RBC surface. AIHA should be suspected in patients with anemia, reticulocytosis, elevated lactate dehydrogenase, low haptoglobin, and indirect hyperbilirubinemia. According to its immunochemical characteristics, AIHA can be classified as warm (WAIHA) or cold (CAIHA), based on the results of direct agglutinin test (DAT). If the DAT is positive for immunoglobulin G (IgG) alone, or IgG and complement 3d, then the diagnosis is most probably WAIHA. Conversely, if the DAT is positive for complement 3d alone, then CAIHA is most likely the diagnosis. In WAIHA, polyclonal IgG (sometimes IgA) autoantibodies, usually directed toward Rh antigens, form a light coat over RBCs that are removed by the spleen. The peak incidence of WAIHA occurs between the ages of 40 and 70, but it can occur at any age. In children, the disease is often self-­limited, occurring after a viral infection and resolving in 2 to 3 months. Initial treatment with corticosteroid therapy (prednisone; 1 mg/kg per day) usually results in improved hemoglobin levels within several days, and remission occurs in 80% of patients. Children generally respond better to steroid therapy than adults. The steroid dose is tapered gradually to the lowest dose needed to control hemolysis. More recently, rituximab, a monoclonal antibody directed against the CD20 antigen on the surface of mature

B lymphocytes, has replaced the other second-­line medications and it is often associated with a complete response in both the adults and children. Splenectomy should be reserved to older patients who are cortico-­resistant, for cortico-­dependent patients who require high doses of corticosteroids, for patients who fail to achieve remission within 3 weeks, or for those who cannot maintain acceptable hemoglobin levels. A response rate of 60% to 80% is usually seen within the first 2 weeks after surgery. Approximately 50% of patients will require postoperative low-­dose steroids (15 mg/d) to maintain adequate hemoglobin concentrations. In CAIHA, monoclonal IgM autoantibodies are active at lower temperatures, usually between 4℃ and 25℃, resulting in intravascular complement-­mediated hemolysis. In these patients, RBCs are removed by the liver, rather than the spleen. CAIHA makes up 15% to 25% of AIHA. This disorder is usually associated with infections, such as Epstein-­Barr virus, or with lymphoproliferative disorders. Treatment consists of avoiding cold temperatures, which can prevent an acute hemolytic crisis. Corticosteroid therapy is ineffective and contraindicated because of the risk of infections. Alkylating agents such as chlorambucil and cyclophosphamide have been successfully used for treatment along with plasmapheresis. Splenectomy is also not indicated because of the intravascular location of the hemolysis. Currently, rituximab is the only treatment that offers prolonged response. 

Hereditary Spherocytosis Hereditary spherocytosis (HS) is characterized by the presence of spherocytes on peripheral blood smear, hemolytic anemia, and increased RBC clearance by the spleen. HS is the most common congenital anemia, prompting splenectomy with a prevalence of 1 in 5000 people in Europe and North America. The disorder is also common in Japanese and African populations. HS, also known as Minkowski-­Chauffard disease, is an inherited disease resulting from a genetic mutation encoding red cell membrane components: alpha 1-­spectrin (A1SPT), β-­spectrin (BSPT), ankyrin (ANK1), and band 3 anion transport protein (B3ATP). However, some less common variants are inherited through an autosomal recessive pattern (protein 4.2). Membrane protein mutations lead to the destabilization of the lipid bilayer, with subsequent release of lipids from the membrane surface and consequent sphering of the RBCs. Spherocytes have decreased deformability, which impairs their passage through the splenic pulp and increases osmotic fragility; therefore, spherocytes are prematurely destroyed in the spleen. HS may manifest as a mild or severe form. In mild forms, patients may be asymptomatic or suffer only mild jaundice. Patients with more severe forms may present with anemia, jaundice, splenomegaly, and cholelithiasis. Peripheral blood smear demonstrates spherocytes and reticulocytes. Splenectomy is curative for the majority of patients with severe forms and it is indicated in the presence of growth retardation, skeletal changes, symptomatic hemolytic disease, anemia-­induced organ dysfunction, leg ulcers, or development of extramedullary hematopoietic tumors. 605

606

Splenectomy for Hematologic Disorders

BOX 1  Hematologic Disorders for Which Splenectomy May Be Indicated Red Cell Membrane and Hemolytic Disorders Autoimmune Disorders Autoimmune hemolytic anemia  Structural Abnormalities Hereditary spherocytosis Hereditary elliptocytosis Hereditary pyropoikilocytosis Hereditary stomatocytosis Hereditary xerocytosis  Hemoglobinopathies Thalassemia Sickle cell anemia  Enzymopathies Pyruvate kinase deficiency Glucose-­6-­phosphate dehydrogenase deficiency  Platelet Disorders Immune thrombocytopenia Thrombotic thrombocytopenic purpura  Lymphoproliferative and Myeloproliferative Disorders Hodgkin’s lymphoma Non-­Hodgkin’s lymphoma Hairy cell leukemia Chronic lymphocytic leukemia Chronic myelogenous leukemia Primary myelofibrosis  Miscellaneous Disorders Gaucher’s disease Amyloidosis Sarcoidosis Felty’s syndrome   

Preoperative abdominal ultrasonography should be performed for patients undergoing surgery, and cholecystectomy should be taken into account for patients with gallstones. Splenectomy is usually delayed until the age of 5 years to decrease the risk of overwhelming postsplenectomy infection (OPSI). There may be a role for partial splenectomy in younger children as some studies have shown clinical improvements with splenic function preservation. 

Hereditary Elliptocytosis Hereditary elliptocytosis (HE) is a rare disorder that results from mutation of the RBC membrane skeleton proteins spectrin, protein 4.1R, and glycophorin C. HE has a prevalence of 3 to 5 per 10,000 in the United States. Inheritance usually follows an autosomal dominant pattern, and the disorder is more common in people of African and Mediterranean descent. The true incidence is unknown because of the wide variety of clinical presentations. Most patients with the dominant inheritance are asymptomatic with a mild compensated anemia or no anemia at all. Affected cells are morphologically characterized by biconcave elliptocytes, or rod-­shaped cells. These cells are much more deformable than spherocytes, and patients have a less severe clinical course. In contrast, the rare autosomal recessive form can lead to severe hemolysis. Patients with mild HE, who are asymptomatic and without evidence of hemolysis, do not require treatment. Patients with chronic hemolysis may require blood transfusions and daily folic acid. Splenectomy is curative and indicated for patients with symptomatic anemia. 

Hereditary Pyropoikilocytosis Hereditary pyropoikilocytosis (HP) is an autosomal recessive hemolytic anemia with micropoikilocytosis and thermal instability. HP represents a subtype of HE, arising from the same molecular defects. Patients with HP usually have a common HE mutation from one parent and a milder subclinical defect in spectrin from the other parent. The disease usually is seen as anemia and jaundice in newborns and infants. Splenectomy is curative and indicated for patients with severe anemia. 

Hereditary Stomatocytosis (Hydrocytosis) and Xerocytosis (Desiccytosis) Hereditary stomatocytosis and xerocytosis are rare autosomal dominant hemolytic anemias characterized by a variable clinical course. In stomatocytosis, the underlying defect leads to increased erythrocyte permeability and volume. Stomatocytes have a mouth-­shaped area with central pallor. In xerocytosis, there is a decrease in intracellular cation content and cell volume, and patients presents with target cells and spiculated cells on peripheral blood smear. For cases of severe hemolysis, splenectomy may improve symptoms but does not fully correct the hemolysis. In these patients, the role for splenectomy should be considered carefully because they can develop severe complications, such as hypercoagulability, leading to catastrophic thrombotic episodes, and chronic pulmonary hypertension. Fortunately, most patients have a mild clinical course and do not require surgical intervention. 

Thalassemia Thalassemia is an autosomic recessive genetic disease characterized by insufficient production of alpha-­globin protein for alpha-­ thalassemia and beta-­ globin protein for beta-­ thalassemia. The clinical manifestations associated with thalassemia arise from quantitatively imbalanced accumulation of globin subunits and inadequate hemoglobin production. The beta subtype is the most common form of thalassemia in the United States and occurs mainly in patients of Italian and Greek descent. Patients who have the heterozygous form of β-­thalassemia (thalassemia minor) are usually asymptomatic with microcytosis and mild anemia. The homozygous form (thalassemia major or Cooley’s anemia) is much more severe. Patients are usually asymptomatic until age 6 months because of the presence of fetal hemoglobin, they then develop severe hemolytic anemia, abdominal swelling, growth retardation, irritability, jaundice, pallor, splenomegaly, pigmented gallstones, and skeletal abnormalities. Laboratory values show a severe microcytic anemia with nucleated RBCs, anisocytosis, and poikilocytosis. Patients may also have mild neutropenia and thrombocytopenia. Treatment requires monthly lifelong transfusions of RBCs, in association with iron chelators to avoid iron overload. Splenectomy is reserved to patients with increased blood transfusions in the setting of hypersplenism. Patients requiring more than 180 to 200 mL/kg/yr of RBCs are possible candidates for splenectomy. Usually, a 25% to 60% reduction in transfusion requirements is expected after splenectomy. 

Sickle Cell Anemia Drepanocytosis (sickle cell anemia) is another autosomal recessive genetic disease resulting from an anomaly of the hemoglobin β-­chain that leads to the formation of hemoglobin S, which polymerizes under hypoxemic stress, inducing a characteristic sickle deformation in the shape of red cells. Sickle cells cause stasis and vascular occlusion, leading to tissue ischemia, severe pain, and chronic organ tissue damage. Exacerbations of symptoms are referred to as sickle cell crises. Management of this disease relies on general measures, such as hydration and transfusions, to prevent vaso-­occlusive events and related complications. Patients who are homozygous for the disorder

SPLEEN

have sickle cell disease, and many suffer autosplenectomy by an early age as a result of multiple infarcts. Therefore, splenectomy is rarely indicated but should be considered in the following situations: (1) after a major acute sequestration crisis, which is an absolute life-­ threatening emergency that requires transfusion; (2) hypersplenism, causing abdominal pain and increased transfusion requirements; and (3) splenic abscess, a rare but classic complication, enhanced by repeated infarctions. Acute splenic sequestration has high mortality, up to 15%, and it is characterized by massive splenomegaly, acute exacerbation of anemia, and hypovolemia. This is primarily treated with restoration of blood volume and RBC mass, but recurrence is common. Splenectomy should be considered to prevent further episodes. 

Pyruvate Kinase Deficiency Pyruvate kinase deficiency (PKD) is the most common genetic defect causing congenital nonspherocytic hemolytic anemia. PKD is an autosomal recessive disease that occurs when a defect in the glycolytic pathway results in deficiency of adenosine triphosphate. RBCs are less deformable and often are destroyed in the spleen, leading to splenomegaly. Hemolysis can be exacerbated by acute infections and pregnancy. Patients with PKD have mild to severe anemia and splenomegaly. Clinically, these patients have mild symptoms resulting from the elevated levels of 2,3-­DPG in RBCs, which result in a right shift of the hemoglobin-­oxygen dissociation curve. This means that affected individuals have an increased capacity to release oxygen into tissues, enhancing oxygen delivery. Treatment is symptomatic and splenectomy is only rarely indicated for patients with the severe hemolytic variants of PKD. 

Glucose-­6-­Phosphate Dehydrogenase Deficiency Glucose-­ 6-­ phosphate dehydrogenase deficiency is a widespread (about 400 million people worldwide are carriers) X-­linked genetic disease of the glutathione pathway, which leads to damage of RBCs by toxic oxygen products. It causes acute hemolytic accidents after oxidative stress (acute infections, oxidant medications, and fava beans), or, rarely, chronic hemolytic anemia. In adult patients, glucose-­6-­ phosphate dehydrogenase deficiency does not usually require transfusion, except during severe hemolytic crises; management focuses on the avoidance of trigger food and medications. Splenectomy is not indicated, except in patients with severe hypersplenism, which requires transfusion dependency.  nn PLATELET

DISORDERS

Immune Thrombocytopenia Immune thrombocytopenia (ITP; formerly idiopathic/immune thrombocytopenic purpura) is an autoimmune disease caused by one or several antiplatelet antibodies directed against platelet glycoprotein complexes (GPIIb/IIIa, GPIb/IX). ITP leads to platelet destruction by the reticuloendothelial system in the spleen. In addition to humoral immunity, there is a component of cell-­mediated immunity involved in this process. This condition is characterized by isolated thrombocytopenia with a platelet count that falls below 100,000/mm3. Specific therapy is indicated when platelet count is less than 30,000/mm3, but is also related to bleeding risk, presence of comorbidities, and risks of trauma. Primary ITP is a diagnosis of exclusion as other illnesses, such as human immunodeficiency virus infection, systemic lupus erythematosus, antiphospholipid antibody syndrome, hepatitis C virus, and lymphoproliferative disorders, can cause secondary ITP. In addition, some medications and drugs may elicit similar immune-­mediated platelet destruction (cocaine, antibiotics, antihypertensives, anti-­ inflammatories, heparin, quinidine, and abciximab). The prevalence of ITP in adults is about 5 per 100,000 people, occurring nearly twice as frequently in women. There is an approximately fourfold increase

607

in prevalence in adults older than 55 years of age. Most patients with ITP are asymptomatic, whereas symptoms occur when platelet counts drop below 30,000/mm3. Symptoms include bruising, purpura, petechiae, bleeding from the oral mucosa, epistaxis, menorrhagia, and gastrointestinal bleeding. Less than 1% of patients experience intracerebral hemorrhage, the most severe complication. The prevalence of ITP in children is approximately 12 per 100,000 in girls and 9 per 100,000 in boys. Children may present with a sudden onset of petechiae or purpura, usually days to weeks after an infectious disease. In the pediatric population, ITP is usually self-­limited, with more than 70% remission within 6 months; the risk of intracerebral hemorrhage is less than 0.2%. Observation is a possible treatment option as long as the platelet count is greater than 20,000/mm3. Conservative treatment is also reasonable in adults with platelet counts above 30,000/ mm3. However, patients who exhibit persistent thrombocytopenia or platelet counts below 30,000/mm3 (20,000/mm3 in children), should start corticosteroid therapy. First-­line treatment consisting of oral corticosteroids (typically prednisone) leads to complete remission in 10% to 30% of patients; high-­dose corticosteroids (usually dexamethasone) leads to complete remission in 60% to 80% of patients. The standard initial dose is 1 to 2 mg/kg/d of prednisone for 2 to 4 weeks followed by a steroid taper. Several second-­line therapy options are available including azathioprine, cyclosporine, cyclophosphamide, danazol, or dapsone, none of which has been shown to be superior to other therapies. More recently, new medications have been introduced. Rituximab, by its immunosuppressant effect, decreases the production of antiplatelet antibodies leading to a response in approximately 60% of patients and complete remission in about 40% of patients. Synthetic thrombopoietin receptor agonists have been recently developed. The two main medications used today, romiplostim and eltrombopag, lead to high rates of complete response, but relapses are frequent. Intravenous immunoglobulin (1 mg/kg/d for 1 to 2 days) can be considered for patients who would benefit from a rapid increase in platelet count (i.e., in the setting of bleeding or in preparation for an invasive procedure) or for those who are unable to tolerate steroids. Splenectomy has long been considered as the treatment of choice in the wake of corticosteroid failure because patients respond in 80% of cases with a durable 5-­year response, without additional treatment in 66% of patients. Splenectomy is indicated for refractory thrombocytopenia, relapses requiring multiple cycles of therapy, or in patients who have experience side effects of medical treatment. If perioperative platelet transfusion is needed, transfusion should be withheld until the splenic artery has been ligated. Several studies have reported that splenectomy can be safely performed, with minimal bleeding risk, even in patients with platelet count below 10,000/mm3. 

Thrombotic Thrombocytopenic Purpura Thrombotic thrombocytopenic purpura (TTP) is a severe form of thrombotic microangiopathy characterized by the association of mechanical hemolytic anemia, peripheral platelet consumption, and microthrombosis involving different organs. TTP is a disorder in which a deficiency of the ADAMS13 protein leads to increased platelet aggregation and subsequent microvascular thrombosis. The interaction between von Willebrand factor (vWF) and platelets is usually controlled by the ADAMS13 protein, which cleaves von Willebrand factor and prevents platelet aggregation. TTP may occur spontaneously but it is often precipitated by different factors such as chemotherapeutic agents (gemcitabine, mitomycin C, or calcineurin inhibitors), quinine, cyclosporine, clopidogrel, ticlopidine, hematopoietic stem cell transplantation, or pregnancy. The annual incidence of TTP is 4 to 10 cases per million. TTP is characterized by microangiopathic hemolytic anemia, severe thrombocytopenia, fever, neurologic complications, and renal failure. Patients have petechiae (most commonly on the lower extremities), fever, myalgia, and fatigue. Neurologic symptoms include headache, mental status changes, seizures, and even coma. Patients can develop congestive heart failure or

608

Splenectomy for Hematologic Disorders

cardiac arrhythmias. TTP is suspected with microangiopathic hemolytic anemia and thrombocytopenia in the setting of elevated lactate dehydrogenase and bilirubin, a negative Coombs test, and a peripheral blood smear demonstrating schistocytes, nucleated RBCs, and basophilic stippling. Initial therapy consists of daily plasma exchange. Plasmapheresis is carried out with a goal of exchanging 1.5 plasma volumes. Approximately 70% of patients will respond to this therapy. Platelet transfusions are generally not recommended because of the risk of severe clinical deterioration. Rituximab (anti-­CD20 antibody) and glucocorticoids are second-­line therapies. Until the 1970s, splenectomy was the only treatment modality for TTP. Now, splenectomy is reserved for refractory thrombocytopenia or relapses. When combined with high-­dose steroid therapy, splenectomy has been shown to improve disease-­free interval; however, the response rate is only 40%.  nn LYMPHOPROLIFERATIVE AND

MYELOPROLIFERATIVE DISORDERS

Hodgkin’s Lymphoma Hodgkin’s lymphoma is a malignant neoplasm of lymphoreticular cell origin that usually affects young adults in their second and third decades of life. Primary treatment consists of chemotherapy and/or radiation. Historically, splenectomy was performed as part of a staging laparotomy that included lymph node sampling and liver biopsy. Now, staging laparotomy has been replaced by imaging modalities such as computed tomography (CT) and positron emission tomography. Splenectomy may be beneficial for patients who develop splenomegaly and related thrombocytopenia. 

Non-­Hodgkin’s Lymphoma Non-­Hodgkin’s lymphoma represents the most common primary neoplasm with splenic involvement occurring in 65% to 80% of cases. Splenectomy is indicated for symptoms related to massive splenomegaly and cytopenia resulting from splenic sequestration. Splenectomy may be helpful to assist with diagnosis and to determine appropriate therapy. This may occur in situations in which patients have failed therapy or when inadequate tissue is available for proper histologic or cytometric analysis. There are some subtypes of non-­Hodgkin’s lymphoma that involve the spleen more than others. Splenic marginal zone lymphoma represents a particular entity of indolent B-­cell lymphoma in which the spleen is often the only organ macroscopically involved. Marginal zone lymphomas require strict surveillance and the indication for splenectomy should be discussed in symptomatic patients. However, splenectomy in these patients has been shown to lead to partial or complete remission because the spleen is the site of lymphoma origin. The discovery of isolated splenomegaly without any obvious etiology calls for complete work-­up, which leads to two possible scenarios: (1) splenomegaly remains isolated without a clear diagnosis of hematologic disease and splenectomy should be envisioned for diagnostic purposes. The alternative to splenectomy is a CT-­guided biopsy, which is diagnostic in 80% to 90% of cases; (2) when definitive diagnosis has been established, splenectomy might be part of the therapeutic strategy. However, in case of splenic mantle cell lymphoma, follicular lymphoma or diffuse large B-­cell lymphoma, the indications for splenectomy are quite limited. 

Hairy Cell Leukemia Hairy cell leukemia is a rare disease, representing 2% of all leukemias. Patients have fatigue, left upper quadrant abdominal pain, fever, infection, and/or coagulopathy. The disease is characterized by B lymphocytes that possess cytoplasmic projections from the cell membrane (“hairy cells”). This is an indolent disease that commonly occurs in the fifth decade of life with splenomegaly (80% to 90% of patients), pancytopenia, neoplastic peripheral mononuclear cells, and bone marrow infiltration. Pancytopenia is caused by hypersplenism and replacement of bone marrow by leukemic cells. In the past,

splenectomy was considered the standard of care with a 40% to 70% improvement in the hematologic cell lines for up to 10 years. Recently, treatment with purine analogs, such as pentostatin and cladribine, has replaced splenectomy as primary therapy. These agents have proven response rates of 92%, with a complete remission rates of 80% and a 90% 10-­year survival. Splenectomy is currently reserved for cases of incomplete response to first-­line therapy, persistent splenomegaly in the absence of bone marrow involvement, atraumatic splenic rupture, and severe bleeding. 

Chronic Lymphocytic Leukemia Chronic lymphocytic leukemia represents a B-­cell leukemia in which there is progressive accumulation of functionally impaired lymphocytes. Chronic lymphocytic leukemia usually arises after the fifth decade of life and is more common in men than in women. Splenic infiltration is common and can lead to severe splenomegaly and cytopenias from hypersplenism. Splenectomy is indicated to relieve symptoms associated with massive splenomegaly, such as abdominal pain, distension, and early satiety. Splenectomy for the treatment of severe thrombocytopenia and anemia, in the setting of secondary ITP or AIHA, has a 60% to 70% hematologic response rate and has been shown to improve survival. 

Chronic Myelogenous Leukemia Chronic myelogenous leukemia (CML) is a disorder resulting from an abnormal proliferation of granulocytes. Ninety-­five percent of CML patients have a chromosomic translocation between chromosomes 9 and 22 [t(9;22)] leading to fusion of the breakpoint cluster region and Abelson leukemia virus gene. CML may occur in childhood but it is mainly found in adults with a mean age of 65 years. Diagnosis is commonly made during the chronic phase, which is characterized by splenomegaly in 40% of patients. Despite medical therapy, the disease can progress to an accelerated phase with development of fever, night sweats, weight loss, bone pain, increased white blood cell count, and splenomegaly. An acute blastic crisis with splenomegaly and hypersplenism can occur, resulting in severe anemia, bleeding, and infections. Current first-­line therapy is with imatinib, a tyrosine kinase inhibitor. Bone marrow transplantation or interferon-­ alpha can be used in cases of poor response or relapse. Splenectomy has not shown any survival benefit in the early chronic phase or before bone marrow transplantation, but it may be offered as palliative treatment in patients with severe symptoms due to splenomegaly and hypersplenism. 

Primary Myelofibrosis (Myelofibrosis With Myeloid Metaplasia) Primary myelofibrosis (PMF) is a chronic malignant hematologic disorder that results in hyperplasia of abnormal myeloid precursor cells leading to marrow fibrosis and extramedullary hematopoiesis in the liver and spleen. This can lead to significant splenomegaly, cytopenia, and portal hypertension secondary to venous thrombosis. PMF is prevalent in patients with history of radiation or toxic industrial agent exposure. It is more common in men than women, with an average age of 65 years. Splenectomy is indicated for patients who develop hemolysis requiring significant transfusions, thrombocytopenia, symptomatic splenomegaly, recurrent splenic infarctions, hypercatabolic symptoms (anorexia, fatigue, fever, night sweats, weight loss), and portal hypertension with refractory ascites and variceal hemorrhage. Splenectomy in progressive multifocal leukoencephaly has a considerable risk of morbidity (15%–30%) and mortality (10%) and should be performed in select patients only. Splenectomy in patients with progressive multifocal leukoencephaly has been associated with hemorrhage, infection, leukocytosis, severe thrombocytosis (18%–50%), progressive hepatomegaly (12%–29%), fatal hepatic failure (7%), and leukemic

SPLEEN

609

transformation (11%–20%). The appropriate use of palliative splenectomy in PMF can result in improved quality of life for patients who are unresponsive to conventional treatment. Postoperative use of platelet-­lowering agents such as hydroxyurea, interferon-­alpha, aspirin, and anagrelide has been shown to reduce thrombotic complications. Ligation of the splenic vein at its confluence with the superior mesenteric vein has been described to improve laminar flow and decrease portal vein thrombosis. Compensatory massive hepatic enlargement can be treated with low-­level radiation and chemotherapy.  nn MISCELLANEOUS

DISORDERS

Amyloidosis Amyloidosis is a common disorder that results in extracellular deposition of insoluble fibrillar proteins in tissues and organs. Hepatosplenomegaly may occur in 25% of patients, and severe splenomegaly is seen in approximately 10% of individuals. Splenectomy is indicated for symptomatic splenomegaly. In addition, patients with severe hepatic dysfunction may develop coagulopathy associated with factor X deficiency. In these patients, splenectomy may improve factor X levels. Perioperative administration of factor VIIa is important to control bleeding in patients undergoing surgery. However, splenectomy does not modify the ultimate course of the disease. 

Gaucher’s Disease Gaucher’s disease is a glycolipid storage disease resulting from a deficiency of beta-­glucosidase (glucocerebrosidase). This leads to deposition of glucocerebroside in the reticuloendothelial system with severe organomegaly, pulmonary infiltrates, and bone marrow infiltration. Patients can have anemia, thrombocytopenia, osteopenia, bone pain, osteonecrosis, and massive hepatosplenomegaly. Splenectomy is indicated for severe and symptomatic splenomegaly and refractory cytopenia. Partial splenectomy has been advocated in children with Gaucher’s disease to preserve splenic function. The spleen is a reservoir for storage material; therefore, splenectomy can result in redistribution and deposition in other organs causing severe bone disease (tenfold increased risk of osteonecrosis) and worsening of lung and kidney function. 

Felty’s Syndrome Felty’s syndrome includes rheumatoid arthritis, neutropenia, and splenomegaly. In 85% of cases this disease is associated with an HLA DR4 antigen. Patients presents with chronic infections as a result of neutropenia, especially when neutrophils are below 0.5 × 109/mm3. First-­line treatment consists of low-­dose methotrexate or disease-­ modifying antirheumatic drugs. Granulocyte colony-­ stimulating factor may be used for treatment failures, in cases of increased infection risk, or before surgery. Splenectomy is indicated when medical treatment fails, and usually results in 80% hematologic response rate. Unfortunately, infectious complications may still recur and do not always correlate with granulocyte counts. 

Sarcoidosis Sarcoidosis is a noncaseating granulomatous disease. Although 90% of patients have primary lung involvement, it can affect every organ in the body. Primary splenic sarcoidosis is very rare, and splenic involvement is often found as part of a multiorgan disease. Up to 40% of patients with sarcoidosis have splenomegaly and 3% have massive splenomegaly. Treatment is mostly conservative and includes corticosteroids or methotrexate. Indications for splenectomy include: splenomegaly and hypersplenism, intractable pain, and exclusion of a neoplastic process. Splenectomy does not alter the course of sarcoidosis but has been shown to improve refractory hypercalcemia. 

FIG. 1  Accessory spleen.

Idiopathic Splenomegaly In the setting of splenomegaly without a clear cause, splenectomy has a diagnostic and therapeutic role. Studies have revealed a 40% to 70% occurrence of lymphoma in this patient population. Most of these patients do not exhibit any signs of malignancy or lymphadenopathy. Tissue obtained through splenectomy may be the only way to perform appropriate histopathologic and cytologic diagnosis. When hypersplenism is present, splenectomy can alleviate symptoms of splenomegaly and correct cytopenia.

Preoperative Considerations Preoperative imaging with ultrasound and/or CT scan is critical for operative planning. CT provides information regarding anatomic relationships, spleen size, vascular anatomy, presence of accessory spleens (Fig. 1), perisplenic lymphadenopathy, and inflammation. The normal spleen measures about 11 cm in length. Moderate splenomegaly, from 11 to 25 cm, should be noted in preoperative planning. Massive splenomegaly, greater than 25 cm length, may change preoperative and intraoperative strategy (Fig. 2). Although not indicated for a normal-­sized spleen, preoperative splenic artery embolization can be useful in patients with massive splenomegaly, to prevent excessive blood loss in the setting of severe thrombocytopenia, or in patients who do not wish to receive blood transfusions. In addition, embolization helps reduce the spleen size before laparoscopic resection. Timing is important because patients can develop significant pain from infarcted splenic tissue; we suggest performing angioembolization within 24 hours before surgery. A broad-­spectrum antibiotic prophylaxis should be administered at the time of induction to anesthesia and continued postoperatively for at least 24 hours. Low-­ molecular-­weight heparin should be administered subcutaneously before induction of anesthesia and should be continued postoperatively for up to 1 month as prophylaxis for splanchnic thrombosis. The use of an orogastric or nasogastric tube can reduce gastric distension and improve visualization and dissection of the short gastric vessels along the greater curvature of the stomach. Blood products must be available intraoperatively, especially platelets for patients with severe thrombocytopenia. Prophylactic platelet transfusions are typically given only when the platelet count is below 50,000 and the platelets are administered only after the splenic artery has been ligated. If patients have been treated with chronic corticosteroids, stress dose steroids should be administered with a rapid taper postoperatively. In elective cases, it is recommended to vaccinate patients against encapsulated organisms (Haemophilus influenzae B, polyvalent pneumococcus, and meningococcus vaccines) 2 weeks before splenectomy. If splenectomy is emergent, the patient should be vaccinated postoperatively. 

610

Splenectomy for Hematologic Disorders

12-mm camera port

10-12-mm camera port 5-mm camera port 5-mm optional port

FIG. 2  Massive splenomegaly in a patient who underwent splenectomy for complications of a lymphoproliferative disorder.

Surgical Procedure Laparoscopic splenectomy has become the standard approach for performing splenectomy in patients with hematologic disorders. The laparoscopic approach provides the advantages of shorter length of stay, decreased postoperative pain, and decreased morbidity. Recent studies have shown a trend toward shorter operative times that are comparable to open splenectomy in cases of normal or moderately enlarged spleens. Most data show comparable detection of accessory spleens that can result in disease recurrence in cases of autoimmune hematologic disorders. However, laparoscopic splenectomy is not a forgiving procedure. Methodical control of the hemostasis is the key for success. The splenic parenchyma is fragile, has a rich blood supply, and is particularly vulnerable to capsular tear and hemorrhage. Understanding the variation of splenic anatomy is essential for a safe intraoperative management. Much of the controversy surrounding laparoscopic splenectomy involves the size of the spleen. The normal adult spleen measures up to 11 cm in length and weighs approximately 80 to 300 g. Moderate splenomegaly generally is defined as a spleen that is 11 to 25 cm, and massive splenomegaly represents a spleen that is more than 25 cm. We believe that laparoscopic splenectomy can be performed safely in patients with splenomegaly. Factors to be considered should include medical comorbidities, indication for surgery, blood counts, coexisting coagulopathy, and history of previous splenic irradiation. The hand-­assisted laparoscopic approach may be useful for inexperienced surgeons to shorten the learning curve and to allow rapid control of hilar vessels and assistance with retraction. Nevertheless, open surgery should never be considered a failure and may be the safest approach in some cases. Laparoscopic splenectomy can be performed with the patient in lateral decubitus position or supine. A bean bag can be used to facilitate positioning for surgeons who prefer the right lateral decubitus approach. A split-­leg bed can be helpful when the patient is supine; this also allows the surgeon to stand between the patient’s legs. For the anterior approach, port placement generally includes a 12-­mm periumbilical camera port, and three to four additional ports in a V-­shaped placement adjacent to the left upper quadrant, with the initial port for the camera at the base of the V. One line of the V extends from the initial port to the xiphoid process; the other line of the V extends from the initial port to the most lateral left subcostal region. Two dissection ports are placed, one near the midline and one along the lateral V line. A 10-­to 12-­mm port at this location may be preferable because it is likely the port for introduction of an endoscopic stapler or Endoclip device. An additional port for retraction is placed further lateral at the anterior axillary line (Fig. 3). If a fifth port is necessary for retraction, it is placed in the subxiphoid region. Access to the abdomen can be

FIG. 3  Laparoscopic splenectomy, anterior approach.

gained using a Veress needle, an optical trocar, or an open approach depending on the surgeon’s preference. The patient is then placed in reverse Trendelenburg and tilted slightly to the right. For patients with large liver or splenomegaly, a self-­retaining liver retractor, such as the Nathanson device with fast-­clamp, can facilitate visualization. For the lateral approach, the position is similar to that used for posterolateral thoracotomy and/or laparoscopic left adrenalectomy. Patients are initially positioned supine on a beanbag. Once general anesthesia is established and the airway is secured, the operative team repositions the patient in lateral decubitus with the right side down. The kidney rest is raised, and the operating table is flexed. The goal is to maximize the working space between the left costal margin and the left anterior superior iliac spine. The umbilicus is avoided, and the first port is positioned approximately one-­third the distance from the umbilicus to the splenic hilum. After securing access to the peritoneal cavity, typically three additional ports are placed along the costal margin. Depending on the spleen size and body habitus, it may be necessary to place the trocars inferiorly or medially. A 10-­to 12-­mm port, capable of accommodating an endostapler or large Endoclip device, is typically placed in the left subcostal anterior axillary line. A 5-­mm port is placed in the left subcostal region in the midaxillary line. A fourth port, usually 5 mm, is placed in the far-­left lateral subcostal position. Occasionally, an additional port is required for retraction toward the midline, near the xiphoid process (Fig. 4). The abdomen is explored, paying careful attention to identify any accessory spleens. The liver should also be inspected for signs of cirrhosis. The splenocolic ligament is mobilized and divided with an energy device. This allows for further mobilization and inferior retraction of the splenic flexure of the colon. The gastrosplenic ligament and the short gastric vessels then are divided using an ultrasonic energy device, endoscopic metallic clips, or bipolar energy device. This dissection should be carried up to the level of the left crus, and the stomach can be retracted to the right. The splenorenal ligament then is dissected to identify the splenic artery and splenic vein within the splenic hilum (Fig. 5). These structures then are divided using a vascular load on an endoscopic linear stapling device. The splenophrenic ligament is divided last because it maintains cephalad/lateral retraction of the spleen during dissection of the hilar vessels. The spleen is then placed into an endoscopic bag and morcellated. After extraction, the splenic bed, hilum, and greater curvature of the stomach should be inspected thoroughly to ensure hemostasis. At this point, the abdomen should be examined again for splenunculi or accessory spleens. The most common locations for splenunculi are the gastrosplenic ligament and greater omentum. For open splenectomy, a midline or left subcostal incision may be used. The midline incision may be preferable in patients with massive splenomegaly or with a narrow costal margin. 

SPLEEN

12-mm camera port

10-12-mm camera port 5-mm camera port 5-mm optional port

FIG. 4  Laparoscopic splenectomy, lateral approach.

611

BOX 2  Guidelines for OPSI Prophylaxis Incidence of OPSI • First 2 years after splenectomy: adults (0.9%) children (5.0%) • Risk factors: Age 25% devascularization

V

Shattered spleen Hilar laceration with splenic devascularization

Modified from Moore EE, Gogbill TH, Jurkovich GJ, et al. Organ injury scaling: spleen and liver (1994 revision). J Trauma. 1995;38:323–24.

A

617

or higher injuries, angioembolization may be beneficial if a nonoperative strategy is pursued in these patients (Fig. 1). Similarly, though not explicitly delineated in the AAST scale, the presence of vascular findings such as active contrast extravasation, or pseudoaneurysms on initial imaging increase the risk of failure and generally benefit from more aggressive intervention. Hypotension unresponsive to transfusion, peritonitis, or significant transfusion requirement are generally considered failure, requiring salvage angiographic or operative intervention. Transfusion thresholds vary in practice; there are those who set a threshold of any transfusion needed as an indication for operative intervention. Typically, transfusion as a threshold for operative intervention range from more than 2 to 4 units of packed red cells. Failure rate of nonoperative management for injuries of any grade is approximately 15%, and can be as high as 75% in grade V injuries. Selective management practice for splenic injuries varies greatly between adult and pediatric populations, likely secondary to the greater concern for the development of OPSI in children. Pediatric patients are more likely to undergo greater transfusion amounts than adults in the attempt to salvage the injured spleen.

Angioembolization Splenic angioembolization has become a mainstay of selective management of blunt splenic injury, allowing avoidance of operative morbidity and preservation of splenic tissue while decreasing vascular inflow to control hemorrhage. Hemorrhage control is achieved in 85% to 95% of patients referred for angioembolization; however, hybrid operative-­fluoroscopic suites notwithstanding, the presence of florid hemodynamic instability is a contraindication to angioembolization because of limited resuscitative capabilities in most angiographic suites. Additional considerations include resource availability, operator expertise, and anatomic suitability; vascular injury such as intimal dissection or perforation during angiography is an uncommon but potentially catastrophic complication of endovascular therapy. Nonetheless, this strategy, when appropriate, allows avoidance of the morbidities of laparotomy as well as limiting asplenic or hyposplenic consequences.

B FIG. 1  Axial (A) and coronal (B) views of an American Association for the Surgery of Trauma grade 3 splenic laceration.

618

SPLENIC SALVAGE PROCEDURES

A

B

FIG. 2  (A) Angiographic demonstration of splenic artery pseudoaneurysm within splenic parenchyma following abdominal packing at laparotomy. (B) The same patient status after coil embolization.

A

B

FIG. 3  (A) Combined proximal and distal coil embolization of the splenic arterial supply. (B) Axial computed tomography images of the same patient postembolization demonstrating splenic infarctions.

Embolization is typically performed with either metal coils or gelatin sponge (Fig. 2). Conflicting data exist regarding the superiority of embolization material, but tend to favor coils when differences are noted with regard to both success rate and subsequent complications, both hemorrhagic and infectious. Splenic artery angioembolization can be performed proximally at the main splenic artery or more distally or a combination of both. Proximal embolization reduces global parenchymal perfusion of the spleen (that is, theoretically decreases arterial pressure) to facilitate thrombus formation, whereas distal embolization more definitively abrogates arterial inflow to the affected region. Selection of technique is commonly driven by operator expertise and the location or extent of the injury. Failure, as defined by need for operative intervention appear to be comparable between the two strategies; rates of nonoperative rebleeding are similar as well. Distal embolization unsurprisingly incurs increased risk of infarction, and late complications thereof are uncommon but primarily consist of splenic

abscesses and cysts that may necessitate subsequent operative intervention (Fig. 3). Assessments of splenic function following proximal and distal embolization consistently demonstrate superiority over splenectomy in multiple hematologic and immunologic parameters. Disparities between the two techniques have been subtle, in part because of limited study sizes, but trend toward favoring distal embolization for greater preservation of function when any difference is detected. This finding, in the context of partial splenic infarction after distal embolization, suggests that preserved tissue after distal embolization remains functionally intact, whereas proximal embolization may compromise global function without radiographic evidence of infarction. The data remain far from definitive. Additionally, no strong recommendations can as yet be made to the need for vaccines against encapsulated bacteria following splenic embolization, so practice remains varied.

SPLEEN

619

Gelfoam Dacron felt strips

Stomach Mattress sutures approximate a deep laceration

FIG. 4  Demonstration of topical hemostatic agent and technique of pledgetted repair of splenic laceration. (From Cameron JL, Sandone C. Atlas of Gastrointestinal Surgery, vol 1, 2nd ed. Shelton, CT: People’s Medical Publishing House; 2007.)

In children, the use of angiography is held for those with evidence of ongoing bleeding but is not recommended prophylactically for patients, even those with pseudoaneurysm or blush. 

Operative Techniques Although operative management of splenic injury incurs the morbidity of abdominal access and a potential for secondary hemorrhage, operative salvage techniques retain the benefits of preserved splenic function. Use of these techniques, however, should be considered in the context of the scenario mandating operation (e.g., primary operation vs previous failure of nonoperative management, timing, concomitant injuries). Penetrating abdominal trauma, for example, may mandate abdominal exploration that precludes imaging and nonoperative strategies. Because the decision to pursue operative intervention frequently stems from an acuity that makes operative salvage higher risk or time-­consuming in an often-­times sensitive situation, and must be weighed against the more definitive hemorrhage control of splenectomy, particularly in patients at risk for coagulopathy and those who will be intolerant of even transient hypoperfusion such as in traumatic brain injury. After abdominal access, exposure of the spleen is performed in a manner paralleling that for splenectomy. Provided more urgent injuries have been addressed, the omentum is mobilized and the splenic hilum identified. The short gastric, splenocolic, splenophrenic, and splenorenal ligaments are divided if not already dissected by hematoma. The spleen can then be medialized on its main vascular pedicle adjacent to the tail of the pancreas, either manually or with the assistance of packing along the posterior peritoneum to elevate the spleen into the operative field. Care should be taken to avoid injury to the pancreas during this

dissection. Once the spleen is delivered into the field, salvage techniques or splenectomy may then be selected based on the extent and location of injury. In more minor injuries, direct thermal cautery may suffice with the argon beam (the standard electrocautery device usually does not work by itself). There are a variety of topical hemostatic agents commercially available as adjuncts to direct pressure and packing. Available as powders, foams, gels, and sheets of various dimensions, these products may contain either active or passive hemostatic agents, most commonly cellulose or kaolin. Impregnated gauzes in particular offer the ability to maintain adherence to injured tissue and simultaneously allow application of direct pressure to promote hemostasis, though not all products are appropriate to leave in the abdomen outside of damage control situations. Pledgeted suture splenorrhaphy offers more stability when anatomically practical, most commonly deeper linear lacerations (Fig. 4). As with similar repairs for hepatic injuries, pledgets are commonly fashioned from polyethylene terephthalate, fibrocellulose strips, or other hemostatic materials and secured with large mattress sutures, 0-0 or 2-­0, with care taken to avoid further lacerating parenchyma while securing the suture. If the injury extent is not amenable to topical or suture repair but is limited to one anatomic aspect of the spleen, partial splenectomy allows for splenic salvage of remaining viable tissue. (Fig. 5) After mobilization of the spleen, the distal branches of the splenic artery are isolated. The vessel supplying the injured pole can be clamped or encircled with a vessel loop to assess for adequate hemostasis on vessel occlusion; if not, an alternate technique or splenectomy should be used. The vessel is then suture ligated and sharply divided, and a demarcation of ischemic tissue identified. Depending on tissue dimensions, a linear stapler with a vascular load or cautery may be

620

Splenic Salvage Procedures

Mikulicz pad on spleen

Splenic a.

Splenic v. branch to lower pole

FIG. 5  Technique of partial splenectomy. (From Cameron JL, Sandone C. Atlas of Gastrointestinal Surgery, vol 1, 2nd ed. Shelton, CT: People’s Medical Publishing House; 2007.)

used for the parenchymal resection. The transected surface is then oversewn with running suture or can be reinforced with mattressed pledgets or omentum. When the injury pattern is not amenable to pledgeted repair or partial splenectomy, mesh splenorrhaphy remains an alternative for tamponade, but has somewhat fallen out of favor because the technique can be relatively time consuming, implants a foreign body with concomitant infectious concerns, and requires an organ sufficiently injured to preclude alternative strategies while remaining viable. A mesh of polyglactin or similar material is sized and trimmed, then a purse-­string suture is initiated hemicircumferentially. After fully mobilizing the spleen and achieving vascular control of the hilum, the mesh is wrapped around the lateral surface and then bunched around the hilar vessels with the sutured perimeter posterior. The purse string can then be completed along the unsecured circumference to the appropriate tension to achieve hemostasis in the parenchyma. Care should be taken to avoid strangulation of hilar vasculature when tightening the purse string or exerting excessive pressure on splenic tissue causing ischemia. Although damage control laparotomies have become commonplace for managing polytrauma, caution should be exercised if considering salvage in the form of abdominal packing for all but the most trivial splenic injuries. Even though prompt reexploration limits the morbidity of temporary closure, splenectomy at initial operation can be performed rapidly for definitive hemorrhage control and may facilitate correction of physiologic derangements during the resuscitative phase. 

Intensive Care Unit Considerations Patients with splenic injuries managed by salvage techniques should be serially monitored for potential rebleeding. Hemoglobin

measurements are typically obtained in intervals ranging from every 2 to 6 hours during the first 12 to 24 hours, then spaced out as the patient demonstrates hematologic and hemodynamic stability. After 24 to 48 hours, and depending on injury severity and selected management strategy, level of care can then be deescalated, but clinicians should be cognizant of the possibility for delayed splenic rupture. Delayed splenic rupture typically occurs in fewer than 5% of patients and carries a higher mortality in the 10% range, likely from delayed recognition. Although case reports for delayed rupture have described events sometimes months removed from initial injury, the first week is the typical window for this complication. Leukocytosis and thrombocytosis are commonplace following both splenectomy and splenic salvage as cells normally removed from circulation by the spleen are allowed to persist, and the relative proportions of different hematologic lineages may vary over time. Splenectomy typically produces more pronounced leukocytosis than embolization and similar degrees of thrombocytosis. However, unlike splenectomy, these alterations often normalize by between 4 and 6 weeks postintervention for angioembolization. Distinguishing expected leukocytosis from infection can be challenging, but existing literature for splenectomy suggests a white blood cell count greater than 15,000/μL and a platelet/white cell ratio less than 20 on or after postoperative day 5 are indicative of ongoing infection. Similar thresholds for angioembolization and other salvage techniques have not been established, but trends in the appropriate direction should raise the index of suspicion in these patients. Anticoagulation, most commonly as prophylaxis against venous thromboembolism, must be weighed against hemorrhagic risk following splenic salvage. Though not studied rigorously, the available evidence suggests prophylactic low-­molecular-­weight heparin can be initiated within 24 to 48 hours with comparable outcomes and transfusion requirements. Meanwhile, delay of prophylaxis in a trauma population significantly increases risk of venous thromboembolism, with reports of a significant increase after 72 hours. Optimal timing of therapeutic anticoagulation when indicated remains unclear. 

Follow-­up Late complications such as arteriovenous fistula or pseudoaneurysm formation, though rare, may occur, particularly when observation without operative or angiographic intervention is used. As the initial postinjury thrombus resorbs, pseudoaneurysms previously not appreciated may be identified. Though there are no formal recommendations regarding repeat imaging in these scenarios, a venous-­ contrast abdominal computed tomography (CT) scan is commonly selected when any imaging modality is used. Timing of repeat examinations is also controversial, and have ranged from 48 hours to 1 week in the literature depending on the center. If any abnormalities are identified at this point, angioembolization is recommended. The utility in the outpatient setting is even more unclear, often falling under the purview of primary care physicians, with Doppler ultrasound or CT being described. Injury grade is proportional with rate of healing. Most well studied in children and adolescents, specific time to radiographic healing has varied between reports, ranging from 1 to 4 months for grade I injuries and 5 to 11 months for grade IV. Consequently, return to physical activity is similarly contentious, but light activity at 1 month postinjury and full activity after 3 months has been suggested in the general population. Imaging as a tool to guide activity restrictions is not supported in existing data. Though vaccination against encapsulated organisms has been used in the past with splenic salvage, immunoglobulin titers and peripheral smear findings affirm the preservation of splenic immunologic function and suggests this is likely unnecessary in most patients after splenic salvage. Volumetric assessments of residual splenic tissue have not been well correlated with immunocompetency. The decision to vaccinate therefore should reflect the clinician’s assessment of anticipated splenic function given the nature of the injury and selected salvage technique. 

SPLEEN

nn SUMMARY Splenic salvage offers multiple benefits over splenectomy and should be attempted in patients with sufficient viable tissue and appropriate clinical risk. With injury patterns that do not mandate primary abdominal exploration, low grade (I–II) injuries can often be observed, whereas higher grade (III–V) injuries should be offered angioembolization. The decision to pursue open intervention does not preclude salvage but is generally higher risk given the underlying process mandating exploration.

Suggested Readings Alvarado AR, Udobi K, Berry S, et  al. An opportunity for improvement in trauma care: 8-­week booster vaccination adherence among patients after trauma splenectomy. Surgery. 2018;163:415–418.

621

Dionne B, Dehority W, Brett M, Howdieshell TR. The asplenic patient: post-­ insult immunocompetence, infection, and vaccination. Surg Infect (Larchmt). 2017;18:536–544. Gates RL, Price M, Cameron DB, et al. Non-­operative management of solid organ injuries in children: an American Pediatric Surgical Association Outcomes and Evidence Based Practice Committee systematic review. J Pediatr Surg. 2019. [Epub ahead of print]. Moore EE, Gogbill TH, Jurkovich GJ, et al. Organ injury scaling: spleen and liver (1994 revision). J Trauma. 1995;38:323–324. Schimmer JA, van der Steeg AF, Zuidema WP. Splenic function after angioembolization for splenic trauma in children and adults: a systematic review. Injury. 2016;4:525–530. Schnüriger B, Inaba K, Konstantinidis A, et al. Outcomes of proximal versus distal splenic artery embolization after trauma: a systematic review and meta-­analysis. J Trauma. 2011;70:252–260. Sims CA, Wiebe DJ, Nance ML. Blunt solid organ injury: do adult and pediatric surgeons treat children differently? J Trauma. 2008;65:698–703.

Hernia

Management of Inguinal Hernia Gina L. Adrales, MD, MPH, FACS, and Bethany C. Sacks, MD, MEd

I

nguinal hernia is a prevalent surgical disease accounting for more than 20 million annual repairs worldwide (HerniaSurge Group, 2018). Although such repairs are common within the field of general surgery, the debilitating nature of chronic groin pain that can occur after 10% to 15% of open inguinal hernia repairs and the high number of cases entailed in ascending the learning curve for minimally invasive inguinal hernia repair point to the importance of surgical precision and preparation in delivering an effective repair. nn INDICATIONS AND TIMING

OF REPAIR

Symptomatic inguinal hernias in male patients should be repaired. Watchful waiting can be safely applied to male patients with asymptomatic or minimally symptomatic inguinal hernias with a low rate of emergent repair (1.8 per 1000 patient years); however, there is a time-­ dependent risk of development of nonemergent symptoms, such as inguinal pain or progressive hernia enlargement, which will lead to a majority of patients seeking eventual surgical repair (Fitzgibbons, 2006). Other considerations that should influence the timing of repair include the increased complexity of repair of larger hernia defects and the greater risk of perioperative complications in the emergent setting particularly for patients with significant comorbidities or bleeding disorders. There is currently no available watchful waiting trial in female patients. Although inguinal hernias are much less common among female patients, the risk of hernia emergencies associated with observation of inguinal hernias in female patients is unknown but appears to be higher than for males. Thus, there is little evidence-­based guidance to offer female hernia patients in decision making regarding repair of asymptomatic inguinal hernias. The comparative increased difficulty in identifying an inguinal hernia in the female patient may lead to a delay in diagnosis and more advanced disease at presentation. When imaging is needed to evaluate female patients with nonacute groin pain, ultrasound may be inadequate and pelvic magnetic resonance imaging is considered the best imaging modality for this patient population.  nn CONSIDERATIONS

IN SELECTING THE OPERATIVE APPROACH

A variety of surgical repairs has been developed. Selection of the appropriate technique is largely surgeon-­dependent based on experience, with the best outcomes associated with the repair most frequently performed. Patient and hernia characteristics should guide

decision making in procedural selection. Although there are no absolute contraindications to laparoscopic repair, those patients who are not candidates for general anesthesia are better approached with an open technique that can be performed with spinal or local anesthesia with sedation. Although laparoscopic inguinal hernia repair can be performed under epidural anesthesia, inadequate muscular relaxation limits the working space with this approach so that the large majority of these repairs are conducted with general endotracheal anesthesia. The choice of total extraperitoneal repair (TEP) versus transabdominal preperitoneal repair (TAPP) may also be dictated by the patient characteristics as well as surgeon experience. Patients who have previously undergone surgery may have adhesions that make laparoscopy more difficult, in particular, patients who have had a prostatectomy have usually obliterated the preperitoneal space required for minimally invasive techniques. There is compelling evidence to support laparoscopic inguinal hernia repair for patients with bilateral inguinal hernias as well as recurrent inguinal hernias after prior open inguinal hernia repair. For the majority of inguinal hernia patients, open and minimally invasive mesh repairs are equivalent.  nn OPEN

INGUINAL HERNIA REPAIR

Tissue Repair Although direct comparisons are limited and available randomized controlled trials are of low quality and subject to bias, the Shouldice repair is the best nonmesh inguinal hernia repair technique with supporting level IA evidence. Compared with other tissue repairs, Shouldice imparts lower recurrence, hematoma, and chronic pain rates. Surgeon experience and meticulous technique have significant impact on outcomes. From a 2012 Cochrane systematic review, the recurrence rate for Shouldice was higher than mesh Lichtenstein repair. There was no significant difference in chronic pain, surgical site infection, or seroma. Because there was inclusion of heterogenous studies with questionable external validity, the comparison was not robust. Given the risk of mesh repair in contaminated cases and strangulated hernias as well as increasing patient concerns about mesh, it is important to have a primary tissue repair in one’s surgical armamentarium. 

Lichtenstein Repair The most common open mesh repair is the Lichtenstein, which was the original tension-­free repair that reinforces the inguinal floor with a prosthetic mesh. Initially described in 1989, the repair involves fixation of mesh using a permanent suture to the anterior rectus sheath, internal oblique aponeurosis, the pubic tubercle, and the shelving edge of the inguinal ligament. The approach is via an incision inferior and medial to the anterior superior iliac spine which is extending medially for approximately 6 to 8 cm. The dissection is carried down to the aponeurosis of the external oblique muscle which is opened in the direction of the muscle fibers. Using Metzenbaum scissors, this incision in the aponeurosis is extended through the external inguinal 623

624

MANAGEMENT OF INGUINAL HERNIA

ring. Upon opening the aponeurosis, the ilioinguinal nerve will be visualized and should be preserved. The interior oblique fibers are dissected bluntly from the overlying external oblique muscle flaps that reveal the shelving edge of the inguinal ligament. In addition, the iliohypogastric and ilioinguinal nerves are identified and preserved. The spermatic cord, along with an indirect hernia sac if present, should be identified and encircled with a Penrose drain. The hernia sac will usually be seen on the anteromedial surface of the cord structures. At this point, the genital nerve is visualized along the inferolateral surface of the cord adjacent to the external spermatic vein. The floor of the inguinal canal should be carefully assessed for weakness indicating a direct hernia. The medial aspect of the mesh is secured to the aponeurotic tissue overlying the pubic tubercle with an overlap of approximately 2 cm. The mesh should be secured using a running nonabsorbable monofilament suture, which is continued laterally along the inferior edge of the mesh to the shelving edge of the inguinal ligament to just lateral to the inguinal ring. The mesh is split to encircle the spermatic cord, with two-thirds of the mesh above and one-third below. The two tails are secured with a single nonabsorbable suture to the shelving edge to recreate a new internal ring. Interrupted absorbable monfilament sutures loosely secure the superior edge of the mesh to the internal oblique aponeurosis and muscle with avoidance of entrapment of the iliohypogastric nerve. The layers of the abdominal wall are then closed, starting with the aponeurosis of the external oblique, Scarpa’s fascia, and skin. A modification of the Lichtenstein repair uses the addition of a mesh plug that is placed through the internal ring in the case of an indirect inguinal hernia. For a direct hernia, the plug is sutured to Cooper’s ligament, the inguinal ligament, and the internal oblique aponeurosis. However, the use of mesh plugs is discouraged due to the risk of migration, mesh contraction, and chronic pain. 

Local Anesthetic Block Local anesthetic block is a safe and cost-­effective method of delivery of anesthesia for open inguinal repair and postoperative analgesia. Regardless of whether general anesthesia is used, preemptive local anesthetic injection before the surgical incision may decrease postoperative pain by reducing the stimulation of nociceptors. A 50:50 mixture of 1% lidocaine and 0.5% bupivacaine is used with addition of epinephrine. Subdermal infiltration under the planned incision is performed transversely with approximately 5 mL followed by intradermal injection to raise a wheal at the incision. Deeper subcutaneous vertical injection is performed. Proximal ilioinguinal nerve anesthesia is applied with injection medial to the anterior superior iliac spine though there is risk of direct nerve injury with this blind technique. Injecting 10 mL of the local anesthetic directly under the exposed external oblique aponeurosis bathes the peripheral inguinal nerves with additional solution reserved for injection at the spermatic cord and pubic tubercle and final subaponeurotic and subcutaneous injection before closure.  nn MINIMALLY

INVASIVE INGUINAL HERNIA REPAIR

Transabdominal Preperitoneal Repair The transabdominal preperitoneal inguinal hernia repair is performed with the patient in the Trendelenburg position with tilt toward the contralateral side to allow the viscera to fall away from the operative field. The abdomen is accessed via a curvilinear incision at the inferior umbilical rim and placement of a 10-­mm port. There is often a concomitant umbilical hernia; this should be assessed by examination preoperatively and at the time of incision. The TAPP approach is well suited to the patient with an umbilical hernia because it allows use of this abdominal wall defect for midline port placement after clearance of the hernia contents, avoids the risk of leak of insufflation into the peritoneal cavity that could be seen with the total extraperitoneal approach, and provides for repair of the umbilical and inguinal hernias. Visual inspection of both inguinal regions allows

5 mm

TAPP ports

5 mm

TEP ports

FIG. 1  Proper patient and trocar positioning for total extraperitoneal and transabdominal preperitoneal repair. Trocars should be positioned with sufficient space between them to allow for adequate hand motion. TAPP, Transabdominal preperitoneal; TEP, total extraperitoneal.

for port placement adjustment to triangulate the instruments to the target space. A 5-­mm 30-­degree laparoscope is preferred. Bilateral hernias are approached via two 5-­mm working ports placed about 6 cm lateral to and slightly below the umbilical port. For the unilateral repair, the ipsilateral 5-­mm trocar is placed at the level of the umbilicus, whereas the contralateral 5-­mm trocar is placed several centimeters lower than the umbilicus for ergonomic optimization (Fig. 1). Robotic TAPP can also be performed, in which case the three robotic ports are placed at or above the level of the umbilicus at least 15 to 20 cm from the superior extent of the operative field, with one midline port and two working ports 8 to 10 cm on each side of the midline. Wide exposure of the myopectineal orifice is critical to the operation. The peritoneum is opened from just lateral to the anterior superior iliac spine to the medial umbilical ligament and dissected down to well over the psoas and iliac vessels leaving the bladder down (Fig. 2). Although the size of the opening in the peritoneum may be reduced after gaining experience with the technique to limit the time for peritoneal closure, the preperitoneal dissection should be ample to allow for identification of all hernia defects at the myopectineal orifice and for wide mesh coverage. A wider opening allows the peritoneal flap to lay down more fully and this visual advantage can be helpful early in the surgeon’s experience. Lateral and the most medial dissection are performed first to allow the peritoneum to fall away and improve operative exposure prior to dissection of the indirect hernia sac. The inner edge of the peritoneum should be at minimum 1 to 2 cm away from the inferior edge of the mesh to avoid folding of the mesh when the peritoneum is closed. The lipoma of the direct hernia defect should be fully reduced and removed. The white edge of the attenuated transversalis fascia can be seen at the intersection with the lipoma and denotes the plane of dissection (Fig. 3). Caution is exercised to identify and avoid injury to the corona mortis, the sometimes-­present aberrant vessels from the epigastric vessels to the obturator vessels located below the direct defect at Cooper’s ligament. The external iliac vein should be visible. If there is overlying fat, this suggests a femoral hernia with lipoma that should be reduced. The hernia sac at the indirect space is dissected from the cord structures during cephalad retraction of the hernia sac. The TAPP approach provides an advantage during complex hernia cases, such as the scrotal inguinal hernia, by allowing visualization from both the preperitoneal and peritoneal spaces of herniated viscera. The fat and tissue should be cleared between the vas deferens and Cooper’s ligament, and the hernia sac and peritoneum should be separated fully

HERNIA

625

Direct space Indirect space Femoral space Iliac crest Bogros’ space Obturator space

Retzius’ space

Median umbilical ligament (reflected)

Medial umbilical ligament (reflected)

Reflected peritoneum

FIG. 2  Preperitoneal inguinal anatomy.

A

B

from the cord structures well over the iliac vessels and psoas. The indirect lipoma is identified just lateral to the vas and gonadal vessels and this is retracted cephalad and fully reduced. This may appear small but can be larger beyond the inguinal ring. Leaving a residual lipoma may result in continued groin pain. Meticulous attention should be paid to fascial closure and other aspects that can reduce the risk of hernia recurrence. The weakened transversalis fascia of the direct hernia sac should be inverted and sutured to reduce the hernia recurrence risk for larger direct hernias and likely decrease the seroma rate. A wide mesh should be used for ample coverage of the myopectineal orifice (Fig. 4). The mesh should lay smoothly against the abdominal wall extending below the pubis and

FIG. 3  (A–B) Dissection of the hernia sac. Apply gentle traction on the hernia sac and carefully dissect away from the spermatic structures.

Cooper’s ligament and anterior to the bladder so that the mesh will not shift when the bladder fills. Fixation of the mesh may not be needed routinely; however, if there is a large hernia defect, fixation is recommended whether by sutures, a minimal number of tacks, or self-­gripping mesh to avoid shifting of the mesh because of seroma or eventration of the mesh. The mesh should be completely reperitonealized. The peritoneum closure should be complete without gaps. Sutured closure with absorbable 2-­0 or 3-­0 suture is preferred. A self-­locking knot is placed on the suture before inserting the suture reduces operative time. Caution is urged regarding barbed suture to ensure it is well covered and exposure minimized to avoid bowel adherence to the barbs. 

626

MANAGEMENT OF INGUINAL HERNIA 12 cm.

At least 10 cm. dissection

Reflected peritoneum

15 cm. FIG. 4  Preperitoneal mesh placement with wide coverage of the myopectineal orifice.

Total Extraperitoneal Repair Patient preparation is similar for both TEP repair and TAPP; the patient is supine with arms out or tucked and the abdomen and suprapubic areas are clipped if necessary. Patients are required to void before surgery; however, a urinary catheter is used by many surgeons to ensure bladder emptying. Entry is via an infraumbilical 12-­mm port through which the anterior rectus sheath is incised just off the midline. After opening the sheath, the rectus muscle is split to access the posterior sheath, and a balloon trocar can facilitate the dissection of the preperitoneal space. By inflating the balloon under direct vision, the epigastric vessels can be visualized along the anterior abdominal wall, and the hernia is usually identified as well. Inadvertent entry into the peritoneal cavity will be readily apparent and may occasionally require conversion to a TAPP or open repair. Once the balloon has expanded the preperitoneal space, it is removed and the space is insufflated to a pressure of 12 to 15 mm Hg. A 10-­mm camera is used, preferably a 45-­or at least 30-­degree angled scope. Two 5-­mm ports are placed in the midline, allowing for several fingerbreadths between the lowest port and the pubic symphysis as well as between the ports themselves. The lateral abdominal wall is identified first, by bluntly dissecting the filmy avascular tissue anterior to the hernia sac but inferior to the epigastric vessels, which should be identified and protected. Medially, Cooper’s ligament is identified and cleared of any overlying fat while taking care to avoid the venous circle of Bendavid as well as the femoral canal inferiorly. The spermatic cord and hernia sac can be grasped and retracted cephalad while taking care to reduce an adherent cord lipoma as well. If the peritoneum is inadvertently torn during dissection, which more commonly occurs in patients who have had prior surgery, the defect should be closed, if possible, with an endoscopic clip or with an Endoloop. In addition, the resulting pneumoperitoneum may decrease the working space significantly and the pneumoperitoneum can be decompressed via a Veress needle or 5-­mm port in the left upper quadrant of the abdomen.  nn TAPP VERSUS TEP? The International Guidelines for Groin Hernia Management published by the HerniaSurge Group and endorsed by multiple major international hernia societies offer statements and recommendations regarding comparison of laparoscopic TAPP and TEP techniques. These are largely equivalent operations with similar operative time and cost, complication

risk, pain incidence, and recurrence rates. Robotic-­assisted TAPP adds additional cost. TEP confers a higher conversion rate than TAPP but the incidence is low. The Guidelines note that although very rare, there is a trend for more visceral injuries in TAPP and more vascular injuries in TEP. Meticulous fascial closure at the umbilical port site in TAPP is needed. Though the risk of port-­site hernia is very low with TAPP, there is a known increased incidence of port-­site hernia for umbilical sites. Because the surgical outcomes are similar, the choice should be based on the surgeon’s skills, training, and experience. There are certain situations that are better suited for one minimally invasive approach over another. The recurrent hernia after laparoscopic repair is better addressed by TAPP or open repair rather than TEP. TEP holds an advantage for the bilateral inguinal hernia without a large scrotal component as the technique does not require peritoneal closure. Patients with prior lower abdominal surgery, particularly by midline laparotomy incision, scrotal hernia particularly with incarceration, prior prostatectomy, prior inguinal hernia plug and patch repair, and those with uncertain diagnosis in which diagnostic laparoscopy would be beneficial, should undergo laparoscopic or robotic-­assisted TAPP rather than TEP. The enhanced TEP techniques with access to the preperitoneal space higher or more lateral in the abdominal wall addresses some of these shortcomings with traditional TEP, allowing an extraperitoneal approach in cases in which lower abdominal surgery would otherwise be prohibitive.

Risk Factors for Inguinal Hernia Development and Recurrence Patients with family or personal history of inguinal hernia, older age (direct hernias), male sex, collagen disorder with decreased mature type I collagen, and history of prostatectomy are risk factors for inguinal hernia development. Interestingly, in contrast to risk factors for ventral incisional hernia, obesity is protective in inguinal hernia development, with lower body mass index associated with higher inguinal hernia incidence. As outlined, surgical technical factors contribute to the risk of hernia recurrence, including decreased mesh overlap. Lower case volume is associated with a higher recurrence risk. Surgical outcomes are poorer for female patients. Direct and sliding hernias are associated with a higher rate of recurrence.  nn CHRONIC

PAIN AFTER INGUINAL HERNIA REPAIR

The incidence of chronic pain after groin repair is higher than the risk of recurrence. There is a higher risk of pain associated with open repair. The ilioinguinal, iliohypogastric, and genitofemoral nerves should be identified during open repair. Retraction of the nerves and disruption of the perineurium should be avoided. The genitofemoral nerve is often the hardest to identify by the spermatic vessels. Neurectomy proximal to the site of nerve injury should be performed at the time of repair if an injury is suspected. The nerve should be allowed to retract into the muscular bed. Chronic groin pain is a small but known risk after laparoscopic inguinal hernia repair. The lateral femoral cutaneous nerve may be seen in the fatty layer below the iliopubic tract and this should be left undisturbed with avoidance of disruption of the perineurium. The femoral branch of the genitofemoral nerve is seen more medially. The genitofemoral nerve will be found medially near the spermatic cord vessels. In the female patient, the genitofemoral nerve will join the round ligament close to the inguinal canal. When performed, division of the round ligament should be performed close to the peritoneum to avoid nerve injury. If mesh fixation tacks are used, care is taken not to press too deeply into the abdominal wall to avoid a nerve injury superficially. The number of tacks used should be minimized because the risk of chronic pain increases resulting from potential nerve injury. Fixation should be performed medially to the epigastric vessels and anteriorly with avoidance of the lateral space, particularly below the iliopubic tract (Fig. 5). 

HERNIA

627

Epigastric vessels

Iliopubic tract "Triangle of Pain"

Corona mortis vessels

Lat. femoral cutaneous n. Femoral brs., genitofemoral n.

"Triangle of Doom"

Femoral n.

Genital br., genitofemoral n.

FIG. 5  Triangle of pain.

nn CONCLUSION Elective repair of symptomatic inguinal hernias is recommended; however, observation is appropriate for those without symptoms or who do not wish to undergo surgery. The choice of minimally invasive or open repair is primarily surgeon preference, and the use of prosthetic mesh will significantly reduce the risk of recurrence. The major risks of operative repair are usually minimal, but some patients may develop chronic pain from nerve injury and potential mesh complications. These morbidities must be considered during the preoperative discussion regarding repair, particularly in those patients with minimal symptoms.

Suggested Readings Amato B, Moja L, Panico S, et  al. Shouldice technique versus other open techniques for inguinal hernia repair. Cochrane Database Syst Rev. 2012;4:CD001543.

Management of Recurrent Inguinal Hernia Brian Jacob, MD, and Kathryn Ely Pierce Chuquin, MD

I

nguinal hernias are common, with a lifetime risk of 27% in men and 3% in women. More than 500,000 inguinal hernia repairs are performed each year in the United States. With many diverse ways of performing an inguinal hernia repair and no consensus proving that one method is more optimal than another, the published data often conflict regarding a recommended technique. As such, hernia surgeons and general surgeons continue to perform the technique that they are best trained to perform. Recurrences will happen regardless of whether the repair was performed via open nonmesh (tissue)–based,

Amid PK. The Lichtenstein open “tension-­free” mesh repair of inguinal hernias. Surgery Today. 1995;25(7):619–625. Daes J, Felix E. Critical view of the Myopectineal Orifice. Ann Surg. 2017;266(1):e1–e2. Fitzgibbons RJ, Giobbie-­Hurder A, Gibbs JO, et al. Watchful waiting vs. repair of inguinal hernia in minimally symptomatic men: a randomized clinical trial. JAMA. 2006;295:285–292. Hernia Surge Group. International guidelines for groin hernia management. Hernia. 2018;22(1):1–165. Hu QL, Chen DC. Approach to the patient with chronic groin pain. Surg Clin North Am. 2018;98(3):651–665. Lichtenstein IL, Shulman AG, Amid PK, et al. The tension-­free hernioplasty. Am J Surg. 1989;157:188–193. McCormack K, Scott NW, Go PM, et  al. Laparoscopic techniques versus open techniques for inguinal hernia repair. Cochrane Database Syst Rev. 2003;1:CD001785.

open mesh–based, or minimally invasive technique (with or without mesh). Inguinal hernia recurrence remains frequent, with many large studies reporting a recurrence rate as high as 15%. Even the most experienced surgeon will encounter recurrences. When it comes to the topic of outcomes after inguinal hernia repair, recurrences are just one outcome metric that is important. Other important postoperative issues, such as chronic groin pain, will be discussed in other chapters. nn DEFINING

HERNIAS

RECURRENT INGUINAL

Recurrent inguinal hernias that happen in the first 3 to 6 months after the surgery are called early recurrences and are generally considered a result of a technical error during the initial repair, although this is not always the case. Technical factors that may contribute to hernia recurrence are discussed in the following sections. Patients with a chronic cough who have a coughing attack during the first few weeks after a repair, or violent extubation immediately after a repair, are also at risk for developing a recurrence despite having a solid repair the first time.

HERNIA

627

Epigastric vessels

Iliopubic tract "Triangle of Pain"

Corona mortis vessels

Lat. femoral cutaneous n. Femoral brs., genitofemoral n.

"Triangle of Doom"

Femoral n.

Genital br., genitofemoral n.

FIG. 5  Triangle of pain.

nn CONCLUSION Elective repair of symptomatic inguinal hernias is recommended; however, observation is appropriate for those without symptoms or who do not wish to undergo surgery. The choice of minimally invasive or open repair is primarily surgeon preference, and the use of prosthetic mesh will significantly reduce the risk of recurrence. The major risks of operative repair are usually minimal, but some patients may develop chronic pain from nerve injury and potential mesh complications. These morbidities must be considered during the preoperative discussion regarding repair, particularly in those patients with minimal symptoms.

Suggested Readings Amato B, Moja L, Panico S, et  al. Shouldice technique versus other open techniques for inguinal hernia repair. Cochrane Database Syst Rev. 2012;4:CD001543.

Management of Recurrent Inguinal Hernia Brian Jacob, MD, and Kathryn Ely Pierce Chuquin, MD

I

nguinal hernias are common, with a lifetime risk of 27% in men and 3% in women. More than 500,000 inguinal hernia repairs are performed each year in the United States. With many diverse ways of performing an inguinal hernia repair and no consensus proving that one method is more optimal than another, the published data often conflict regarding a recommended technique. As such, hernia surgeons and general surgeons continue to perform the technique that they are best trained to perform. Recurrences will happen regardless of whether the repair was performed via open nonmesh (tissue)–based,

Amid PK. The Lichtenstein open “tension-­free” mesh repair of inguinal hernias. Surgery Today. 1995;25(7):619–625. Daes J, Felix E. Critical view of the Myopectineal Orifice. Ann Surg. 2017;266(1):e1–e2. Fitzgibbons RJ, Giobbie-­Hurder A, Gibbs JO, et al. Watchful waiting vs. repair of inguinal hernia in minimally symptomatic men: a randomized clinical trial. JAMA. 2006;295:285–292. Hernia Surge Group. International guidelines for groin hernia management. Hernia. 2018;22(1):1–165. Hu QL, Chen DC. Approach to the patient with chronic groin pain. Surg Clin North Am. 2018;98(3):651–665. Lichtenstein IL, Shulman AG, Amid PK, et al. The tension-­free hernioplasty. Am J Surg. 1989;157:188–193. McCormack K, Scott NW, Go PM, et  al. Laparoscopic techniques versus open techniques for inguinal hernia repair. Cochrane Database Syst Rev. 2003;1:CD001785.

open mesh–based, or minimally invasive technique (with or without mesh). Inguinal hernia recurrence remains frequent, with many large studies reporting a recurrence rate as high as 15%. Even the most experienced surgeon will encounter recurrences. When it comes to the topic of outcomes after inguinal hernia repair, recurrences are just one outcome metric that is important. Other important postoperative issues, such as chronic groin pain, will be discussed in other chapters. nn DEFINING

HERNIAS

RECURRENT INGUINAL

Recurrent inguinal hernias that happen in the first 3 to 6 months after the surgery are called early recurrences and are generally considered a result of a technical error during the initial repair, although this is not always the case. Technical factors that may contribute to hernia recurrence are discussed in the following sections. Patients with a chronic cough who have a coughing attack during the first few weeks after a repair, or violent extubation immediately after a repair, are also at risk for developing a recurrence despite having a solid repair the first time.

628

MANAGEMENT OF RECURRENT INGUINAL HERNIA

Late recurrences occurring after 6 to 12 months are generally from patient factors (discussed later); however, late recurrence can also be due to the same technical factors seen after early recurrences, as well as from mesh shrinkage or migration. When considering repair of a recurrent inguinal hernia, it is important to know how and when the initial hernia was repaired. Because the previous repair always dictates the optimal approach to repair of a recurrent inguinal hernia, all efforts should be made to obtain the previous operative report before a rerepair is planned. Patients who had a previous inguinal hernia repair as a child likely had an open nonmesh repair. Patients can also be divided into having a previous open (anterior) repair with or without mesh or into having a previous laparoscopic or robotic (posterior) repair with or without mesh. If a previous anterior repair was performed with mesh, it is important to know what type of mesh was used (e.g., Lichtenstein, plug and patch, PHS system).  nn RISK

FACTORS FOR RECURRENT INGUINAL HERNIAS

Technical Factors The most important technical factor contributing to hernia recurrence is failure to achieve a tension-­free repair. Mesh-­free repairs such as the Bassini, Halsted, and McVay techniques have been shown to have a 50% to 75% higher recurrence rate than mesh repairs from tension on the approximated tissue. The use of mesh to achieve tension-­free repair is now considered the gold standard. There has been some debate over the use of heavyweight versus lightweight mesh, but meta-­analyses have shown no difference in recurrence rate based on the weight of mesh used. Additionally, no difference in recurrence rates has been found when comparing self-­gripping mesh to suture-­or tack-­secured meshes. Other technical factors that contribute to hernia recurrence include incomplete dissection of the myopectineal orifice from anterior superior iliac spine to symphysis, insufficient reduction of the hernia sac, insufficient reduction of lipomas, inadequate mesh implant size or material, mesh migration, or mesh folding (clamshelling). Failure to resect a cord lipoma may also result in pseudo-­ recurrence, so care should be taken to reduce or excise these lipomas. Complete and masterful dissection following the steps as outlined by Daes and Felix and originally described in the International Hernia Collaboration closed Facebook group will minimize early recurrence rates. That said, a recurrence does not mean that the procedure was performed incorrectly or that a technical error was experienced during the surgery. Hernia recurrence is a known risk and should always be mentioned in the informed consent. One source of debate is whether a laparoscopic inguinal hernia repair is associated with higher recurrence rates than open hernia repair. Studies vary significantly, with many studies pointing toward an increased recurrence rate with laparoscopic repair; however, several large meta-­analyses did not find a significant long-­term difference between recurrence rates in laparoscopic and open approaches. Other studies have demonstrated that surgeon experience with laparoscopic repair has a significant impact on recurrence rates; those performing a higher volume of laparoscopic repairs experience a lower recurrence rate, eventually equaling that of open repair. 

Patient Factors Any risk factor for poor wound healing contributes to the risk of hernia recurrence, including immunosuppression, smoking, diabetes, and obesity. Smoking not only contributes to tissue hypoxia, but also alters collagen synthesis and remodeling, leading to weaker tissue. Obesity puts undue strain on the initial repair and increases the risk of wound infection. Additionally, a large meta-­analysis found that being female, initial operation for direct inguinal hernia, and operation for recurrent hernia were all risk factors for recurrence. Female sex risk is hypothesized to be due to the higher incidence of femoral hernias that

may be missed at initial open operation. The reason for a higher recurrence after initial operation for direct inguinal hernia is unclear, but may be due to inherent tissue weakness. Although this study did not find family history of recurrent hernias to be a significant risk factor, it is well documented elsewhere that collagen synthesis disorders such as Marfan’s syndrome and Ehlers-­Danlos syndrome carry a higher risk of hernia and poor wound healing. The recovery from the original surgery may also play a role in recurrence rates. It is important to note any postoperative infections, hematomas, or seromas, especially if the mesh implant shifts or migrates secondary to that process. 

Mesh Factors We, as authors, do not intrinsically blame the mesh for a recurrence. We believe that the products on the market have been adequately studied such that causation of recurrences cannot be attributed to the mesh alone; thus, when referring to mesh factors, we are describing what happens to the mesh after it is implanted. Shrinking of the mesh can lead to recurrence, as well as pain from the product pulling on surrounding structures. Some three-­dimensional shaped mesh (plugs, PHS systems, and others) placed during open repairs can also shrink, migrate, or are designed to resorb, leading to a space for a recurrence, and often chronic pain. Mesh can also migrate or fold, which usually happens very early after it is placed. Additionally, it can conform to the space it was bridging. Imaging is indicated when a mesh issue is suspected. We usually prefer investigation with a computed tomography (CT) scan; however, a normal CT scan does not necessarily rule out a mesh factor. Mesh issues related to recurrences after a laparoscopic repair, in our opinion, can only be truly confirmed with diagnostic laparoscopy.  nn DIAGNOSING

HERNIAS

RECURRENT INGUINAL

Diagnosing a recurrent inguinal hernia involves taking a careful history, performing a focused physical examination, and considering imaging. Most patients are asymptomatic and do not realize they are experiencing a recurrence; therefore, not all inguinal hernia recurrences will need immediate repair. When patients are symptomatic, they will sometimes state that they had a previous inguinal hernia repair with an interval without symptoms, followed by the onset of vague, temporary, discomfort located in just one spot near the inguinal canal. As the size of the hernia grows, patients will often complain of an intermittent bulge with or without localized nonradiating pain. Nevertheless, there is no single way a recurrent inguinal hernia presents. The symptoms are usually present or exacerbated when standing, exercising, or lifting. Once symptomatic, a repair is typically indicated. When a patient with a history of inguinal hernia repair presents with groin pain, it is vital that the surgeon not assume that the pain is from a recurrent inguinal hernia until all other possible etiologies have been ruled out and documented. Other causes of groin pain are numerous (Table 1). Often, the history alone will help the most in determining the cause of the pain. In general, pain from a recurrent hernia is intermittent and elicited directly over the bulge or defect. Pain from a recurrent hernia should not generally be made worse with movements of the hip joint or leg or with palpation of the symphysis or pubic tubercle. Acute onset of pain after an event (sport or traumatic) is a musculoskeletal injury until proven otherwise. Cyclic pain can be from an endometrioma or endometriosis. Constant pain can be due to old sutures, mesh, fractures, infections, or other causes, but is usually not from a recurrent inguinal hernia. If history suggests a hernia recurrence, then the workup may also include sonography, CT, or magnetic resonance imaging (MRI) with an athletic pubalgia protocol. Once other etiologies have been ruled out, an inguinal hernia repair can be safely planned. Sometimes, recurrent inguinal hernias are not palpable and are not seen on imaging, which can be especially true for tiny femoral hernias or tiny indirect inguinal hernias (as well as for small lipomas).

HERNIA

TABLE 1  Causes of Groin Pain by System

629

Repair of the Recurrent Inguinal Hernia Repair of a recurrent inguinal hernia is often more difficult than the initial operation because of distorted tissue planes. Repeat operation carries a higher rate of complications, including chronic pain, damage to cord structures, ischemic orchitis, and a higher risk of re-­ recurrence. Before rerepair of an inguinal hernia, patient-­controlled risk factors should be optimized, particularly smoking cessation and weight loss, and counselling regarding the previously mentioned risks. Watchful waiting for asymptomatic inguinal hernia is a well-­ established option in primary inguinal hernias, but has not been specifically studied for recurrent hernia. This is, however, included as an option for recurrent inguinal hernia in many of the guidelines. Surgical repair is specifically recommended for the symptomatic recurrent hernia. The only true contraindication to repair of a recurrent inguinal hernia is active infection because mesh should be used in all repairs and use of prosthetic mesh is contraindicated in infection. To minimize risks, as discussed in The International Guidelines for Groin Hernia Management, the main tenet of recurrent inguinal hernia repair is to approach the hernia through undisturbed tissue planes. Thus, the approach will be dictated by the prior surgery as discussed in the following section. A summary of our approach to a recurrent inguinal hernia is found in Fig. 2. 

System

Causes

Musculoskeletal

Avascular necrosis Avulsion fracture Osteoarthritis Bursitis Muscle strain or tear Tendonitis Piriformis syndrome Athletic pubalgia Referred or radiating back pain

Genitourinary/gynecologic

Epididymitis Orchitis Retractile testicle Hydrocele Varicocele Testicular/ovarian torsion Endometrioma/endometriosis Kidney stone

Gastrointestinal

Inguinal hernia Colitis/enteritis

Recurrent Inguinal Hernia After a Previous Open (Anterior) Repair

Neurologic

Sciatica Nerve entrapment

Infectious

Lymphadenopathy Urinary tract infection Sexually transmitted infection

Following an open inguinal hernia repair, recurrence is more likely to occur in the direct space, most often from inadequate mesh overlap at the pubic tubercle. These cases should almost always be approached by minimally invasive surgery (MIS) posterior method (laparoscopy or robotic) to avoid distorted and scarred anterior tissue planes. Reoperating through previously dissected tissue planes increases the risk of nerve entrapment, chronic pain, and testicular atrophy. In addition, the MIS approach has been associated with earlier return to work and lower incidence of chronic groin pain. Either transabdominal preperitoneal repair (TAPP) or total extraperitoneal repair (TEP) is reasonable. The choice of approach depends on surgeon preference and whether the surgery will be done laparoscopically or robotically. Although a minimally invasive approach is recommended, if expertise in MIS techniques is not available, an open posterior approach is acceptable as well. 

  

These are termed hidden hernias or occult hernias. All types can be symptomatic and are best identified during a diagnostic laparoscopy.

Role of Physical Examination and Imaging With a Recurrent Inguinal Hernia The diagnosis of a recurrent inguinal hernia is essentially the same as for a primary inguinal hernia and, again, is largely a clinical diagnosis, based on history and physical examination. Imaging is useful when the diagnosis is unclear. Nevertheless, it is important to again state that inguinal hernias may be missed on imaging; therefore, normal imaging does not exclude the presence of a recurrent inguinal hernia. Ultrasound is an inexpensive tool with a sensitivity of greater than 90% and a specificity of approximately 85%. This method allows for the groin to be examined both at rest and under a Valsalva maneuver. CT of the abdomen and pelvis has a role in the diagnosis of recurrent inguinal hernia because it can detect meshomas, metallic tack location, and recurrences that may not be readily palpable on examination. CT has a sensitivity less than 85%, with a specificity of only 65% to 85%. When pain is part of the complaint, we strongly recommend that a radiopaque marker be placed at the site of the pain while obtaining the CT scan because this can more precisely define the location of the structure involved. An MRI, on the other hand, though more expensive and time consuming, has a very high sensitivity, greater than 95%, and can also be performed at rest and with Valsalva. Although not essential for the workup of most inguinal hernias, MRI is recommended by Miller et al. in the workup of an occult hernia because it demonstrates the highest sensitivity, specificity, and negative predictive value compared with ultrasound and CT. The authors of that study state that MRI scans resulted in correct identification of an inguinal hernia in 91% of the cases of a false-­negative CT scan. Representative images of an ultrasound, CT scan, and MRI of inguinal hernia are shown in Fig. 1. 

Recurrence After a Previous Laparoscopic or Robotic (Posterior) Repair Following a minimally invasive hernia repair, recurrence is more likely to occur in the indirect space or inferiorly near the femoral canal, possibly resulting from incomplete reduction of the hernia sac or herniation of the sac around the mesh. Also, retained spermatic cord lipomas are not uncommon. Guidelines still recommend an anterior open repair of a recurrent inguinal hernia initially repaired using an MIS technique to avoid the distorted posterior tissue planes and mesh. With growing experience and comfort, however, laparoscopists are increasingly able to identify and repair a recurrence via MIS approach, even after a previous TEP or TAPP technique. The major advantage in MIS rerepair is that the surgeon possesses a view of all potential sites of recurrences, which is especially helpful when the location of the recurrence (direct, indirect, or femoral) is unclear. Because the dissection of the preperitoneal tissue plane is more difficult after a prior posterior repair, an attempt at TEP often results in tearing of the peritoneum, thus forcing conversion to TAPP. Any surgeon who is attempting a MIS rerepair of a recurrent hernia after a prior MIS repair must be comfortable with both TEP and TAPP techniques. 

Recurrence After a Previous Anterior and Posterior Repair For complex cases in which a patient is presenting with more than one recurrence, in which case both the anterior and posterior planes have

630

MANAGEMENT OF RECURRENT INGUINAL HERNIA

A

B M2

M2

VALS

C

Left GROIN Trans

3.0cm

VALS ...

Left GROIN Trans

bpm

3.0cm

...

bpm

FIG. 1  (A) Magnetic resonance imaging of an inguinal hernia. (B) Computed tomography image of an inguinal hernia. (C) Ultrasound of a left inguinal hernia at rest and with Valsalva maneuver.

Recurrent Inguinal Hernia

Asymptomatic

Watchful waiting

Symptomatic

Prior tissue repair

Prior anterior repair

Prior posterior repair

Prior posterior and anterior repairs

Anterior or posterior repair

Posterior MIS repair

Anterior repair, consider repeat posterior MIS repair

Repeat posterior MIS repair versus IPOM

FIG. 2  Algorithm for management of recurrent inguinal hernia. IPOM, Intraperitoneal onlay mesh; MIS, minimally invasive surgery.

HERNIA

been violated and mesh exists in the anterior and/or posterior space, the procedure of choice will depend on the surgeon experience. In our group, we prefer laparoscopy (or a robotic approach) in most of these cases initially as a diagnostic tool, to identify precisely the location and etiology of the recurrence. Often the prior MIS-­placed mesh can be taken down to identify the femoral space, the direct space, and the indirect space sufficiently to make a diagnosis and perform an MIS repair. In cases in which the peritoneum is destroyed during the dissection, an intraperitoneal onlay mesh repair can be safely performed. The mesh can be sutured to the anterior abdominal wall or to the old mesh, as well sutured to the retroperitoneum. No tacks should be deployed in the region of the nerves below the inguinal ligament.  nn CONCLUSIONS Recurrent inguinal hernias should not be taken lightly. Careful review of the history, a focused physical examination, selective imaging, and using a mastered technique will help optimize the outcomes. Having a detailed history, including the previous operative reports, not only dictates the approach for the recurrent hernia, but gives the surgeon an idea of what to expect at reoperation. Some operations for recurrent inguinal hernias will be straightforward, but some of them will be quite difficult. As with the index operation, the approach will ultimately depend on surgeon skill, comfort, and preference. It is unreasonable to expect a surgeon who is uncomfortable with laparoscopy to use an MIS approach to address a recurrent hernia solely because it was repaired using an open technique previously. Conversely, it is acceptable for a surgeon who is comfortable and proficient with MIS hernia repair to approach a recurrence using MIS techniques even if the initial hernia was repaired laparoscopically or robotically. For any inguinal hernia, including the recurrent hernia, the surgeon should be sure to look for and resect any cord lipoma because a cord lipoma alone can present as an inguinal bulge that creates the same symptoms as a true hernia. Even if a hernia itself is repaired or rerepaired, if a cord lipoma is left behind, the patient may still be symptomatic.

Incisional, Epigastric, and Umbilical Hernias Julie L. Holihan, MD, MS, and Mike K. Liang, MD

V

entral hernias are a common clinical problem with nearly 400,000 ventral hernia repairs performed each year in the United States. Despite this, ventral hernia repair is associated with a high rate of surgical complications, such as infection and hernia recurrence. To obtain optimal outcomes, there are a number of important clinical decisions that should be considered when approaching a patient with a ventral hernia. This chapter discusses patient selection and optimization, surgical techniques, and complex settings (including emergent repairs and cases with contamination). Primary (e.g., umbilical or epigastric hernias) and incisional ventral hernias have considerable differences in anatomy, pathophysiology, and outcome; therefore, they are discussed separately. nn PATIENT

SELECTION

For elective cases, careful patient selection is critical. Nonoperative management of ventral hernias is safe and feasible and should be the strategy of choice in certain patients. Significant comorbidities, including metastatic cancer, advanced cirrhosis, severe cardiopulmonary disease, and other conditions causing a life expectance less than 2 years are relative contraindications to elective surgery.

631

Finally, the steps for MIS inguinal hernia repair, as outlined by Daes and Felix in their paper “Critical View of the Myopectineal Orifice” (2017), applies to recurrent hernias and should be thought of as the “nine commandments” of inguinal hernia repair. These rules describe wide and safe dissection of the myopectineal orifice, adequate visualization of all potential hernia defects (direct, indirect, femoral), sufficient reduction of hernias and cord lipomas, and secure placement of the mesh. If these rules are followed by the surgeon during MIS inguinal hernia repair, fewer recurrences can be expected. Inguinal hernia is a prevalent surgical problem, and with recurrence rates of approximately 15%, every general surgeon will encounter recurrences in their career. Defining and diagnosing the recurrent inguinal hernia, as well as ruling out other causes of inguinal pain, are of paramount importance before rerepair. Surgical approach to repair is dictated by the method of the index surgery as well as the surgeon’s comfort with minimally invasive surgical techniques. In all, general surgeons should have a thorough understanding of inguinal anatomy from both anterior and posterior approaches and be comfortable with both minimally invasive and open repair of the inguinal hernia.

Suggested Readings Burcharth J, Pommergaard H, Bisgaard T, Rosenberg J. Patient-­related risk factors for recurrence after inguinal hernia repair. Surg Innov. 2014;22(3):303–317. Daes J, Felix E. Critical view of the myopectineal orifice. Ann Surg. 2017; 266(1). Fingerhut A, Khoursheed M. Recurrent inguinal hernia: the best approach. The SAGES Manual of Hernia Repair. 2012:167–180. HerniaSurge Group. The international guidelines for groin hernia management. Hernia. 2018;22(1):1–165. Miller J, Cho J, Michael MJ, Saouaf R, Towfigh S. Role of imaging in the diagnosis of occult hernias. JAMA Surgery. 2014;149(10):1077. Niebuhr H, Köckerling F. Surgical risk factors for recurrence in inguinal hernia repair – a review of the literature. Innov Surg Sci. 2017;2(2):53–59.

Patients with certain modifiable risk factors should also be initially managed nonoperatively. These patients can benefit from medical optimization before elective surgery. If and when they are able to meet the goals set out for them, they can be offered elective repair. For example, patients who smoke should not be offered elective surgery because of an increased risk of wound complications and hernia recurrence. Rather, they should be educated on smoking cessation and offered resources to aid with successful smoking cessation. In addition, obese patients should be counselled on weight loss before elective surgery. Elective surgery is considered for patients with a body mass index (BMI) less than 30. For patients with a BMI greater than 40, we opt for nonoperative management until the patient is able to lose weight. For patients with a BMI between 30 and 40, we recommend preoperative weight loss with individualized patient goals. Referrals to a physical therapist and a dietician can be offered to help patients with weight loss. Patients with poorly controlled diabetes should be referred to their primary care physician to better control diabetes before scheduling elective surgery. At our institution, a hemoglobin A1c less than 8.0% is recommended. Finally, asymptomatic or oligosymptomatic patients should be considered for nonoperative management because these patients are unlikely to see any improvement in quality of life following repair. All patients managed nonoperatively should be counseled on signs and symptoms that should prompt a visit to the emergency department. Symptoms may include increased pain at the hernia site, obstructive symptoms such as nausea, vomiting, distension, or obstipation. Signs may include changes in hernia characteristics such as increased firmness, new inability to reduce the hernia, and changes in surrounding skin color. 

HERNIA

been violated and mesh exists in the anterior and/or posterior space, the procedure of choice will depend on the surgeon experience. In our group, we prefer laparoscopy (or a robotic approach) in most of these cases initially as a diagnostic tool, to identify precisely the location and etiology of the recurrence. Often the prior MIS-­placed mesh can be taken down to identify the femoral space, the direct space, and the indirect space sufficiently to make a diagnosis and perform an MIS repair. In cases in which the peritoneum is destroyed during the dissection, an intraperitoneal onlay mesh repair can be safely performed. The mesh can be sutured to the anterior abdominal wall or to the old mesh, as well sutured to the retroperitoneum. No tacks should be deployed in the region of the nerves below the inguinal ligament.  nn CONCLUSIONS Recurrent inguinal hernias should not be taken lightly. Careful review of the history, a focused physical examination, selective imaging, and using a mastered technique will help optimize the outcomes. Having a detailed history, including the previous operative reports, not only dictates the approach for the recurrent hernia, but gives the surgeon an idea of what to expect at reoperation. Some operations for recurrent inguinal hernias will be straightforward, but some of them will be quite difficult. As with the index operation, the approach will ultimately depend on surgeon skill, comfort, and preference. It is unreasonable to expect a surgeon who is uncomfortable with laparoscopy to use an MIS approach to address a recurrent hernia solely because it was repaired using an open technique previously. Conversely, it is acceptable for a surgeon who is comfortable and proficient with MIS hernia repair to approach a recurrence using MIS techniques even if the initial hernia was repaired laparoscopically or robotically. For any inguinal hernia, including the recurrent hernia, the surgeon should be sure to look for and resect any cord lipoma because a cord lipoma alone can present as an inguinal bulge that creates the same symptoms as a true hernia. Even if a hernia itself is repaired or rerepaired, if a cord lipoma is left behind, the patient may still be symptomatic.

Incisional, Epigastric, and Umbilical Hernias Julie L. Holihan, MD, MS, and Mike K. Liang, MD

V

entral hernias are a common clinical problem with nearly 400,000 ventral hernia repairs performed each year in the United States. Despite this, ventral hernia repair is associated with a high rate of surgical complications, such as infection and hernia recurrence. To obtain optimal outcomes, there are a number of important clinical decisions that should be considered when approaching a patient with a ventral hernia. This chapter discusses patient selection and optimization, surgical techniques, and complex settings (including emergent repairs and cases with contamination). Primary (e.g., umbilical or epigastric hernias) and incisional ventral hernias have considerable differences in anatomy, pathophysiology, and outcome; therefore, they are discussed separately. nn PATIENT

SELECTION

For elective cases, careful patient selection is critical. Nonoperative management of ventral hernias is safe and feasible and should be the strategy of choice in certain patients. Significant comorbidities, including metastatic cancer, advanced cirrhosis, severe cardiopulmonary disease, and other conditions causing a life expectance less than 2 years are relative contraindications to elective surgery.

631

Finally, the steps for MIS inguinal hernia repair, as outlined by Daes and Felix in their paper “Critical View of the Myopectineal Orifice” (2017), applies to recurrent hernias and should be thought of as the “nine commandments” of inguinal hernia repair. These rules describe wide and safe dissection of the myopectineal orifice, adequate visualization of all potential hernia defects (direct, indirect, femoral), sufficient reduction of hernias and cord lipomas, and secure placement of the mesh. If these rules are followed by the surgeon during MIS inguinal hernia repair, fewer recurrences can be expected. Inguinal hernia is a prevalent surgical problem, and with recurrence rates of approximately 15%, every general surgeon will encounter recurrences in their career. Defining and diagnosing the recurrent inguinal hernia, as well as ruling out other causes of inguinal pain, are of paramount importance before rerepair. Surgical approach to repair is dictated by the method of the index surgery as well as the surgeon’s comfort with minimally invasive surgical techniques. In all, general surgeons should have a thorough understanding of inguinal anatomy from both anterior and posterior approaches and be comfortable with both minimally invasive and open repair of the inguinal hernia.

Suggested Readings Burcharth J, Pommergaard H, Bisgaard T, Rosenberg J. Patient-­related risk factors for recurrence after inguinal hernia repair. Surg Innov. 2014;22(3):303–317. Daes J, Felix E. Critical view of the myopectineal orifice. Ann Surg. 2017; 266(1). Fingerhut A, Khoursheed M. Recurrent inguinal hernia: the best approach. The SAGES Manual of Hernia Repair. 2012:167–180. HerniaSurge Group. The international guidelines for groin hernia management. Hernia. 2018;22(1):1–165. Miller J, Cho J, Michael MJ, Saouaf R, Towfigh S. Role of imaging in the diagnosis of occult hernias. JAMA Surgery. 2014;149(10):1077. Niebuhr H, Köckerling F. Surgical risk factors for recurrence in inguinal hernia repair – a review of the literature. Innov Surg Sci. 2017;2(2):53–59.

Patients with certain modifiable risk factors should also be initially managed nonoperatively. These patients can benefit from medical optimization before elective surgery. If and when they are able to meet the goals set out for them, they can be offered elective repair. For example, patients who smoke should not be offered elective surgery because of an increased risk of wound complications and hernia recurrence. Rather, they should be educated on smoking cessation and offered resources to aid with successful smoking cessation. In addition, obese patients should be counselled on weight loss before elective surgery. Elective surgery is considered for patients with a body mass index (BMI) less than 30. For patients with a BMI greater than 40, we opt for nonoperative management until the patient is able to lose weight. For patients with a BMI between 30 and 40, we recommend preoperative weight loss with individualized patient goals. Referrals to a physical therapist and a dietician can be offered to help patients with weight loss. Patients with poorly controlled diabetes should be referred to their primary care physician to better control diabetes before scheduling elective surgery. At our institution, a hemoglobin A1c less than 8.0% is recommended. Finally, asymptomatic or oligosymptomatic patients should be considered for nonoperative management because these patients are unlikely to see any improvement in quality of life following repair. All patients managed nonoperatively should be counseled on signs and symptoms that should prompt a visit to the emergency department. Symptoms may include increased pain at the hernia site, obstructive symptoms such as nausea, vomiting, distension, or obstipation. Signs may include changes in hernia characteristics such as increased firmness, new inability to reduce the hernia, and changes in surrounding skin color. 

632

INCISIONAL, EPIGASTRIC, AND UMBILICAL HERNIAS

nn TREATMENT

HERNIAS

OF PRIMARY VENTRAL

Umbilical and Epigastric Hernias Umbilical hernias are a defect through the umbilical ring not associated with a prior incision, whereas epigastric hernias are primary ventral hernias located along the linea alba and superior to the umbilical ring. Primary ventral hernias are an extremely common problem, present in nearly one-­half of the population on radiologic imaging or physical examination. In low-­risk patients with a BMI less than 30 and small defects (10 cm), we consider placement of an epidural to aid in pain control. In cases where drains are placed, they are removed after 2 consecutive days with less than 20 mL output per day.

Suggested Readings Helgstrand F, Rosenberg J, Kehlet H, Bisgaard T. Outcomes after emergency versus elective ventral hernia repair: a prospective nationwide study. World J Surg. 2013;37(10):2273–2279. Holihan JL, Alawadi ZM, Harris JW, et al. Ventral hernia: patient selection, treatment, and management. Curr Probl Surg. 2016;53(7):307–354. Kaufmann R, Halm JA, Eker HH, et al. Mesh versus suture repair of umbilical hernia in adults: a randomised, double-­blind, controlled, multicentre trial. Lancet. 2018;391(10123):860–869. Liang MK, Holihan JL, Itani K, et al. Ventral hernia management: expert consensus guided by systematic review. Ann Surg. 2017;265(1):80–89. Muysoms FE, Antoniou SA, Bury K, et al. European Hernia Society guidelines on the closure of abdominal wall incisions. Hernia. 2015;19(1):1–24.

BOX 1  Eponyms for Critical Anatomic Structures Semilunar Line Linea semilunaris Spigelian line Semilunar line of Douglas  Arcuate Line Linea semicircularis Douglas’ line Arcuate line of Douglas Arcuate line of the abdomen Linea arcuata vaginae musculi recti abdominis Semicircular line  Spigelian Hernia Hernias of the semilunar line Hernias through the conjoint tendon Spontaneous lateral ventral hernias   

636

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

fascia, which is a small strip of aponeurosis of the transversus abdominis muscle bounded laterally by the linea semilunaris and medially by the lateral margin of the rectus abdominis muscle, is most important (Fig. 1). This fascia was described by Adriaan van den Spiegel (1578–1625), a Flemish anatomist. The first description of a hernia through this area was by Klinkosch in 1764. Spigelian hernias usually are located at the junction of the arcuate line and the semilunar line in the Spigelian fascia. However, the exact location of the arcuate line is variable, which gives rise to the concept of the “Spigelian hernia belt,” which is located between horizontal lines drawn just below the umbilicus cranially and at the anterior superior iliac spines caudally and is usually about 6 cm in length. Spigelian hernias can occur anywhere along the Semilunar line, but 90% occur in the Spigelian hernia belt.

1 2 3 4

5

Clinical Presentation The hernia, which is also called a spontaneous lateral ventral hernia or hernia of the semilunar line, can be congenital or acquired. Associated conditions include collagen disorders, aging, rapid weight loss, chronic obstructive pulmonary disease, trauma, and a history of surgery. The reported incidence ranges from 0.1% to 2%, with a slight predilection in women. They are generally small in diameter, typically measuring 1 to 2 cm but can be as large as 14 cm. They tend to develop during the fourth to seventh decade of life. Patients may be completely symptom free, with the diagnosis only being made with imaging studies done for another indication. Symptomatic patients may have a localized palpable mass along the semilunar line with or without chronic pain, which may be intermittent. The pain tends to be worse with movement. The diagnosis can be elusive when there is no mass, which is common because the hernia is commonly interparietal and does not penetrate the external oblique fascia and therefore is difficult to feel (Fig. 2). Cross-­sectional imaging or ultrasonography can be helpful for patients with localized pain in the Spigelian belt and no palpable mass but are not definitive because false-­negative results have been reported. Laparoscopy or laparotomy is the only completely reliable diagnostic method. Approximately 20% will have an acute presentation with bowel obstruction. Because of this, a strategy of watchful waiting is not recommended when they are found incidentally. In children, the hernia has been associated with an ipsilateral undescended testicle. The differential diagnosis includes an arcuate line hernia, which is an unusual internal hernia where a fold of peritoneum invaginates medial to the arcuate line beneath the posterior lining of the rectus muscle. Another unusual hernia, which can be a source of confusion, is a spontaneous posterior rectus sheath hernia, which protrudes through the posterior rectus sheath, sometimes forming a mass in the rectus muscle. 

7

Treatment Open Approach, Nonprosthetic 6

FIG. 1  View of the anterior abdominal wall with the external oblique, internal oblique, and rectus abdominus muscles peeled away on the left. (1) Transverses muscle, (2) semilunar line, (3) posterior rectus sheath, (4) Spigelian aponeurosis, (5) arcuate line, (6) anterior superior iliac spine, (7) Spigelian hernia belt. (From Salameh JR. Primary and unusual abdominal wall

hernias. Surg Clin North Am. 2008;88:45-­60.)

The initial step is to ascertain the location of the hernia. This can be problematic if the hernia is not palpable. In the past, a midline or paramedian laparotomy has been recommended to avoid an extensive preperitoneal dissection to find the hernia. Today laparoscopy would be preferred. However, with modern imaging techniques the hernia can usually be precisely identified. The skin can be marked with ultrasound guidance, which accurately identifies the location of the incision. An oblique incision is made, and the subcutaneous tissue is explored if the hernia has penetrated the external oblique fascia. Because these hernias are commonly intraparietal, the next step is to incise the fascia of the external oblique muscle (Fig. 2). The hernia sac is then dissected free from surrounding structures and its contents reduced. If the contents of the sac are incarcerated, it may be necessary to incise the neck of the hernia, and this is best done medially toward the rectus muscle. The sac can then be either opened and

Aponeurosis of external oblique muscle Internal oblique muscle FIG. 2  (A) Typical location of a Spigelian hernia along the Spigelian belt. (B) Cross-­sectional diagram showing the interparietal nature of the hernia with the external oblique fascia intact. (From Fitzgibbons RJ Jr, Cemaj S, Quinn TH. Abdominal

wall hernias. In: Greenfield’s Surgery: Scientific Principles & Practices, 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2011.)

Skin

Transversus abdominis muscle

A

Intestine

B

Fat Anterior rectus sheath Rectus abdominis muscle Transversalis fascia

637

HERNIA

External oblique aponeurosis

Internal oblique muscle

Hernia sac

Hernia

A

Hernia defect

Lateral border of rectus abdominus muscle

B

External oblique aponeurosis

Mesh Sutures

C

External oblique aponeurosis

D

FIG. 3  (A) A Spigelian hernia in the right lower quadrant. (B) Hernia sac with bowel contents. (C) Prosthetic mesh is beneath the internal oblique muscle and secured circumferentially to provide an underlap. (D) External oblique muscle is repaired.

excised or reduced into the abdominal cavity. If there is any suspicion that the contents of the hernia have been compromised, the sac must be opened to allow inspection. Because these hernias can be multiple with Swiss cheese–type defects, the surrounding abdominal wall should be bimanually palpated either through the abdominal cavity if the sac has been excised or through the invaginated sac if it has only been reduced. A primary repair of the separated transversus and internal oblique muscles can be accomplished with either running or interrupted suture. The external oblique aponeurosis is then repaired followed by skin closure. 

Open Approach, Prosthetic Most surgeons today prefer to reinforce the repair with prosthetic material. There are three basic types of repair. The first uses a preshaped polypropylene cone plug. Using the same approach as described above, the hernial sac is dissected free and excised or inverted. If the sac has been excised, the peritoneum is closed with running suture, and the mesh plug is then inserted into the defect and sutured to its edges with interrupted sutures. The musculoaponeurotic layers are then closed over the plug. The second method uses a flat sheet of prosthetic material placed between the repaired internal oblique muscle and the external oblique, with a 2-­to 3-­cm underlap. This may be the best method in cases where there is too much tension to close the internal oblique. The mesh then serves to bridge

the separated internal oblique with the external oblique closed over it (Fig. 3). The third technique requires the development of a preperitoneal space to allow for the placement of a large prosthesis with wide underlap. The mesh is secured with sutures or tacks placed circumferentially around the prosthesis. If the preperitoneal space has been dissected widely enough to allow at least a 5-­cm overlap of the defect, no fixation is required. This is considered obsolete by some because the same procedure can be performed more effectively laparoscopically, as will be described below. Once the prosthesis is in place the transversus and internal and external oblique muscles are reapproximated as in the nonprosthetic technique. The advantage of this is that a relatively inexpensive prosthesis such as polypropylene mesh can be used. A simpler approach is to place the prosthesis intraabdominally, but this requires the use a more expensive product with antiadhesive properties approved for intraabdominal placement. Some examples are listed in Table 1. 

Laparoscopic Approach Spigelian hernias are readily visible at laparoscopy (Figs. 4 and 5). If the hernia is small, it can be closed primarily with one of the many available trocar site fascial closing devices. One or two 5-­mm laparoscopic cannulas are placed in the contralateral lower quadrant opposite the hernia. The trocar site closure device is introduced percutaneously as demonstrated in Fig. 6. All but the smallest defects

638

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

TABLE 1  Examples of Prosthetic Products Acceptable for Intraabdominal Placement Brand Name (Manufacturer)

Prosthesis

Adhesion Barrier

Parietex Composite (Covidien)

Polyester

Collagen-­polyethylene glycol-­glycerol

C-­Qur (Atrium)

Lightweight polypropylene

Omega-­3 fatty acid

Proceed (Ethicon)

Lightweight polypropylene+ polydioxanone + poliglecaprone-­25

Oxidized regenerated cellulose

Sepramesh (Genzyme)

Lightweight polypropylene

Sodium hyaluronate, carboxymethylcellulose, and polyethyleneglycol (hy)

Composix EX (Bard)

Heavyweight polypropylene

ePTFE

Composix L/P (Bard)

Lightweight polypropylene

ePTFE

Dulex (Bard/Davol)

ePTFE

The prosthesis is approved for intraabdominal use

DualMesh, DualMesh plus (W.L. Gore)

ePTFE

The prosthesis is approved for intraabdominal use

Ventralight ST Mesh (Bard)

Medium weight monofilament polypropylene mesh

Based on Seprafilm technology

ePTFE, Expanded polytetrafluoroethylene.

FIG. 4  Typical appearance of an incarcerated Spigelian hernia from a laparoscopic perspective. (From Mederos R, Lamas JR, Alvarado J, et al.

Laparoscopic diagnosis and repair of Spigelian hernia: a case report and literature review. Int J Surg Case Rep. 2017;31:184-­187.)

should be reinforced using an intraperitoneal onlay mesh technique (IPOM). This requires the edition of a third cannula, usually 10 to 12 mm, because this size is necessary to accommodate the mesh. The prosthesis should be large enough to widely overlap the defect in all directions. Sutures are attached circumferentially around the prosthesis with long tails. The device is rolled tightly along its long axis and introduced into the abdominal cavity through the largest cannula in the configuration. It is then unrolled in the abdomen and secured to the abdominal wall using a trocar site fascial closure device. The long suture tails are brought out of the abdomen through separate myofascial/peritoneal sites but the same skin puncture. This allows for full-­thickness myofascial attachment with the knot residing in the subcutaneous tissue. Alternatively, tacks or fibrin glue can be used. For larger defects, a combination of these fixation techniques is advised, for example, sutures plus tacks. The mesh material should not have the tendency to erode into bowel because it is being placed intraabdominally. Expanded polytetrafluoroethylene (ePTFE) is acceptable, but more popular now is polypropylene or polyester that has one of several types of adhesion barriers incorporated onto the side facing the viscera (Table 1). Increasingly surgeons are moving away from IPOM techniques for any ventral hernia because of concern about leaving prosthetic material of any type in the abdomen. A totally extraperitoneal approach can also be used that has the advantage of isolating the abdominal contents

from the prosthesis. The procedure is similar to a totally extraperitoneal (TEP) inguinal herniorrhaphy, but the balloon dissector is passed more laterally to create the space needed for mesh fixation. An infraumbilical incision is made to accommodate a larger cannula for the balloon dissector and mesh insertion. The rectus sheath is opened on the side of the hernia and the balloon dissector introduced into the space between the rectus muscle and the posterior rectus sheath under direct vision with an indwelling 0-­degree telescope. The balloon is directed toward the hernia rather than the pubic tubercle, as would be done for an inguinal hernia repair. Once the space is large enough, two additional 5-­mm cannulas are placed along the midline between the symphysis pubis and the umbilicus. The dissection then can be completed using standard laparoscopic instruments, reducing the sac and providing for wide overlap with a mesh prosthesis after the defect is closed primarily as was described above with the IPOM approach. The procedure can also be performed using a transabdominal preperitoneal technique (TAPP), which is particularly well suited for robotics. Cannula positions are similar to the IPOM, which allows the peritoneum above the hernia defect to be incised transversely, followed by a dissection of the preperitoneal space as in the TEP. Regardless of the technique chosen, an adequate mesh overlap of more than 3 cm is required.  nn LUMBAR

HERNIA

Lumbar hernias develop through a weakening of the posterolateral abdominal wall and can contain intraperitoneal or extraperitoneal contents. They occur in an area bordered superiorly by the twelfth rib, inferiorly by the iliac crest, medially by the lateral edge of rectus envelope, and laterally by the psoas muscle. There are three types: (1) superior, (2) inferior, and (3) diffuse. Superior and inferior lumbar hernias occur through classical anatomic triangles described below and are referred to as primary or spontaneous. This makes up only about 10% of lumbar hernias, with fewer than 300 reported cases since 1980. Diffuse lumbar hernias are much more common, usually the result of trauma, infection, or poor healing of flank incisions for renal or adrenal procedures. Previous flank incisions can also cause a “pseudo hernia” because of the division of lower thoracic nerves resulting in paralyzed muscle with loss of muscle control and tone, but there is no actual fascial defect.

Anatomy Primary lumbar hernias occur through one of the two lumbar triangles (Figs. 7 to 9). The superior lumbar triangle, also known as the

HERNIA

A

639

B

FIG. 5  Thoracoabdominal view of the defect after the incarceration has been reduced. The small diameter of the defect is typical and probably accounts for the high rate of strangulation on presentation. (From Mederos R, Lamas JR, Alvarado J, et al. Laparoscopic diagnosis and repair of Spigelian hernia: A case

report and literature review. Int J Surg Case Rep. 2017;31:184-­187.)

FIG. 6  Closure of a small Spigelian hernia defect using a laparoscopic trocar site closure device. In step 1, the closure device with a suture is passed through a 3-­mm skin incision and brought into the abdominal cavity on one side of the hernia defect. In step 2, the device is brought back out of the abdominal cavity leaving the suture in the abdomen. The device is the redirected at a different angle to enter the abdomen 180 degrees on the other side of the hernia defect. In step 3, the original suture is pulled out the abdominal wall so both tails are now extracorporeal. For step 4, the suture is tied closing the hernia defect, and the knot is allowed to retract back into the subcutaneous tissue and the skin closed over it. (From Fitzgibbons RJ Jr, Cemaj S, Quinn TH.

Abdominal wall hernias. In: Greenfield’s Surgery: Scientific Principles & Practices, 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2011.)

Grynfeltt triangle, is bounded laterally by the posterior border of the internal oblique muscle, the medial border of the quadratus lumborum and superiorly by the twelfth rib. The floor of the triangle is the transversalis fascia, and the roof is made up of the external oblique and latissimus muscle. The inferior lumbar triangle, also known as Pettit’s triangle, is bounded by the anterior border of the external oblique muscle, the anterior border of the latissimus dorsi muscle, and the inferior border of the iliac crest. The superficial fascia and the internal oblique muscle make up the roof and the floor, respectively. 

Clinical Presentation Congenital lumbar hernias occur from birth mainly as the result of defective development of the musculoskeletal system or the posterior

abdominal wall. These hernias are usually associated with other congenital abnormalities, including undescended testes, bilateral renal agenesis, and abnormalities of the spine, ribs, and other viscera. Congenital hernias are more common in the superior lumbar triangle. Lumbocostovertebral syndrome is a rare disorder characterized by congenital absence of ribs, myelomeningocele, scoliosis, hemivertebrae, and hypoplasia of the truncal and abdominal wall and is associated with an increase incidence of lumbar hernias. The syndrome may also be associated with vertebral defects, anal atresia, cardiac defects, tracheoesophageal fistula, renal abnormalities, and limb abnormalities. Malrotation of the left kidney and hydronephrosis can also be associated with an increased risk of congenital lumbar herniation. Acquired primary lumbar hernias are spontaneous, with an increased risk with older age, rapid weight loss, muscular atrophy,

640

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

12th rib

Superior lumbar triangle

Latissimus dorsi

Erector spinae External oblique

Internal oblique

Inferior lumbar triangle (Petit's triangle)

Quadratus lumborum

Iliac crest

FIG. 7  Pertinent anatomy related to superior and inferior primary lumbar hernias. (From Sundaramurthy S, Suresh HB, Anirudh AV, et al. Primary lumbar her-

nia: a rarely encountered hernia. Int J Surg Case Rep. 2016.)

12th rib Aponeuroses of the oblique and transverse m. Quadratus lumborum m. Latissimus dorsi m. Posteroinferior serratus m.

* FIG. 8  Cross-­sectional anatomy of a superior lumbar hernia (asterisk). (From Lumbar, pelvic, and uncommon hernias. In: Yeo,

Thoracolumbar fascia Erector spinae m.

C, editor. Shackelford’s Surgery of the Alimentary Tract, 8th ed, vol I. Philadelphia: Elsevier; 2019.)

*

FIG. 9  Cross-­sectional anatomy of an inferior lumbar hernia (asterisk). (From Lumbar, pelvic, and uncommon hernias. In: Yeo,

C, editor. Shackelford’s Surgery of the Alimentary Tract, 8th ed, vol I. Philadelphia: Elsevier; 2019.)

Aponeuroses of the oblique and transverse m. Iliac crest Latissimus dorsi m. Quadratus lumborum m. Thoracolumbar fascia Erector spinae m.

HERNIA

641

Treatment

FIG. 10  Axial computed tomography image showing the typical appearance of a traumatic lumbar hernia. (From Bathla L, Davies E, Fitzgibbons RJ

Jr, Cemaj S. Timing of traumatic lumbar hernia repair: is delayed repair safe? Report of two cases and review of the literature. Hernia. 2011;15:205-­9.)

chronic bronchitis, extremes of body mass, chronic debilitating disease, and strenuous physical activity. Petit triangle hernias have been reported to be more common in young, athletic women. Body habitus is believed to play a role because short, obese people with wide hips and more horizontal ribs will have larger anatomic triangles. Approximately one-­third will present with incarceration and 10% with bowel or urinary obstruction necessitating emergency treatment. A delay in diagnosis is not uncommon because lumbar hernia is not considered when caring for patients with low back pain. They occur more commonly in men between the ages of 50 and 70 years and are more frequent on the left side than on the right. Patients are usually symptom free at first but then may have complaints of lower back pain and a pulling sensation or symptoms like renal colic. In many cases the patient does not feel the need to mention anything to their doctor because at times it may disappear and only be present when straining. When more symptomatic, the patient seeks treatment for a “lump in the flank” and reports dull, heavy, localized pain. Primary lumbar hernias tend to become larger and have a high rate of incarceration, making a correct diagnosis important. A mass may be palpable within the area of the lumbar triangle and may have audible bowel sounds if the hernia contains intestinal contents. It may be difficult to detect these masses in obese patients. A computed tomography (CT) scan or magnetic resonance imaging (MRI) can show the anatomy of the destructive layer of the musculature and delineate the contents of the hernia as intraabdominal viscera or retroperitoneal fat. Lumbar hernias located in the superior lumbar triangle may be imaged as an impingement on the boundaries of the inner latissimus dorsi muscle. Lumbar hernias of the inferior lumbar triangle will be found superior to the lateral iliac crest. Lumbar hernias should be differentiated from soft tissue tumors, abscesses, hematomas, and renal tumors. Blunt trauma may cause a sudden rise in intraabdominal pressure, which can lead to a rupture of the posterior abdominal wall musculature. The result is a diffuse traumatic lumbar hernia most often caused by motor vehicle accidents, deceleration injuries, and crushing injuries and do not necessarily emanate from the two classical triangles described above (Fig. 10). Commonly a significant portion of the lateral abdominal wall is avulsed from the iliac crest. When they do correspond to the triangles, the superior is the more common site. Trauma associated with fractures of the pelvic or lumbar spine can also lead to lumbar hernias. Lumbar hernias also occur after certain procedures, including total hip arthroplasties, iliac bone graft donor sites after spinal fusion surgery, and autologous latissimus dorsi flap breast reconstructions. 

Lumbar hernias require surgical intervention because they may become incarcerated or strangulated or grow in size, leading to the need for emergency surgery. On making a diagnosis, it is recommended that the patient undergo surgery as soon as possible because the larger the hernia, the more difficult the repair. Laparoscopic and open procedures have been described, but in general laparoscopy is reserved for defects 5 cm or less. For an open approach the patient is placed in the lateral decubitus position with a lumbar roll. A transverse incision is made and the hernia sac identified, which may be either excised or reduced into the preperitoneal space. Small hernias can be repaired primarily with suture, but larger hernias require the placement of mesh between the muscle and the peritoneum. For the Petit hernia the mesh is placed below the latissimus dorsi and external oblique muscles and the iliac crest; for Grynfeltt hernias it is placed below the erector spinae and internal oblique muscles and the twelfth rib. A retromuscular preperitoneal space is developed using blunt dissection. The space is created for at least 5 cm beyond the margins of the defect on all sides. An adequately sized polypropylene or polyester mesh to provide wide overlap of the entire dissected space is placed in this space. Fixation can be accomplished with sutures or tacks or no fixation at all if the overlap is sufficient. Bone anchors into iliac crest are sometimes useful for the Petit hernia. The muscles surrounding the defect are closed over the mesh if there is not too much tension. For laparoscopic surgery, the patient is placed in the lateral decubitus position with a lumbar roll. Three laparoscopic cannulas are placed in the midline equidistant between the xyphoid process and the symphysis pubis. The contents of the hernia are reduced, and the colon and the kidney are mobilized away from the hernia defect to place a mesh. The mesh can be placed in the retroperitoneal space by opening the peritoneum and creating a large enough space to accommodate it. Alternatively, an IPOM approach with an adhesion resistant prosthesis can be used. An acquired lumbar hernia that occurs because of previous flank incision presents a unique challenge. The anatomy corresponds to the muscle(s) denervated and not the triangles seen with a primary hernia. These hernias have very little risk of incarceration or strangulation because there is no true fascial defect. Thus a strategy of watchful waiting is entirely acceptable. However, when large, these can be painful and associated with a significant cosmetic deformity. Laparoscopy can be useful as an adjuvant to help define the borders of the hernia, but the results with a purely laparoscopic approach are poor. For the open approach the patient is placed in the lateral decubitus position with the iliac crest just below the flexion point of the operating table with a lumbar roll in place to increase the space between the costal margin and the iliac crest. It is a good idea to mark the edges of the bulge in the preoperative area preferably in the standing position. This is because the defect may be difficult to discern when in the lateral decubitus position, which is the preferred position. A transverse incision is then made over the area that had been marked before surgery. The hernia sac is dissected free from surrounding structures. Overlapping and imbricating suture repairs can be performed, but the results are commonly short lived. If the area of weakness is small, the sac can then be reduced and a preperitoneal plane developed providing a minimum of 5 cm underlap for a mesh. The attenuated musculofascial element is then imbricated over the mesh. More commonly the denervated area is large, and a simple prosthetic underlay repair is destined to fail because the prosthesis will balloon through the denervated muscle and the patient will feel the hernia has returned. The CT image and patient in Fig. 11 are typical of an extensive denervation injury after a flank incision for a right partial nephrectomy. In this case, the sac was opened so that U sutures could be placed around the ninth through the twelfth ribs with the aid of a trocar site closure device under direct vision. The tails of the sutures were left long for the eventual attachment of the mesh along the costal margin. The dissection of the skin and

642

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

FIG. 11  Flank lumbar hernia as the result of a surgical incision.

subcutaneous tissue extended caudally to the iliac crest, where several bone anchors were placed again with long suture tails. The lateral edge of the rectus muscle and the psoas muscle were also exposed. Next the peritoneum was closed completely, isolating the intraabdominal viscera from the abdominal wall. A heavyweight polypropylene mesh was secured to the iliac crest with the bone anchors. It was stretched tightly across the closed peritoneum and attached to the previously placed rib sutures. Laterally the mesh was attached to the psoas muscle and medially to the lateral edge of the rectus envelope with continuous suture. The superficial muscles, which were quite attenuated, were closed over the prosthesis in an imbricating fashion to minimize any residual bulge.  nn OBTURATOR

HERNIA

Obturator hernias are rare, most commonly occurring in thin, elderly women, and for this reason it is sometimes referred to as the “little old lady’s hernia.” The mortality rate for repair is high, especially when complicated due in no small part to the frailty of the patients. Thus prompt diagnosis and treatment is crucial for a successful outcome but is often difficult to achieve because of the lack of the obvious physical findings that occur with other abdominal wall hernias.

Anatomy The obturator canal is a 1-­cm wide and 2-­to 3-­cm long opening in the superior part of the obturator membrane covering the foramen formed by the union of the pubic bone and ischium, through which the obturator nerve, artery, and vein pass. It starts in the pelvis and follows an oblique downward course, exits the pelvis, and ends in the obturator region of the thigh. An obturator hernia will occur through the obturator canal (Fig. 12). The sac usually follows the course of the obturator nerve, which has an anterior and posterior branch (Fig. 13). Visualizing this laparoscopically, the defect is inferior to the iliopubic tract and the pectineal (Cooper) ligament and quite medial to the femoral canal (Fig. 14). This should not be confused with a femoral hernia, which also occurs below the iliopubic

tract but is superior to the Cooper ligament, with the defect immediately medial to the femoral vein. 

Clinical Presentation Obturator hernia is an infrequent hernia characteristically difficult to diagnose on examination. The incidence has historically been noted to be between 0.07% and 0.1% of all abdominal wall hernias; however, in a more recent retrospective review of 3269 patients undergoing laparoscopic repair of inguinal hernias, obturator canal defects were discovered in 0.4% of patients, 97% of whom were women. The typical patient presenting with obstruction due to incarceration is female, in her 70s or 80s, and with a body mass index below 25. The hernia is more frequent on the right and is occasionally bilateral. Presentation and diagnosis are often made acutely in a thin, elderly woman presenting with bowel obstruction. Nonacute symptoms are often vague, characterized by intermittent lower abdominal pain, transient episodes of obstruction with spontaneous resolution, or nonspecific gastrointestinal complaints, such as early satiety, anorexia, and weight loss. The hernia may compress the obturator nerve, causing obturator neuralgia, which is neuropathic pain in the region of the medial thigh that radiates behind the knee. Physical examination findings are also nonspecific and unlikely to identify the hernia because the associated mass is concealed beneath the adductor muscles of the thigh. The Howship-­Romberg sign, neuropathic pain originating from compression of the obturator nerve elicited on the ipsilateral inner thigh with extension, adduction, or medial rotation and relieved by hip flexion, is said to be pathognomic. The diagnosis may also be made by palpation of the hernia mass on vaginal or rectal examination, or rarely a groin mass may be palpable with the patient supine, with the thigh flexed and the hip adducted and laterally rotated. But differentiating this from the more common femoral hernia or large inguinal hernia is difficult. Plain roentgenograms of the abdomen may confirm bowel obstruction but are not diagnostic of an obturator hernia as the cause. CT scan or MRI of the abdomen/pelvis will likely identify a transition point near the obturator canal and possibly an overt herniation in cases of obstruction but is less reliable in nonobstructed patients. The hernia most often contains small bowel, but the adnexa, appendix, mesentery, colon, and preperitoneal fat can also be involved. 

HERNIA

643

Obturator artery Obturator nerve Loop of ileum Obturator canal

A

Pectineus Sartorius

Obturator externus

Adductor longus

Sac of obturator hernia

Pectineus Obturator externus muscle

Adductor longus

B

FIG. 12  (A) Pertinent anatomy of the obturator canal. (B) Obturator hernia.

Treatment The hernia should be repaired at diagnosis in patients at acceptable risk because of the significant incidence of complications if untreated. The choice of the type of repair is dependent on clinical factors, as well as surgeon experience and preference. In elective cases and even urgent cases when there is minimal concern of potential contamination caused by prolonged bowel strangulation, a robotic or pure laparoscopic TAPP would be the approach of choice. Essential in the operation is reduction of hernia content with evaluation for nonreversable ischemia, a complete reduction of the peritoneum and associated preperitoneal fat, and then closure of the defect with either direct suture closure using permanent ligatures or placement of prosthetic mesh with closure of the peritoneal defect overlying the fixated mesh. A laparoscopic TEP would also be an option if no concern existed for ischemia with the advantage of not opening the peritoneum and potentially avoiding intraabdominal adhesions but with the disadvantage of not visualizing for potential compromised hernia

content. A Richter-­type antimesenteric incarceration of small bowel with potential for later tissue compromise should always be a consideration. In either case, if a prosthetic mesh is selected and adequate visualization obtained, use of a large prosthesis to cover the inguinal, femoral, and obturator orifices is optimal. Urgent repair of an obturator hernia associated with bowel obstruction is mandatory because mortality rates between 11% and 30% have been reported. The typical patient has significant comorbidities and a high frailty index. A minimally invasive approach can minimize these associated risk factors but in the case of overt sepsis, strangulation, high grade obstruction with distended edematous bowel, and factors typically preempting safe laparoscopic surgery an open operation would be deemed best. The open approach is best performed through a low midline incision. The hernia content is reduced, which sometimes requires dividing the obturator membrane in the direction of the Cooper ligament. Incarcerated hernia contents are inspected with appropriate management of ischemia. Primary closure is performed

644

MANAGEMENT OF SPIGELIAN, LUMBAR, AND OBTURATOR HERNIATION

Pubic ramus

Pectineus muscle

Pelvic fascia (partial layer) Adductor longus muscle Obturator nerve

Anterior division of obturator nerve

Obturator artery Obturator membrane

Adductor brevis muscle

Obturator internus muscle Obturator exterrnus muscle

Posterior division of obturator nerve

Peritoneum Adductor magnus muscle

Ischium

FIG. 13  Coronal view of the obturator canal demonstrating that the hernia sac can follow either the anterior or the posterior branch of the obturator nerve. (From Stamatiou D, Skandalakis LJ, Zoras O, Mirilas P. Obturator hernia revisited: surgical anatomy, embryology, diagnosis, and technique of repair. Am

Surg. 2011;77:1147-­57.)

covered with peritoneum, either ePTFE or one of the newer adhesion barrier prosthesis should be used to avoid later bowel complications (Table 1). In the event of contamination, a biologic mesh can be considered. Care needs to be taken in the closure of the defect or fixation of mesh to not incorporate the obturator nerve branches. Recurrence rates are low after obturator hernia repair.

Pe (Co ctine op al lig er’ s) ame n

t

Suggested

Right obturator foramen Right external iliac vessels

FIG. 14  Laparoscopic view of an obturator hernia defect. (From Conti L

Baldini E, Capelli P, Capelli C. Bowel obstruction in obturator hernia: a challenging diagnosis. Int J Surg Case Rep. 2018;42:154-­157.)

if adjacent tissue can be approximated, either muscle to muscle or muscle to periosteum of the pubic bone or a patch formed to cover the defect with omentum, bladder or other local tissue. Alternatively, a mesh is placed over the defect, with fixation to the Cooper ligament superiorly and the obturator muscle inferiorly. If the defect cannot be

readings

Bathla L, Davies E, Fitzgibbons Jr RJ, Cemaj S. Timing of traumatic lumbar hernia repair: is delayed repair safe? Report of two cases and review of the literature. Hernia. 2011;15(2):205–209. Beffa LR, Margiotta AL, Carbonell AM. Flank and lumbar hernia repair. Surg Clin North Am. 2018;98(3):593–605. Lee y, Monson J. Lumbar and pelvic hernias. In: Yeo CJ, et al., eds. Shackelford’s Surgery of the Alimentary Tract. 7th ed. Vol. I. Philadelphia: Elsevier; 2012. Salameh JR. Primary and unusual abdominal wall hernias. Surg Clin North Am. 2008;88(1):45–60. Skandalakis PN, Zoras O, Skandalakis JE, Mirilas P. Spigelian hernia: surgical anatomy, embryology, and technique of repair. Am Surg. 2006;72(1):42–48. Stamatiou D, Skandalakis L, Zoras O, Mirilas P. Obturator hernia revisited: surgical anatomy, embryology, diagnosis, and technique of repair. Am Surg. 2011;77(9):1147–1157.

HERNIA

Athletic Pubalgia: The “Sports Hernia” Bruce Ramshaw, MD, FACS, and Lauren Grimsley, MD, MBA

T

he diagnosis and management of athletic pubalgia is a complex undertaking. Even the terminology to describe the condition has been controversial, primarily because the most commonly used term, sports hernia, is not an accurate description of the underlying disease for most cases. For most patients who are diagnosed with a sports hernia, no true hernia defect is present. More recently, the term core muscle injury has been applied to this condition. The management can be a long and frustrating process, and the decision to proceed with more aggressive treatment, including surgery, can be complex. The basic strategies and techniques for surgical management also vary, and although most patients do well with surgical management from those surgeons who are experienced treating this condition, failures still exist regardless of the approach. This chapter discusses the background, diagnosis, and management of athletic pubalgia, and it presents a concept for learning and improving care for treating complex conditions such as athletic pubalgia. nn BACKGROUND Athletic pubalgia is defined as chronic groin pain (lasting more than 6–8 weeks without improvement, or the patient not able to return to normal athletic activities) in an individual who frequently engages in athletic and/or strenuous activity. It is diagnosed in the absence of a palpable hernia and when other causes of inguinal pain are excluded. Because of the lack of a visible or palpable hernia defect, this syndrome has not been well understood by many general practitioners. With the advent of social media and increasing news coverage about professional athletes whose careers have been interrupted or ended as a result of this condition, attempts to better understand the complex nature of this problem have been made. Although this condition has been popularized by the experience of professional athletes, it also affects athletes in college and high school, as well as recreational athletes and people who perform strenuous activities. In 1980, Gilmore recognized chronic groin pain in professional athletes, mostly soccer players, and undertook surgically repairing the disruption. This condition was subsequently referred to as Gilmore’s groin. A similar condition was also reported in Europe and Australia. Many investigators have described the chronic lower abdominal and groin pain in professional athletes as a syndrome secondary to muscular injury or incipient inguinal hernias. In 1991, Taylor and colleagues concluded that the majority of groin pain in athletes was caused by muscle strains, inguinal hernias, or subclinical abdominal wall defects without herniation. In a smaller subset of cases, however, there was no palpable hernia or any other cause of groin pain. Following the terminology used in the European literature, they referred to this condition as pubalgia.  nn ANATOMY, PATHOPHYSIOLOGY,

AND DIFFERENTIAL DIAGNOSIS

Etiology Gilmore identified torn external oblique aponeurosis, torn conjoint tendon, and dehiscence between the torn conjoint tendon and inguinal ligament as the underlying causes of chronic groin pain. Meyers and colleagues suggested that hyperextension injury, with the pivot point being the anterior pelvis or pubic symphysis, is the most likely etiology of athletic pubalgia, and used the term core muscle injury.

645

Tendons of the rectus abdominis and adductor longus insert on the pubic symphysis and antagonize each other. This anatomic apparatus, when combined with sport activities that involve rapid pelvic movements such as hyperextension, can potentially injure weaker abdominal wall muscles. This group has documented more than 20 distinct anatomic defects from magnetic resonance imaging (MRI) results, with surgical recommendations based on each abnormality (Table 1). A different perspective on the mechanism of this condition focuses on the notion that the syndrome is an incipient hernia, with the defect being in the transversalis fascia, which comprises the posterior wall of the inguinal canal. Polglase and colleagues found substantial derangement in the posterior wall of the inguinal canal in 61 of 72 professional athletes (mostly Australian rules football players who presented with chronic groin pain and underwent surgical exploration of the inguinal canal). Although a consensus on the complex nature of chronic groin pain in an athlete is yet to be established, a third viewpoint combines both mechanisms of muscle tear and incipient hernia, or groin floor weakness, and categorizes this condition as one component of a broader pattern of groin disruption injury. According to this school of thought, this problem consists of an occult or incipient direct hernia, which is derived from pelvic instability. This underlying mechanism also gives rise to other groin syndromes such as osteitis pubis, conjoint tendinopathy and/or tear, adductor tendinopathy and/or tear, and nerve entrapment and/or irritation of the variety of nerves in the groin and pelvic area. 

People at Risk Chronic groin pain is more likely to occur in athletes who engage in sports that require sudden turns and pelvic movements and/or those that require the athlete to push hard against resistance. It is thought that the highest prevalence occurs in athletes who are professional soccer players, ice hockey players, and in athletes who play American or Australian rules football and rugby. There seems to be an increase in the number of female athletes experiencing chronic groin pain as more women engage in competitive sports. It has also been postulated that gender-­specific anatomic variation may play a role in predisposing male athletes, more than female athletes, to developing chronic groin pain. Women typically have a broader pelvis and a wider pubic symphysis that allows for more surface area of muscle insertion and more stability than the narrower, less stable male pelvis. Brophy and colleagues demonstrated that male soccer players tend to generate a greater activation of the iliacus muscles in the kicking limb; this difference may contribute to a higher susceptibility to developing chronic groin pain. Other risk factors include limb length discrepancy, poor pelvic muscle balance, reduced hip range of motion, and pelvic instability. 

Presentation and Differential Diagnosis Athletes with chronic groin pain often present with lower abdominal pain on exertion. Most can recall a distinct injury during exertion before the onset of the pain; however, this condition can sometimes occur gradually from overuse and chronic repetitive injury. The pain is usually located in the inguinal region around the insertion point of the rectus abdominis muscle on the pubis or along the course of the ilioinguinal nerve. Most patients experience unilateral pain, although in some athletes, their symptoms begin with unilateral pain and progress to bilateral pain. The pain may also radiate to the lower abdomen, perineum, scrotum, or thigh. At rest, there is usually minimal or no pain. Valsalva maneuvers such as coughing or bearing down may sometimes reproduce the pain. Adduction against resistance can exacerbate the pain in some patients. The majority of these patients report that their groin pain preceded their adduction pain. Patients will experience these symptoms for many months, and some present

646

ATHLETIC PUBALGIA: THE “SPORTS HERNIA”

TABLE 1  Surgical Management of Core Muscle Injuries Structure or Syndrome

Incidence (%)

Defect

Possibly Indicated Procedure

Unilateral rectus abdominis/ unilateral adductor

22

Tear and compartment syndrome

Repair and compartmental decompression

Adductor longus

16

Pectineus

22

Adductor brevis

8

Pure adductor syndromes

21

Usually compartment syndrome

Compartmental decompression

Unilateral rectus abdominis

16

Tear

Repair

Bilateral rheumatoid arthritis

15

Tears

Repair

Severe osteitis variant

8

Usually tears, compartment syndrome and bone edema

Repair, compartmental decompression, and steroid injection

Iliopsoas variant

4

Impingement and bursitis

Lengthening procedure

Baseball pitcher/hockey goalie syndrome

4

Adductor tear and adductor muscle belly compartment syndrome

Compartmental decompression

Spigelian hernia

4

Tear

Repair

Rectus femoris variant

3

Impingement

Compartmental decompression

High rectus abdominis variant

2

Tear

Repair

Female variant

2

Medial disruption with lateral thigh compensation

Repair and compartmental decompression

Round ligament syndrome

1

Inflammation with tear

Repair and excision

Dancer’s variant

25 beats/min (not ventilated) or >12 L/min (ventilated)

Thrombocytopenia (applicable only 3 days after initial resuscitation)

25% increase in insulin requirements in 24 hours

>200 mg/dL (without treatment) or insulin resistance: >7 IU/hr intravenous insulin

Inability to continue enteral feedings >24 hours

Abdominal distension, enteral feeding intolerance (residual >150 mL/hr), ­uncontrollable diarrhea

Abdominal distension, enteral feeding intolerance (residual >2× feeding rate), uncontrollable diarrhea

From Greenhalgh DG, Saffle JR, Holmes JH 4th, et al. American Burn Association consensus conference to define sepsis and infection in burns. J Burn Care Res. 2007; 28:776–790. ABA, American Burn Association; SD, standard deviation.

nn BURN

METABOLISM AND NUTRITION SUPPORT

Severe burns result in a massive stress response that induces profound, protracted catabolism that sometimes can persist for years

100 90 Percentage above normal

if needed, and careful surveillance for invasive pulmonary infections, which are commonplace. Inhalation injury is devastating to airway host defenses (e.g., mucociliary activity), thus the increased risk. N-­acetylcysteine (a putative antioxidant) and aerosolized heparin (to possibly maintain lung microcirculatory integrity by preventing the fibrin microthrombi that entrap circulating neutrophils and platelets, amplifying inflammation and coagulopathy) have also been used to prevent fibrin cast formation, but are unproven. Carbon monoxide poisoning, which manifests as an elevated carboxyhemoglobin concentration (COHb) and impaired oxygenation, and sometimes also as neuropathy or encephalopathy, should be treated with 100% oxygen until the COHb is less than 5%. One-­ hundred percent oxygen shortens the elimination half-­life of COHb from 120 minutes to 30 minutes. Hyperbaric oxygen therapy may be considered for high COHb, critical dysoxia, or neurologic findings, if the patient can withstand the 2 to 3 hours of limited inability to intervene (none, if urgent) while pressurized during a typical “dive.” Accurate history and persistent lactic acidosis, especially with dysoxia, can suggest cyanide toxicity, which may result from combustion of synthetic materials including building supplies, furniture, and fabrics or form direct exposure to hydrogen cyanide gas or soluble salts. The primary critical care concerns are hemodynamic instability and cerebral edema. Treatment options, which should be administered without laboratory confirmation, include the hydroxycobalamin (vitamin B12) cyanide antidote kit, which binds to this mitochondrial toxin and sodium thiosulfate/nitrite. Bore is intravenous. After treatment with hydroxycobalamin, red-­pigmented urine may be excreted, which is not myoglobinuria. It is not necessary to increase urine output for the red pigmentation or myoglobinuria unless there are other potential etiologies, such as with electrical or crush injury, for rhabdomyolysis, fascial compartment syndrome, or hyperpyrexia. Superimposed inhalation and burn injuries can compound the inflammatory response, thereby complicating resuscitation with increased fluid requirements above those estimated by formula. However, patients with inhalation injury are at inherent risk for ARDS and ventilator-­associated pneumonia. As such, resuscitation for patients with combined burn/inhalation injury are based on available burn resuscitation formulas with adjustments for urine output and mean arterial pressure as necessary. 

x

80 70

Sepsis Burn Trauma

60 50 40

x

30 20 10 0 1 1 day week

1 month

1 year

FIG. 5  Hypermetabolic response after severe burns, trauma, and sepsis. (Modified from Clark A, Imran J, Madni T, et al. Nutrition and metabolism in burn patients. Burns & Trauma. 2017;5:11.)

after the initial burn (Fig. 5). Typically, patients with major burns have higher caloric/metabolic requirements when compared with patients with sepsis or multiple trauma. The acute response to burn injury typically manifests two distinct states. The shock, or ebb phase, is typically brief in duration and hypodynamic in nature resulting in depressed cardiac function, decreased metabolism, and reduced tissue perfusion. This state can result in hyperglycemia and elevated lactate production. Unlike the acute phase, the chronic response to burn injury, or flow phase, is a hyperdynamic, catabolic state that may last days to weeks after initial burn injury. The metabolic rate and Q are markedly elevated, resulting in increased whole-­body oxygen consumption and resting energy expenditure >10% above normal. The importance of adequate early nutritional support cannot be overstated in mitigating the stress response of burn injury. Failure to support fully the metabolic needs of the burn patient can lead to loss of lean tissue, poor wound healing, sepsis, multiple organ dysfunction syndrome, and death. Early enteral nutritional support (within 24 hours of injury) attenuates the hypermetabolic/catabolic response to severe burn injury. The caloric requirement for burn patients can be estimated by a variety of formulas (Table 5). Regardless of which formula is used, the primary goal of nutritional support is to match the metabolic demand caused by the hypermetabolic state, while maintaining safety

1304

Medical Management of the Burn Patient

TABLE 5  Common Formulas Used to Calculate Caloric Needs of Burn Patients Adult formulas

Kcal/day

Comments

Harris Benedict

Men: 66.5 + 13.8 (weight in kg) + 5 (height in cm) − 6.76 (age in years) Women: 655 + 9.6 (weight in kg) + 1.85 (height in cm) – 4.68 (age in years)

Estimates basal energy expenditure; can be adjusted by both activity and stress factor, multiply by 1.5 for common burn stress adjustment

Toronto formula

−4343 + 10.5 (TBSA) + 0.23 (calorie intake in last 24 hr) + 0.84 (Harris Benedict estimation without adjustment) + 114 (temperature) − 4.5 (number of postburn days)

Useful in acute stage of burn care; must be adjusted with changes in monitoring parameters

Davies and Lilijedahl

20 (weight in kg) + 70 (TBSA)

Overestimates caloric needs for large injuries

Ireton-­Jones

Ventilated patient: 1784 − 11 (age in years) + 5 (weight in kg) + (244 if male) + (239 if trauma) + (804 if burn) Nonventilated patient: 629 − 11 (age in years) + 25 (weight in kg) − (609 if obese)

Complex formula which integrates variables for ventilation and injury status

Curreri

Age 16–59: 25 (weight in kg) + 40 (TBSA) Age >60: 20 (weight in kg) + 65 (TBSA)

Often overestimates caloric needs

PEDIATRIC FORMULAS Galveston

0–1 year: 2100 (body surface area) + 1000 (body surface area × TBSA) 1–11 year: 1800 (body surface area) + 1300 (body surface area × TBSA) 12–18 years: 1500 (body surface area) + 1500 (body surface area × TBSA)

Focuses on maintaining body weight

From Clark A, Imran H, Madni T, Wolf SE. Nutrition and metabolism in burn patients. Burns & Trauma. 2017;5:11. https://doi.org/10.1186/[AU76]8-­017-­0076-­x.

(principally, from either pulmonary aspiration of gastric contents/ enteral feeding solution or use of a malpositioned tube) and avoiding overfeeding. The macronutrients in the formulation of burn nutrition solutions include carbohydrates, proteins, and lipids. Some research suggests that given the inhibition of lipolysis in the acute response to injury, lipids should be limited as a source of calories (fatty acid deficiency is a rarity that takes weeks to develop even absent any intake of fat). Comparison of high-­carbohydrate and protein, low-­fat enteral formula with low carbohydrate and protein, high-­fat formula in a systematic way showed no clear benefit to either formula, although the risk of pneumonia may have been lower with the high-­carbohydrate formula. Protein appears to be an essential macronutrient for wound healing, and protein requirements in burn patients may be 50% higher than in healthy individuals. Protein delivery should be 1.5 to 2 g/kg body weight daily. Parenteral nutrition in burn patients is typically reserved for patients whose gastrointestinal function is impaired. Parenteral nutrition requires central venous access, is associated with higher rates of complications in critically ill patients (e.g., hepatic dysfunction, catheter-­related blood stream infection), and does not support gut mucosa directly, as does enteral feeding. Immunonutrition (use of micronutrients that modify the immune response during critical illness), has not gained popularity despite supporting evidence. Immunonutrients that have been administered to burn patients include omega-­3 fatty acids (fish oil), arginine, and glutamine. Arginine is a conditionally essential amino acid that serves as a precursor to glutamate, promotes T-­cell proliferation, stimulates growth hormone and insulin secretion, and promotes wound healing. Glutamine is believed to be a conditionally essential amino acid that provides fuel for immune cells and enterocytes, serves as a precursor for the antioxidant glutathione, and potentially reduces insulin resistance. Whereas glutamine has shown promise as a nutritional supplement in the critical care population, a recent

review of immunonutrition in the burn population found insufficient evidence to recommend its use. Evidence-­based practice guidelines for the use of micronutrients in burn patients have not been developed. Intuitively, diminished gastrointestinal absorption, increased urinary losses, altered distribution, and altered carrier protein (e.g., albumin, lactoferrin) concentrations following severe burn will lead to a deficiency in many micronutrients if not supplemented. There is little evidence for pharmacologic doses of any micronutrient in burn patients. In fact, the majority of patients achieve the recommended supplemental micronutrients in sufficient amounts through standard nutritional therapy.  nn NEW TRENDS

IN BURN INJURY MANAGEMENT

As described earlier, the burn-­induced stress response stimulates secretion of catecholamines, glucocorticoids, and peptide hormone mediators that are believed to be the primary effectors of hypermetabolism after severe burns. Nonspecific β-­blockade with propranolol, a nonspecific beta-­1, beta-­2 adrenergic receptor antagonist, has shown good results in blunting the response to burn injury. Low-­dose propranolol (0.5–1.0 mg/kg) to patients with severe burns reduces myocardial oxygen consumption without affecting oxygen transport adversely. Higher doses (2 mg/kg/day for 5 days) cause a 20% reduction in heart rate. Propranolol also decreases resting energy expenditure, hepatic steatosis, and myolysis. The benefit of propranolol lasts demonstrably for at least 12 months after injury. Anabolic steroid production is decreased after severe burn injury and can remain below normal for up to 3 years postinjury. Restoring serum testosterone in severely burned patients can decrease lean tissue loss and improve protein synthesis. Oxandrolone, an analogue of testosterone, is well absorbed orally and is preferred to the authentic hormone for repletion of women and prepubescent boys. In skeletal

T R AU M A A N D E M E R G E N C Y C A R E

muscle, oxandrolone binds to intracellular androgen receptors and stimulates anabolism via protein synthesis. Porro et al. demonstrated that oxandrolone 0.1 mg/kg by mouth twice a day for 1 year postburn provides benefits that persist for up to five years post-­injury with few deleterious side effects. Other anticatabolic agents have also been used to improve muscle mass. Recombinant human growth hormone (rhGH) has shown promise as an agent for blunting catabolism, rhGH ameliorates the hypermetabolic response to major stress, and enhances immune function, protein synthesis, and wound healing, after burn injury. Unfortunately, Takala et al. found that rhGH increases mortality in critically ill adults. Concomitant use of rhGH and propranolol attenuated hypermetabolism, peripheral lipolysis, inflammation, and avoids the problematic effects of using rhGH alone. Insulin is a classic anabolic hormone, but stimulates weight gain primarily through increased fat stores. Vitamin C (ascorbic acid) is an antioxidant that acts to quench oxygen free radicals that cause tissue injury via oxidation of cell membrane lipids. Much of the imbalance in predicted fluid requirements and required administration in burn patients has been attributed to lipid peroxidation and dysregulation of scavenging systems for reactive nitrogen species, altering endothelial integrity and metabolism as the result of endothelial cell injury. Several laboratory and clinical studies demonstrated that adjunctive ascorbic acid in burn resuscitation decreases the exogenous fluid volume needed to restore and maintain adequate perfusion. Tanaka et al. randomized 37 patients with more than 30% TBSA burns to receive isotonic crystalloid and colloid fluid resuscitation with or without high-­dose intravenous vitamin C (66 mg/kg per hour) for the first 24 hours after admission. Patients who received ascorbic acid infusion in the first 24 hours post-­burn demonstrated improved partial pressure of oxygen in arterial blood:fraction of inspired oxygen with fewer ventilator days. Patients requiring escharotomies or fasciotomies were fewer in the ascorbic acid group. Kahn et al. conducted a 3-­year retrospective analysis of patients treated (2007–2009). The control group were resuscitated according to the Parkland formula, whereas the experimental group was resuscitated by the Parkland formula plus the volume required to administer an intravenous vitamin C dose of 66 mg/kg per hour. Twenty-­four-­ hour fluid requirements were 25% less in the vitamin C–treated subjects. Although overall resuscitation volume was decreased, the study demonstrated no improvements in respiratory function or mortality. Complications of high-­ dose vitamin C infusions include an increase in postburn AKI as a result of osmotic diuresis. This, in turn, leads to decreased fluid resuscitation requirements and worsening renal function without changing resuscitation endpoints. Additionally, oxalate nephropathy (obstructive uropathy secondary to concretions) has been described in postmortem examinations of burn patients treated with high-­dose vitamin C infusion. Hyperoxaluria can worsen existing kidney injury or delay kidney recovery. More research is needed before widespread adoption.  nn SUMMARY The medical management of burn resuscitation and continuing burn care has evolved over the past several decades. Judicious, balanced use of crystalloid fluid with colloid administration has resulted in minimizing “fluid creep” and its sequelae. Advances in monitoring and measurement of burn depth have also led to more precise estimation of fluid requirements. Adjuncts to care including nutrition support, prevention of complications, and the use of novel pharmacotherapies have all played a role in reducing the morbidity and mortality of burn injury.

Suggested Readings Acute Respiratory Distress Syndrome Network, Brower RG, Matthay MA, et al. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342(18):1301–1308. Attias O B-­JG. Abdominal compartment syndrome in children. In: Wheeler D, Wong H, Shanley T, eds. Pediatric Critical Care Medicine. Springer, London.

1305

Barie PS, Schubl SD, Narayan M. Multiple organ dysfunction syndrome. In: Acute Care Surgery. 2nd ed. 2018. Baxter CR. Fluid volume and electrolyte changes of the early postburn period. Clin Plast Surg. 1974;1(4):693–703. Bittner EA, Shank E, Woodson L, Martyn JAJ. Acute and perioperative care of the burn-­injured patient. Anesthesiology. 2015;122(2):448–464. Caironi P, Tognoni G, Masson S, et al. Albumin replacement in patients with severe sepsis or septic shock. N Engl J Med. 2014;370(15):1412–1421. Carter JE, Neff LP, Holmes JH. Adherence to burn center referral criteria: are patients appropriately being referred? J Burn Care Res. 2010;31(1):26–30. Cartotto R, Greenhalgh D. Colloids in acute burn resuscitation. Crit Care Clin. 2016;32(4):507–523. Cartotto R, Li Z, Hanna S, et  al. The Acute Respiratory Distress Syndrome (ARDS) in mechanically ventilated burn patients: an analysis of risk factors, clinical features, and outcomes using the Berlin ARDS definition. Burns. 2016;42(7):1423–1432. Caruso DM, Matthews MR. Monitoring end points of burn resuscitation. Crit Care Clin. 2016;32(4):525–537. Clark A, Imran J, Madni T, Wolf SE. Nutrition and metabolism in burn patients. Burn Trauma. 2017;5(1):11. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. BMJ. 1998;317(7153):235–240. Dezman ZDW, Comer AC, Smith GS, Narayan M, Scalea TM, Hirshon JM. Failure to clear elevated lactate predicts 24-­hour mortality in trauma patients. J Trauma Acute Care Surg. 2015;79(4):580–585. Du GB, Slater H, Goldfarb IW. Influences of different resuscitation regimens on acute early weight gain in extensively burned patients. Burns. 1991;17(2):147–150. Dünser MW, Takala J, Brunauer A, Bakker J. Re-­thinking resuscitation: leaving blood pressure cosmetics behind and moving forward to permissive hypotension and a tissue perfusion-­based approach. Crit Care. 2013;17(5):326. Eljaiek R, Heylbroeck C, Dubois M-­J. Albumin administration for fluid resuscitation in burn patients: a systematic review and meta-­analysis. Burns. 2017;43(1):17–24. Ferrada P, Anand RJ, Whelan J, et al. Qualitative assessment of the inferior vena cava: useful tool for the evaluation of fluid status in critically ill patients. Am Surg. 2012;78(4):468–470. Finfer S, Bellomo R, Boyce N, et  al. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med. 2004;350(22):2247–2256. Freiburg C, Igneri P, Sartorelli K, Rogers F. Effects of differences in percent total body surface area estimation on fluid resuscitation of transferred burn patients. J Burn Care Res. 2007;28(1):42–48. Greenhalgh DG, Saffle JR, Holmes JH, et al. American burn association consensus conference to define sepsis and infection in burns. J Burn Care Res. 2007;28(6):776–790. Haberal M, Abali AE, Karakayali H. Fluid management in major burn injuries. Indian J Plast Surg. 2010;43(3):29. Hagstrom M, Wirth GA, Evans GRD, Ikeda CJ. A review of emergency department fluid resuscitation of burn patients transferred to a regional, verified burn center. Ann Plast Surg. 2003;51(2):173–176. Hébert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med. 1999;340(6):409–417. Held JM, Litt J, Kennedy JD, et al. Surgeon-­performed hemodynamic transesophageal echocardiography in the burn intensive care unit. J Burn Care Res. 2016;37(1):e63–e68. Herndon DN. Total Burn Care. Philadelphia: Elsevier; 2012. Charpentier J, Mira ESG JP. Study. efficacy and tolerance of hyperoncotic albumin administration in septic shock patients: the EARSS Study. Intensive Care Med. 2011;37(suppl 1). Jaskille AD, Shupp JW, Jordan MH, Jeng JC. Critical review of burn depth assessment techniques: part I. Historical review. J Burn Care Res. 2009;30(6):937–947. Jeschke MG, Gauglitz GG, Kulp GA, et al. Long-­term persistence of the pathophysiologic response to severe burn injury. Androulakis IP, ed. PLoS One. 2011;6(7):e21245. Jeschke MG, Patsouris D, Stanojcic M, et  al. Pathophysiologic response to burns in the elderly. EBioMedicine. 2015;2(10):1536–1548. Jones LM, Deluga N, Bhatti P, Scrape SR, Bailey JK, Coffey RA. TRALI following fresh frozen plasma resuscitation from burn shock. Burns. 2017;43(2):397–402. Kahn SA, Beers RJ, Lentz CW. Resuscitation after severe burn injury using high-­dose ascorbic acid: a retrospective review. J Burn Care Res. 2011;32(1):110–117.

1306

Management of Frostbite, Hypothermia, and Cold Injuries

Kahn SA, Bernal N, Mosier MJ. Pearls from the national burn repository. J Burn Care Res. 2018;39(4):626–627. Mokline A, Abdenneji A, Rahmani I, et al. Lactate: prognostic biomarker in severely burned patients. Ann Burns Fire Disasters. 2017;30(1):35–38. Narayan M, Scalea T. Cardiovascular disease and monitoring. In: Peitzman AB, Fabian TC, Rhodes M, et al., eds. The Trauma Manual: Trauma and Acute Care Surgery. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2013. Navickis RJ, Greenhalgh DG, Wilkes MM. Albumin in burn shock resuscitation. J Burn Care Res. 2016;37(3):e268–e278. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000;85(4):599–610. Orgill DP. Excision and skin grafting of thermal burns. N Engl J Med. 2009;360(9):893–901. Palmieri TL, Holmes JH, Arnoldo B, et al. Transfusion requirement in burn care evaluation (TRIBE). Ann Surg. 2017;266(4):595–602. Papavramidis T, Pliakos I, Papavramidou N, Marinis A, Kesisoglou I. Abdominal compartment syndrome. Intra-­abdominal hypertension: defining, diagnosing, and managing. J Emerg Trauma Shock. 2011;4(2):279. Porro LJ, Herndon DN, Rodriguez NA, et al. Five-­year outcomes after oxandrolone administration in severely burned children: a randomized clinical trial of safety and efficacy. J Am Coll Surg. 2012;214(4):489–502. Pruitt BA. Protection from excessive resuscitation: “pushing the pendulum back”. J Trauma. 2000;49(3):567–568. Rae L, Fidler P, Gibran N. The physiologic basis of burn shock and the need for aggressive fluid resuscitation. Crit Care Clin. 2016;32(4):491–505.

Management of Frostbite, Hypothermia, and Cold Injuries Paul N. Manson, MD

nn IMMERSION

FOOT AND HAND

Immersion foot and hand are seen after prolonged exposure to cold but not freezing water. After recovery from the acute episode, major nerve paralysis may be seen in addition to chronic vasospastic cold sensitivity and pain. Pain in the affected area and paresthesias are commonly seen after all types of cold injury.  nn CHILBLAINS

T

he several types of localized cold injuries may be classified according to the temperature that produces them: (1) nonfreezing or (2) freezing temperatures. Trench foot, immersion hand and foot, and chilblains are produced by cold but not freezing temperatures. Frostbite is produced by freezing temperatures. Trench foot and immersion foot and hand are seen principally in military populations, whereas chilblains and frostbite are seen more commonly in civilian populations. Further, cold injuries may be divided into (1) localized cold injuries (i.e., frostbite) and (2) generalized cold injuries (i.e., hypothermia). Frequently, localized cold injuries such as frostbite do not co-­exist with systemic hypothermia. Localized cold injuries have in common the fact that they are produced by exposure to cold stimuli and that they occur at the extremities of circulation. Localized cold injuries may be seen in the cheeks, nose, ears, and face but are seen primarily in the hands and feet. Cold injuries in the face tend to be superficial because of the blood supply. Serious cold injuries are confined almost exclusively to the extremities, in which there is a small margin of difference between the injuries that produce superficial versus deep injury. nn TRENCH

Reed SF, Britt RC, Collins J, Weireter L, Cole F, Britt LD. Aggressive surveillance and early catheter-­directed therapy in the management of intra-­ abdominal hypertension. J Trauma. 2006;61(6):1359–1363. Reems MM, Aumann M. Central venous pressure: principles, measurement, and interpretation. Compend Contin Educ Vet. 2012;34(1):E1. Resch TR, Drake RM, Helmer SD, Jost GD, Osland JS. Estimation of burn depth at burn centers in the United States: a survey. J Burn Care Res. 2014;35(6):491–497. Rizzo JA, Rowan MP, Driscoll IR, Chung KK, Friedman BC. Vitamin C in burn resuscitation. Crit Care Clin. 2016;32(4):539–546. Rojas Y, Finnerty CC, Radhakrishnan RS, Herndon DN. Burns: an update on current pharmacotherapy. Expert Opin Pharmacother. 2012;13(17):2485– 2494. Suresh MR, Dries DJ. Critical care update burn care: resuscitation and respiratory care. Air Med J. 2018;37:12–15. Takala J, Ruokonen E, Webster NR, et  al. Increased mortality associated with growth hormone treatment in critically Ill adults. N Engl J Med. 1999;341(11):785–792. Tan HB, Danilla S, Murray A, et al. Immunonutrition as an adjuvant therapy for burns. Cochrane database Syst Rev. 2014;12:CD007174. Wearn C, Lee KC, Hardwicke J, et  al. Prospective comparative evaluation study of laser doppler Imaging and thermal imaging in the assessment of burn depth. Burns. 2018;44(1):124–133. Zuo KJ, Medina A, Tredget EE. Important developments in burn care. Plast Reconstr Surg. 2017;139(1):120e–138e.

FOOT

Trench foot is usually seen in military populations and occurs with exposure to above-­freezing temperatures, generally over a prolonged period of time. The presence of moisture is very important in its pathogenesis. Chronic symptoms produced after recovery from the acute injury are those of pain, paresthesia, and a particular susceptibility to further cold injury. 

Chilblains represent the mildest form of cold injury and occur after prolonged exposure to cold and wet conditions. The symptoms consist of burning and itching and are associated with a mild dermatitis. Vesicles and hemorrhagic lesions may be seen in the acute period. The chronic condition is characterized by cold sensitivity, itching, paresthesias, and skin eruptions, which may be reddish lesions, vesicles, or superficial ulcers. The chronic condition may be treated by protection from cold and heat to avoid production of dermatitis symptoms and pain. The role of sympathetic denervation in the management of the chronic condition has been suggested but not established. Some feel it may be helpful in chronic, well-­established symptoms that require treatment. Chilblains do not produce tissue loss, and thus they require no reconstruction.  nn FROSTBITE Frostbite occurs from exposure to freezing temperatures. The period of exposure required for its production may be short or long, depending on environmental conditions, wind, and protection. Frostbite has been classified into degrees of injury depending on the depth of damage. Often several degrees of injury will be seen in the same extremity, with the damage increasing as injury progresses from proximal to distal.

First-­Degree Frostbite First-­ degree frostbite is a superficial skin injury characterized by numbness, edema, and erythema. The injury is similar to a

1306

Management of Frostbite, Hypothermia, and Cold Injuries

Kahn SA, Bernal N, Mosier MJ. Pearls from the national burn repository. J Burn Care Res. 2018;39(4):626–627. Mokline A, Abdenneji A, Rahmani I, et al. Lactate: prognostic biomarker in severely burned patients. Ann Burns Fire Disasters. 2017;30(1):35–38. Narayan M, Scalea T. Cardiovascular disease and monitoring. In: Peitzman AB, Fabian TC, Rhodes M, et al., eds. The Trauma Manual: Trauma and Acute Care Surgery. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2013. Navickis RJ, Greenhalgh DG, Wilkes MM. Albumin in burn shock resuscitation. J Burn Care Res. 2016;37(3):e268–e278. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000;85(4):599–610. Orgill DP. Excision and skin grafting of thermal burns. N Engl J Med. 2009;360(9):893–901. Palmieri TL, Holmes JH, Arnoldo B, et al. Transfusion requirement in burn care evaluation (TRIBE). Ann Surg. 2017;266(4):595–602. Papavramidis T, Pliakos I, Papavramidou N, Marinis A, Kesisoglou I. Abdominal compartment syndrome. Intra-­abdominal hypertension: defining, diagnosing, and managing. J Emerg Trauma Shock. 2011;4(2):279. Porro LJ, Herndon DN, Rodriguez NA, et al. Five-­year outcomes after oxandrolone administration in severely burned children: a randomized clinical trial of safety and efficacy. J Am Coll Surg. 2012;214(4):489–502. Pruitt BA. Protection from excessive resuscitation: “pushing the pendulum back”. J Trauma. 2000;49(3):567–568. Rae L, Fidler P, Gibran N. The physiologic basis of burn shock and the need for aggressive fluid resuscitation. Crit Care Clin. 2016;32(4):491–505.

Management of Frostbite, Hypothermia, and Cold Injuries Paul N. Manson, MD

nn IMMERSION

FOOT AND HAND

Immersion foot and hand are seen after prolonged exposure to cold but not freezing water. After recovery from the acute episode, major nerve paralysis may be seen in addition to chronic vasospastic cold sensitivity and pain. Pain in the affected area and paresthesias are commonly seen after all types of cold injury.  nn CHILBLAINS

T

he several types of localized cold injuries may be classified according to the temperature that produces them: (1) nonfreezing or (2) freezing temperatures. Trench foot, immersion hand and foot, and chilblains are produced by cold but not freezing temperatures. Frostbite is produced by freezing temperatures. Trench foot and immersion foot and hand are seen principally in military populations, whereas chilblains and frostbite are seen more commonly in civilian populations. Further, cold injuries may be divided into (1) localized cold injuries (i.e., frostbite) and (2) generalized cold injuries (i.e., hypothermia). Frequently, localized cold injuries such as frostbite do not co-­exist with systemic hypothermia. Localized cold injuries have in common the fact that they are produced by exposure to cold stimuli and that they occur at the extremities of circulation. Localized cold injuries may be seen in the cheeks, nose, ears, and face but are seen primarily in the hands and feet. Cold injuries in the face tend to be superficial because of the blood supply. Serious cold injuries are confined almost exclusively to the extremities, in which there is a small margin of difference between the injuries that produce superficial versus deep injury. nn TRENCH

Reed SF, Britt RC, Collins J, Weireter L, Cole F, Britt LD. Aggressive surveillance and early catheter-­directed therapy in the management of intra-­ abdominal hypertension. J Trauma. 2006;61(6):1359–1363. Reems MM, Aumann M. Central venous pressure: principles, measurement, and interpretation. Compend Contin Educ Vet. 2012;34(1):E1. Resch TR, Drake RM, Helmer SD, Jost GD, Osland JS. Estimation of burn depth at burn centers in the United States: a survey. J Burn Care Res. 2014;35(6):491–497. Rizzo JA, Rowan MP, Driscoll IR, Chung KK, Friedman BC. Vitamin C in burn resuscitation. Crit Care Clin. 2016;32(4):539–546. Rojas Y, Finnerty CC, Radhakrishnan RS, Herndon DN. Burns: an update on current pharmacotherapy. Expert Opin Pharmacother. 2012;13(17):2485– 2494. Suresh MR, Dries DJ. Critical care update burn care: resuscitation and respiratory care. Air Med J. 2018;37:12–15. Takala J, Ruokonen E, Webster NR, et  al. Increased mortality associated with growth hormone treatment in critically Ill adults. N Engl J Med. 1999;341(11):785–792. Tan HB, Danilla S, Murray A, et al. Immunonutrition as an adjuvant therapy for burns. Cochrane database Syst Rev. 2014;12:CD007174. Wearn C, Lee KC, Hardwicke J, et  al. Prospective comparative evaluation study of laser doppler Imaging and thermal imaging in the assessment of burn depth. Burns. 2018;44(1):124–133. Zuo KJ, Medina A, Tredget EE. Important developments in burn care. Plast Reconstr Surg. 2017;139(1):120e–138e.

FOOT

Trench foot is usually seen in military populations and occurs with exposure to above-­freezing temperatures, generally over a prolonged period of time. The presence of moisture is very important in its pathogenesis. Chronic symptoms produced after recovery from the acute injury are those of pain, paresthesia, and a particular susceptibility to further cold injury. 

Chilblains represent the mildest form of cold injury and occur after prolonged exposure to cold and wet conditions. The symptoms consist of burning and itching and are associated with a mild dermatitis. Vesicles and hemorrhagic lesions may be seen in the acute period. The chronic condition is characterized by cold sensitivity, itching, paresthesias, and skin eruptions, which may be reddish lesions, vesicles, or superficial ulcers. The chronic condition may be treated by protection from cold and heat to avoid production of dermatitis symptoms and pain. The role of sympathetic denervation in the management of the chronic condition has been suggested but not established. Some feel it may be helpful in chronic, well-­established symptoms that require treatment. Chilblains do not produce tissue loss, and thus they require no reconstruction.  nn FROSTBITE Frostbite occurs from exposure to freezing temperatures. The period of exposure required for its production may be short or long, depending on environmental conditions, wind, and protection. Frostbite has been classified into degrees of injury depending on the depth of damage. Often several degrees of injury will be seen in the same extremity, with the damage increasing as injury progresses from proximal to distal.

First-­Degree Frostbite First-­ degree frostbite is a superficial skin injury characterized by numbness, edema, and erythema. The injury is similar to a

T R AU M A A N D E M E R G E N C Y C A R E

1307

cycle may involve actual freezing and thawing of tissue. Circulation virtually ceases as the vasoconstriction increases, creating marked ischemia of the tissue. 

Direct (Freezing) Damage to Cells Freezing produces extracellular and then intracellular ice crystals. Extracellular ice crystals produce direct damage to cell walls and also cause a leaching out of electrolytes and water from the intracellular compartment, resulting in intracellular dehydration followed by cell death. Intracellular ice crystals produce direct damage to important cell structures. 

Progressive Vascular Thrombosis

In second-­degree frostbite, partial-­thickness skin injury occurs that is characterized by numbness, edema, erythema, and vesiculation. The vesicles may be filled with either clear or bloody fluid. The partial-­ thickness skin injury heals in 2 to 4 weeks (Fig. 2). The quality of the regenerated skin depends on the depth of the injury to the dermis and parallels thermal burn injuries in that deep second-­degree injuries heal with thin atrophic skin that has a reduced number of skin appendages. 

Frostbite produces progressive dermal ischemia, similar to the process seen in thermal burns or in the no-­reflow phenomenon. The role of edema and endothelial injury in the subsequent arrest of dermal blood flow has implicated various inflammatory mediators such as thromboxanes, prostaglandins, histamine, and bradykinin; these observations have led to the use of inhibitors of these mediators in an attempt to ameliorate the harmful tissue reaction and improve tissue survival. Tissue injury occurs both from direct damage to cells and from the vasoconstriction anoxia, which results in vascular stasis and thrombosis. After the state of decreased circulation, a state of reactive hyperemia occurs, which may or may not be associated with a no-­reflow phenomenon, in which capillary thrombosis occurs in the zone of stasis. The pathology of tissue repair is similar to that of burn injuries in that epithelium migrates to cover the wound from the surviving remnants of sweat glands, hair follicles, and margins of the living wound. The quality of the skin produced is inversely proportional to the depth of damage. Deep injuries have necrotic nerve, muscle, and bone and are healed by scar after the shedding of dead tissue. Frequently, autoamputation occurs after ischemic necrosis and gangrene. Left untreated, the process is generally dry and progresses to autoamputation without infection over several months. 

Third-­Degree Frostbite

Environmental Influences

Third-­degree frostbite represents full-­thickness skin loss. After the injury, a nonviable segment of full-­thickness skin loss is observed; this may be seen initially as a gray-­blue patch, or death of the skin may follow an initial period of reactive hyperemia after 24 to 72 hours. Eventually, a black eschar forms that generally separates slowly in 1 to 3 months unless infection occurs (Fig. 3). 

A number of behavioral and environmental factors may influence the production of cold injuries. Two of the most significant environmental factors are the ambient temperature and the wind. The effects of temperature may be modified by wearing protective clothing, which provides insulation proportional to its thickness and weight. The effects of temperature are modified by the presence of wind and wet conditions, which accelerate heat loss. Wet conditions increase heat conduction to the environmental air, whereas wind accelerates the loss of heat in the air. Siple has developed a “wind chill index” to reflect the magnitude of the contribution of wind in heat loss. The heat loss would be the same at 20°F with a wind of 45 miles per hour as at –40°F with a wind of 2 miles per hour. The mean outdoor temperature recorded in frostbite injury series is –29°C. Frostbite is seen isolated to the upper extremities in 19% of cases, isolated to the lower extremities in 47% of cases, in both upper and lower extremities in 31% of cases, and with no extremity involvement in only 3% of cases. Clothes provide insulation proportional to their thickness (one-­ quarter of an inch equals one clothing unit) and weight. They should be light to allow activity and trap air in multiple layers to be effective. Sweating wets the clothing, reducing the insulation value. The importance of light clothing permitting work (heat production) is emphasized. The proper use of protective clothing is important. In one study, 65% of those who suffered frostbite had inadequate protective clothing, whereas 20% had adequate clothing but were wearing it improperly. Only 15% of frostbite victims had adequate clothing and were wearing it properly. Moisture accelerates heat loss. Behavioral factors also predispose to cold injuries. Alcohol and drug intoxication; smoking; accidents such as car, plane, skiing, and vehicular failure; homelessness; high altitude; and outdoor pursuits

FIG. 1  First-­and second-­degree frostbite. Epidermal skin loss and

peeling are seen.

first-­degree burn in that it heals spontaneously (in terms of the epithelium) in 1 to 2 weeks. Superficial desquamation may occur, and regeneration is usually complete with decreased but adequate skin appendages (Fig. 1). 

Second-­Degree Frostbite

Fourth-­Degree Frostbite Fourth-­degree frostbite signifies necrosis of all deep tissue parts down to and sometimes including bone. Black, mummified tissues are present with the initial episode. If the mummified area becomes infected, it softens and becomes swollen and macerated at the margin with viable tissue. 

Pathophysiology of Frostbite Two pathophysiologic mechanisms account for the production of frostbite injury. One is vasoconstriction and damage to the microcirculation in the zone of vascular stasis, which results in progressive vascular thrombosis. The second is direct damage to the cells or cellular toxicity from freezing. Experts disagree on the importance of these two mechanisms, and the importance of each mechanism varies according to the amount of tissue freezing that occurs and the local conditions, such as circulation.

“Hunting” Reaction The body responds to cold with an alternating cycle of vasoconstriction and vasodilatation (called the hunting reaction) in an attempt to conserve heat loss from the skin. As the injury becomes worse, the

1308

Management of Frostbite, Hypothermia, and Cold Injuries

A

B

FIG. 2  (A) Second-­degree frostbite in the fingers. Vesicles and swelling are seen. (B) Second-­degree frostbite in the feet. Vesicles and swelling are

seen.

skin warming. Military experience has emphasized the importance of working and keeping active so as to increase heat production. The importance of avoiding sweating to avoid wetness and replacing wet with dry clothing also is emphasized. 

Superficial Versus Deep Frostbite

FIG. 3 Third-­and fourth-­degree frostbite with tissue death and

demarcation beyond the interphalangeal joint.

predispose to frostbite, as does a previous cold weather injury. Medical conditions that decrease circulation (peripheral arterial disease), neuropathy, and diabetes all decrease the body’s ability to adapt to colder temperatures; psychiatric illness is observed much more frequently in patients who suffer frostbite. The average age of a frostbite victim is 30 to 49 years, with men outnumbering women by 10 to 1. 

Prevention of Frostbite There are two ways to prevent cold injury: (1) by increasing heat production and (2) by decreasing heat loss. Heat loss may be decreased by avoiding wetness or contact with metal (which accelerates heat loss) and by wearing adequate protective clothing. The extremities have a large surface-­to-­mass ratio, and thus represent prime sites for heat loss. Other factors may affect the circulation to extremities, such as the presence of arterial occlusive disease (some feel that the presence of a frostbite injury should prompt an examination for an arterial occlusive lesion or conditions such as diabetic neuropathy, which do not permit reflex vasoconstriction and vasodilatation). The importance of inactivity and immobility in reducing heat production, producing orthostatic edema, and decreasing circulation has been emphasized in studies on military frostbite. Malnutrition, hemorrhage, anemia, and the use of tobacco and alcohol all have been implicated in the increased susceptibility to frostbite injuries. Acclimatization and cold tolerance probably occur. African Americans have increased susceptibility to cold injuries, probably because of less frequent waves of cold-­induced vasodilatation, and thus less effective

Frostbite in the head and neck area is generally superficial. The face is not subject to the same vasoconstrictive phenomena as are the extremities. The drying potential of cold in the facial area is manifested in chapped lips, nose tip, ears, and mucous membranes. Facial frostbite is generally superficial, whereas serious frostbite usually is confined to the extremities. Mills believes that the differentiation of frostbite into first-­ , second-­, third-­, and fourth-­degree injuries is cumbersome and not clinically useful. He believes that one can only classify frostbite as superficial (tissue remains soft) or deep (tissue is hard). It is initially difficult to tell the depth of the injury, and differentiation can be accomplished only after rewarming and a period of observation. Cauchy (2001) believes that the extent of the initial lesion and the result of three-­phase bone scanning can predict the ultimate result: if the initial lesion is confined to the distal digit, the probability of amputation is 1%; the probability increases to 31% for the middle digit, 67% for the proximal phalanx, 98% for the metacarpal and metatarsal, and 100% for the carpal and tarsal areas. In the face, a small white patch of tissue may be seen, and this clears over the course of a week. Medically, frostbite heals with no therapy, therapy with dressings alone, or surgical debridement amputation and dressings. Twenty-­ five percent of patients with frostbite are seen more than 48 hours after the injury. In the extremities, mild frostbite is manifested by pallor, paresthesias, and a dull yellow color of the skin. Ice crystals may be observed. The area is numb, and after rewarming a prickly, itchy sensation or aching pain occurs. After rewarming, reactive hyperemia is observed superficially; in deeper injuries, hypersensitivity and paresthesias are observed. Deep frostbite is differentiated by the absence of circulation; on rewarming, the progression (progressive vascular thrombosis) to full-­thickness tissue loss (eschar formation) occurs. The tissue may remain insensitive after rewarming, presenting as a blue-­ gray patch with absent circulation. Burning pain, paresthesias, and thick-­walled blisters containing blood may follow rewarming in full-­ thickness tissue injury. The history of the injury is important in predicting tissue loss. Important factors include duration of the exposure, temperature, protective clothing worn, contact with metal or moisture, and the presence of previous symptoms that would indicate reduced arterial

T R AU M A A N D E M E R G E N C Y C A R E

BOX 1 Therapy for Frostbite • Correct systemic hypothermia (temperature 1000 V) or low voltage (90 beats/min

Respiratory rate

>20 breaths/min or partial pressure of carbon dioxide 12,000 cells/mm3, 10% bands

Two criteria need to be present to diagnose a patient with systemic inflammatory response syndrome. From Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/ SCCM Consensus Conference Committee. American College of Chest ­Physicians/Society of Critical Care Medicine. Chest. 1992;101(6):1644-­1655.

BOX 2  Six Parameters Composing the National Early Warning System for Sepsis • Respiratory rate • Oxygen saturation • Temperature • Systolic blood pressure • Pulse rate • Level of consciousness/confusion

  From Keep JW, Messmer AS, Sladden R, et al. National early warning score at emergency department triage may allow earlier identification of patients with severe sepsis and septic shock: a retrospective observational study. Emerg Med J. 2016; 33:37-­41.

illustrates the limited ability of SIRS criteria to diagnose sepsis as well as the tremendous genetic variability of responses to systemic infections. The Third International Consensus Definitions for Sepsis and Septic Shock (aka Sepsis 3) proposed the quick Sequential Sepsis-­related Organ Failure Assessment (qSOFA) as a rapid and simple bedside tool to help diagnose sepsis. The components of this score compose hypotension, tachypnea, and altered mental status; derangements in any two of the three should prompt consideration of sepsis. Criticisms of the qSOFA stem from the fact that the patient must manifest end-­organ damage before the tool suggests the diagnosis of sepsis; thus, the sensitivity of qSOFA may be too low and overreliance on this instrument may expose patients to increased mortality due to delays in diagnosis and treatment. Novel early warning scores for sepsis such as the National Early Warning Score or Modified Early Warning Score consider six clinical parameters that aim to identify patients at risk for deterioration; however, these scores have not yet been validated (Box 2). It is likely that no scoring system, combination of biomarkers, or other clinical decision tool will emerge as the benchmark for diagnosing sepsis because of genetic heterogeneity and the variability of individual manifestations of whole-­body immune-­inflammatory activation. The importance of a detailed and thorough history and physical when encountering a patient who may be septic cannot be overstated, but this process may be adversely affected by sepsis itself. Of all the organ systems impacted by sepsis, none is more important than the central nervous system (CNS). The sensitivity of the CNS to dysfunction in sepsis is why nearly every sepsis diagnostic tool or scoring system considers alterations in sensorium or level of consciousness as one of the indicators of sepsis. Septic encephalopathy, a global CNS dysfunction in response to systemic infection, is a common manifestation with an incidence between 30% and 70%. The etiology of this

1428

Septic Response and Management

dysfunction is not well elucidated, but it is thought to be due to the body’s inflammatory response to systemic infection. Unless there are localizing signs, imaging studies are exceedingly low yield and are not routinely indicated to workup septic encephalopathy. Previously, septic encephalopathy was thought to be transient and reversible; however, emerging data suggest that there may be lifelong neurocognitive deficits in survivors of sepsis. This long-­term postsepsis CNS dysfunction is one of the central features of postintensive care syndrome (PICS). It appears that PICS may be more common following ICU discharge for sepsis compared with poly-­trauma or other causes of critical illness. There are no therapies or interventions to treat septic encephalopathy other than supportive care, so its prevention through the timely diagnosis and treatment of sepsis is the best option. Similarly, efforts to avoid ICU delirium appear to be the best intervention to prevent PICS. The interplay between delirium and encephalopathy is complex and not well understood but both likely contribute to PICS. A bundle of ICU-­based efforts known as the ABCDEF (Assess and treat pain, awakening and spontaneous Breathing trials, Choice of drugs, Delirium avoidance, Early mobility, and Family engagement) bundle appears to be a very effective strategy to prevent delirium and PICS and decrease overall ICU mortality (Table 2). As with the CNS, the heart is also sensitive to dysfunction in sepsis, particularly in cases of septic shock. The initial response of the cardiovascular system to systemic infection is to increase cardiac output through tachycardia. This typically results in a hyperdynamic state but as the sepsis progresses, 20% to 50% of patients will manifest sepsis-­ related myocardial dysfunction. Circulating cytokines and other inflammatory mediators promote capillary leak and loss of intravascular fluid into the interstitium, sometimes referred to as the third space. Myocardial dysfunction is compounded by this loss of preload and cardiac output falls due to decreased stroke volume. One of the mainstays of therapy for septic patients is the administration of intravenous crystalloid to restore an effective circulating volume. Nearly two decades ago, a landmark trial showed that aggressive fluid administration to achieve predetermined endpoints of resuscitation within 6 hours in patients with severe sepsis or septic shock yielded lower mortality than standard therapy. The results of this breakthrough study have had far-­reaching effects on the management of septic patients and have rendered the resuscitation of septic patients more uniform. Subsequent prospective trials have failed to show a benefit of additional intravenous fluid resuscitation following an effective resuscitation. The interpretation of this finding should not be that fluid administration is not as beneficial in sepsis, as previously shown, but rather that the standard of care in managing septic shock now universally includes early and aggressive fluid resuscitation and that supra-­physiologic resuscitation is not helpful. Three decades ago, a similar finding was observed when surgical patients who were resuscitated and exhibited supra-­normal parameters had better outcomes compared with those with standard parameters. The survival benefit of supra-­normal resuscitation was lost when compared with surgical patients who were adequately

resuscitated before randomization to an induced supra-­physiologic group. Considering these two seminal papers, the goals of resuscitation in septic patients ought to be to administer enough fluid to restore tissue and organ perfusion, while avoiding overresuscitation.  nn FLUID

RESUSCITATION

The definition of adequate fluid resuscitation has always been controversial and despite a myriad of surrogate endpoints, there is no ideal marker or endpoint of resuscitation in sepsis. It is important that clinicians remember this because overreliance on any one single sign is likely to lead to more adverse outcomes than good. Central venous pressure (CVP) has had varying degrees of popularity over the years as a marker of resuscitation. Early sepsis guidelines suggested that patients be resuscitated to a CVP of 8 to 12 mm Hg; however, CVP does not correlate with responsiveness to fluid administration. CVP has limited ability to estimate cardiac preload or left ventricular end diastolic pressure but may help in differentiating certain types of shock states, a high CVP is observed in cardiogenic shock, whereas a low CVP is observed in hypovolemic shock. It must be remembered that CVP is not precise enough to fine tune the amount of fluid to administer to a septic patient. Similarly, elevated lactic acid is touted as both an indicator of sepsis and lactate clearance is proposed to be a marker of adequate fluid administration. The kinetics of lactic acid metabolism are variable based on patient genetics, muscle mass, and microcirculation and may be dramatically affected by certain medications. Overreliance on lactic acid determination is likely to result in excess fluid administration, particularly if the lactic acid is due to other causes besides anaerobic tissue metabolism as is seen in so called type B lactic acidosis. Lactate clearance has been extensively studied in sepsis and the preponderance of data suggests that those patients who clear their lactic acidosis will have lower mortality than those who do not thus there is some utility in trending this parameter. Although sepsis guidelines do not include any recommendations for pulmonary artery catheter (PAC) use in the management of septic patients, it still has an important role in care of the septic patient. The PAC measures cardiac output, systemic and pulmonary vascular resistance, and mixed venous oxygen saturation and facilitates calculation of oxygen delivery and consumption. Of all the measurements and parameters that can be determined from PAC, none is more useful that the mixed venous saturation (SVO2). SVO2, the oxygen saturation of the sum of all the venous return in the body, is a global indicator of oxygen delivery and consumption. If the SVO2 is above 40%, it can be assumed that all vascular beds are receiving adequate oxygen (and blood flow) but, as with all measurements, there are pitfalls to relying on a single number. SVO2 is susceptible to sampling errors and may be falsely elevated in some medical conditions where an arteriovenous shunt exists, such as portal hypertension or in those patients with a hemodialysis fistula. In addition to the morbidity of central access and the catheter itself, PAC use is associated with higher complications, thus its routine use in sepsis is not recommended. Earlier

TABLE 2  ICU Liberation: ABCDEF Bundle Symptoms: Pain, Agitation, Delirium Guidelines

Monitoring Tools

Care: ABCDEF Bundle

Pain

Critical-­Care Pain Observation Tool Numeric Rating Scale Behavioral Pain Scale

Agitation

Richmond Agitation-­Sedation Scale Sedation-­Agitation Scale (SAS)

Delirium

Confusion Assessment Method for the Intensive Care Unit Intensive Care Delirium Screening Checklist

A: Assess, prevent, and manage pain B: Both spontaneous awakening trials (SAT) and spontaneous breathing trials C: Choice of analgesia and sedation D: Delirium: assess, prevent, and manage E: Early mobility and exercise F: Family engagement and empowerment

Modified from Ely EW. The ABCDEF bundle: science and philosophy of how ICU liberation serves patients and families. Crit Care Med. 2017;45(2):321-­330.

SURGIC AL CRITIC AL C ARE

iterations of the sepsis guidelines recommended monitoring of the mixed venous saturation of blood obtained from the superior vena cava (ScvO2) to guide resuscitation of the septic patient. Although some of the morbidity of having a catheter residing inside the patient’s heart was avoided, ScvO2 monitoring is associated with many of the same pitfalls that were seen with SVO2 monitoring. SVO2 or ScvO2, if available, can aid in the decision to transfuse a septic patient, especially in cases where cardiac output is fixed however restrictive transfusion practices are generally safe in septic patients. A transfusion “trigger” of 7 mg/dL is reasonable. Systolic cardiac failure is less common in sepsis than diastolic dysfunction but in extreme cases, the heart will manifest both forms of cardiac failure. This biventricular type of infection-­induced cardiac failure is termed septic cardiomyopathy. Patients with septic cardiomyopathy typically exhibit severely diminished cardiac contractility and a low cardiac output. This type of cardiac failure does not respond to additional fluid administration; these patients require inotropic support to boost their cardiac output. This population typically requires a PAC to guide management. Dobutamine is the most commonly used agent to increase cardiac contractility but it can worsen hypotension via its vasodilatory effect on certain vascular beds. Epinephrine, a second-­line agent, can augment both cardiac output and blood pressure, but it is arrhythmogenic and can wreak havoc on glucose metabolism. The preferred vasoactive agent in septic shock is norepinephrine because it greatly increases vasomotor tone and blood pressure as well as increases cardiac output through its lesser inotropic effects. Dopamine should not be used and is associated with higher mortality than norepinephrine in the treatment of septic cardiomyopathy. Patients with hypotension refractory to norepinephrine benefit from treatment with the nonadrenergic agent vasopressin, which increases vasomotor tone and blood pressure but has no effect on cardiac output or contractility. Interestingly, during a recent shortage of norepinephrine, sepsis mortality was noted to rise in large national databases when phenylephrine was substituted, thus this agent should not be used in the management of septic shock.  nn CURRENT

RECOMMENDATIONS

Current recommendations are to administer 30 mL/kg of balanced crystalloid (e.g., lactated Ringer’s) to septic patients exhibiting hypotension and/or increased serum lactate, over 3 hours, which is about 2 L of fluid for an average 70-­kg patient. Classic biomarkers such as urinary output, heart rate, blood pressure, and skin perfusion should be continually assessed and monitored for responsiveness to fluid administration and can roughly guide resuscitation. There are conflicting data about the risks and benefits of colloids in the resuscitation of septic patients. Albumin is costly and does not have a mortality benefit or improved organ function when compared to crystalloid resuscitation. Albumin is 64,000 Daltons, a relatively small molecule, and passes freely from the vasculature into the interstitium in septic patients. This fact is in direct opposition to the often-­cited dictum that albumin is a large molecule that tends to stay in blood vessels and as such is falsely believed to be a superior resuscitation fluid. The use of hydroxy ethyl starches in septic patients is associate with the need for renal replacement therapy and should not be used in the resuscitation of septic patients. Current sepsis guidelines have moved away from static measurements of serum lactate or global markers of perfusion and now consider fluid responsiveness in identifying those patients requiring further resuscitation. Passively raising a patient’s legs 45 degrees by raising the lower half of the hospital bed such that the legs are above the heart is a simple bedside test that can guide fluid resuscitation. Patients who exhibit an increase in systolic blood pressure within 60 seconds of the maneuver are very likely to benefit from additional fluid administration; however, this test can be invalid if the patient has elevated intraabdominal pressures, ascites, etc. Mechanically ventilated patients exhibiting a varying baseline on their arterial tracing are likely intravascularly depleted. As the ventilator cycles, intrathoracic pressure

1429

rises, and cardiac preload falls with a resultant narrowing of the pulse pressure. Many bedside monitors are now able to detect and calculate this pulse pressure variation (PPV). Patients with a PPV of 12% or more will benefit from additional intravenous fluid while those with a PPV less than 8% are likely euvolemic. A PPV between 8% and 12% is indeterminant and cannot accurately guide fluid administration. Bedside echocardiography performed by intensivists is rapidly becoming the modality of choice to help guide fluid resuscitation in septic and critically ill patients. As with PPV, variation in the diameter of the inferior or superior venae cavae during the respiratory cycle can identify those patients who are likely hypovolemic. Patients with small contracted cardiac ventricles require further fluid resuscitation and papillary apposition or “kissing ventricles” strongly indicates that the patient will be fluid responsive. Like PPV, stroke volume variation is another indicator that can guide resuscitation and patients with a stroke volume variation of 12% or more are very likely to be fluid responsive. Narrowing of the outflow tract during systole implies left ventricular underfilling and can also guide fluid management. It is best if several echocardiographic findings are considered and a global picture is constructed indicating the patient’s volume status and likelihood of responding to additional fluid rather than fixating on any one single echocardiographic parameter. Although echocardiography is an excellent bedside tool to help guide the resuscitation of septic patients, it has not yet been validated as the gold standard to govern volume resuscitation in sepsis.

Cardiac System The cardiac system is affected several ways in sepsis and it is not uncommon for a septic patient to manifest elevations in troponin in the bloodstream. Elaboration of troponin from the myocardium is a sign of inflammation rather than a marker of coronary atherosclerotic disease and the treatment of this damage should be to obtain timely source control, begin fluid resuscitation and antibiotic therapy. It is inappropriate to delay effective source control to focus on the troponin elevation and cardiology consultations introduce delays in definitive therapy. Even significant sepsis-­related cardiac failure will resolve in a few days if the basic doctrines of sepsis management are observed and long-­term cardiac dysfunction is not common. Troponin elevation during sepsis is a true biomarker of sepsis related injury and the degree of elevation appears to directly correlate with mortality in sepsis. As with troponin elevations, atrial fibrillation (AF) is another common finding in sepsis. The most likely explanation for AF in septic patients is that it is due to the presence of multiple inflammatory mediators that makeup septic response. The incidence of AF rises as patients are treated with fluid and vasoactive substances to achieve higher mean arterial pressures (MAPs) rather than lower MAPs. Nearly one in five patients with sepsis will manifest AF and its incidence appears to be higher in those patients with preexisting cardiac comorbidities. Analogous to troponin leaks, the development of AF in sepsis portends a higher mortality. β-­blocker therapy in the treatment of sepsis-­related AF confers a survival benefit, but this does not appear to be the case with other rate or rhythm controlling agents such as calcium channel blockers or amiodarone. Amiodarone is preferred over β-­blockers for AF in patients with a compromised or failing left ventricle. It is likely that some of β-­blocker’s survival benefit is derived from widespread antiinflammatory protective effects and that the survival benefit is not solely derived from its effects on the cardiovascular system. Unlike nonsepsis-­related AF, anticoagulation to prevent stroke is associated with increased bleeding events. Although evidence to guide clinicians is limited, it appears that the increased bleeding complications of anticoagulation in sepsis outweighs it benefits, particularly in those patients with incisions, anastomoses, and other surgical interventions. 

Lungs Unlike the heart, the lungs are both a frequent home of septic foci as well as a target of the inflammatory products of sepsis. The initial

1430

Septic Response and Management

response of the pulmonary system in sepsis is to increase respiratory rate promoting a respiratory alkalosis to compensate for worsening metabolic acidosis. Ironically, this increase in work of breathing contributes to metabolic acidosis as well because ischemic respiratory muscles elaborate lactic acid. Capillary leak and fluid resuscitation increase lung water and alveolar congestion, which hinders oxygen diffusion far more than that of carbon dioxide, resulting in hypoxemia. Respiratory collapse is common in sepsis so early intubation and mechanical ventilation should be done electively rather than in response to respiratory arrest. Etomidate causes adrenal insufficiency through a direct inhibition of one of the hydroxylases necessary for the creation of cortisol thus other induction agents such as midazolam or ketamine are preferred. Ventilator management should follow the lung protective strategy initially outlined by the Acute Respiratory Distress Syndrome (ARDS) Network consortium. Tidal volumes should be targeted at 6 mL/kg of ideal body weight and respiratory acidosis is tolerated if the patient’s pH is 7.2 or greater. This ventilator strategy is termed lung protective or open lung ventilation and aims to minimize volutrauma in patients with ARDS and sepsis-­related respiratory failure. Similarly, positive end expiratory pressure (PEEP) should be increased to raise mean airway pressure and further limit stretching of the alveoli. The increase in mean airway pressure also improves oxygenation through opening of partially filled alveoli in a process termed recruitment. Inspired fractions of oxygen above 60% are associated with oxygen toxicity and further lung damage, so increased PEEP may also improve oxygenation while limiting toxicity. Lung recruitment improves as PEEP is increased but so do adverse hemodynamic effects, so ventilator management must be tailored to balance the optimization of these two systems in septic patients. Hypovolemic patients are particularly sensitive to the negative hemodynamic effects of increased PEEP, so it is advisable to allow fluid resuscitation to occur before making significant increases in PEEP. Patients with severe ARDS may benefit from several additional measures such as prone positioning, advanced ventilator modes, neuromuscular blockade and in severe cases, extra-­corporeal membrane oxygenation (ECMO). Prone positioning is labor intensive and exposes the patient to complications such as pressure ulcers, dislodgment of medical devices, and limits physical examination, but it can dramatically improve oxygenation in patients with refractory hypoxemia. Several studies have shown that prone positioning confers a mortality benefit when compared to supine ventilation in patients with life-­threatening ARDS. In extreme cases, patients may be placed into prone position even if they have an open abdomen but it is prudent to reinforce the temporary abdominal dressing in these patients. Patients failing conventional ventilator management may benefit from advanced modes such as airway pressure release ventilation, bilevel, and inverse pressure control ventilation. Patients with profound ventilator desynchrony or poor pulmonary compliance with hypoxemia should be paralyzed with neuromuscular blockers (NMBs). NMBs have conferred a mortality benefit in patients with ARDS, but only if the degree of hypoxemia is severe and life-­threatening. NMB has an added benefit of maximizing oxygen delivery to the brain and vital organs, whereas oxygen consumption by skeletal muscle is minimized; however, NMB use is associated with ICU myopathies and muscle atrophy. ECMO is emerging as an effective rescue strategy for life-­threatening ARDS and many patient conditions that were formerly believed to be a contraindication to this modality have been disproven. ECMO can accomplish the gas exchange that failing lungs cannot as well as provide cardiovascular support while allowing the lungs to rest and recover from the toxic effects of high pressures, volumes, and inspired oxygenation. 

Kidneys Perhaps no organ system is more adversely affected in sepsis than are the kidneys since nearly half of septic patients will exhibit acute kidney injury (AKI). AKI increases septic mortality by six to eight

times and leads to chronic kidney disease in survivors. Even minor AKI negatively affects survival in sepsis and can have lifelong adverse effects. Sepsis impairs renal perfusion at both the macroscopic and microscopic levels, and septic patients who are normotensive may still develop AKI. As with other organ systems, circulating toxins, oxidative products, cytokines, and other inflammatory mediators that are part of the septic milieu adversely affecting the renal system. The action of these circulating factors can be compounded by hypoperfusion of the kidney; thus, effective fluid resuscitation and management is critical in reducing sepsis-­related AKI. The ideal MAP to reduce the occurrence of AKI has not been determined. As previously noted, lactated Ringer’s is the preferred fluid because normal saline resuscitation yields higher rates of AKI. Again, the goal should be to restore an effective circulating volume without overresuscitating. Excess fluid has detrimental effects on many organ systems, especially the lungs, and in the late stage of sepsis, positive fluid balance contributes to excess mortality. For patients who require renal replacement therapy, there is no clear-­cut directive governing the type or timing of this therapy despite a multitude of clinical trials. Continuous renal replacement therapy is the practical choice for patients unable to tolerate the negative hemodynamic effects of standard hemodialysis but there appears to be no benefit to early, preemptive continuous renal replacement therapy. 

Liver, Pancreas Unlike the kidneys, sepsis-­induced liver failure is a rare complication, and its appearance is typically a perimortal event. The liver is intimately involved in the body’s response to sepsis and performs a whole host of critically important activities ranging from phagocytosis and clearance of bacteria from the bloodstream, production of countless crucial proteins, regulation of iron metabolism to modulation of the host’s inflammatory response. Patients with preexisting liver dysfunction are particularly vulnerable to sepsis that is borne out by the fact that nearly one-­third of cirrhotic patients will require hospitalization for bacterial infections. Like the liver, outright pancreatic failure is a very uncommon finding in septic patients; however, most septic patients will manifest some degree of pancreatic dysfunction. Elevations in serum amylase and lipase are common as are some disturbances in the exocrine function of the pancreas. Exocrine insufficiency can adversely affect patient’s nutritional status and can lead to malabsorption of fat-­soluble vitamins resulting in clinical deficiency syndromes, especially in long-­term ICU patients. Stools should be examined for signs of fat malabsorption and if suspected, replacement therapy with enteral pancreatic enzymes is an effective treatment. While pancreatic endocrine function is minimally affected by sepsis, dysregulation of glucose metabolism is exceedingly common. The gut has been described as the “motor” of multiple system organ failure, highlighting its importance as both a cause of sepsis as well as a target for sepsis-induced failure. The bowel lymphoid tissue is the largest aggregate of immune cells in the human body and is necessary since the gut is home to more than 100 trillion bacteria known as the microbiome. Although previously thought to be innocuous, the microbiome is being implicated in a wide range of pathologic conditions and is a very active area of investigation. As with the lung, the gut may be the site of a septic focus as well as negatively affected by sepsis from a remote location resulting in an ileus. In the absence of an ileus, the gut ought to be used; several benefits of enteral over parenteral nutrition have been described. Critically ill patients can be tube fed into the stomach and the utility of checking gastric residuals has been disproven, but clinicians must remain vigilant for feeding intolerance. It remains controversial whether transpyloric feedings are safer than gastric feeding, but massive aspiration is less likely, and gastroparesis limiting feeding is avoided with the transpyloric route. Vasopressors are a relative contraindication to enteral nutrition; however, the threshold for withholding feedings has not been established. In practical terms, patients with escalating requirements for vasopressor support or those requiring a second vasopressor or

SURGIC AL CRITIC AL C ARE

inotrope should not be enterally fed. Patients who are intolerant of enteral nutrition or in those that cannot have their needs met enterally ought to receive parenteral nutrition and prolonged periods of starvation must be avoided in critically ill patients. 

Glucose Control Like many other forms of organ dysregulation, sepsis-­ induced hyperglycemia is both a marker of disease severity as well as a mediator of further injury. Cytokines, cortisol, catecholamines, and other inflammatory agents cause alterations in the hormonal control of insulin as well as engender a state of peripheral insulin resistance. Although there have been few studies specifically looking at glucose targets in sepsis, there have been several major clinical trials aiming to identify the optimal blood glucose level in critically ill patients. After many iterations, it appears that the ideal glucose value in ICU patients lies between 140 and 180 mg/dL and the trigger for glucose management should commence at 180 mg/dL. Hypoglycemia is to be avoided because this has been shown several times to be an independent risk factor for death in critically ill patients. Patients with preexisting diabetes are at several-­fold increased risk of sepsis but surprisingly are not at increased risk for death when admitted to the ICU with sepsis, and those with poor premorbid glucose control have even lower death rates, yielding a “diabetes paradox.” This contradiction only applies to ICU patients because diabetic patients admitted to other levels of care in the hospital have worse outcomes than those without diabetes. One of the greatest controversies in the management of critically ill patients centers around the role of the adrenal glands and corticosteroids in patients with shock. After several conflicting large clinical trials, the current recommendation is that patients requiring vasopressors for blood pressure support ought to be given 200 mg of hydrocortisone in divided doses. There is no need for adrenal stimulation studies or for coadministration of pure mineralocorticoid agent such as fludrocortisone. Interestingly, pure glucocorticoid agents such as dexamethasone appear to treat relative adrenal insufficiency despite a lack of any mineralocorticoid activity; however, this drug is not used to treat adrenal insufficiency. “Stress dose” steroids do not need to be weaned and should be discontinued once the patient is adequately resuscitated and vasopressors are no longer required to maintain MAP. Steroids shorten the time that patients require vasopressors, which has practical benefits, but they do not confer a mortality benefit. Steroid therapy can worsen sepsis-­induced hyperglycemia and many patients will require exogenous insulin to keep their blood glucose in the desired range. Curiously, despite their well-­established immunosuppressive actions, stress dose steroids are not associated with increases in infectious complications. 

Immune System The complexity of the immune system and its alterations and dysregulation in response to sepsis is profoundly intricate and is why there is not a single approved medication to reduce mortality in sepsis despite the performance of well over 100 large clinical sepsis trials. Every single component of the humoral and innate immune systems is altered in response to sepsis and these changes generally follow a similar pattern. Initial immune activation is typically followed by a period of hyperinflammation, resolution to homeostasis, or immune suppression but any one of the conditions may predominate at any given time. In some ways, obtaining source control diminishes the trigger that sets off the immune-­inflammatory cascade. Complementary to this is the timely administration of appropriate antibiotics directed against the most likely pathogens. Clinical studies have repeatedly shown that the longer the interval is between the onset of sepsis to dosing of appropriate antibiotics, the worse the outcomes. A retrospective review of the Surviving Sepsis database found that in-­hospital septic mortality increased for every hour that antibiotic

1431

therapy was delayed beginning 2 hours after patients were diagnosed with sepsis. This rise in mortality increased up to a 1.5 times the odd ratio of death when antibiotic therapy was delayed by 6 hours. Much like failure to obtain prompt source control, delays in administering antibiotics to a septic patient contributes to excess mortality. 

Antibiotic Selection Similarly, choosing the wrong empiric antibiotic usually increases the mortality of sepsis by a factor of two to five times. It is imperative that clinicians select antibiotics based on the most likely bacteria for each given septic focus and consider local resistance patterns and patient comorbidities. Blood and other cultures should be obtained at the start of therapy, but antibiotics should never be withheld so that cultures can be done. It may be necessary to use more than one agent if the most likely microbes causing sepsis cannot be treated with monotherapy. The possibility that the septic response is due to infections caused by yeast or viruses should not be overlooked, particularly in patients taking immune modulating agents for inflammatory conditions or immunosuppression following transplantation. Considering the morbidity of delayed or inappropriate antibiotic selection in the treatment of sepsis, clinicians must resist the urge to use overly broad antibiotics or powerful top-­tier agents. Doing so will lead to opportunistic infections, antibiotic-­resistant infections, and further compilations. Accordingly, antibiotic therapy ought to tailor based on culture results as these become available, typically in 2 or 3 days, and the broad-­spectrum empiric antibiotics switched to the narrowest possible agent. This process is known as antibiotic deescalation and it minimizes antibiotic resistance, opportunistic infections, and waste of resources. The duration of antibiotic therapy ought to be limited based on available evidence in the literature and there is no role for long courses of broad-­spectrum “empiric” antibiotics even for the sickest patients. Patients with peritonitis who achieve source control can be treated with a 4-­day duration of antibiotics based on class one data. Similarly, ventilator-­associated pneumonia can be treated with 8 days of therapy for all bacterial infections except those caused by the Enterobacteriaceae family. Occasionally, a patient will manifest ongoing sepsis for several days despite negative culture data and no obvious source. This conundrum is particularly vexing if the patient is elderly or has serious medical comorbidities. In this situation, all antibiotics should be stopped, and new cultures obtained from blood and other likely sites of infection. An exhaustive physical examination should be performed, and computed tomography scanning is usually required to assess for occult infections in the peritoneal and/or thoracic cavities. Surgeons must be humble and should consider that the source of sepsis may be a complication of their surgical intervention. If a septic source is identified, vigorous efforts to achieve source control should be undertaken and reoperative surgery is often necessary. Although stopping antibiotics in someone with ongoing sepsis seems risky, another way of looking at this is that the patient is not responding to the antibiotics that he or she is receiving so stopping them can do no further harm.  nn SUMMARY Sepsis remains a challenging and costly problem for society, and surgeons continue to play a central role in its therapy. The diagnosis of sepsis remains difficult because of its variable presentation, so it is mandatory that clinicians have a high index of suspicion that a deteriorating patient could be septic. Overreliance on anyone clinical parameter, biomarker, or laboratory value is likely to lead to complications and higher mortality. Therapeutic interventions and management strategies need to be continually reevaluated because what seems like an effective or beneficial therapy is often found to be untrue when rigorously studied. Sepsis mortality has fallen as the principles of source control, fluid resuscitation, and timely antibiotics have become widely embraced and surgeons need to stay involved in this process if the best outcomes are to be achieved.

1432

Multiple Organ Dysfunction and Failure

Suggested Readings Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign Guidelines Committee Including the Pediatric Subgroup. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41:580–637. Holst LB, Haase N, Wetterslev J, et al. Triss Trial Group, Scandinavian Critical Care Trials Group. lower versus higher hemoglobin threshold for transfusion in septic shock. N Engl J Med. 2014;371:1381–1391. Mouncey PR, Osborn TM, Power GS, et  al. ProMISe Trial Investigators. Trial of early, goal-­directed resuscitation for septic shock. N Engl J Med. 2015;372:1301–1311.

Multiple Organ Dysfunction and Failure Mitchell Ryan Ladd, MD, PhD, and Arthur Jason Vaught, MD

C

ritically ill patients are complex and require a systematic approach to ensure excellent care. They are often managed by thinking about each system in turn so that all of their maladies are considered and appropriately treated. Unfortunately, despite a physician’s best efforts, patients may deteriorate to the point that multiple organs are dysfunctional and heading toward failure. When this occurs, the acuity and complexity of these patients escalates and considerable attention to the details of each system is paramount. After a short history of the term multiple organ failure (MOF), this chapter will briefly discuss the end-­organ dysfunction and failure often encountered in each organ system in critically ill surgical patients. Appropriate grading systems for the organ systems and patients as a whole will be discussed, as well as a few special populations that deserve mention. Finally, the importance of when to consider end of life and palliative care is included. nn HISTORY The term multiple organ failure was first coined by Arthur Baue in 1975. In the years following, numerous attempts have been made to ascribe an etiology and pathophysiology to this entity while recognizing that each organ system may suffer some dysfunction without being in frank failure. Organ dysfunction that occurs short of failure has now been termed multiorgan dysfunction syndrome (MODS). Decades of research have ensued in an effort to effectively treat and reverse MODS before it becomes MOF, and many advances have been made. Aside from correcting the underlying etiology, however, there has been no single treatment strategy that can be used to cure all patients. Moreover, salvage therapies are lacking for severe MODS that progresses to MOF.  nn DYSFUNCTION AND

SYSTEM

FAILURE BY ORGAN

Neurologic The neurologic function of critically ill patients is often abnormal and can be affected by numerous factors. Mental status change is a common early sign of MODS and shock; however, delirium, stroke, traumatic brain injury, and exposure to sedative medicines also cause altered mental status and should remain on the differential. When a patient’s mental status deteriorates, a prompt workup should ensue to determine the underlying cause and the change should not be attributed to medications alone. Effects of medications on mental status, while important to consider, should essentially be

Peake SL, Delaney A, Bailey M, et  al. ARISE Investigators, ANZICS Clinical Trials Group. Goal-­directed resuscitation for patients with early septic shock. N Engl J Med. 2014;371:1496–1506. Rivers E, Nguyen B, Havstad S, et al. Early goal-­directed therapy collaborative group. early goal-­directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–1377. Yealy DM, Kellum JA, Huang DT, et  al. ProCESS Investigators. A randomized trial of protocol-­based care for early septic shock. N Engl J Med. 2014;370:1683–1693.

a diagnosis of exclusion. The easiest and most often used assessment of global neurologic function in the initial evaluation is the Glasgow Coma Score (GCS; Table 1). Patients with a GCS of 8 or less have a mental status change significant to prevent them from protecting their airway and intubation should be considered. Further, GCS is included as part of the Sequential Organ Failure Assessment (SOFA) score (Table 2) and the Marshall MOD Score (Table 3; more on the SOFA and quick SOFA score appear in the section on scoring systems) and thus is essential in evaluating a newly critically ill patient. Pain control is also a parameter used to assess neurologic dysfunction. Patients with seemingly increased or out of proportion pain should be assessed for an underlying etiology. The nidus of the pain could be one of the many causes of MODS or MOF (i.e., abscess, sepsis, bowel perforation, and bleeding). If these underlying etiologies have been ruled out, a multimodal pain regimen should be implemented that adequately controls pain while limiting the risk of delirium. More recently, intensive care units (ICUs) have adapted pain, agitation, and delirium protocols to help better treat these conditions in a more real-­time way using bedside nursing assessment to drive management. Pain services should be used for patients requiring increasing pain medications, and neuroaxial blockade should be considered if appropriate. There is evidence that ICU patients may have brain dysfunction and depression that lingers beyond their recovery. It is well-­recognized that to prevent changes in consciousness and reduce delirium in the ICU, we should strive to maintain sleep-­wake cycles, reduce sedation as much as is possible in mechanically ventilated patients (by providing sedation vacations), avoid benzodiazepines, and be aggressive with physical therapy and daily mobilization. The ultimate central nervous system dysfunction or failure is brain death. Brain death can be brought on from any of the causes mentioned previously if they ultimately result in global ischemia of the brain and subsequently the absence of blood flow to the brain. This typically occurs as a result of increased intracranial pressure such that it decreases cerebral perfusion pressure to the point that the brain completely dies. Once this happens, the damage is irreversible. The legal declaration of brain death varies from state to state, but typically requires at least two examinations by two providers with an interval observation period. In addition, to be declared brain-­dead, metabolic causes of brain dysfunction must be excluded. These include being normothermic, having a negative toxicology screen, normal electrolytes, and normal acid-­base status. 

Cardiovascular Dysfunction of the cardiovascular system, in the form of hypotension and shock, is perhaps the most frequent cause of ICU admission. Cardiovascular abnormalities can have numerous causes and in the noncardiac surgical patient, is often a secondary insult from a different primary disease, for example sepsis. Cardiac complications from surgery do occur and cardiovascular collapse can be due to postoperative myocardial infarctions, submassive or massive pulmonary

1432

Multiple Organ Dysfunction and Failure

Suggested Readings Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign Guidelines Committee Including the Pediatric Subgroup. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41:580–637. Holst LB, Haase N, Wetterslev J, et al. Triss Trial Group, Scandinavian Critical Care Trials Group. lower versus higher hemoglobin threshold for transfusion in septic shock. N Engl J Med. 2014;371:1381–1391. Mouncey PR, Osborn TM, Power GS, et  al. ProMISe Trial Investigators. Trial of early, goal-­directed resuscitation for septic shock. N Engl J Med. 2015;372:1301–1311.

Multiple Organ Dysfunction and Failure Mitchell Ryan Ladd, MD, PhD, and Arthur Jason Vaught, MD

C

ritically ill patients are complex and require a systematic approach to ensure excellent care. They are often managed by thinking about each system in turn so that all of their maladies are considered and appropriately treated. Unfortunately, despite a physician’s best efforts, patients may deteriorate to the point that multiple organs are dysfunctional and heading toward failure. When this occurs, the acuity and complexity of these patients escalates and considerable attention to the details of each system is paramount. After a short history of the term multiple organ failure (MOF), this chapter will briefly discuss the end-­organ dysfunction and failure often encountered in each organ system in critically ill surgical patients. Appropriate grading systems for the organ systems and patients as a whole will be discussed, as well as a few special populations that deserve mention. Finally, the importance of when to consider end of life and palliative care is included. nn HISTORY The term multiple organ failure was first coined by Arthur Baue in 1975. In the years following, numerous attempts have been made to ascribe an etiology and pathophysiology to this entity while recognizing that each organ system may suffer some dysfunction without being in frank failure. Organ dysfunction that occurs short of failure has now been termed multiorgan dysfunction syndrome (MODS). Decades of research have ensued in an effort to effectively treat and reverse MODS before it becomes MOF, and many advances have been made. Aside from correcting the underlying etiology, however, there has been no single treatment strategy that can be used to cure all patients. Moreover, salvage therapies are lacking for severe MODS that progresses to MOF.  nn DYSFUNCTION AND

SYSTEM

FAILURE BY ORGAN

Neurologic The neurologic function of critically ill patients is often abnormal and can be affected by numerous factors. Mental status change is a common early sign of MODS and shock; however, delirium, stroke, traumatic brain injury, and exposure to sedative medicines also cause altered mental status and should remain on the differential. When a patient’s mental status deteriorates, a prompt workup should ensue to determine the underlying cause and the change should not be attributed to medications alone. Effects of medications on mental status, while important to consider, should essentially be

Peake SL, Delaney A, Bailey M, et  al. ARISE Investigators, ANZICS Clinical Trials Group. Goal-­directed resuscitation for patients with early septic shock. N Engl J Med. 2014;371:1496–1506. Rivers E, Nguyen B, Havstad S, et al. Early goal-­directed therapy collaborative group. early goal-­directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345:1368–1377. Yealy DM, Kellum JA, Huang DT, et  al. ProCESS Investigators. A randomized trial of protocol-­based care for early septic shock. N Engl J Med. 2014;370:1683–1693.

a diagnosis of exclusion. The easiest and most often used assessment of global neurologic function in the initial evaluation is the Glasgow Coma Score (GCS; Table 1). Patients with a GCS of 8 or less have a mental status change significant to prevent them from protecting their airway and intubation should be considered. Further, GCS is included as part of the Sequential Organ Failure Assessment (SOFA) score (Table 2) and the Marshall MOD Score (Table 3; more on the SOFA and quick SOFA score appear in the section on scoring systems) and thus is essential in evaluating a newly critically ill patient. Pain control is also a parameter used to assess neurologic dysfunction. Patients with seemingly increased or out of proportion pain should be assessed for an underlying etiology. The nidus of the pain could be one of the many causes of MODS or MOF (i.e., abscess, sepsis, bowel perforation, and bleeding). If these underlying etiologies have been ruled out, a multimodal pain regimen should be implemented that adequately controls pain while limiting the risk of delirium. More recently, intensive care units (ICUs) have adapted pain, agitation, and delirium protocols to help better treat these conditions in a more real-­time way using bedside nursing assessment to drive management. Pain services should be used for patients requiring increasing pain medications, and neuroaxial blockade should be considered if appropriate. There is evidence that ICU patients may have brain dysfunction and depression that lingers beyond their recovery. It is well-­recognized that to prevent changes in consciousness and reduce delirium in the ICU, we should strive to maintain sleep-­wake cycles, reduce sedation as much as is possible in mechanically ventilated patients (by providing sedation vacations), avoid benzodiazepines, and be aggressive with physical therapy and daily mobilization. The ultimate central nervous system dysfunction or failure is brain death. Brain death can be brought on from any of the causes mentioned previously if they ultimately result in global ischemia of the brain and subsequently the absence of blood flow to the brain. This typically occurs as a result of increased intracranial pressure such that it decreases cerebral perfusion pressure to the point that the brain completely dies. Once this happens, the damage is irreversible. The legal declaration of brain death varies from state to state, but typically requires at least two examinations by two providers with an interval observation period. In addition, to be declared brain-­dead, metabolic causes of brain dysfunction must be excluded. These include being normothermic, having a negative toxicology screen, normal electrolytes, and normal acid-­base status. 

Cardiovascular Dysfunction of the cardiovascular system, in the form of hypotension and shock, is perhaps the most frequent cause of ICU admission. Cardiovascular abnormalities can have numerous causes and in the noncardiac surgical patient, is often a secondary insult from a different primary disease, for example sepsis. Cardiac complications from surgery do occur and cardiovascular collapse can be due to postoperative myocardial infarctions, submassive or massive pulmonary

1433

SURGIC AL CRITIC AL C ARE

embolism, or atrial fibrillation. Distinguishing the cause of cardiovascular collapse is important because the physiology surrounding cardiovascular dysfunction will dictate its treatment. Bedside and formal echocardiography may be helpful in determining the cause of cardiovascular dysfunction if the diagnosis is unclear. Echocardiography may also be helpful in tracking the adequacy of resuscitation both by evaluating the ventricular function, assessing valvular abnormalities, atrial filling, and fullness of the inferior vena cava. In the setting of distributive shock (e.g., sepsis, anaphylaxis) there is vasodilation, capillary leak, and resulting hypotension. In hypovolemic/hemorrhagic shock, there is vasoconstriction in an attempt to preserve blood pressure; however, because of the low blood volume, hypotension persists. Cardiogenic shock is due to injury to the heart itself and results in pump dysfunction or failure, thus there is poor blood distribution to end organs resulting in vasoconstriction, cold extremities, and ultimately hypotension resulting from inadequate pumping. Obstructive shock causes a central cardiac outflow obstruction that results in hypotension. Neurogenic shock results in venodilation, reduced preload, and frequently bradycardia all of which work together to cause hypotension. Different types of shock are treated with different algorithms; it is important to recognize the type of shock a patient has and treat accordingly. Cardiovascular dysfunction sets off a cascade of injury to other organ systems because of inadequate perfusion. It is therefore imperative to support the cardiovascular system, regardless of the cause of dysfunction, to prevent further injury to end organs. This is often done with a combination of fluid resuscitation and vasopressors depending on the type of shock present. In general, these interventions are targeted to maintain a mean arterial pressure (MAP) of 65 and lactate less than 2 mmol/L to ensure adequate end-­organ perfusion. Details of management for each of type of shock can be

Table 1  Glasgow Coma Scale Component

Score

EYE OPENING  Spontaneously

4

  To voice

3

  To pain

2

  No response

1

VERBAL  Oriented

5

 Disoriented

4

 Inappropriate

3

  Incomprehensible sounds

2

  No response

1

MOTOR   Following commands

6

  Localizing to pain

5

  Withdrawing to pain

4

  Abnormal flexion

3

  Abnormal extension

2

  No response

1

  

TABLE 2  Sequential Organ Failure Assessment (SOFA) Score Organ System

0

1

2

3

4

Respiratory: PaO2/FiO2

>400

90 mmHg after 20 weeks

Proteinuria Protein >300 mg/24-­hr urine collection Protein/creatinine ratio > 0.3 Dipstick reading of 1+ (only if other methods not available)

In the absence of proteinuria, new-­onset hypertension with new onset of any of the following: Platelet count 1.1 mg/dL or doubling of serum creatinine concentrations Elevated blood concentration of liver transaminases twice normal Pulmonary edema Cerebral or visual symptoms Eclampsia

Presence of new-­onset grand mal seizure in a woman with preeclampsia

HELLP syndrome

Platelets 8 g/dL) and cardiac arrest (levels >10 g/dL), calcium chloride 1 g intravenously or calcium gluconate 3 g intravenously should be administered. Although HELLP syndrome is highly morbid, it is by definition limited to pregnancy and should resolve within 72 hours after delivery. In the event maternal improvement does not occur, ICU staff should consider alternative diseases. Differential diagnosis should include atypical hemolytic uremic syndrome, thrombotic thrombocytopenia purpura, and other autoimmune disease such as systemic lupus erythematosus. These diseases may be quiescent throughout life and may be newly triggered by pregnancy. 

usually are above 500 IU/L and can often be above 1000 IU/L. There are also other signs of liver failure including unexplained hypoglycemia, acidosis, an elevated direct bilirubin and altered mental status from hyperammonemia. Upon imaging, the liver shows steatosis. Commonly, women with ALFP have coagulopathy and DIC and pose a high transfusion risk. Women with AFLP should be at centers with an efficient blood bank at time for delivery for risk of massive transfusion. Although patients with AFLP can progress to complete liver failure with renal injury, the disease is usually limited to pregnancy and after delivery patients tend to have full liver recovery within 1 week. Supportive measures such as transfusion for bleeding and DIC, mechanical ventilation for respiratory failure, extensive glucose supplementation, and ammonia clearing agents are warranted during the recovery. For patients who have liver recovery, neither transplant nor biopsy is warranted; however, if recovery is not seen, these options should be considered to rule out other forms of failure. Because of the rarity of the disease, other forms of liver failure should be considered in reproductive age women: acetaminophen toxicity, drug induced from other medications, infections, newly diagnosed HIV, and hepatitis to name a few. 

Acute Fatty Liver of Pregnancy

nn PALLIATIVE

Acute fatty liver of pregnancy (AFLP) is an incredibly rare, morbid disorder characterized by quickly progressive liver failure. Although AFLP affects 1 in 10,000 pregnancies, maternal mortality from this process can be as high as 5% to 10% in a tertiary care center. Women with AFLP have liver abnormalities but usually to its most severe form. In AFLP, both aspartate aminotransferase and alanine aminotransferase

In a culture of medicine that focuses on curing disease, palliative care remains an underused service in many institutions. Patients with intractable or incurable preexisting conditions are at most risk of diseases causing MODS-­like sepsis. The most common preexisting condition associated with palliative care is metastatic cancer; however, critical care providers should be aware the scope of palliative care can

Obese Patients in the Critical Care Unit Obesity continues to be a huge medical problem in the United States. Obesity is associated with cardiovascular disease, coronary artery disease, and diabetes as well as increased health care cost. Although obesity can cause increased complications, superobese patients (body mass index >50) may pose different challenges to critical care practitioners. Patients with superobesity and obesity are more likely to have longer length of stays in the ICU when compared with nonobese patients. They are also more likely to need noninvasive ventilation and succumb to pulmonary disease. Postoperatively, superobesity carries risk of higher perioperative morbidity including infections, which include sepsis, septic shock, and urinary tract infections. Patients with superobesity are also more likely to have prolonged ventilatory days, unplanned or emergent intubations, renal insufficiency, and pulmonary embolism. Dosing of weight-­based medications such as antibiotics and anticoagulation also pose a problem and put these patients at risk for critical illness secondary to subtherapeutic levels. 

CARE AND ETHICS

SURGIC AL CRITIC AL C ARE

be used in severe heart failure, pulmonary hypertension, as well as end-­stage pulmonary disease. Palliative care consultation is also not just for preexisting conditions. In cases with intractable septic shock, inoperable bleeding, or organ failure without the ability to transplant, palliative care should be considered. A multidisciplinary team of the primary surgeon, intensivist, the patient or family, as well as a provider well versed in palliative medicine should discuss and determine what is “success” for these patients. Palliative care should be considered and integrated early in a diagnosis of serious illness to help facilitate treatment and patient goals. Despite being advocated by the World Health Organization, many patients do not have palliative care initiated until the final days of life. This is likely secondary to the negative stigma that palliative care focuses on death, dying, and hopelessness. Palliative medicine is rooted in symptom control, pain management, and improved quality of life. In patients undergoing end-­of-­life care, it is important for clinicians to discern what is futile in their treatment. Futile interventions are those that simply cannot accomplish the intended physiologic goal, and clinicians should not provide futile interventions. In clinical settings where there is concern for futile treatment, ethics committees should be consulted if available.  nn SUMMARY This chapter attempted to provide an overview of multiple organ dysfunction that is often seen in the ICU and to describe each organ in turn. MOF is the final common pathway for many ICU patients and especially those that develop severe MODS that is not expeditiously corrected. We have described commonly used grading systems to help discern the severity of the organ dysfunction. These grading systems can be helpful in gauging the severity of and in some cases the predicted mortality for patients with MOD. We have also described select special patient populations that are increasingly encountered in the ICU. Finally, we have emphasized the importance of involving palliative care early in the course of patients that experience MODS and MOF and in necessary cases, ethics consultants when the ICU practitioner has reached the point that they deem further curative-­ intent care to be futile.

Suggested Readings Acosta LF, Garcia CR, Dugan A, Marti F, Davenport D, Gedaly R. Impact of super obesity on perioperative outcomes after hepatectomy: the weight of the risk. Surgery. 2017;162(5):1026. Acute Kidney Injury Network. http://akinet.org/. American College of Obstetricians and Gynecologists. Task Force on hypertension in pregnancy. hypertension in pregnancy. report of the American College of Obstetricians and Gynecologists’ Task Force on hypertension in pregnancy. Obstet Gynecol. 2013;122(5):1122. Bunchorntavakul C, Reddy KR. Acute liver failure. Clin Liver Dis. 2017;21(4):769. Cardoso FS, Marcelino P, Bagulho L, Karvellas CJ. Acute liver failure: an up-­ to-­date approach. J Crit Care. 2017;39(25).

Antibiotics for Critically Ill Patients Kelly A. Boyle, MD, and David J. Milia, MD

I

nfection remains a significant source of morbidity and mortality in the intensive care unit (ICU). As in polytrauma, acute myocardial infarction, and stroke, early identification and appropriate management of sepsis reduces mortality and improves outcomes. Although initiatives such as the Surviving Sepsis Campaign have been shown

1439

Creanga AA, Syverson C, Seed K, Callaghan WM. Pregnancy-­related mortality in the United States, 2011-­2013. Obstet Gynecol. 2017;130(2):366. Delano MJ, Ward PA. The immune system’s role in sepsis progression, resolution, and long-­term outcome. Immunol. Rev. 2016;274:330. Ferguson ND, Fan E, Camporota L, Antonelli M, Anzueto A, Beale R, et al. The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material. Intensive Care Med. 2012;38(10):1573. George JN, Nester CM, McIntosh JJ. Syndromes of thrombotic microangiopathy associated with pregnancy. Hematology. 2015;2015(1):644. Gheorghiţă V, Barbu AE, Gheorghiu ML, Căruntu FA. Endocrine dysfunction in sepsis: A beneficial or deleterious host response? Germs. 2015;5(1):17. Goren O, Matot I. Perioperative acute kidney injury. Br J Anaesthsia. 2015;115:ii3. Guerin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. NEJM. 2013;68(23). Huerta A, Arjona E, Portoles J, Lopez-­Sanchez P, Rabasco C, Espinosa M, et  al. A retrospective study of pregnancy-­associated atypical hemolytic uremic syndrome. Kidney Int. 2018;93(2):450. Intensive versus conventional glucose control in critically ill patients. NEJM. 2009;360(13):1283. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818. Kon AA, Shepard EK, Sederstrom NO, Swoboda SM, Marshall MF, Birriel B, et al. Defining futile and potentially inappropriate interventions: a policy statement from the Society of Critical Care Medicine ethics committee. Crit. Care Med. 2016;44(9):1769. Kumar SI, Doo K, Sottilo-­Brammeier J Lane C, Liebler JM. Super obesity in the medical intensive care unit. J. Intensive Care Med. 2018: 885066618761363. Lauzier F, Cook D, Griffith L, Upton J, Crowther M. Fresh frozen plasma transfusion in critically ill patients. Crit Care Med. 2007;35(7):1655. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med. 1995;23(10):1638. Napolitano LM. Anemia and red blood cell transfusion advances in critical care. Crit Care Clin. 2017;33(2):345. NHLBI ARDS Network. Tools. http://www.ardsnet.org/tools.shtml. Oxford CM, Ludmir J. Trauma in pregnancy. Clin Obstet Gynecol. 2009;52(4):611. Papazian L, Forel JM, et al. Neuromuscular blockers in early acute respiratory distress syndrome. NEJM. 2010;63(12). Pollock W, Rose L, Dennis C-­L. Pregnant and postpartum admissions to the intensive care unit: a systematic review. Intensive Care Med. 2010;36(9):1465. Reintam Blaser A, Jakob SM, Starkopf J. Gastrointestinal failure in the ICU. Curr Opin Crit Care. 2016;22(2):128. Saad AF, Roman J, Wyble A, Pacheco LD. Pregnancy-­associated atypical hemolytic-­uremic syndrome. AJP Rep. 2016;6(1):e125. Thakar CV. Perioperative acute kidney injury. Adv Chronic Kidney Dis. 2013;20(1):67. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, et al. The SOFA (Sepsis-­related Organ Failure Assessment) score to describe organ dysfunction/failure. on behalf of the working group on sepsis-­related problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707. Zarychanski R, Houston DS. Assessing thrombocytopenia in the intensive care unit: the past, present, and future. Hematol Am Soc Hematol Educ Progr. 2017;1:660.

to improve outcomes, the most fundamental tenet in the management of the septic patient is source control. In surgical patients, procedural interventions are the mainstay for achieving source control in the appropriate clinical setting. To accompany this, proper antibiotic choice and duration can augment surgical control through local and systemic effects (Tables 1 and 2). nn APPROACH TO ANTIBIOTIC

IN THE CRITICALLY ILL

USE

Sepsis in the surgical patient continues to be a challenge for surgical intensivists. Early identification of septic patients, timely implementation of therapies, and identifying the source of infection are

SURGIC AL CRITIC AL C ARE

be used in severe heart failure, pulmonary hypertension, as well as end-­stage pulmonary disease. Palliative care consultation is also not just for preexisting conditions. In cases with intractable septic shock, inoperable bleeding, or organ failure without the ability to transplant, palliative care should be considered. A multidisciplinary team of the primary surgeon, intensivist, the patient or family, as well as a provider well versed in palliative medicine should discuss and determine what is “success” for these patients. Palliative care should be considered and integrated early in a diagnosis of serious illness to help facilitate treatment and patient goals. Despite being advocated by the World Health Organization, many patients do not have palliative care initiated until the final days of life. This is likely secondary to the negative stigma that palliative care focuses on death, dying, and hopelessness. Palliative medicine is rooted in symptom control, pain management, and improved quality of life. In patients undergoing end-­of-­life care, it is important for clinicians to discern what is futile in their treatment. Futile interventions are those that simply cannot accomplish the intended physiologic goal, and clinicians should not provide futile interventions. In clinical settings where there is concern for futile treatment, ethics committees should be consulted if available.  nn SUMMARY This chapter attempted to provide an overview of multiple organ dysfunction that is often seen in the ICU and to describe each organ in turn. MOF is the final common pathway for many ICU patients and especially those that develop severe MODS that is not expeditiously corrected. We have described commonly used grading systems to help discern the severity of the organ dysfunction. These grading systems can be helpful in gauging the severity of and in some cases the predicted mortality for patients with MOD. We have also described select special patient populations that are increasingly encountered in the ICU. Finally, we have emphasized the importance of involving palliative care early in the course of patients that experience MODS and MOF and in necessary cases, ethics consultants when the ICU practitioner has reached the point that they deem further curative-­ intent care to be futile.

Suggested Readings Acosta LF, Garcia CR, Dugan A, Marti F, Davenport D, Gedaly R. Impact of super obesity on perioperative outcomes after hepatectomy: the weight of the risk. Surgery. 2017;162(5):1026. Acute Kidney Injury Network. http://akinet.org/. American College of Obstetricians and Gynecologists. Task Force on hypertension in pregnancy. hypertension in pregnancy. report of the American College of Obstetricians and Gynecologists’ Task Force on hypertension in pregnancy. Obstet Gynecol. 2013;122(5):1122. Bunchorntavakul C, Reddy KR. Acute liver failure. Clin Liver Dis. 2017;21(4):769. Cardoso FS, Marcelino P, Bagulho L, Karvellas CJ. Acute liver failure: an up-­ to-­date approach. J Crit Care. 2017;39(25).

Antibiotics for Critically Ill Patients Kelly A. Boyle, MD, and David J. Milia, MD

I

nfection remains a significant source of morbidity and mortality in the intensive care unit (ICU). As in polytrauma, acute myocardial infarction, and stroke, early identification and appropriate management of sepsis reduces mortality and improves outcomes. Although initiatives such as the Surviving Sepsis Campaign have been shown

1439

Creanga AA, Syverson C, Seed K, Callaghan WM. Pregnancy-­related mortality in the United States, 2011-­2013. Obstet Gynecol. 2017;130(2):366. Delano MJ, Ward PA. The immune system’s role in sepsis progression, resolution, and long-­term outcome. Immunol. Rev. 2016;274:330. Ferguson ND, Fan E, Camporota L, Antonelli M, Anzueto A, Beale R, et al. The Berlin definition of ARDS: an expanded rationale, justification, and supplementary material. Intensive Care Med. 2012;38(10):1573. George JN, Nester CM, McIntosh JJ. Syndromes of thrombotic microangiopathy associated with pregnancy. Hematology. 2015;2015(1):644. Gheorghiţă V, Barbu AE, Gheorghiu ML, Căruntu FA. Endocrine dysfunction in sepsis: A beneficial or deleterious host response? Germs. 2015;5(1):17. Goren O, Matot I. Perioperative acute kidney injury. Br J Anaesthsia. 2015;115:ii3. Guerin C, Reignier J, Richard JC, et al. Prone positioning in severe acute respiratory distress syndrome. NEJM. 2013;68(23). Huerta A, Arjona E, Portoles J, Lopez-­Sanchez P, Rabasco C, Espinosa M, et  al. A retrospective study of pregnancy-­associated atypical hemolytic uremic syndrome. Kidney Int. 2018;93(2):450. Intensive versus conventional glucose control in critically ill patients. NEJM. 2009;360(13):1283. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818. Kon AA, Shepard EK, Sederstrom NO, Swoboda SM, Marshall MF, Birriel B, et al. Defining futile and potentially inappropriate interventions: a policy statement from the Society of Critical Care Medicine ethics committee. Crit. Care Med. 2016;44(9):1769. Kumar SI, Doo K, Sottilo-­Brammeier J Lane C, Liebler JM. Super obesity in the medical intensive care unit. J. Intensive Care Med. 2018: 885066618761363. Lauzier F, Cook D, Griffith L, Upton J, Crowther M. Fresh frozen plasma transfusion in critically ill patients. Crit Care Med. 2007;35(7):1655. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med. 1995;23(10):1638. Napolitano LM. Anemia and red blood cell transfusion advances in critical care. Crit Care Clin. 2017;33(2):345. NHLBI ARDS Network. Tools. http://www.ardsnet.org/tools.shtml. Oxford CM, Ludmir J. Trauma in pregnancy. Clin Obstet Gynecol. 2009;52(4):611. Papazian L, Forel JM, et al. Neuromuscular blockers in early acute respiratory distress syndrome. NEJM. 2010;63(12). Pollock W, Rose L, Dennis C-­L. Pregnant and postpartum admissions to the intensive care unit: a systematic review. Intensive Care Med. 2010;36(9):1465. Reintam Blaser A, Jakob SM, Starkopf J. Gastrointestinal failure in the ICU. Curr Opin Crit Care. 2016;22(2):128. Saad AF, Roman J, Wyble A, Pacheco LD. Pregnancy-­associated atypical hemolytic-­uremic syndrome. AJP Rep. 2016;6(1):e125. Thakar CV. Perioperative acute kidney injury. Adv Chronic Kidney Dis. 2013;20(1):67. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, et al. The SOFA (Sepsis-­related Organ Failure Assessment) score to describe organ dysfunction/failure. on behalf of the working group on sepsis-­related problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707. Zarychanski R, Houston DS. Assessing thrombocytopenia in the intensive care unit: the past, present, and future. Hematol Am Soc Hematol Educ Progr. 2017;1:660.

to improve outcomes, the most fundamental tenet in the management of the septic patient is source control. In surgical patients, procedural interventions are the mainstay for achieving source control in the appropriate clinical setting. To accompany this, proper antibiotic choice and duration can augment surgical control through local and systemic effects (Tables 1 and 2). nn APPROACH TO ANTIBIOTIC

IN THE CRITICALLY ILL

USE

Sepsis in the surgical patient continues to be a challenge for surgical intensivists. Early identification of septic patients, timely implementation of therapies, and identifying the source of infection are

1440

Antibiotics for Critically Ill Patients

TABLE 1 Antibiotic Classes and Agents Clinically Important in the Critically Ill: β-­Lactam Antibiotics β-­Lactam antibiotics

Attach to penicillin-­binding proteins in the cell membrane, interfering with peptide chain cross-­linkages in the bacterial cell wall peptidoglycan, leading to cell lysis

Groups

Examples

Comments

PENICILLINS Natural penicillins

Penicillin G

Very effective against Streptococcus spp., nonresistant Enterococcus faecalis, and ­Clostridium perfringens

Penicillinase-­resistant penicillins

Oxacillin Nafcillin

Very effective against MSSA and Streptococcus spp.

Aminopenicillins

Ampicillin Ampicillin-­sulbactam

Active against Streptococcus spp., MSSA, E. faecalis; variable activity against VRE Adding BLI extends activity against gram-­negative bacilli and Bacteroides fragilis Pseudomonas is resistant Escherichia coli resistance is rising; coverage is variable and unreliable

Ureidopenicillins

Piperacillin-­tazobactam

Broad-­spectrum activity including against Pseudomonas and anaerobes

CEPHALOSPORINS First generation

Cefazolin

• Active against MSSA, Streptococcus spp., E. coli, Klebsiella, Proteus mirabilis

Second generation

Cefuroxime

• Increased aerobic gram-­negative activity, including against Haemophilus influenzae

Second generation (cephamycins)

Cefoxitin Cefotetan

• Extended activity against B. fragilis; resistance increasing • Methylthiotetrazole side chain inhibits vitamin K activation, resulting in rise in international normalized ratio

Third generation

Cefotaxime Ceftizoxime Ceftriaxone

• Increased spectrum of activity against aerobic gram-­negative bacilli • Vulnerable to inactivation by extended-­spectrum β-­lactamases and AmpC cephalosporinases

Third-­generation ­antipseudomonal

Ceftazidime  Ceftazidime-­avibactam Ceftolozane-­tazobactam

• Active against Pseudomonas, less active against Staphylococcus • Adding BLI extends activity against resistant aerobic gram-­negative bacilli

Fourth generation

Cefepime

• Very effective against aerobic gram-­negative bacilli, including Pseudomonas • Able to avoid destruction by β-­lactamases by rapid penetration through the cell wall • Maintains activity against Staphylococcus and Streptococcus

Fifth generation

Ceftaroline

• Broad-­spectrum activity, including against MRSA and aerobic gram-­negative bacilli • Inactive against Pseudomonas

OTHER β-­LACTAMS Carbapenems

Ertapenem

• Broad-­spectrum activity against gram-­positive cocci, gram-­negative bacilli, and anaerobes; readily penetrate cell membranes of gram-­negative bacilli, have high affinity for penicillin-­binding proteins, and are resistant to hydrolysis by β-­lactamases • Ertapenem is not active against Pseudomonas or Enterococcus

AP carbapenems

Imipenem-­cilastatin Doripenem Meropenem

• Seizure risk is greater than with other β-­lactam antibiotics • Active against Pseudomonas • Doripenem is indicated for complicated intraabdominal infections and urinary tract infections but not for pneumonia

Monobactam

Aztreonam

• Active against aerobic gram-­negative bacilli, including Pseudomonas

BLI, β-­lactamase inhibitor; MRSA, methicillin-­resistant Staphylococcus aureus; MSSA, methicillin-­susceptible Staphylococcus aureus; VRE, vancomycin-­resistant Enterococcus.

essential to sepsis management. The practice of indiscriminate pan-­ culturing and initiation of broad-­spectrum empiric antibiotics on all patients is falling out of favor. Instead, surgical intensivists should be taking a good history and performing a thorough physical examination to guide necessary laboratory tests, imaging studies, or interventions. Once potential sources are identified, directed culturing allows for tailored therapy. Depending on the site of infection and the clinical status of the patient, empiric antibiotics may be beneficial while cultures are pending or other interventions are being arranged. In

addition to treating the infections that prompted admission to the ICU, the surgical intensivist will have to juggle the many complications that can arise in critically ill patients. Many of these complications are healthcare-­ associated infections related to indwelling devices, including ventilator-­ associated pneumonia, central line– associated bloodstream infection, catheter-­associated urinary tract infection, and surgical site infection. It is imperative that every practitioner faced with the decision to start antibiotics approaches this responsibility with good clinical stewardship. This means starting

1441

SURGIC AL CRITIC AL C ARE

TABLE 2 Antibiotic Classes and Agents Clinically Important in the Critically Ill: Non–β-­Lactam Antibiotics Groups

Examples

Comments

Glyco-­lipopeptides

Vancomycin Telavancin Oritavancin Dalbavancin

• Disrupts bacterial cell wall synthesis by impairing peptidoglycan synthesis • Active against gram-­positive cocci, including MRSA • Daptomycin and oritavancin active against VRE • Oral vancomycin effective against Clostridium difficile • Oritavancin and dalbavancin have very long half-­lives, allowing single dosing (oritavancin) or weekly × 2 dosing (dalbavancin) • Daptomycin is inactivated by surfactant, not used for pneumonia; also associated with myopathy; statins should be stopped; check CPK weekly

Polymyxins

Polymyxin B Colistin (polymyxin E)

• Cationic polypeptides that act as detergents to disrupt the bacterial cell membrane • Active against Enterobacteriaceae, Acinetobacter, Pseudomonas, and Stenotrophomonas • Cause significant renal toxicity

Aminoglycosides

Gentamicin Tobramycin Amikacin

• Inhibit bacterial protein synthesis by binding irreversibly to the 30S ribosomal subunit • Active against gram-­negative bacilli, including Pseudomonas (gentamicin less so than tobramycin or amikacin) • Synergistic activity with penicillins and vancomycin against Enterococcus and Staphylococcus • Exhibit concentration-­dependent activity and a significant postantibiotic effect • Significant incidence of renal toxicity and ototoxicity (auditory and vestibular) • Once-­daily dosing improves effectiveness and reduces toxicity • Limiting usage to 5 days reduces toxicity

Tetracyclines

Doxycycline Minocycline

• Inhibit bacterial protein synthesis by binding to the 30S ribosomal subunit • Active against Staphylococcus, including MRSA; variable activity against VRE and Streptococcus

Glycylcycline

Tigecycline

• Broad-­spectrum agent active against gram-­positive cocci (MRSA and VRE), gram-­ negative bacilli, and anaerobes • Inactive against Pseudomonas • US Food and Drug Administration black box warning, should not be used as first-­line therapy

Macrolides, lincosamides, chloramphenicol

Erythromycin  Clarithromycin Azithromycin Fidaxomicin Clindamycin Chloramphenicol

• Inhibit bacterial protein synthesis by reversibly binding to 50S ribosomal subunit at overlapping sites; use of 2 or more of these classes results in antagonism • Variable activity against MSSA and Streptococcus spp.; active against atypical bacterial • Fidaxomicin only active against C difficile • Azithromycin and clarithromycin active against Haemophilus influenzae • Active against Staphylococcus and Streptococcus, 66% to 75% of Bacteroides fragilis • Potentially additive in clostridial and streptococcal gangrene for its ability to inhibit toxin production • Active against Staphylococcus (MRSA), Streptococcus, E. coli, Klebsiella, H. influenzae, anaerobes • Variable activity against VRE • Limited use because of bone marrow suppression and rare aplastic anemia

Streptogramins

Quinupristin-­dalfopristin • Inhibits bacterial protein synthesis by binding to 50S ribosomal subunit • Active against gram-­positive cocci, including MRSA and VRE • Inactive against Enterococcus faecalis • Causes severe arthralgia, myalgia, and phlebitis; infuse through central line

Oxazolidinones

Linezolid Tedizolid

• Inhibit bacterial protein synthesis by binding to the 23S ribosomal RNA of the 50S subunit • Active against gram-­positive cocci including MRSA and VRE, resistance to Enterococcus is rising • High bioavailability: oral is as effective as intravenous • Bone marrow suppression with thrombocytopenia may occur after 14 days of treatment with linezolid, reversible after cessation • Tedizolid indicated for acute bacterial skin and skin structure infections for 6 days • Weak monoamine oxidase inhibitors, 1%–3% incidence of serotonin syndrome when given with monoamine oxidase inhibitors or selective serotonin reuptake inhibitors • Long-­term use associated with reversible optic neuropathy and irreversible peripheral neuropathy Continued

1442

Antibiotics for Critically Ill Patients

TABLE 2 Antibiotic Classes and Agents Clinically Important in the Critically Ill: Non–β-­Lactam Antibiotics—cont’d Groups

Examples

Comments

Fluoroquinolones

Ciprofloxacin Levofloxacin Moxifloxacin

•  Inhibit DNA gyrase and topoisomerase IV • Active against broad spectrum of nonresistant gram-­positive cocci and gram-­negative bacilli, including Pseudomonas (not moxifloxacin) • Moxifloxacin active against 85% of B. fragilis, but does not enter urine and cannot treat UTIs • Concentration-­dependent antibacterial effect • High bioavailability: oral is as effective as IV • Not approved for children 50% platelet fall and platelet nadir ≥20

Platelet count fall by 30%–50% or nadir 10–19

Platelet count fall by