Textbook of Gastrointestinal

Radiology

Textbook of Gastrointestinal

RadiologyFourth Edition

Richard M. Gore, MDChief of Gastrointestinal Radiology

NorthShore University HealthSystemEvanston, Illinois

Professor of RadiologyPritzker School of Medicine at the University of Chicago

Chicago, Illinois

Marc S. Levine, MDChief of Gastrointestinal Radiology

Hospital of the University of PennsylvaniaProfessor of Radiology and Advisory Dean

Perelman School of Medicine at the University of PennsylvaniaPhiladelphia, Pennsylvania

1600 John F. Kennedy Blvd.Ste 1800Philadelphia, PA 19103-2899

TEXTBOOK OF GASTROINTESTINAL RADIOLOGY, 4TH EDITION ISBN: 978-1-4557-5117-4Volume 1 Part No. 9996090353Volume 2 Part No. 9996090299

Copyright © 2015, 2008, 2000, 1994 by Saunders, an imprint of Elsevier Inc.

All rights reserved. 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).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability 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.

Library of Congress Cataloging-in-Publication Data

Textbook of gastrointestinal radiology / [edited by] Richard M. Gore, Marc S. Levine. — Fourth edition. p. ; cm. Gastrointestinal radiology Includes bibliographical references and index. ISBN 978-1-4557-5117-4 (hardcover, 2 v. set : alk. paper) I. Gore, Richard M., editor. II. Levine, Marc S., editor. III. Title: Gastrointestinal radiology. [DNLM: 1. Gastrointestinal Diseases—diagnosis. 2. Gastrointestinal Diseases—radiography. 3. Digestive System—pathology. WI 141] RC804.R6 616.3′407572—dc23

2014034661

Content Strategist: Helene CaprariContent Development Specialist: Joanie MilnesPublishing Services Manager: Patricia TannianSenior Project Manager: Claire KramerDesign Direction: XiaoPei Chen

Printed in the United States

Last digit is the print number: 9 8 7 6 5 4 3 2 1

To Margaret and our children,

Diana, Elizabeth, and George

RICHARD M. GORE

To Deborah, my beautiful and amazing wife

and better half for 37 years and counting …

MARC S. LEVINE

vi

Jalil Afnan, MD, MRCSRadiologistDepartment of RadiologyLahey Clinic Hospital and Medical CenterBurlington, MassachusettsAssistant Professor of RadiologyTufts UniversitySchool of MedicineBoston, Massachusetts

Jeffrey A. Alexander, MDAssociate Professor of MedicineMayo Clinic School of MedicineRochester, Minnesota

Lauren F. Alexander, MDAssistant Professor of Abdominal ImagingDepartment of Radiology and Imaging SciencesEmory UniversityAtlanta, Georgia

Surabhi Bajpai, MBBS, DMRDResearch FellowDepartment of RadiologyDivision of Abdominal Imaging and InterventionMassachusetts General HospitalBoston, Massachusetts

Mark E. Baker, MDProfessor of RadiologyCleveland Clinic Lerner College of MedicineCase Western Reserve UniversityImaging InstituteCleveland ClinicStaff RadiologistImaging Institute, Digestive Disease Institute, Cancer InstituteCleveland ClinicCleveland, Ohio

Stephen R. Baker, MD, MPHILProfessor and ChairmanDepartment of RadiologyRutgers New Jersey Medical School ChiefDepartment of RadiologyThe University HospitalNewark, New Jersey

Aparna Balachandran, MDAssociate ProfessorDiagnostic ImagingThe University of Texas MD Anderson Cancer CenterHouston, Texas

CONTRIBUTORS

Dennis M. Balfe, MDProfessor of RadiologyWashington University School of MedicineSt. Louis, Missouri

Emil J. Balthazar, MDProfessor EmeritusDepartment of RadiologyNew York University School of MedicineAttending ConsultantDepartment of RadiologyBellevue HospitalNew York, New York

Stuart A. Barnard, MA, MB, BS, MRCS, FRCRRadiologistDepartment of RadiologyMiddlemore HospitalCounties Manukau HealthAuckland, New Zealand

Ahmed Ba-Ssalamah, MDMedical University of ViennaDepartment of Biomedical Imaging and Image-Guided

TherapyVienna, Austria

Genevieve L. Bennett, MDAssistant Professor of RadiologyDepartment of RadiologyDivision of Abdominal ImagingNew York University School of MedicineAssistant Professor of RadiologyDepartment of RadiologyDivision of Abdominal ImagingNew York University Langone Medical CenterNew York, New York

Senta Berggruen, MDDepartment of RadiologyNorthwestern Memorial HospitalChicago, Illinois

Jonathan W. Berlin, MDClinical Professor of RadiologyDepartment of Diagnostic RadiologyNorthShore University HealthSystemEvanston, Illinois

George S. Bissett III, MDProfessor of Radiology and PediatricsVice-ChairmanDepartment of RadiologyDuke University School of MedicineDurham, North Carolina

F

Contributors vii

Roi M. Bittane, MDRadiology ResidentDepartment of RadiologyWinthrop University HospitalMineola, New York

Michael A. Blake, MB, MRCPI, FRCR, FFR(RCSI)Associate Professor of RadiologyHarvard Medical SchoolFellowship DirectorDivision of Abdominal ImagingMassachusetts General HospitalBoston, Massachusetts

Peyman Borghei, MDClinical Assistant Professor of RadiologyUniversity of California at IrvineChief of Interventional RadiologyVA HospitalLong Beach, California

Kevin P. Boyd, DOAssistant Professor of RadiologyChildren’s Hospital of WisconsinMedical College of WisconsinMilwaukee, Wisconsin

Warren M. Brandwein, MDFellowBody and Musculoskeletal Imaging SectionDepartment of RadiologyNorthwestern UniversityChicago, Illinois

David H. Bruining, MDAssociate Professor of MedicineMayo ClinicRochester, Minnesota

James L. Buck, MDProfessorDepartment of Diagnostic RadiologyUniversity of Kentucky College of MedicineLexington, Kentucky

Carina L. Butler, MDAssistant ProfessorDepartment of Diagnostic RadiologyUniversity of Kentucky College of MedicineUniversity of Kentucky Chandler Medical CenterLexington, Kentucky

Selim R. Butros, MDFellow in Abdominal Imaging and Interventional

RadiologyDepartment of RadiologyMassachusetts General HospitalBoston, Massachusetts

Laura R. Carucci, MDProfessor of RadiologyDirector of Computed Tomography and Magnetic Resonance

ImagingAbdominal Imaging SectionDepartment of RadiologyVirginia Commonwealth University Medical CenterRichmond, Virginia

Wei-Chou Chang, MDDepartment of RadiologyUniversity of CaliforniaSan Francisco, California

Raj R. Chinnappan, MDClinical AssistantAbdominal and Interventional RadiologyHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

Byung Ihn Choi, MDProfessor of RadiologyDepartment of RadiologySeoul National University College of MedicineSeoul National University HospitalSeoul, Republic of Korea

Peter L. Cooperberg, OBC, MDCM, FRCP(C), FACR, FFR(RCSI)honProfessor Emeritus of RadiologyUniversity of British ColumbiaVancouver, British Columbia, Canada

Abraham H. Dachman, MDProfessor of RadiologyDirector, Fellowship ProgramsDepartment of RadiologyThe University of Chicago Medical CenterChicago, Illinois

Alexander Ding, MD, MSDepartment of RadiologyDivision of Abdominal Imaging and InterventionHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

Carolyn K. Donaldson, MD, RPVIAssistant Professor of RadiologyUniversity of ChicagoNorthShore University HealthSystemEvanston, Illinois

Ronald L. Eisenberg, MD, JDProfessorDepartment of RadiologyHarvard Medical SchoolRadiologistBeth Israel Deaconess Medical CenterBoston, Massachusetts

viii Contributors

Sukru Mehmet Erturk, MD, PhDAssociate Professor of RadiologyAttending RadiologistAdministrative DirectorSisli Etfal Training and Research HospitalDepartment of RadiologyIstanbul, Turkey

Thomas A. Farrell, MB, FRCR, MBASection Head, Interventional RadiologyNorthShore University Health SystemClinical Assistant ProfessorUniversity of Chicago Pritzker School of MedicineChicago, Illinois

Kate A. Feinstein, MD, FACRProfessor of Radiology and SurgeryDepartment of RadiologyUniversity of Chicago Pritzker School of MedicineChicago, Illinois

Sandra K. Fernbach, MDProfessor of Radiology (Retired)University of Chicago Pritzker School of MedicineChicago, Illinois

Hector Ferral, MDSenior Clinical EducatorDepartment of RadiologySection of Interventional RadiologyNorthShore University HealthSystemEvanston, Illinois

Florian J. Fintelmann, MD, FRCPCClinical Assistant in RadiologyDepartment of RadiologyMassachusetts General HospitalBoston, Massachusetts

Elliot K. Fishman, MD, FACRProfessor of RadiologyDepartments of Oncology and SurgeryJohns Hopkins HospitalBaltimore, Maryland

Joel G. Fletcher, MDProfessor of RadiologyMayo ClinicRochester, Minnesota

Kathryn J. Fowler, MDAssistant ProfessorDirector of Body Magnetic Resonance ImagingDepartment of RadiologyWashington UniversitySt. Louis, Missouri

Aletta A. Frazier, MDDepartment of Diagnostic Radiology and Nuclear

MedicineUniversity of Maryland School of MedicineBaltimore, Maryland

Ann S. Fulcher, MDProfessor and ChairmanDepartment of RadiologyVirginia Commonwealth University Medical CenterRichmond, Virginia

Helena Gabriel, MDAssociate Professor of RadiologyDepartment of RadiologyNorthwestern UniversityChicago, Illinois

Ana Maria Gaca, MDClinical AssociateDepartment of RadiologyDuke University Medical CenterDurham, North Carolina

Kirema Garcia-Reyes, MDDepartment of RadiologyDuke University Medical CenterDurham, North Carolina

Gabriela Gayer, MDClinical ProfessorDepartment of RadiologyStanford Medical CenterStanford, CaliforniaDepartment of Nuclear MedicineSheba Medical CenterRamat Gan, Israel

Gary G. Ghahremani, MD, FACRClinical Professor of RadiologyUniversity of California Medical CenterSan Diego, CaliforniaEmeritus Professor of RadiologyNorthwestern UniversityChicago, Illinois

Seth N. Glick, MDClinical Professor of RadiologyUniversity of PennsylvaniaPenn Presbyterian Medical ImagingPhiladelphia, Pennsylvania

Margaret D. Gore, MDClinical Assistant Professor of RadiologyDepartment of Diagnostic RadiologyNorthShore University HealthSystemEvanston, Illinois

Richard M. Gore, MDChief of Gastrointestinal RadiologyNorthShore University HealthSystemEvanston, IllinoisProfessor of RadiologyPritzker School of Medicine at the University of ChicagoChicago, Illinois

F

Contributors ix

Sofia Gourtsoyianni, MD, PhDConsultant RadiologistGuy’s and St Thomas’ National Health Service

Foundation TrustLondon, United Kingdom

Nicholas C. Gourtsoyiannis, MDProfessorDepartment of RadiologyUniversity of Crete Medical SchoolChairmanDepartment of RadiologyUniversity Hospital of HeraklionHeraklion, Crete, Greece

Jared R. Green, MDAssistant ProfessorDepartment of Medical ImagingAnn & Robert H. Lurie Children’s Hospital of ChicagoChicago, Illinois

Gianfranco Gualdi, MDProfessor of RadiologyDirector of DEA Radiology DepartmentSapienza UniversityRome, Italy

Rajan T. Gupta, MDAssistant Professor of RadiologyDirector of the Abdominal Imaging Fellowship ProgramDepartment of RadiologyDuke University Medical CenterDurham, North Carolina

Ravi Guttikonda, MDBody Imaging FellowNorthwestern Memorial HospitalHudson, Ohio

Robert A. Halvorsen, MDProfessor of RadiologyMedical College of Virginia HospitalsVirginia Commonwealth UniversityRichmond, Virginia

Nancy A. Hammond, MDAssociate Professor of RadiologyDirector of the School of UltrasoundNorthwestern UniversityChicago, Illinois

Mukesh G. Harisinghani, MDProfessor of RadiologyHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

Sandeep S. Hedgire, MDDivision of Abdominal Imaging and InterventionMassachusetts General HospitalBoston, Massachusetts

Frederick L. Hoff, MDAssociate ProfessorDepartment of RadiologyNorthwestern Univeristy Feinberg School of MedicineChicago, Illinois

Caroline L. Hollingsworth, MD, MPHAssistant ProfessorDepartment of RadiologyDivision of Pediatric RadiologyDuke University School of MedicineDurham, North Carolina

Karen M. Horton, MDProfessorRussell H. Morgan Department of Radiology and

Radiological ScienceJohns Hopkins Medical InstitutionsBaltimore, Maryland

Steven Y. Huang, MDAssistant ProfessorDepartment of Interventional RadiologyThe University of Texas MD Anderson Cancer CenterHouston, Texas

James E. Huprich, MDEmeritus Associate Professor of RadiologyMayo Clinic RochesterRochester, Minnesota

Aleksandar M. Ivanovic, MDAssistant ProfessorCenter for Radiology and Magnetic Resonance

ImagingClinical Center of SerbiaFaculty of MedicineBelgrade, Serbia

Jill E. Jacobs, MDProfessor of RadiologyNew York School of MedicineNew York, New York

Bruce R. Javors, MDProfessor of Clincal Radiology (Retired)Albert Einstein College of MedicineNew York, New York

Bronwyn Jones, MB, BS, FRACP, FRCRProfessor of RadiologyJohns Hopkins University School of MedicineBaltimore, Maryland

Naveen Kalra, MDProfessor of RadiologyPostgraduate Institute of Medical Education and

ResearchChandigarh, India

x Contributors

Avinash Kambadakone, MD, FRCRAssistant ProfessorDepartment of RadiologyHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

Mariam M. Kappil, MD, BS, DABRPediatric RadiologistAnn & Robert H. Lurie Children’s Hospital of ChicagoChicago, Illinois

Ana L. Keppke, MDRadiologistKettering Network Radiologists, Inc.Kettering, Ohio

David H. Kim, MD, FACRProfessor of RadiologyVice-Chair of EducationResidency Program DirectorUniversity of Wisconsin School of Medicine and Public HealthSection of Abdominal ImagingMadison, Wisconsin

Stanley Taeson Kim, MDInstructorDepartment of RadiologyNorthwestern University Feinberg School of MedicineInterventional RadiologistDepartment of RadiologyNorthwestern Memorial HospitalInterventional RadiologistDepartment of Medical ImagingChildren’s Memorial HospitalChicago, Illinois

Douglas R. Kitchin, MDClinical InstructorDepartment of RadiologyUniversity of WisconsinMadison, Wisconsin

Michael L. Kochman, MDWilmott Family Professor of MedicineVice-Chair of Medicine for Clinical Services Center for

Endoscopic Innovation Research and TrainingUniversity of PennsylvaniaPhiladelphia, Pennslyvania

Dow-Mu Koh, MD, MBBS, FRCRConsultant Radiologist in Functional ImagingDepartment of RadiologyRoyal Marsden HospitalSurrey, United Kingdom

J. Satheesh Krishna, MDPostgraduate Institute of Medical Education and ResearchChandigarh, India

Naveen Kulkarni, MDDepartment of RadiologyHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

John C. Lappas, MD, FACRProfessorDepartment of Radiology and Imaging SciencesIndiana University School of MedicineIndianapolis, Indiana

†Igor Laufer, MDProfessor of RadiologyPerelman School of Medicine at the University of PennsylvaniaPhiladelphia, Pennsylvania

Fred T. Lee, Jr, MDRobert Turrell Professor of Imaging ScienceDepartment of RadiologyUniversity of Wisconsin School of Medicine and Public

HealthMadison, Wisconsin

Jeong Min Lee, MDAssociate ProfessorDepartment of RadiologySeoul National UniversitySeoul National University HospitalSeoul, Republic of South Korea

Marc S. Levine, MDChief of Gastrointestinal RadiologyHospital of the University of PennsylvaniaProfessor of Radiology and Advisory DeanPerelman School of Medicine at the University of

PennsylvaniaPhiladelphia, Pennsylvania

Angela D. Levy, MDProfessor of RadiologyDepartment of RadiologyMedstar Georgetown University HospitalWashington, District of Columbia

Jennifer E. Lim-Dunham, MDAssociate ProfessorDepartments of Radiology and PediatricsLoyola University Medical CenterStritch School of MedicineMaywood, Illinois

Mark D. Little, MDAssistant ProfessorDepartment of RadiologyUniversity of AlabamaBirmingham, Alabama

†Deceased.

F

Contributors xi

Russell N. Low, MDMedical DirectorSharp and Children’s MRI CenterSharp Memorial HospitalSan Diego, California

Dean D. T. Maglinte, MDDistinguished ProfessorDepartment of Radiology and Imaging SciencesIndiana University School of MedicineIndianapolis, Indiana

Abdullah Mahmutoglu, MDAttending RadiologistDepartment of RadiologySisli Etfal Training and Research HospitalIstanbul, Turkey

Maria A. Manning, MDAssistant Professor of Diagnostic RadiologyUniversity of Maryland School of MedicineSection Chief of Gastrointestinal RadiologyAmerican Institute of Radiologic PathologyBaltimore, Maryland

Charles S. Marn, MDProfessor of Radiology and GastroenterologyChair of the Quality Assurance CommitteeDepartment of RadiologyMedical College of WisconsinMilwaukee, Wisconsin

Gabriele Masselli, MD, PhDConsultant RadiologistAssociate Professor in Radiology and Nuclear MedicineDepartment of RadiologySapienza UniversityRome, Italy

Shaunagh McDermott, MB, BCh, BAODepartment of Abdominal ImagingMassachusetts General HospitalBoston, Massachusetts

Alec J. Megibow, MD, MPH, FACRProfessor of RadiologyNew York University Langone Medical CenterNew York, New York

Uday K. Mehta, MDAssistant Professor of RadiologyDepartment of RadiologyNorthShore University HealthSystemEvanston, Illinois

Vincent M. Mellnick, MDAssistant Professor of RadiologyMallinckrodt Institute of RadiologyWashington University School of MedicineSt. Louis, Missouri

Christine O. Menias, MDProfessor of RadiologyMayo Clinic College of MedicineScottsdale, Arizona

Joseph Meranda, MDRadiologistAbdominal ImagingRiverside Radiology and Interventional AssociatesColumbus, Ohio

James M. Messmer, MD, MEd, FACRProfessor Emeritus of RadiologyVirginia Commonwealth University School of MedicineRichmond, Virginia

Arthur B. Meyers, MDAssistant Professor of RadiologyChildren’s Hospital of WisconsinMedical College of WisconsinMilwaukee, Wisconsin

Morton A. Meyers, MD, FACR, FACGDistinguished ProfessorDepartment of Radiology and Internal MedicineStony Brook School of MedicineStony Brook, New York

Frank H. Miller, MDProfessor of RadiologyNorthwestern Univeristy Feinberg School of MedicineChief of Body Imaging Section and Fellowship ProgramChief of Gastrointestinal RadiologyMedical Director of Magnetic Resonance ImagingNorthwestern Memorial HospitalChicago, Illinois

Tara Morgan, MDAssistant ProfessorDepartment of Radiology and Biomedical ImagingUniversity of CaliforniaSan Francisco, California

Koenraad J. Mortele, MDAssociate Professor of RadiologyHarvard Medical SchoolDirector of the Division of Clinical Magnetic

Resonance ImagingStaff RadiologistDepartment of RadiologyBeth Israel Deaconess Medical CenterBoston, Massachusetts

Peter R. Mueller, MDProfessor of RadiologyMassachusetts General HospitalBoston, Massachusetts

Brian P. Mullan, MDAssistant Professor of RadiologyDepartment of RadiologyMayo ClinicRochester, Minnesota

xii Contributors

Vamsi Narra, MD, FACR, FRCRProfessor of RadiologyChief of Abdominal Imaging SectionChief of RadiologyBarnes-Jewish West County HospitalWashington UniversitySt. Louis, Missouri

Albert A. Nemcek, Jr, MDProfessorDepartment of RadiologyNorthwestern University Feinberg School of MedicineStaff Interventional RadiologistNorthwestern Memorial HospitalChicago, Illinois

Geraldine Mogavero Newmark, MDVice ChairmanOutpatient ImagingDepartment of RadiologyNorthShore University HealthSystemEvanston, Illinois

Jennifer L. Nicholas, MD, MHA, MAPediatric RadiologistMedical ImagingAnn & Robert H. Lurie Children’s HospitalAssistant ProfessorDepartment of RadiologyNorthwestern University Feinberg School of MedicineChicago, Illinois

Paul Nikolaidis, MDAssociate Professor of RadiologyNorthwestern UniversityChicago, Illinois

David J. Ott, MDProfessor EmeritusDepartment of RadiologyWake Forest University Medical CenterWinston-Salem, North Carolina

Joseph Owen, MDDepartment of RadiologyWashington University School of MedicineSt. Louis, Missouri

Orhan S. Ozkan, MDProfessor of RadiologyDepartment of RadiologyUniversity of Wisconsin School of Medicine and Public HealthMadison, Wisconsin

Nickolas Papanikolaou, PhDAffiliated ResearcherDepartment of Magnetic Resonance ImagingHuddinge HospitalKarolinska InstituteStockholm, Sweden

Mikin V. Patel, MD, MBADepartment of RadiologyUniversity of Chicago Pritzker School of MedicineChicago, Illinois

Pritesh Patel, MDAssistant Professor of RadiologyUniversity of ChicagoChicago, Illinois

Erik K. Paulson, MDProfessor of RadiologyChairmanDepartment of RadiologyDuke University School of MedicineDurham, North Carolina

Christine M. Peterson, MDAssociate Professor of RadiologyMilton S. Hershey Penn State Medical CenterHershey, Pennsylvania

Perry J. Pickhardt, MDProfessor of RadiologyChief of Gastrointestinal ImagingUniversity of Wisconsin School of Medicine and Public

HealthMadison, Wisconsin

Aliya Qayyum, MBBS, MRCP, FRCRProfessor of RadiologySection Chief of Abdominal ImagingThe University of Texas MD Anderson Cancer CenterHouston, Texas

David N. Rabin, MDAssistant Professor of RadiologyUniversity of Chicago Pritzker School of MedicineNorthShore University HealthSystemEvanston, Illinois

Siva P. Raman, MDAssistant ProfessorDepartment of RadiologyJohns Hopkins UniversityBaltimore, Maryland

Peter M. Rodgers, MB, BS, FRCRConsultant RadiologistLeicester Royal InfirmaryUniversity Hospitals of Leicester National Health Service TrustLeicester, United Kingdom

Pablo R. Ros, MD, MPH, PhDRadiologist-in-ChiefUniversity Hospitals Health SystemTheodore J. Castle UniversityProfessor and ChairmanDepartment of RadiologyProfessor of PathologyCase Western Reserve UniversityUniversity Hospitals Case Medical CenterCleveland, Ohio

F

Contributors xiii

Stephen E. Rubesin, MDProfessor of RadiologyDepartment of RadiologyHospital of the University of PennsylvaniaPhiladelphia, Pennsylvania

Tara Sagebiel, MDAssistant ProfessorDepartment of Diagnostic RadiologyThe University of Texas MD Anderson Cancer CenterHouston, Texas

Dushyant V. Sahani, MDDirector of Computed TomographyAssistant RadiologistDepartment of RadiologyMassachusetts General HospitalAssociate Professor of RadiologyDepartment of RadiologyHarvard Medical SchoolBoston, Massachusetts

Sanjay Saini, MDProfessor of RadiologyHarvard Medical SchoolVice-Chair for FinanceMassachusetts General HospitalBoston, Massachusetts

Martha Cotsen Saker, MDDepartment of Medical ImagingAnn & Robert H. Lurie Children’s Hospital of ChicagoDepartment of Medical ImagingShriners Hospitals for ChildrenChicago, Illinois

Riad Salem, MDAssociate ProfessorDepartment of RadiologyDivision of Interventional RadiologyNorthwestern UniversityChicago, Illinois

Kumar Sandrasegaran, MDAssociate ProfessorDepartment of RadiologyIndiana University School of MedicineIndianapolis, Indiana

Rupan Sanyal, MDAssistant ProfessorDepartment of RadiologyUniversity of Alabama at BirminghamBirmingham, Alabama

Christopher D. Scheirey, MDRadiologistDepartment of RadiologyLahey Clinic Hospital and Medical CenterBurlington, MassachusettsAssistant Professor of RadiologyTufts University School of MedicineBoston, Massachusetts

Francis J. Scholz, MDRadiologistDepartment of RadiologyLahey Clinic Hospital and Medical CenterBurlington, MassachusettsProfessor of RadiologyTufts University School of MedicineBoston, Massachusetts

Adeel R. Seyal, MDDepartment of RadiologyNorthwestern University Feinberg School of MedicineChicago, Illinois

Martin J. Shelly, MB, BCh, BAO, MRCSI, FFRRCSIConsultant RadiologistCavan Monaghan HospitalRoyal College of Surgeons in Ireland Healthcare GroupCounty Westmeath, Ireland

Linda C. Sherbahn, MD, MS, BAClinical Assistant Professor of RadiologyNorthShore University HealthSystemEvanston, Illinois

Ali Shirkhoda, MD, FACRClinical Professor of RadiologyUniversity of California at IrvineAttending RadiologistVeterans Affairs HospitalLong Beach, California

Ana Catarina Silva, MDRadiology AssistantDepartment of RadiologyUnidade Local de Saúde de MatosinhosPorto, Portugal

Paul M. Silverman, MDDepartment of RadiologyDivision of Diagnostic ImagingThe University of Texas MD Anderson Cancer CenterHouston, Texas

Stuart G. Silverman, MDProfessorDepartment of RadiologyHarvard Medical SchoolDirector, Abdominal Imaging and InterventionDirector, CT ScanDirector, Cross-Sectional Imaging ServiceDepartment of RadiologyBrigham and Women’s HospitalBoston, Massachussetts

xiv Contributors

Robert I. Silvers, MDClinical Assistant ProfessorDepartment of RadiologySection Chief of Body ImagingNorthShore University Health SystemAssistant Professor of RadiologyUniversity of Chicago Pritzker School of MedicineChicago, Illinois

Ajay K. Singh, MDAssociate DirectorDivision of Emergency RadiologyMassachusetts General HospitalBoston, Massachusetts

Jovitas Skucas, MDProfessor EmeritusDepartment of Imaging SciencesUniversity of Rochester Medical CenterRochester, New York

Gail S. Smith, MDClinical Assistant ProfessorDepartment of Diagnostic RadiologyNorthShore University HealthSystemEvanston, Illinois

Sat Somers, MB, ChB (Sheffield), FRCPC, FFRRCSI(Hon.), FACRProfessorDepartment of RadiologyMcMaster UniversityHamilton, Ontario, Canada

Anthony W. Stanson, MDProfessor Emeritus of RadiologyMayo Clinic College of MedicineDepartment of RadiologyMayo ClinicRochester, Minnesota

Allison L. Summers, MDDepartment of RadiologyNorthwestern Memorial HospitalChicago, Illinois

Richard A. Szucs, MDChairman of RadiologyBon Secours St. Mary’s HospitalRichmond, Virginia

Mark Talamonti, MDProfessor and ChairmanDepartment of SurgeryNorthShore University HealthSystemEvanston, Illinois

Andrew J. Taylor, MDProfessorDepartment of RadiologyUniversity of MinnesotaMinneapolis, Minnesota

Darshit J. Thakrar, MD, DNB, DABRAttending RadiologistAnn & Robert H. Lurie Children’s Hospital of ChicagoChicago, Illinois

Kiran H. Thakrar, MDClinical Assistant ProfessorDepartment of Diagnostic RadiologyNorthShore University HealthSystemEvanston, Illinois

Yee Liang Thian, MBBS, FRCRConsultantDepartment of Diagnostic ImagingNational University HospitalSingapore

Ruedi F. Thoeni, MDProfessor of RadiologyChief of Abdominal ImagingSan Francisco General HospitalDepartment of Radiology and Biomedical ImagingUniversity of CaliforniaSan Francisco, California

Stephen Thomas, MDAssistant Professor of RadiologyUniversity of ChicagoChicago, Illinois

William Moreau Thompson, MD, BAProfessor and Vice ChairDepartment of RadiologyUniversity of New MexicoAlbuquerque, New Mexico

Temel Tirkes, MDAssistant Professor of RadiologyDivision of Diagnostic RadiologyUniversity of Indiana School of MedicineIndianapolis, Indiana

Mary Ann Turner, MDProfessor and Vice-ChairDepartment of RadiologyDirector and Chief of Gastrointestinal RadiologyVirginia Commonwealth University Medical CenterRichmond, Virginia

Jennifer W. Uyeda, MDClinical AssistantAbdominal and Interventional RadiologyHarvard Medical SchoolMassachusetts General HospitalBoston, Massachusetts

Fauzia Q. Vandermeer, MDAssistant Professor of Diagnostic RadiologyDepartment of Diagnostic Radiology and Nuclear

MedicineUniversity of Maryland School of MedicineBaltimore, Maryland

F

Contributors xv

Robert L. Vogelzang, MDProfessor of RadiologyNorthwestern Feinberg School of MedicineChicago, Illinois

Patrick M. Vos, MD, FRCPCClinical Associate ProfessorDepartment of RadiologyUniversity of British ColumbiaVancouver, British Columbia, Canada

Natasha Wehrli, MDAssistant Professor of RadiologyWeill-Cornell Imaging at New York Presbyterian HospitalNew York, New York

Daniel R. Wenzke, MDClinical Assistant Professor of RadiologyNorthShore University HealthSystemUniversity of Chicago Pritzker School of MedicineEvanston, Illinois

Ellen L. Wolf, MDProfessor of Clinical RadiologyDepartment of RadiologyAlbert Einstein College of MedicineMontefiore Medical CenterBronx, New York

Jade J. Wong-You-Cheong, MD, MBChB(Hons)ProfessorDepartment of Diagnostic Radiology and Nuclear MedicineUniversity of Maryland School of MedicineBaltimore, Maryland

Cecil G. Wood III, MDClinical lnstructorDepartment of RadiologyNorthwestern University Feinberg School of MedicineChicago, Illinois

Michael A. Woods, MDAssistant Professor of RadiologyDepartment of RadiologyUniversity of Wisconsin School of Medicine and Public

HealthMadison, Wisconsin

Vahid Yaghmai, MD, MSProfessorDepartment of RadiologyNorthwestern University Feinberg School of MedicineMedical Director of Computed Tomography ImagingDepartment of RadiologyNorthwestern Memorial HospitalChicago, Illinois

Benjamin M. Yeh, MDProfessor of RadiologyRadiology and Biomedical ImagingUniversity of CaliforniaSan Francisco, California

xxii

38 ComputedTomographyEnterographyVideo 38-1: Enterography Demonstrating Both Jejunal and Ileal Crohn’s Enteric Inflammation

VIDEO CONTENTS

53 ComputedTomographyColonographyVideo 53-1: CTC Interpretation

Video 53-2: 2D Detection Pitfall

Video 53-3: 2D Detection Pitfall

Video 53-4: 3D Detection Pitfall

Video 53-5: Characterization Pitfall

General RadiologicPrinciples

SECTION

I

F

3

1 Imaging Contrast Agents and PharmacoradiologyJOVITAS SKUCAS

CHAPTER OUTLINE

Intravascular Contrast AgentsIodinated Water-Soluble AgentsIodinated OilOther Agents

Gastrointestinal Contrast AgentsBarium SulfatePharyngographic AgentsUpper Gastrointestinal Tract StudiesSmall Bowel StudiesBarium EnemaWater-Soluble Contrast AgentsNegative Contrast AgentsGastrointestinal Computed Tomography AgentsAdverse Reactions

Cholangiographic Contrast Agents

Magnetic Resonance Imaging Contrast AgentsIntravascular AgentsGastrointestinal Agents

PharmacoradiologyVasoconstrictorsVasodilatorsGastrointestinal Agents That Produce HypotoniaGastrointestinal Agents That Increase Bowel MotilityMixed Action AgentsDrugs Affecting the Biliary Tract and Pancreas

anattempttodeliverthegreatestiodineconcentrationwiththeleast toxicity. From a simplistic viewpoint, the intravascularcontrastagentscanbeviewedmerelyasvehiclesfordeliveringiodinetoabloodvesselorstructure.Thesewater-solubleintra-vascular contrast agents canbe subdivided into the followingcategories: (1) ionic, high osmolality, roughly five times theosmolalityofblood;(2)nonionic,lowosmolality,roughlytwiceorslightlymorethantheosmolalityofblood;and(3)isotonicagents—nonionic dimers. The basic structures and physico-chemical characteristics of available contrast agents are notcoveredhere;thesetopicsarediscussedinappropriatespecial-izedpublications.1

Atthex-rayenergiesusedinCT,themassattenuationcoef-ficientforiodineisconsiderablygreaterthanthatofsurround-ingsoft tissuesandblood.After the intravascular injectionofiodinated contrast, initial CT images reveal aortic and majorarterial enhancement, followedbya capillaryorparenchymal“blush”andeventualvenousopacification.Therateofcontrastinjection and timing of subsequent CT scans determine thestructuresenhancedonanyoneimage.Comparedwithearlierscanners,multidetectorCT(MDCT)scanners require shorterinjectionratesbecauseoftheshortscanningtimes;asaresult,fasterinjectionratesofmoreconcentratedcontrastagentsarenecessary,andacontrast’sviscosityhasadominantrole.Varioustechniques of intravascular contrast agent administration arediscussedelsewhereinthistext.

Anumberofdrugs,especiallymoreacidicones,areincom-patibletomixingwithcontrastagents,anincompatibility lessevidentwithnonionicagents.Nevertheless,asageneralsafetyprecaution,adrugprobablyshouldnotbemixedwithacon-trast agent, and a catheter should be flushed if used for bothdrugandcontrastinjection.

Ionic AgentsAcetylationofaminotriiodobenzoicacidandfurtherstructuralchangesledtothedevelopmentofioniccontrastagents.Theseagentsareformulatedassaltsandconsistofacationandanion.Twocommonlyusedcationsaresodiumandmeglumine.Theanionportionofthemoleculeconsistsofabenzeneringcon-tainingiodinesubstitutedatpositions2,4,and6,plusanumberofothersidechains.Thesesidechainsdeterminewatersolubil-ityandindirectlyaffectresultanttoxicity.Thebenzeneringcanbe viewed as a scaffold for attaching iodine and side chains.Whenthemoleculedissociates,threeiodineatomsareavailableforeverytwoparticlesinsolution,oraratioof1.5:1.

Further refinements of ionic contrast media consist of theattachment of two monomer triiodinated benzene rings atoneofthesidegroups.Suchadimer,containingtwobenzenerings,witheachhavingthreeiodineatomsandonlyonecationparticle, has six iodine atoms per two particles, or a ratio of3:1.

Only a basic introduction and overview about contrast agentuse in medical imaging are presented in this introductorychapter. These agents can be subdivided into intravascularcontrastagentsforcomputedtomography(CT)andangiogra-phy,intraluminalgastrointestinal(GI)tractagents,cholangio-graphicagentsandauniquegroupofagentsusefulinmagneticresonanceimaging(MRI).Thecurrentlyactiveresearchfieldofcontrastagent–assisted,image-guidedtherapywillbediscussedbriefly.Inaddition,thereisasectiononpharmacologicagentsusefulinGIradiology.

Intravascular Contrast AgentsIODINATED WATER-SOLUBLE AGENTS

Basic PropertiesWith the exception of MRI, all current intravascular contrastagents use iodine for x-ray absorption. Theoretically, sodiumiodideisideal,butitstoxicityandiodismprecludeitsuse.Thecomplexdeliverymoleculesdevelopedovertheyearsrepresent

4 SECTION I General Radiologic Principles

nonionic agents have similar extravascular diffusion rates.Extensive literature is available on relative time-dependentblood iodine concentrations and renal excretion of variouscontrast agents.

Dynamic CT scanning after a single bolus injection relieson the enhancement of vascular structures above baseline.Correctarterialphasetimingisobtainedbyinjectinganinitialtest dose or using automatic bolus tracking. One mg iodineper gram of tissue corresponds roughly to an increase of 30Hounsfieldunits(HU),whichisaboutthelimitfordetection.In general, it is desirable to have sufficient iodine concentra-tion in the vascular structures of interest to elevate themabovebaselinebyupto100HU.Withthisdegreeofenhance-ment, major vessel thrombi are detected and vascular fistulasandrelatedconditionsevaluated.Whetherearlydynamicscan-ning (arterial phase) is superior to portal venous phase oreven delayed scanning after contrast equilibration dependson the organ in question and information sought. For theliver, thesearterialandportalphasetimewindowareroughly20to30seconds.Theseshorttimeintervalsarereadilyachievedwith MDCT.

A typical CT examination consists of a precontrast scan,followedbyscanningaftertheinitialbolusreachesthestructureofinterest.Arelativelylarge-calibervenouscatheterandpowerinjectorprovidereproducibleinjectionrates,keepinginmindthatpredictionofbolusarrival is somewhat empiricbecause,amongotherfactors,decreasedcardiacoutputcanprolongvas-cularflowtimes.

Intravascularcontrastagentscrosstheplacenta,areexcretedinbreastmilk, andaffect fetal and infant thyroid function. Iffeasible, alternate studies should be considered during preg-nancy. Breast-feeding should be stopped for 1 to 2 days aftercontrastinjection.

Acute Adverse ReactionsOnlyabriefsummaryofcontrastreactionsandtheir therapyareprovidedhere.MoredetailedinformationisavailablefromACR2andEuropeanSocietyofUrogenitalRadiology(ESUR)3publications.

The nonionic contrast agents have a considerably lowerosmolality than the ionic agents; adverse reactions caused byhyperosmolalityarethereforereducedwiththenonionicagents.Hyperosmolalityisalsorelatedtovasodilationoftheinvolvedcapillaries. Nonionic contrast agents induce less hypotensionthanionicagents.

Riskfactorsforacuterenalfailureincludediabetesmellituswithdecreasedrenalfunction,renalinsufficiency,dehydration,anduseofahighdose.Atopyconfersanincreasedrisktocon-trastreactions,butallergiestoshellfishappearnottoincreasetheriskofreaction.4Iodinatedcontrastagentsarecontraindi-cated in patients with obvious hyperthyroidism. Also, theseagentsshouldbeavoidedfor2monthspriortothyroidisotopeimagingorradioactivethyroidtherapy.4

Types of Reactions. Acutereactionsvaryfromminoreffectstosevereandlife-threatening.Sensationsofwarmth,nausea,andvomitingappeartobeadirectsideeffecttocontrast.Reactionssuchasmildchanges inbloodpressureormildwheezingareoftenself-limitingormayprogresstomoreseverereactions.Anarbitrarybutusefulgradingofcontrastreactionsismild,mod-erate,severe,andfatal.Compilationsofreactionstoioniccon-trast agents in the 1970s and 1980s revealed a risk of severe

Ionic contrast agents are hypertonic at the concentrationsused for vascular opacification; considerable effort has beenspentinanattempttodecreasetheirosmolality.Ingeneral,theviscosityofa sodiumsalt is less than thatofacorrespondingmeglumine salt, but the sodium salt tends to be more toxic.Toxicity and viscosity limitations during intra-arterial injec-tionsarenotasrelevantfortheintravenous(IV)injectionsusedwithCT.

Nonionic AgentsIf the carboxyl group in position 1 on the benzene ring isreplacedwithastablesidegroup,themoleculenolongerdis-sociateswheninsolution,andeachparticleinsolutionhasthreeiodineatoms,oraratioof3:1.Adimerstructurecanalsobeachievedbylinkingtwotriiodobenzoicacidmolecules(ioxaglicacid),formulatedasamegluminesodiumsalt,anothercontrastagentwitharatioof3:1.Acontrastagentwitharatioof6:1hasalsobeendeveloped(iodixanol)andisoftenreferredtoasbeingiso-osmolar.

Various manufacturers have taken different approaches tothetypeofsidechainsusedwith ionicandnonioniccontrastagents.Asaresult, thesecompoundsdiffer in theirviscositiesandotherproperties.Theinteractionwithothermoleculesalsodiffersbetweenionicandnonionicagentsandisaffectedbythetypeofsidebranchespresent.Withinlimits,however,foreachgroup of contrast agents, viscosity varies directly with iodineconcentration.

Commercial contrast agents also contain chelating agents,usually calcium edetate disodium, to chelate impurities andbuffering agents to achieve an acceptable pH. Their action ismoreimportantduringcontrastagentmanufacturethanduringclinicaluse.

The American College of Radiology (ACR) has publishedcriteriaforsituationsinwhichnonionicagentsarepreferred,2although in many practices this is a nonissue because non-ionic agents are used almost exclusively. The nonionic agentsare associated with less patient discomfort and thus resultinlessmotionartifacts;thisisanevidentadvantage,especiallywith complex examinations such as three-dimensional (3D)reconstruction.

PharmacokineticsAfterabolus intravascular injection, the initialplasma iodineconcentrationisdeterminedbycontrastagent iodineconcen-trationandinjectedvolume.Ionicandnonioniccontrastagentsareeventuallydistributedthroughouttheextravascular,extra-cellularspace,withintravascularandextravascularequilibriumachievedwithin10minutesafterintravascularinjection.Theyareexcretedmostlybyrenalglomerularfiltration.

After injection, the relative plasma iodine concentrationin a particular vessel depends on dilution by blood, extra-vasculardiffusion,andrenalexcretion; thefirst factor ismostimportant during arterial and venous phase imaging, withextravasculardiffusionplayinga largerroleduringtheparen-chymal phase. In theory, a contrast agent can be designed tohave fast or slow extravascular diffusion and rapid or slowrenal clearance; an ideal blood pool agent should have slowextravasculardiffusion. In practice,withanequivalent iodinedose, nonionic agents achieve greater initial peak vascularenhancement than ionic agents, but subsequent blood iodineconcentrations and parenchymal opacification are similar forthetwotypesofagents(exceptrenalvisualization).Ionicand

F

1  Imaging Contrast Agents and Pharmacoradiology 5

a previous reaction to a contrast agent. Regimens that havebeen proposed range from 3 days to immediately before ascan. At the University of Rochester, we recommend thatpatientswhohavehada significantprior reaction to IVcon-trast agents be pretreated with 50 mg of prednisone orallyevery 12 hours, for a total of three doses, with the last dosegiven approximately 1 hour before the examination, and 25to50 mgofdiphenhydraminehydrochloride(Benadryl)orally,2 hours before the examination.

The prevalence of seizures after IV contrast injection isincreased in patients with brain metastases. Capillaries inbrain metastases do not exhibit normal blood-brain barrierintegrity and are permeable to a contrast agent. To decreasethe risk of seizures, it has been suggested that these patientsbe premedicated with diazepam, 5 mg IV, before contrastadministration.9

Treatment  of  Reactions.  Any physician injecting a contrastagentintravascularlycanexpecttoencounterabroadspectrumofreactions,frommildtosevere,andmustbepreparedtodealwiththem.Ingeneral,mildreactionssuchasflushingormildurticaria require no treatment, and most reactions resolvespontaneously.Similarly,nauseaandvomitingrequiregeneralsupportandobservationonly.Ifsymptomsoccurbeforeallthecontrast agent has been administered, the rate of injectionshould be slowed or the injection postponed until symptomsclear.

Early IV access should be established. The catheter usedforcontrast injectionshouldbekept inplace,ensuring intra-vascular access until the possibility of a reaction has passed.Withprogressivehypotension,itbecomesincreasinglydifficultto cannulate a peripheral vein.

Moderateurticariadeveloping in theabsenceofother sig-nificantsymptomscanbetreatedwithdiphenhydramine,25to50mg, orally or injected. With more severe urticaria, oneshouldalsoconsideranH2blockingagentsuchascimetidine(Tagamet),300mginjectedslowly(diluted)IV.Forsevereurti-caria,epinephrine,0.1to0.3mL(1:1000)shouldbegivensub-cutaneouslyunlesscontraindicated.Ifneeded,thedosecanberepeated in 15 minutes. Epinephrine should be used withcaution in older patients who have underlying cardiovasculardisease;electrocardiographicmonitoringshouldbeconsideredforthesepatients.

Severereactions,suchasseverebronchospasm,convulsions,orsignificantcardiopulmonaryreactions,requirepromptandvigorous therapy. Bronchospasm and laryngeal edema gener-ally respond to subcutaneous epinephrine. If needed, theepinephrine dose can be repeated. Diphenhydramine andcorticosteroids, such as hydrocortisone, 100 to 300 mg IV,are also often used. Oxygen should be administered by maskornasalcannula.Betaagonistinhalersalonemaybebeneficialfor mild bronchospasm or can be used in conjunction withaminophylline therapy. With refractory bronchospasm, ami-nophylline, 250 to 400 mg diluted in dextrose and water,can be administered IV over a 10- to 20-minute period.Aminophylline should be used with caution because it couldexacerbatecoexistinghypotension.Trachealintubationshouldbe considered early in the course of these symptoms; later,severe laryngeal edema may make intubation difficult if notimpossible.

Because the treatment of hypotension in the settings oftachycardiaandbradycardia isdifferent, thepulserateshould

reactiontobe1in1000to4000studies.Typesofreactionsaresimilarwithionicandnonionicagents.Ingeneral, theriskofadversereactionsvarieswithcontrastosmolality,sofewerreac-tions occur with nonionic agents than with ionic agents. Inparticular, the risk of severe adverse reactions is lower withnonionic contrast agents.5 Deaths have occurred with bothionicandnonionicagents.

At times, urticaria and even more severe reactions do notrepresentaclassicantigen-antibodyreactionbutaresecondaryto histamine or serotonin release induced directly by thecontrast agent. However, histamine release is probably notthe only factor involved in serious contrast reactions.Amongothereffects,contrastagentsactivate thecomplementsystem,which acts as a host defense, and is related to coagulationabnormalitiesandbradykininrelease.Overall,onlyaminorityof unpredictable reactions mimics immunoglobulin E (IgE)hypersensitivity, probably secondary to an antigen-antibodyreaction. How to classify the rare bowel wall edema is notclear.6 Iodide mumps is a rare delayed reaction to iodine-containing contrast media.

Somereactionsarediseasespecific.IVcontrastagentsinthepresence of a pheochromocytoma can lead to catecholaminereleaseandacutehypertension.Inthissetting,theonsetofsuchhypertensionshouldsuggestapheochromocytoma.

Over the years, many radiologists have avoided the use ofintravascularcontrastagentsinpatientswithsicklecelldisease,althoughtheprevalenceofadversereactionshasnotbeenestab-lished.Overanextendedperiod,bottledhyperosmolarcontrastagents can leach allergens from rubber stoppers. As a rule,contrast-containing vials and bottles should be stored in anuprightposition.

Premedication. Thespecificallergenresponsibleforiodinatedcontrast sensitization is unknown. It is difficult to prove thatiodine is responsible for hypersensitive contrast reactions, acommon assumption.A myosin protein rather than iodine isbelievedtobetheallergenresponsibleinshellfish.Ratherthanaskapatientaboutiodineallergies,amoreappropriatequestionappearstobewhetherdrugallergiesarepresent.Onapracticallevel,thecauseofanadversereactionisoftennotsought,andthe reaction is simply labeled as allergic, hypersensitive, oranaphylactic.

No reliable blood test detects patients who are allergic tocontrast media. Risk factors associated with a contrast reac-tionincludeasthmaandahistoryofpriorreactiontocontrast.However, even these are unpredictable, and a patient mani-festing a severe reaction may have had prior intravascularcontrast with no adverse reaction. Although patients withurticaria-likereactionshaveincreasedplasmalevelsofprekal-likrein andα2-macroglobulin and lower levels of C1-esteraseinhibitor, their predictive value is limited because of normalvariation.7 Pretesting with a small dose of contrast was oncepopularbuthasbeenabandonedashaving littleornovalue.Acutereactionshavedevelopedafterlessthan1 mLofadmin-istered contrast.

Inamulti-institutionalstudyinvolvingioniccontrastagents,pretreatment with methylprednisolone, 32 mg, 12 hours and2 hours before contrast injection, significantly reduced theriskofreactions.8Withthistwo-doseregimen,thenumberofreactions in patients receiving ionic contrast agents approxi-mated those seen with nonionic agents and no pretreatment.Premedication is often considered for patients who have had

6 SECTION I General Radiologic Principles

Various authors use different definitions of significant, withtheESURguidelinesusingacreatinine level increaseofmorethan25%,or44µmol/L(0.5 mg/dL),within3days.3A tran-sient nonoliguric decrease in renal function lasting up to 3weeksismorecommonthanthemoreominousoliguricmani-festation, which may require hemodialysis.

Risk factors for nephrotoxicity appear to be multifactorialand include preexisting renal insufficiency, diabetes, dehydra-tion,cardiovasculardisease,advancedage,myeloma,hyperten-sion,hyperuricemia,andpossiblycontrastosmolalityanddose.Patientsatthegreatestriskforacuterenalfailurearediabeticswith preexisting renal insufficiency. Precaution is also neces-sary with treatment by agents such as nonsteroidal anti-inflammatory drugs (NSAIDs), aminoglycosides, cyclosporin,orevensulfonamides;underlyingnephrotoxicityisacommonpathway.

Indiabeticswithunderlying cardiovascularor renal insuf-ficiency, metformin, a biguanide antihyperglycemic agent, isassociatedwithlacticacidosisandresultantincreasedmortality.This association appears to be indirect and probably involvesunderlyingrenalinsufficiency.However,enoughpatientstakingmetforminandreceivingIVcontrasthavedevelopedlacticaci-dosistoprompttheU.S.FoodandDrugAdministration(FDA)toissueawarning—metforminshouldbediscontinuedbeforeoratthetimeofcontrastusefor48hoursaftertheprocedureandreinstatedonlyifrenalfunctionremainsnormal.However,substantial inconsistencies exist in the guidelines.11 Adequatehydrationshouldbemaintained.

Because iodinated contrast agents are not protein-bound(except forcholangiographicagents), theycanbedialyzed. Inpatientsonhemodialysis,additionalhemodialysissessionsaregenerallynotnecessary.

Themostimportantpreventivemeasureistoensurethatthepatientiswellhydrated.IfIVhydrationisnecessary,someevi-dencesuggeststhattheIVuseofsodiumbicarbonatehydrationissuperiortosodiumchloride.12,13Otherguidelinesincludetheuseof low-osmolarcontrast,discontinuingnephrotoxicdrugsforatleast24hours,andconsiderationofalternateimaginginhigh-risk patients. The osmotic diuretic mannitol does notprovideanybenefit; this loopdiuretic furosemideexacerbatesrenaldysfunction.2Currently,itisprobablysafetoassumethatdiureticsdonotofferanyprotectiveeffect,andanecdotalevi-denceevensuggeststhatdiureticsshouldbestoppedpriortoacontraststudy.14

N-acetylcysteine, an antioxidant, appears to diminishcontrast-induced renal toxicity,15,16 although some have ques-tioned its renalbenefit.17Manyof these studieshave involvedcoronaryangiographyanddifferinmethodologyfromcontrastCT.Therolefortheophyllineislesswellestablished.

Hemodialysis after contrast use in preexisting renal failurepatients is not thought to be warranted. Hemofiltration inchronicrenalfailurepatients,ontheotherhand,hascausedlesscreatinine level increase than incontrols.18Thecomplexityofthis procedure and the cost of hemofiltration limit its use toselectpatients.

IODINATED OIL

Intra-arterial iodizedpoppyseedoil (EthiodolorLipiodol) isusedforseveralindications:

1. AsaCTdiagnosticagentforlivertumordetection,espe-ciallyhepatocellularcarcinoma.

be monitored. A pulse may not be palpable in a hypotensivepatient;cardiacauscultationorelectrocardiographicmonitor-ingmaybenecessary.

Hypotensionintheabsenceofothermajorsignsofanana-phylactic reaction should initially be treated with oxygen, legelevation, and rapid administration of IV fluids. Epinephrineshouldbeconsidered,keepinginmindthatfluidtherapyalonemaybesufficienttherapy.Althoughsubcutaneousepinephrineinjections are adequate for a mild to moderate reaction, IVadministration isneeded formoderate to severehypotension.For IV administration, epinephrine should be diluted to1:10,000and1.0to3.0mLadministeredslowly.Thedosecanbe repeated in 15 minutes, and the rate of injection can betitratedtoachievethedesiredresult.Avasopressoragentsuchasdopamine,2to5µg/kg/min,canbeaddedtosustainbloodpressure.Forunresponsivehypotension,otheragentsareavail-ablefortreatmentofunderlyingshock.AnH2blockersuchascimetidinecanbeadded,300mgindextroseandwater,infusedslowly.Similarly,diphenhydramine,25to50mg,canbeinjectedIV. Corticosteroids are also often used, with a typical dose ofhydrocortisone being 500mg IV. Steroids probably have noimmediate effect on a reaction; their main use is to decreasedelayedreactions.

Attimes,hypotensioncanbecorrectedwithvigoroushydra-tionalone,keepinginmindthatoverhydrationofpatientswithpossible underlying cardiovascular and/or renal disease alsocarriesarisk.Thus,theinitiationoftherapybyadequatehydra-tion is reasonable, but appropriate pharmacologic therapyshouldbeinstitutedwithoutunduedelay.

Hypotensioninthepresenceofbradycardiasuggestsavaso-vagalreaction.SomepatientsrespondtobeingplacedinaTren-delenburg position. Hypotension in these patients should betreatedwithrapidIVinfusionofisotonicsaline.Oxygenshouldbeadministered.Bradycardiacanbetreatedwithatropine(0.5to 1.0mg IV), with the dose repeated every 5 minutes, to amaximaltotaldoseof3.0mg.

Somepatientsreceivelong-termtherapywithabetablockersuchaspropranolol.Acontrast reaction in thesepatientscanbeconfusingbecause,eveninthesettingofanaphylacticshock,a beta blocker–induced bradycardia can persist. IV glucagon,1.0mg or more, may be useful for bradycardia. Dopamine isalso effective. Doses of epinephrine that are usually adminis-teredmaynotbeeffectiveinreversingthishypotension.

Emergency cardiopulmonary resuscitation is necessary forcardiovascularcollapse.RefractoryseizuresaretreatedwithIVdiazepam(Valium)and/orphenobarbital.

Contrast extravasation is treated by extremity elevation,warm or cold compresses and, if extensive, a plastic surgeryconsultation. Hyaluronidase, an enzyme that breaks downinterstitialbarriers,hasbeeninjectedintotheextravasationsiteby some investigators, but its impact on tissue healing is stillnotclear.

Thisdiscussionisonlymeanttobeaguide.Thetreatmentofallreactionsshouldbeindividualized.

Contrast-Induced Nephropathy. Thepathogenesisofcontrast-induced nephrotoxicity is incompletely understood, but it isbelieved that a number of intrinsic renal events lead to renalmedullaryischemia,usuallyaugmentedbyareducedintravas-cular volume.10 Direct cytotoxicity, oxidative tissue damage,and apoptosis are contributing factors. This nephrotoxicity ismanifested by a significant rise in the serum creatinine level.

F

1  Imaging Contrast Agents and Pharmacoradiology 7

Gastrointestinal Contrast AgentsBARIUM SULFATE

Bariumsulfateisawhitecrystallinepowderhavingamolecularweightof233.Becauseofitsspecificgravityof4.5,patientstendto comment that a cup of barium suspension is“heavy.” Theterms thick and thin should only be used when referring toviscosity.Theyshouldnotbeusedtosignifyradiodensity,whichhasmanyothercausativefactors.

Althoughbariumsulfateitselfisinertanddoesnotsupportbacterial growth, some additives in commercial preparationsare organic. When a container is opened or reconstitutedwith tap water, the suspension should be refrigerated if it isto be kept overnight. Although many commercial formula-tions contain preservatives, bacterial contamination can anddoes occur.

Certain commercial formulations are advertised as beingapplicablethroughouttheGItract. Invariably, theserepresenta compromise. The GI tract varies in pH, in composition ofmucus, and in type of mucosa, and optimal coating in onepart does not mean that a similar coating can be expectedinanother.Coatingthemucosawithbariumorsimplyopaci-fyingthelumenrequiresdifferentbariumformulations.20Thelarge-particle, high-density barium suspensions designed fordouble-contrast use should not be simply diluted and usedforsingle-contraststudies.Ingestingsuchadilutedsuspensioncauses rapid barium particle sedimentation, with the nonde-pendent lumen containing little barium; lesions on the non-dependent wall can therefore be missed. Products designedprimarilyforsingle-contrastexaminations,ontheotherhand,canbedilutedconsiderablybeforeanysettlingoccurs,mainlybecause they contain relatively small barium particles.

Ingestionofabariumsulfatesuspensiontendstobeconsti-pating.Currently,commercialbariumproductshaveadditivesthat minimize this effect, and the formation of a bariolith israre.

PHARYNGOGRAPHIC AGENTS

Pharyngeal radiography was already established in the 1960s,whencineradiographybecamepopular toevaluatedysphagia.Although conventional or digital radiography produces highanatomicresolution,dynamicswallowingisbestevaluatedwithvideo fluoroscopy or cineradiography. This pharyngogram,commonly called a modified barium swallow, evaluates oro-pharynxanatomyandfunctionusingcontrastagentsofvaryingconsistencies.

Afterastroke,appropriatepatientfeedingwithoutinducingaspirationcanbedeterminedbyusingbariumsuspensionsofdifferentviscositiesandbarium-coated food.Contrastconsis-tenciesusedrangefrombarium-coatedcrackers toaviscosityapproachingthatofwater.Toimprovepatientacceptance,someinvestigatorshavedevelopedtheirowncontrastagentssuchasbarium pudding or barium honey. Anatomic detail is beststudiedwithhigh-densitybariumproducts suchas the250%w/vsuspensionsdesignedforgastricdouble-contrastexamina-tions. Fistulas are also best studied with this type of contrastagent.Abariumpastecanalsobeusedtostudyanatomy,butthehighpasteviscositylimitsitsapplicationinfistuladetection.The volume of barium used should be individualized. Thus,withsuspectedaspiration,aseveral-milliliterbolusisswallowed

Oftenusedasagoldstandard,itdetectsmoretumorsthan other imaging modalities. Nevertheless, a study ofexplantedliversrevealedthatpretransplantationiodizedoil CT tumor sensitivity is still rather low. Also, oneshouldkeepinmindthatiodizedoilisretainedbysomebenigntumors,evenhemangiomas.

2. Ethiodolisoftenincludedasachemoembolizationingre-dientwheninjectingintoatumorfeedingartery.

This acts as a chemotherapeutic agent carrier and,because of its high viscosity, is a temporary embolizingmaterial that prolongs chemotherapeutic agent contactwithatumor.Ethiodolremainswithintumorneovascu-laritymuchlongerthaninnormalliverparenchymaandthusactsasamarker.

3. Occasionally, ethiodol is injected during percutaneousradiofrequencyablationofhepatocellularcarcinomas.19

Thisaids intheCTdelineationofextentofcoagula-tionnecrosis.

Foranumberofreasons,intra-arterialiodizedoilisconsid-erablymorepopularintheFarEastthanintheWest.

OTHER AGENTS

Several reports have described the use of gadolinium (Gd)-based contrast agents for CT imaging in patients with renalinsufficiency or prior severe reaction to an iodinated agent.Oneshouldkeep inmind,however, that thepharmacokineticproperties of gadopentetate dimeglumine (Gd-DTPA), withonlyonegadoliniumion,aresimilartoiodinatedagentscon-taining three to six iodine atoms. Also, the toxicity of gado-linium agents, at doses achieving equivalent x-ray stoppingpower, isgreater than thatofnonionic iodinatedagents.Thisis in distinction to the use of approved lower gadoliniumMRI doses, which are insufficient for useful x-ray contrast,buthavenegligiblenephrotoxicity.3TheESURpositionisthatgadolinium-based contrast agents are more nephrotoxic thaniodinated contrast agents in equivalent x-ray attenuationdoses.3

Carbondioxide is aviableangiographic contrast agent forcertaindigitalvascularindicationsintheabdomen.Ithasbeenusedasaguide forvascular interventionalprocedures. Itdis-places blood, forms a gaseous column, and is cleared by thelungs.

Apart fromMR,viableabdominalreticuloendothelialcon-trast agents have eluded clinical development. The first suchagent,thoriumdioxide(Thorotrast),useduntilthe1950s,hasleftapainfulandsorry legacy.Iodinatedoilyemulsionsaccu-mulateintheliver,spleen,bonemarrowand,toalesserextent,inotherorgans, longenoughtopermitCTimaging,buthightoxicityandlowspecificityledtotheirabandonment.Colloidaliodineoremulsifiedperfluorooctylbromideparticlesarealsoincorporatedinreticuloendothelialcells.Moststudiesofemul-sifiedperfluorooctylbromide tookplace in the1980s, and itsusehaslostfavorsincethen.

Liposomes are taken up by the reticuloendothelial system,and considerable effort has been expended to encapsulatewater-soluble iodinated contrast agents inside liposomes.Researchactivitypeakedduringthe1990sbut,inspiteofocca-sionalmorerecentpapers,pronouncedadversereactionslimitthe use of liposomal CT contrast agents in humans. Mostcurrent reticuloendothelial contrast research revolves aroundMRIagents.

8 SECTION I General Radiologic Principles

pressureandmayalsopusha foreignbody into the stomach.Thistechniqueshouldbeperformedwithcaretoavoidperfora-tion. Glucagon has also been proposed to help relieve spasm,althoughit isnotclearwhetherpharmacoradiologyhasasig-nificantroleinacutedysphagia.

Commercialbariumsulfatetabletswithadiameterof12mmareavailableandareusefultoevaluatesubtleesophagealstric-tures.Forresultstobemeaningful,thepatientshouldbeatleastin a 45-degree upright position and at least 60mL of watershouldbeingestedwiththetablet.Tablettransittimethroughthe esophagusnormally is less than 20 seconds. These tabletscontain 650mg of barium sulfate plus additives. The tabletsdissolveintheesophagusorstomach(Fig.1-1).Relativelyfreshtablets should be used because older tablets take longer todissolve.

It has been proposed that barium tablets be administeredroutinelyduringchestradiographybecausetabletretentionintheesophagus isassociatedwith thepresenceof structuralorfunctionalesophagealabnormalities,butthis techniqueisnotwidelyused.

Inpatientswithsuspectedesophagealperforation,studyofthe esophagus with a water-soluble agent may not detect asubtle leak. Thus, administration of higher density bariumsulfate enables detection of leaks that might otherwise bemissed.21Forthispartoftheexamination,bariumsuspensionsvarying from 35% to 80% w/v are used. Residual barium inthe mediastinum does not result in clinically detectable

initially; if no aspiration is detected, the bolus is graduallyincreasedinvolume.

Theoropharynxhandleshigh-andlow-viscosityliquidsdif-ferently, sopharyngeal functionshouldbe studiedwithhigh-and low-viscosity barium suspensions. The low-viscositysuspensionshouldhaveaviscosityapproachingthatofwater,whereas the high-viscosity suspension should be similar to athick milkshake. It should be emphasized that some high-density double-contrast barium products are relatively fluidandarenotapplicableashigh-viscositypreparations.

A tracheoesophageal fistula is easier to detect fluoroscopi-callywith thepatient in the lateralposition.With thepatientin a frontal position, it may not be possible to determinewhetherbariuminthetracheawasaspiratedorflowedthrougha fistula.

UPPER GASTROINTESTINAL TRACT STUDIES

Studies of the esophagus consist of single-contrast, double-contrast, andmucosal relief views, togetherwithfluoroscopicevaluation of motility. Normal esophageal tonicity leads tolumen collapse when a primary peristaltic wave has passed.Therefore,regardlessofwhichmethodisused,thestudymustbeperformedreasonablyquickly.

Somepatientsymptomsarereproducedbyusingacoldcon-trastsuspensionoracidiccontrast.Althoughsomehavefoundacidifiedcontrastuseful,itisnotcommonlyused.

Sufficientairisintroducedintotheesophagusforadouble-contraststudyinsomepatientswithpooresophagealmotilityorthosewithgastroesophagealreflux.Inmostpatients,however,an additional negative contrast agent is necessary to obtaindouble-contrastviews.Solidgas-producingtablets,powder,orliquideffervescentagentsareused.Thesecontainsodiumbicar-bonateandanacid,suchastartaricacidorcitricacid,which,inthepresenceofaliquid,producecarbondioxide.About400to 500mL of gas is necessary for adequate esophageal andgastricdistention.Onetechnique is tohavethepatientdrink,inquicksuccession,firstoneandthenanotherliquidefferves-cent solution, followed immediately by 60 to 120mL of abarium suspension. The two effervescent agents distend theesophagusbyreleasingcarbondioxide,andbariumthencoatstheesophagealmucosa.

The high-density, low-viscosity barium products designedfor the stomachandduodenumalsocoat theesophaguswall.Visualizationoftheesophagusisimpairedifbariumisingestedbeforeeffervescentagents.Ontheotherhand,thequalityofthegastricmucosalcoating is improved if thebariumsuspensionisgivenfirst.Thesequenceof ingestioncanbetailoredtothepatient’s symptoms—if esophageal disease is suspected, theeffervescentagentsaregivenfirst; ifgastroduodenaldisease issuggested,thebariumsuspensionisgivenfirst.

Esophageal varices tend to be more prominent and theirdetectioneasieriftheesophageallumeniscollapsed.Althoughhigh-density, low-viscosity barium products detect largeresophageal varices, commercially available barium pastes arerecommended.Someof thesepastesare tooviscousandtendto flow in a bolus; these should be diluted with water so thepasteviscosityissimilartothatofhoney.

In a patient with acute dysphagia, an esophagram canbe therapeutic. With the patient upright, the weight of abariumcolumncandislodgeaforeignbodyintothestomach.Liquideffervescentagentsincreasetheintraluminalesophageal

Figure 1-1  Barium sulfate tablet proximal to a stricture. A previous esophagram suggested narrowing at this site, and the tablet confirms this finding. (From Schabel SI, Skucas J: Esophageal obstruction following administration of “aged” barium sulfate tablets—a warning. Radiology 122:835–836, 1977.)

F

1  Imaging Contrast Agents and Pharmacoradiology 9

The contrast agents are instilled until a lesion is detected orcontrastreachestherightcolon.Ifneeded,glucagonisadmin-isteredtoinducehypotonia.

Airasasecondenteroclysiscontrastagent,usedmorecom-monlyinJapanandEurope,resultsinconsiderablymoreradio-graphiccontrast thanobtainedwithmethylcellulose.Airdoesnotpropelthebariumaheadofitasquicklyasmethylcellulose;ittendstopercolatethroughbarium-filledloopsofbowel.Nev-ertheless, with overlapping bowel loops, as is common in thepelvis,infusionofairisoftenhelpful.Airbubblescanbeconfus-ing,althoughsomethinkthatbetterdiagnosticresultscanbeachieved with air, even in patients with inflammatory boweldisease.

Severaltubelessenteroclysistechniqueshavebeendescribed.One simple method is to initially perform a conventionalsmall bowel study, but when barium approaches the cecum,the patient swallows additional effervescent agents, turnsprone, and the table is turned into a 20- to 40-degree Tren-delenburg position. Efflux of gas results in a double-contrastsmall bowel study. A tubeless double-contrast small boweltechnique consists of effervescent granules coated with anacid-resistant lacquer. Gas is released directly into the smallbowel lumen.

Barium formulations specifically designed for enteroclysisprovide thebest results andare commercially available.Forasingle-contrast study, a barium suspension having a specificgravityofabout1.25(equivalentto≈35%w/v)istypical.Foradouble-contraststudy,thebariumsuspensionshouldbeintherangeof50%to95%w/v.

Retrograde ileographyconsists initiallyofa single-contrastbariumenema,butinfusionofbariumisthencontinuedretro-grade intothe ileum.Becauseflowiscontrolledbytheexam-iner,theileumcanbereadilystudiedwithoutoverlappingloopsfromthemoreproximalsmallbowel.Bariumisinstilleduntiltheregioninquestionisreached.Premedicationwithglucagonincreasespatientcomfortandalsorelaxestheileocecalvalve.Ifa redundant sigmoid colon obscures part of the small bowel,thebariumenemacanbe followedbya saline enema; suchasolutionpushesbariumaheadofitandresultsinasee-througheffect.A20%to25%w/vbariumsuspensionistypicalforsuchastudy.

It is not unusual to achieve a double-contrast study ofthe terminal ileum during a double-contrast barium enema,especially if glucagon is used. This type of study is usefulinsuspectedilealCrohn’sdiseaseorgynecologicmalignanciesinvolving the ileum.

Aperoralpneumocolon,consistingofantegradeandretro-gradecomponents, isdesigned toevaluate thedistal ileumorright colon. Initially, a conventional antegrade small bowelexaminationisperformed.Whenbariumoutlinestheterminalileum, air is instilled through the rectum to obtain double-contrastviewsofthedistalsmallbowelorproximalcolon.Thisstudycanalsobecombinedwithenteroclysis.Routineuseofahypotonicagentishelpful.

BARIUM ENEMA

Single-contrastanddouble-contrasttechniquesarewellestab-lishedmethodsofevaluatingthecolon.Numerousstudieshavecompared the relative accuracy of a single-contrast versusdouble-contrast barium enema. Some radiologists prefer asingle-contraststudyinolderordebilitatedpatients.

mediastinitis and does not interfere with subsequent radio-graphicevaluation.

High-density,low-viscositybariumpreparationsspecificallydesigned for theupperGI tractproducebestdouble-contrastresults.Avolumeof60to120mLofa250%w/vsuspensionisgenerallysufficient.Agoodbariumformulationshouldresultin routine identification of the areae gastricae. Small cancers,ulcers, gastritis. and duodenitis are readily detected during ahigh-qualityexamination.

Whenappropriatedouble-contrastgastricviewshavebeenobtained, a lower density barium suspension is ingested forsubsequent single-contrast evaluation. For this part of theexamination, barium suspensions varying from 35% to 80%w/vareused.Variousexternal compressionpaddlesareavail-ableandarehelpfulinobtainingmucosalreliefviews.

SMALL BOWEL STUDIES

Anumberoftechniqueshavebeendevelopedtostudythesmallbowel, suchasconventionalantegradestudy,enteroclysis, ret-rograde ileography, peroral pneumocolon, and CT and MRenterography.Thetypeofexaminationperformedvarieswithclinicalindication.Specificcontrastagentshavebeendevelopedforeachtypeofstudy.

An antegrade examination is the simplest and most tradi-tionalwayofstudyingthesmallbowel.Serialsmallbowelradio-graphsareobtainedafterthepatientingestsabariumsuspension.Theprimarycontrastagentrequirementisthatitdoesnotfloc-culateorprecipitateduringtransit.Thebariumdoesnotcoatthe mucosa; visualization is obtained primarily by filling thebowellumenwiththebariumsuspension.A40%to60%w/vsuspensionistypical.Manyradiologistspreferavolumeof500to800mL.

Contraindications to an antegrade barium study are sus-pectedcolonicobstructionorbowelperforation.Anumberofclinicians hesitate to request a barium study in the setting ofsmallbowelobstructionandprefertowaituntiltheobstructionclears. Such an approach is analogous to obtaining a chestradiographonlyafterapneumoniaclears.Smallbowelobstruc-tionisnotacontraindication.Bariumproximaltoasmallbowelobstructioncontinuestostayinsuspension,andbariuminspis-sationdoesnotoccur.Withasmallbowelobstruction,anante-gradestudywithbariumissafeandnotonlycandetectthesiteofobstructionbutalsosuggestacause.

In enteroclysis (small bowel enema), contrast is injectedthroughasteerablecatheterdirectlyintothesmallbowel,sotheflow-limitingfunctionofthepylorusisbypassed.Thebariumsuspensioncanbeinfusedbygravity,hand-heldsyringes,oraninfusion pump. Typical infusion rates are 75 to 100mL/min,although theflowrate shouldbe individualized. If the rate istooslow,excessiveperistalsisresultsinastudysimilartoacon-ventionalantegradesmallbowelexamination.Withaflowratethatistoohigh,overdistentionleadstobowelatoniaandlackofprogression.

It is debatable whether single-contrast or double-contrastenteroclysis yieldsbetter results.For thedouble-contrastpor-tion,manyU.S.investigatorsuseasolutionofmethylcelluloseinwater (≥0.5%).Methylcellulosehelpspropel abariumsus-pensionaheadofit.Watercanbeusedasthesecondcontrastagent,althoughwatertendstowashoffbariumadheringtothemucosa.Thetotalvolumeofthetwocontrastagentsistailoredforeachexamination;insomepatients,upto2Larerequired.

10 SECTION I General Radiologic Principles

perforation,withtherealizationthatsmallperforationscanbemissed.Similarly,walled-offperforationsoraperforationinanareaofspasmcanbedifficulttodetect,anditmaybenecessarytocompletetheexaminationwithbarium.

With a chronic or loculated perforation, the higher radio-graphicvisibilityofbariumoftenyieldsmoreinformationthanthat obtained with water-soluble contrast agents. Thus, achronic abscess or other cavity in continuity with the bowellumencanbesafelystudiedwithbarium.Ifthereisapossiblecommunication with the peritoneal cavity, however, water-solubleagentsarepreferred.

Meconium ileus and meconium plug syndrome can betreatedwithan iodinatedcontrastenema.Thepatientshouldbewell-hydrated.

Somesurgeonstreatpostoperativeadynamicileuswithoral,full-strength, ioniccontrastagents,but studieson these typesofagentshavebeenlimited.Also,ahypertonicfluidproximaltoamechanicalobstructionresultsinfurtherdistention.

Contrast AgentsIn general, to achieve adequate radiographic opacification ofmost GI structures, at least a 60% solution of an ionic con-trast agent is needed. The resultant iodine concentration is282 to292 mg/mL for themorecommonlyusedcommercialproducts, with a resultant osmolality of about 1500 mOsm/kg, or approximately five times that of serum. Because ofthis hyperosmolarity, fluid is drawn into the bowel lumen,anddiarrheaiscommonaftertheiruse.Theseagentsstimulateintestinal peristalsis, so faster visualization of the distal smallbowel can be achieved than with barium sulfate. The needforafasterexaminationshouldbebalancedagainstdecreasedradiographic contrast obtained with these agents. In general,intraluminal dilution leads to poor visualization of the smallbowel.

Some commercial ionic contrast agents designed for oraluse,suchasthediatrizoatemegluminepreparationsGastrogra-finandoralHypaque,containflavoringagents.Thesearepre-ferredtothenonflavoredproducts,whicharedesignedprimarilyforIVuse.

Nonionic contrast agents with an iodine concentrationof approximately 300mg/mL have an osmolality of 600 to710mOsm/kg,whichis lessthan50%ofthatof ionicagents.At this concentration, however, they are still hyperosmolarcomparedwithserum.

Ideally,oneofthenonionicagentsshouldbeusedwheneverawater-solubleagentisindicatedforevaluationoftheGItract;forcertainexaminations,someradiologistsusenonionicagents.Ifaperforationintothepleuralorperitonealcavityissuspectedin an adult and aspiration is not a consideration, nonionicagents probably do not offer any major advantage over theirioniccounterparts.Nevertheless,someleaksarebetterdefinedwithbariumthanwater-solubleagents.InstudiesoftheGItractin infants and children, in whom perforation is not an issue,barium rather than a water-soluble is the preferred contrastagent.

NEGATIVE CONTRAST AGENTS

When performing double-contrast GI studies, the cheapestsecondcontrastagentisair.Excellentdouble-contrastesopha-gealviewscanbeobtainedifthepatientswallowsairtogetherwiththebariumpreparation.

Dryandliquidbariumformulationsarecommerciallyavail-able.Ifdrybarium-prefilledenemabagsareused,theamountof water added and degree of subsequent shaking to achievewettability should be standardized. The level marking on theenemabagshouldnotbeused togauge theamountofwaterneeded; resultantdilutions tend tobeerratic.Liquidbarium-filledenemabagsshouldbekeptontheirsidesbecauseconsid-erablesettlingcanoccurifbagsarestoredbeforeuse.

A 12% to 25% w/v barium suspension is commonly usedfor single-contrast barium enemas. The main requirementof the barium suspension is that it neither flocculates norsettles during the examination. Because the sedimentationrate depends, in part, on the amount and type of additivespresent, some products that are well suspended at higherconcentrations settle readily when diluted. If there is doubtaboutacommercialproduct’ssedimentationrate,aradiographobtained with a horizontal x-ray beam should reveal anysettling tendencies.

Double-contrastbariumenemasuspensionsshouldconsistofrelativelyhighbariumconcentrationsbutshouldstillbesuf-ficientlyfluidtoflowreadilythroughenematubing.Theirvis-cosity is greater than that of the lower concentration bariumformulationsdesignedforsingle-contraststudies.Theresultantmucosal coating should be uniform, without undue artifacts.Thesuspensionshouldnotdryoutwhileanexaminationisinprogress.Thesebariumformulationsaregenerallyintherangeof60%to120%w/v,with85%beingtypical.

Evenwhenall conditionsare standardized, the subsequentmucosalcoatingcanvaryamongpracticesbecauseofvariationsinlocalwaterhardnessandtypeofwaterused(distilledwateror cold or hot tap water). Premixed liquid formulations areavailable to avoid these variations. The barium suspension issimplypouredintoanenemabagwithoutfurtherdilution.

Some radiologists perform colonic lavage before a bariumenema. This lavage invariably results in water retention andsubsequentdilutionofthebariumsuspension.Bariummanu-facturers recognize this difference and market two differentpreparations;theonedesignedtobeusedaftercolonic lavagehasabariumsuspensionwithaslightlygreaterspecificgravity.

Tumorscanbedifficult todetect inasegment involvedbyseverediverticulosis.Thissegmentcanbefurtherstudiedifthedouble-contrastbariumenemaisfollowedbyamethylcelluloseenema.

WATER-SOLUBLE CONTRAST AGENTS

IndicationsWater-solubleorganic iodinecompoundsdesigned for theGItractwereintroducedinthe1950s.Eversince,controversyhassurrounded the relative merit and role of these agents. Thesecompounds do not coat the GI mucosa; rather, they providebowel visualization simply by passive filling of the intestinallumen.

For most bowel examinations, experienced radiologistsprefer barium suspensions. Some surgeons, however, are stillbeingtaughtaboutthepurporteddangersofbariumandinsistontheuseofwater-solubleagents.Stimulationofperistalsisinpostoperative patients and lack of radiographically visiblesequelae of spill from the GI tract are reasons cited by somesurgeonsfortheirpreferenceforwater-solubleagents.

Water-solubleagentsareindicatedifanacuteperforationissuspected. The examination generally confirms or excludes a

F

1  Imaging Contrast Agents and Pharmacoradiology 11

before each scan. More convenient is a high-viscosity, low-concentrationbariumpaste,whichprovidesprolongedesopha-gealcoating.22Mucosaladherencebysuchapasteislongenoughto allow a typical CT examination to be completed. Ingestedwaterisoftenausefulandsatisfactorycontrastagentforevalu-ationoftheGItractduringhelicalCT.

Thetraditionalmethodofopacifyingthestomachandsmallbowel is tohavethepatientdrinkapproximately500mLofadiluteCTcontrastagentseveralhoursbeforetheexamination,with a similar amount ingested immediately before scanning.Ideally, such an agent should differentiate bowel from sur-rounding structures without introducing artifacts. Use of adilute iodine solution or a barium suspension is generally apersonalpreference.A1%to3%w/vbariumsulfatesuspensionor a 2% to 5% solution of Gastrografin or similar iodinatedagentistypical.Onerefinement(granted,littlepracticed)istousea2.0%bariumconcentrationforjejunalopacificationandaslightlylowerconcentrationforpelvicstructures.WithslowerCTunits,aniodinesolutionproducesfewerstreakartifactsthanbarium, a problem of little consequence with multislice CT.Commercial barium suspensions tend to taste better thaniodinesolutions,afactorwhenexaminingchildrenandnausea-pronecancerpatients.

Theiodinetastecanbemaskedbyaddingsugarandvariousfruitextracts;althoughessentiallysugar-freeiodinecontrastisavailable,bariumproducts, ingeneral,containlesssugarthancorrespondingiodinecontrasts.Atthedilutionsused,theiodin-atedsolutionsarehypoosmolar,butsomepatientsstilldevelopdiarrhea.Thepoortasteandhencepooracceptanceof iodin-atedagentsbypatientscanbepartlyovercomebytheempiricadditionofaflavoredjuicesuchasKool-Aid.23Atthedilutionsused in CT, nonionic contrast agents do not have any realadvantageoverionicagents.

With suspected pelvic disease, a contrast agent can beingested the evening before the examination. Even if full-strength Gastrografin is ingested, overnight dilution in thebowel is sufficient to eliminate most streak artifacts. Betterrectosigmoid opacification is obtained after ingestion of sucha full-strength contrast agent than dilute barium, probablybecausehyperperistalsisisinducedbytheiodinatedagent.Nev-ertheless, identifyingthelargebowelis lessofaproblemthanidentifyingfluid-filledloopsofsmallbowelonabdominalCT.

Both CT and MR enterography can detect Crohn’s diseasewith somewhat similar accuracy.24Adequate CTbowelopaci-fication can be achieved by using a 2% flavored bariumsuspension.

CTenteroclysisconsistsofbowelintubationandinstillationof an iodinated contrast agent, dilute barium suspension,or methylcellulose suspension. Whether a positive or water-density agent is superior is not clear; an IV contrast agent toopacify bowel mucosa aids in lesion detection. Negative oralcontrast agents designed specifically for CT enteroclysis havealso become available. Also, with MDCT and coronal recon-structionnowmorereadilyavailable,retainedbowelfluidoftenprovidesasufficientmarker,especiallyindilatedbowelloops.

Iftheimagingstudyisperformedtoevaluatearectallesion,ahigh-viscosity,low-volumebariumpastemaysuffice;approxi-mately100mLofa3.6%w/vcarboxymethylcelluloseand2%w/v barium sulfate paste mixture has been proposed in thisclinicalsetting.25

A basic question when protocoling an examination onmultidetector CT is whether intraluminal water-density or

One commercial preparation incorporated carbon dioxidedirectlyinthebariumsuspension;whenthepatientdrankthis“bubblybarium,”carbondioxidewasreleasedintotheesopha-gusandstomach.Theeffectwas similar to thatofdrinkingabottle of club soda, and this product did not achieve readyacceptance.

Effervescenttablets,granules,andpowdersarecommerciallyavailable.Theyproducecarbondioxideoncontactwithwaterand most are satisfactory in achieving adequate gastric andduodenaldistention.Thereis,however,considerablevariationintheirdissolutiontime.Mostcommercialeffervescentpowdersandgranulescomeinsingle-dosepackages.Inclinicaluse,thepatient places the effervescent agent in the mouth and usessmallamountsofwater towash itdown.This is immediatelyfollowedbyabariumsuspension.Theswallowedgasisusedfordouble-contrastviewsofthestomachandduodenum.

Liquid effervescent agents, consisting of separate acid andbasesolutions,canbepreparedlocallybyahospitalpharmacy.The acid portion consists of citric and tartaric acids and thebaseportionissodiumbicarbonate.Adoseof12to15mLissatisfactoryformostpatients.

Carbondioxideisusedbysomeinplaceofairfordouble-contrastbariumenemasandCTcolonography.Carbondioxideis absorbed faster than air. Whether a gas or air is usedprobably does not influence examination quality although,with all other factors remaining constant, colonic distentionis less with carbon dioxide compared with air.

Some double-contrast preparations result in excessive gasbubbles.Anantifoamagentshouldbeaddedempiricallyiftheseoccur on a regular basis.Although many commercial bariumpreparationsalready includesuchanagent, insomeareas theamount isnotsufficient.Acommonlyusedantifoamagent isdimethylpolysiloxane(simethicone);theadditionof1.5mLofsimethicone(equivalentto100mg)isoftensufficienttoelimi-natebubbles.

GASTROINTESTINAL COMPUTED TOMOGRAPHY AGENTS

The term double-contrast abdominal CT is used by some tosignifytheuseofIVandoralcontrast.However,thisisamisuseof the traditional connotation of double contrast and is bestavoidedtopreventconfusion.

Full-strengthbariumpreparationsshouldnotbedilutedtothe low concentrations needed for CT. The barium particlessettle out after ingestion of such a dilute solution, leading toinhomogeneous bowel lumen opacification. The uppermostpart of a loop of bowel may not contain enough barium forvisualization, and excess barium in the dependent portionresultsinstreakartifacts.

Stable but low-concentration barium formulations specifi-callydesignedforCTarecommerciallyavailable,withmostofthesebrandnamesending in -cat.MostCTbariumproductscontainsmallparticlesthatresistsettling.Additivesselectedalsopreventbariumsedimentation.At the lowbariumconcentra-tionsused,bariumparticlesdonotcoatthemucosabutsimplyprovidelumenopacification.

Anesophagusmarkedwithacontrastagentaidsinevaluat-ingthemediastinumduringchestCT.The low-concentrationCTagentsused in the rest of theGI tract donotopacify theesophagus long enough, although one option is to have thepatient drink small sips of a conventional CT contrast agent

12 SECTION I General Radiologic Principles

multiforme,respiratorycomplications,anaphylaxis,GIangio-edema,andevendeathhaveoccurred.Patientswithasthmaandseverefoodallergiesappeartobeataslightlyincreasedriskforthese reactions, but the average radiologist will probably notencounterahypersensitivityreactiontoabariumproductinalifetimeofpractice.

Thecauseofhypersensitivityreactionsduringmostbariumstudiesisnotknown.Ingeneral,theincriminatingagentisnotsoughtandnotestingperformedinmostpatientswhodevelopareaction.

Esophageal perforation and spillage of barium into themediastinumresultinaninflammatoryreaction,withbariumpersisting in the mediastinum for a prolonged period. Suchpriorextravasationcanoftenberecognizedradiographicallyasdense linear radiopacities, but no strong evidence exists thatthesesequelaehaveanyclinicalsignificanceforthepatient.

Mostperforationsassociatedwithabariumenemaoccurinthe rectum and are not immediately detected by fluoroscopy.Rectalperforationstendtoresultfrominjudiciousinsufflationof an enema balloon. A British survey over a 3-year periodbetween 1992 and 1994 revealed a complication rate of 1 in9000andadeathrateof1in57,000.28Although10%ofpatientswithabowelperforationdied(3of30),themortalitywas56%(9of16)inpatientsdevelopingacardiacarrhythmia.

Spillageofbariumintotheperitonealcavitycanbesecond-ary to a preprocedure perforation, such as in patients withulcers.Someperforations,however,areassociatedwithabariumstudyandcanoccurduringanupperGIexaminationorbariumenema, and even during enteroclysis. Initially, leukocytes aredrawnintotheperitonealcavity,togetherwithaninpouringoffluid.Profoundhypovolemiadevelopsifmassiveinpouringoffluidintotheperitonealcavityisuntreated.Bacterialcontami-nationduringaperforationcanresultinoverwhelmingsepsisandshockwithinhours.

Immediatemanagementofbariumperitonitisincludesinfu-sionoflargevolumesofIVfluid.Antibioticsareadministeredbecause of associated bacterial contamination. Most patientsundergo surgery, with an attempt made to evacuate bariumfromtheperitonealcavity.Invariably,bariumcrystalsembed-ded on the peritoneal surface resist dislodgment.Attempts toremove barium particles with a wet sponge simply inducesdiffuseperitonealbleeding.

Bariumcrystalsinciteaninflammatoryreaction;eventually,these crystals become coated by a fibrin membrane, andextensive fibrosis and granulomatous tissue develop. Densefibrosis can involve adjacent structures and, depending onlocation, subsequent ureteral obstruction or bowel deformityand stenosis develop. Perirectal fibrosis can narrow the rec-tosigmoid lumen and even mimic a carcinoma. Residualbarium is identified with conventional radiography or CT.Noevidence suggests thatbarium in soft tissues is a carcino-gen (Fig. 1-2).

Thebariumintravasationcaninvolvesystemicveinsandtheportalvenoussystem.29,30Somepatientshavenopredisposingfactorstoaccountforintravasation.Overall,bariumintravasa-tionisassociatedwithamortalityrateofmorethan50%.

Water-Soluble AgentsTheriskofsensitivityreactionstooraliodinatedcontrastagentsisconsiderablylessthanwithintravascularinjection.Inyoungchildrenandadultswithhypovolemia,however,theintroduc-tion of large volumes of a hypertonic agent into the GI tract

fat-densitycontrast,orevenagas,issuperiortoapositivecon-trast agent.Positivebowel contrast creates artifacts, especiallywith maximal intensity projection images of vascular struc-tures.Also,bowelwallenhancementbyIVcontrastisusefulfordetecting bowel wall thickening, and a positive intraluminalcontrastcanobscuresubtlelesions.Asaresult,MDCTstudiescanbeperformedwithoralwater-densityorevennegativecon-trast agents.Althoughadequate forgastric andduodenaldis-tentionbecauseofitsabsorptionfromthebowel,ingestedwaterdoesnotreadilydistendthedistalsmallbowel;useofacarboxy-methylcelluloseorpolyethyleneglycolsolutioninhibitsabsorp-tionandimprovesdistention(comparedtowater).Apreliminarystudy of simethicone-coated cellulose (SonoRx, Bracco Diag-nostics,MonroeTownship,NJ),developedfororaluseinupperabdominal ultrasonography, found no significant advantageoveroralwaterinabdominalCT.26

Inthepast,anumberoffatdensityproducts,suchasmineraloil,cornoil,andmilk,andevenaparaffinemulsion,werepro-posedforCTuse,butthesehavehadlimitedapplication.Resid-ualbowelgasoftenservesasamarker,especiallyinthecolon.If a nasogastric tube is in place, air can be injected into thestomach and small bowel. If excessive amounts of gas arepresent,imagingwithwindowsettingsslightlywiderthanusualishelpful.2Dcolonographyand3Dvirtualcolonoscopyrequirecolonicdistentionwitha contrast agent; typically, air isused,andcarbondioxideisusedlessoften.

ADVERSE REACTIONS

Barium SulfateBariumsulfateispoorlysolubleinwater.Theconstipatingten-dencyofbariumproductsiswellknowntomostradiologists.Throughthejudicioususeofadditives,thissideeffectismini-mizedinmostcurrentformulations,althoughbariumimpac-tioninthecolonisstilloccasionallyencountered.

Aspiration of small amounts of commercial barium for-mulationsisoflittleclinicalsignificance.Afterbariumaspira-tion, most is cleared from major bronchi and trachea withinhours, although some is retained in the interstitium and inmacrophages.Thisresidueisgenerallynotvisualizedonradio-graphs.Alveolarizationofbarium,however,canresult inpro-longedretention.Ifaspirationissuspectedclinically,nonionic,sterile,iso-osmolariodinatedcontrastcanbeusedratherthanbarium.

HypersensitivityreactionsduringGIexaminationsarerare,although they have been known to occur. Although bariumsulfate is inert, commercial formulations contain numerousknownproprietaryadditives.27Theseincludestabilizing,flavor-ing, coating, and viscosity-varying agents and range fromnaturalflavorsandgums, suchas lemon,pectin, andguar, tosyntheticproducts,suchasvariousmethylcelluloses.Someradi-ologists may still be familiar with chocolate-flavored bariumproducts,whicharenolongerusedbecauseofcommonaller-giestochocolate.Anaphylaxiscanbecausedbycarboxymethyl.Theroleofeffervescentagentsinthesereactionsisspeculative.Methylparabenandsimilarcompounds,usedaspreservatives,caninducehypersensitivityreactions,butbariummanufactur-ershave replaced themwithmore innocuouspreservatives inmostcommercialbariumproducts.

Reactions appear to be more common during double-contrast than single-contrast studies. Most reactions are mildand consist of urticaria or pruritus, although erythema

F

1  Imaging Contrast Agents and Pharmacoradiology 13

actualpractice,adversereactionsareuncommon.Theyundergoenterohepatic circulation and occasionally delayed reactionsareencountered.

IV cholangiographic agents undergo hepatocyte uptakeand biliary excretion by active transport. They are excretedunchanged or after conjugation with glucuronic acid. Thesehepatocyte-specificlivercontrastagentsconsistoftriiodoben-zene compounds which, by means of benzene ring substitu-tions,havehadtheirhydrophilicitydecreasedtothepointthattheycannowpassthroughmembranes.Iodipamidemeglumine(Cholografin) is theonlyagentavailable in theUnitedStates.Mostoftheinjectedcontrastisboundtoalbumin;biliaryvisu-alizationisgenerallyevidentabout15minutesafterthestartofIVinjection.Onemightassumethatbecausecholangiographicagentsundergohepatocyteuptake,theywouldmakeusefulCTliver contrast agents; in actual practice, however, hepatocyteuptakeistooslowandexcretiontoofastforthemtoservethisfunction.

Toxicity and allergic reactions are more severe with iodip-amide than with more typical intravascular contrast agents.Nephrotoxicityisdosedependent.Inpartbecauseofthistoxic-ity, cholangiography has been supplanted by other imaging,includingMRIanddirectcholangiopancreatography.Thelatterisachievedbypercutaneous transhepaticcholangiography,anendoscopicretrogradeapproach,orinjectionthroughasurgi-callyplacedtube.

Direct cholangiography is generally performed with full-strength contrast concentrations (≈300 mg iodine/mL). Inparticular,whensearchingforsubtle leaks,ahighiodinecon-centration is advantageous. Also, when injecting proximal toanobstruction,ahighiodineconcentrationallowsfordilutionbyresidualbile.Whensearching for stones,however,dilutionof contrast with an equal volume of water appears useful;subtle stones can be missed in markedly opacified bile ducts.Contrast overinjection during direct cholangiography shouldbeavoided.IntheUnitedStates,syringeinjectioniscommon,butinEurope,adripinfusiontechniqueismorepopular.Thestudy should be terminated if the pancreatic duct begins toopacify—acute pancreatitis is a complication of this study.Earlierstudiessuggestedthatmegluminesaltsofioniccontrastagents resulted in less bile duct epithelial damage than cor-responding sodium salts; the use of nonionic contrast makesthis point moot.

IndicationsforCTcholangiographyhavebeenevolving.CTcholecystographyisperformed10to12hoursafteroraladmin-istrationofacholecystographiccontrastagent(iopanoicacid).Anotherapproachisslowinfusionofacholangiographiccon-trastagent,resultinginbiliaryimagessuperiortothoseobtainedwith conventional IV cholangiography. Major intrahepaticductsarevisualizedinmostindividuals.Preliminaryevidencehas suggested that CT cholangiography is somewhat superiortoMRcholangiopancreatography(MRCP)invisualizingsmallbiliary stones. It is not useful in jaundiced patients becauseinsufficientcontrastisexcretedintobileducts.

MRCP is a noninvasive imaging technique for visualizingbiliaryandpancreaticducts.Ithasevolvedasanalternativetodiagnostic endoscopic retrograde cholangiopancreatography(ERCP) and diagnostic percutaneous cholangiography. Twoapproachesarefeasible—anIVcontrast-assistedtechniqueandatechniquewithoutcontrastusingheavilyT2-weightedimagesto make nonflowing fluid hyperintense. The former uses pri-marilyhepatobiliaryMRagents(seelater);contrast-containing

Figure 1-2  Prior colonic perforation during a barium enema. Barium crystals are encased by dense adhesions that also involve bowel. (From Miller RE, Skucas J: Radiographic Contrast Agents. Baltimore, University Park Press, 1977, p 137.)

canresultinhypovolemia,shock,andpossiblydeath.Insuchasetting,adequateintravascularfluidreplacementandtheuseofanonioniccontrastagentshouldbeconsidered.

If aspiration or a tracheoesophageal fistula is suspected,hyperosmolarioniccontrastagentsarecontraindicatedbecausetheycaninducepneumonia,pulmonaryedema,ordeath.Thenonionic agents are reasonable substitutes. In most adults,however,bariumsulfateisthepreferredcontrastofchoice.

Other Contrast AgentsSomereactionsoccurevenbeforeacontrastagent is instilled.Latex, used in some enema balloons, has been implicated insomereactions.Theoffendingantigeninlatexisbelievedtobeawater-soluble,heat-stableprotein.It is foundonthesurfaceof cured latexandprobably is a contaminantofnatural latexwhenitisobtainedfromtheHevea brasiliensistree.Insensitiveindividuals, contact with skin leads to urticaria; contact withmucous membranes can result in more severe anaphylacticreactions.Currently,nonlatexandsyntheticlatexproductsareavailable.

Cholangiographic Contrast AgentsWithnormalrenalfunction,onlyabout1%ofanionicornon-ioniccontrastagentdoseundergoeshepaticexcretion,whichisinsufficientforCTbileductvisualization.Inthesettingofrenalfailure,however,itisnotunusualtodetectgallbladderopacifi-cation,evenwithconventionalradiography.

Cholecystography using oral cholecystographic agents hasbeen supplanted by other imaging modalities and currentlyis rarely performed. Several commercial cholecystographicagentsareavailable.Toensureoralabsorption,theyhavehydro-philic and lipophilic properties; in blood, they are bound toalbumin and, theoretically, should have high toxicity, but in

14 SECTION I General Radiologic Principles

At recommended doses, gadolinium chelates have a loweradverse reactionrate than iodinatedcontrastagents,butana-phylacticreactionsandevencardiopulmonaryarrest,includingfatal ones, do occur. Risk factors for reactions are not welldefined but appear to be similar to those for iodine contrast.Unlike iodine-related reactions, gadolinium-related reactionstendtobedelayed,attimesoccurringeven1hourormorelater.Theseagents are excretedbyglomerularfiltrationand, in theusual doses, nephrotoxicity is uncommon, although nephro-genicsystemicfibrosishasbeenreportedinsomepatientswithrenaldisease.31

In patients on hemodialysis, about 80% of gadolinium isdialyzedafterthefirstandessentiallyallafterthefourthdialysis.Anormaldosehasbeenusedinhemodialysispatients.

Gadolinium agents exhibit poor water relaxivity at highermagneticfields(>4T).Undoubtedly,newMRIagentswillbedevelopedforusewithhighfieldstrengthmagnets.

Reticuloendothelial AgentsLargersuperparamagneticironoxide(SPIO)particlesaretakenupby thereticuloendothelial system(RES)and,amongothereffects, result in decreased liver and spleen parenchymalenhancement on T2-weighted images. Uptake is also presentin lymph nodes and bone marrow. Ferumoxides (Endorem,Guerbet, Villepinte, France) is an SPIO agent consisting of acolloidal mixture of ferrous and ferric oxide. Tissues lackingreticuloendothelial(RES)cells,suchasmetastases,havelittleorno signal loss and thus appear hyperintense to the resultanthypointense, normal RES-containing liver or spleen paren-chyma.Notonlyareknowntumorsbetteridentifiedbut,com-pared with unenhanced MRI sequences, more tumors aredetected.Thisdifferentiationisnotabsolute,however,becausesomewell-differentiatedneoplasmscontainREScellsandthustakeupironoxideparticles.TheseSPIO-inducedchangesdifferamongvarioustumors,potentiallyprovidingtissuecharacter-ization of different tumors. SPIO enhancement is impairedin diffuse liver disease. Nevertheless, imaging interpretationswith these compounds is complex; some of the particles areultrasmall,andT1-weightedgradient-echosequencesresultinimagingpatternssimilartothosewithGdchelatesandultrasmallSPIOparticles.

Theseagentshavealongerintravascularhalf-lifethangado-liniumchelates.AprolongedscanningwindowisavailableoncetheseparticlesarewithintheRES;eventually,freeironisusedinnormal ironmetabolism.Theirrole inclinical imaginghasnotyetbeenclearlydefined.Disadvantagesincludeprolongedscanningtimesandanincreasedfalse-positiverate.

Hepatobiliary Agents (Intracellular Agents)Several hepatobiliary-specific paramagnetic contrast agents,such as Gd-EOB-DTPA (Eovist, Bayer Imaging, Whippany,NJ) and Gd-BOPTA (MultiHance, Bracco Diagnostics) areinitially extracellular and then undergo hepatocyte uptake.Early dynamic perfusion imaging is similar to that with con-ventional gadolinium chelates, but this can be followed bydelayedhepatic imaging; eachphaseyieldsdifferent informa-tion. Image interpretation differs from conventional gadolin-iumscans; thus,withGd-BOPTA,ahepatocellularcarcinomashows early peripheral enhancement, but parenchymal phaseimagesrevealanisointenseorevenhypointensetumor.32

These agents are eliminatedbybiliary and renalpathways.EovistachievesgreaterliverenhancementthanMultiHance,has

bileishyperintenseonT1-weightedsequences.Alimitationofthisprocedureisthatreasonablehepatocytefunctionisrequiredto accumulate enough biliary contrast to be imaged. Thecontrast-lesstechnique,oftensimplycalledMRCP,hasnoneofthese limitations and has become the primary imaging studyforvisualizingbileducts.

Magnetic Resonance Imaging Contrast AgentsINTRAVASCULAR AGENTS

The term contrast agenthasadifferentmeaning inMRI thanusually applied to barium sulfate or iodinated agents. MRIcontrastagentsarenotvisualizeddirectly;rather,theirprimaryfunctionistoalterwaterprotonrelaxationtimes.Theseagentshave considerable variation in diffusion and renal clearance,thus leading to the use of a specific contrast agent for aspecific application. Although MRI contrast agents are morecomplex and serve a more varied function than CT contrastagents, they are used for the same primary purpose—toimprove lesion detection and characterization by increasingcontrast tissuesignal intensitydifferencesbecauseofdifferenteffectsontissueprotonrelaxation.Thesedifferencesvarywithtime and depend on the degree of lesion vascularity. MuchcurrentMRIcontrastagentresearchhasfocusedonimprovingtheir specificity.

MRIcontrastagentsareoftenclassifiedbytheirmetalcom-ponent.Amoreusefulclassificationisbasedontheirdistribu-tion, with the realization that many agents overlap betweencategories;manyareinitiallybloodpoolagentsbutsubsequentdistributiondependsontheirmolecularconfiguration:

1. Conventionalgadoliniumchelates(extracellularagents)2. Macrophage-monocyticphagocytic(reticuloendothelial)

agents3. Primarilyhepatobiliaryagents(intracellularagents)4. PrimarilybloodpoolagentsAllcurrentlyavailableMRIcontrastagentsshortentissueT1

and T2 relaxation times. The paramagnetic gadolinium andmanganese contrast agents primarily shorten T1 and thusincreasesignalintensity(enhancement)ofnormalparenchymaon T1-weighted images. The superparamagnetic iron oxidesprimarily shorten T2, thus decreasing signal intensity onT2-weighted images and, depending on the sequence used,increase theT1 signal.Thesemetal ionsarechelated tootherstructures,suchasDTPA,toreducetheirtoxicity.

Gadolinium ChelatesThe most often used vascular MRI contrast agents are gado-liniumchelates.Thesemostlyhydrophiliccompoundsareche-latedto“mask”thetoxicgadoliniumion.Theirbiodistributionand perfusion-related issues are similar to iodine-containingcontrast agents. To take full advantage of these MRI contrastagents,dynamicimagingmustbeperformedshortlyaftercon-trast injection (arterial to portal venous phases); later, theseagents equilibrate with the extracellular space, and lesionsbecomeisointensetoparenchyma.Livertumordetectionwiththeseagentsreliesondifferencesinbloodflowbetweentumorandnormaltissue.AmorerecentapplicationoftheseagentsisinMRangiography(MRA)asasubstituteforabdominaldigitalsubtraction angiography (DSA). Image processing allows fortheseparationofarteriesandveins.

F

1  Imaging Contrast Agents and Pharmacoradiology 15

Apotentialuseforironoxideagentsistolabelthemwithatarget-specificdrug,suchascholecystokinin,withreceptorsinthepancreas.Developmentofsuchagentsisstillinitsinfancy.

Aseparategroupofbloodpoolcontrastagentsconsistsofgadolinium-based products bound to albumin, dextran, oranother similarly large molecule (also called macromolecularMRI contrast agents); imaging using Gd-BOPTA has alreadybeendiscussed(seeearlier).BiodegradablepolymericMRIcon-trastagentshavebeenreported.35Tightalbuminbindingpro-longs blood pool time. The clinical role of these agentspresumablywillbesimilartothatofUSPIOagents.

CarbondioxideisabloodpoolMRIcontrastagent.Anintra-vascular column of carbon dioxide leads to signal loss andresultsinablackbloodMRangiographytechnique.

MorecomplexapplicationsofMRIcontrastagents involvecombining twoagents(sometimescalleddouble-contrastMRimaging).Forexample,moreinformationisobtainedaboutafocal tumor by combining perfusion data of a gadoliniumchelatewithitsRESstatusobtainedwithaSPIOagentthanispossible with a single agent alone. Such combined contrastagentuseismostlyexperimental.

Another research approach involves excitation-based,frequency-labeled,exchangetransferimagingtoseparatetissuemagnetization transfer contrast components by using a para-magneticchemicalexchangesaturationtransferagent.36MRAofhepaticvesselsappearscomparabletoDSAandsuperiortoportalveinimagesforevaluatingliverarteries.

GASTROINTESTINAL AGENTS

AnoralMRIbowelcontrastagent identifies thebowel lumenand differentiates the normal bowel wall from an abnormalprocess.Boweldistentioncanbeobtainedbyinjectingcontrastvia a nasojejunal catheter (enteroclysis) or having the patientdrinkalargequantityoffluid(enterography).Anoralcontrastagentaids in identifyingasoft tissuetumorinthebowelwallandinadjacentorgans.

OralMRIcontrastagentsaresubdividedintopositivecon-trastagents,whichpredominantlyshortenT1andincreaseMRIsignal intensityonT1-weighted images,andnegativecontrastagents,whichshortenT2anddecreasesignalintensityorsimplylackhydrogenprotonsandwater-densitycontrast.Positivecon-trastagentsconsistofvariousiron,manganese,andgadoliniumparamagnetic compounds; they are useful in detecting sinustracts.Ontheotherhand,theymaskintraluminalcontentandmakebowelwallvisualizationdifficult.

Adistinctionbetweenpositiveandnegativecontrastagentsisnotabsolute,andsomeagentschangepropertieswithdilutionandtheMRIsequenceused.Gadoliniumisapositivecontrastagent that shortens T1 in the smallbowelbut,whenconcen-trated in thecolon, it actsasanegativecontrastagent.Ferricammonium citrate is hyperintense on T1- and T2-weightedimages at concentrations lower than 45mg/mL; at higherconcentrations, and at 10 to 20mg/mL, bowel loops becomehypointenseonT2-weighted images.Amore relevant issue iswhethersuchcontrastimproveslesiondetection;currentresultsarenotclear.

Adilutebariumsulfatesuspensionisausefulnegativeagent.PerfluorocarbonslackhydrogenprotonsanddonotproduceanMRIsignalonT1-orT2-weightedimages,buttheirroleasoralagents isnot established.Methylcellulose,polyethyleneglycol,and dilute magnesium sulfate37 have been evaluated. Air and

abiliaryexcretionrateofabout50%oftheinjecteddose,anda delayed biliary phase is evident. Gd-BOPTA exhibits a T1relaxivity roughly double that of conventional gadoliniumagents, probably secondary to its binding to albumin, whichdecreasesextravascularleakage.Althoughonlyasmallpercent-age is takenupbyhepatocytes, thishasaprolongedeffectonliversignalintensity.

Mn-DPDP(Teslascan,GEHealthcareAS,Buckinghamshire,England)dissociatespartlyinplasma,withfreeMn2+takenupbyhepatocytesandothertissues,includingthepancreas.Non-dissociatedMn-DPDPiseventuallyeliminatedbythekidneys.It is often considered to be a hepatobiliary-specific paramag-neticcontrastagent,havinganeffectlastingforseveralhours,thus permitting delayed imaging. It is injected by slow IVinfusion,sodynamic imagingisnot feasible.OnT1-weightedimages,thisagentselectivelyenhancesnormalliverparenchymaandhepatocyte-containingtumorssuchasfocalnodularhyper-plasia,regenerativenodules,andhepatocellularadenomasandcarcinomas,butshowslittleornoenhancementofmetastases,cysts,andhemangiomas.Thisagentdifferentiateshepatocellu-larcarcinomasfrommetastases,withmetastasesofnonhepato-cyteoriginbecomingmoreconspicuousbecauseofanincreasedsignal fromsurroundingnormal liverparenchyma.Anexcep-tion is metastatic neuroendocrine tumors; occasionally, thesetumorsenhancewithMn-DPDP.ApossibleuseisduringMRI-guidedthermaltumorablationwhenprolongedtumorvisual-izationisbeneficial33;Mn-DPDPidentifiesmorefocal tumorsin cirrhotic and noncirrhotic livers than detected on precon-trast images. It appears useful in defining intrahepatic biliaryanatomic variants, such as in pretransplantation liver lobedonors. Because it does not differentiate between benignandmalignantprimary liverneoplasms, itsdiagnostic impactisnotclear.

Manganese isexcreted inbile,butexcretion is inhibited inbiliary stasis. It is contraindicated in severe liver failure.Mn-DPDP’soverall safetyappears tobesimilar tothatof theotherhepatocytegadoliniumproducts.

Blood Pool AgentsUltrasmall superparamagnetic iron oxide (USPIO) particles,unlikemostMRIcontrastagents,shortenT1andT2relaxationtimes, with a few exceptions. Initially conceived as MRI lym-phographic agents, their current role has evolved mostly intoblood pool agents. They have a blood half-life measured inhours.Onedisadvantageisthesuperimpositionofarteriesandveins, although image processing techniques such as subtrac-tionandphasecontrastcandifferentiatethesestructures.Theseironoxidecontrastagentsappearusefulindifferentiatinghighlyvascular lesions,suchashemangiomas, fromsolidneoplasms.Also, their prolonged reduction of intravascular T1 valuesmakesthemusefulforMRangiographicinterventionalproce-dureswithoutneedforrepeatcontrastinjection.Theirrelaxiv-ity increases at lower field strengths, and they are thereforesuitableforuseinopenmagnets.

Ironoxideparticlessmallerthan10nmpassthroughcapil-lariesandareeventuallyclearedbyliver,spleen,bonemarrow,andlymphnodereticuloendothelialcells,resultinginhomoge-neoussignallossinthesestructures.Theseagentsthereforecanpotentially identify lymph node metastasis independently ofnode size; normal nodes lose signal intensity but nodes (orregionsofnodes)containingmetastasesdonot takeup theseparticles.34

16 SECTION I General Radiologic Principles

Propranolol blocks β-adrenergic vasodilation and, whenused in conjunction with epinephrine, results in mesentericvesselvasoconstriction.

Norepinephrinehasαreceptorstimulationsimilartothatofepinephrine but lacksβ receptor stimulation. It has been lesswellstudiedforimagingusethanepinephrine.

Patientswithapheochromocytomahaveincreasedepineph-rine or norepinephrine levels after contrast injection; it thusappears prudent to premedicate these patients with α- andβ-adrenoceptorantagonists tocontrol symptomsandpreventanadrenergiccrisis,althoughthismaynotbenecessarywhenusingsomenonioniccontrastagents.

Angiotensin,ahormonefamilywithvasoconstrictiveactiv-ity,isapotentvasoconstrictoractingonnormalvesselsmoothmuscle.Similartoepinephrine,ittendstoenhancevisualizationofmalignantneoplasmsbyaselectiveincreaseintumorbloodflow.

Vasopressininpharmacologicdoseshasapressoreffectandconstrictsnormalsplanchnicsmallvessels(includingcapillar-ies), thus decreasing portal blood flow. It has little effect onhepaticarteryflow.Transcathetervasopressininfusioncontrolssome GI hemorrhage, keeping in mind that intra-arterialadministrationmaycausemesentericinfarctionorsmallbowelnecrosis.

Bombesin,agutpeptide,releasesendogenousgastrin,whichactivates gastric mucosal sensory neurons, which in turnincreasegastricmucosalbloodflowandthusprotectthemucosaagainst damage. Somatostatin negates this bombesin-inducedgastroprotection.Althoughbombesinisnotusedinradiology,thesomatostatinanalogue,octreotide,hasbeenusedfortherapyof esophageal and gastric variceal bleeding. Gut neuroendo-crinetumorsalsocontainsomatostatinreceptors,andoctreo-tide is useful for diagnosis and palliation of these tumors.Itsuppressescarcinoidtumorsymptoms.

VASODILATORS

Vasodilators increase blood flow in a selective vascular bed.Theireffectsdifferinnormalandneoplasticvessels,butvaso-dilators are less suitable than vasoconstrictors for outliningsmallneoplasms;increasednormalvascularitytendstoobscurea small tumor. Occasionally helpful is a high-calorie meal,which accentuates superior mesenteric artery and portal veinblood flow.

Tolazoline,anadrenergicalphareceptorblockingagentandsynthetic vasodilator, aids in the angiographic visualizationof small vessels. Direct mesenteric artery injection leads toimprovedvenousvisualization.Itseffectonneoplasmvisualiza-tion is mixed. Whether tolazoline has a role in intra-arterialprovocativemesentericangiographyto identifyGIbleeding isnot clear. Infusion therapy with tolazoline and heparin hasbeen used to treat nonocclusive mesenteric ischemia.41 Intra-arterial verapamil and tolazoline appear to be comparable intheirvasodilatoryefficacy.42

Bradykinin is a nonapeptide, produced from decapeptidekallidin,normallypresentinbloodinaninactiveform.Brady-kinin is a plasma kinin, potent vasodilator, and one of themediators of anaphylaxis released from cytotropic antibody-coatedmastcells.Radiographically,bradykinininjectedintothesuperiormesentericarteryimprovesportalveinvisualization.

Acetylcholine,aparasympathetichormone,isavasodilatorpreviouslyusedmostlytoevaluaterenalarterystenosis.

water are also MR contrast agents. Nonabsorbable water-densityagentsaresimilartothoseusedforCT(seeearlier).

Positivecontrastagentsaccentuatemotionartifacts.However,contrast artifacts are more pronounced with negative agents,butbowelwallabnormalitiesarebetterevaluatedwithnegativeagents. Antiperistaltic pharmacologic agents have been used;they reduce motion artifacts, even with a high field strengthunit.

MR colonography is performed using water, a gadoliniumsolution,orabariumsuspensionasa luminal contrastagent.Evaluation can include surface-rendered, virtual endoscopicendoluminal views, orthogonal sections in three planes, andwater-sensitive images.38 A preliminary report has suggestedthatMRproctography is inferior tobariumproctography fordetectingpelvicfloorabnormalities.39

PharmacoradiologyPharmacologicagentsusefulinGIradiologywillbediscussedinthissection.Includedarethosehavinganintravasculareffect(vasoconstrictorsandvasodilators),thosemodifyinggutmotil-ity, and those affecting bile flow. Excluded are experimentalagents,agentsdesignedformolecularimaging,andtherapeuticmaterials.

Fromaradiologist’sviewpoint,mostgutmotilityagentscanbedividedintothosethatincreaseGItonicityandmotilityandthosethatdecreasethesefunctions.Someagentshavedifferenteffects on different parts of the GI tract; they are discussedseparatelyinthesectiononmixedactionagents.

Bowel tonicity is not the same as peristalsis. In general,however,pharmacologicagentsthatincreaseboweltonicityalsoresult in increased peristalsis. For example, agents inducinggastrichypertoniatendtoresultinfastergastricemptyingandhypertonic small bowel agents result in faster small boweltransit.Hypotonicagentshaveanoppositeeffect.

One exception to this classification is famotidine, whichsuppresses gastric secretions. A preliminary report suggestedthat famotidine may be useful prior to an upper GI exami-nation; the decreased gastric secretion improved the qualityof the examination.40

VASOCONSTRICTORS

Theprimaryuseofvasoactivedrugs inabdominal imagingisto alter blood flow in a way designed to increase diagnosticaccuracy.Somedrugsaiddeliveryofchemotherapeuticagentstoneoplasms.Allsideeffectsmustbeacceptable.Vasoconstric-tors aid primarily in detecting and characterizing neoplasms;theyconstrictnormalbloodvesselsbuthavelittleifanyeffectonmalignantvessels.

Epinephrine, an adrenergic hormone, stimulates α and βreceptors and, depending on specific innervation, results invasoconstrictionordilation.Initiallyused inrenalarteriogra-phy, epinephrine decreases contrast opacification of normalrenalparenchyma,thusaccentuatingrenalcellcarcinomavas-culature.Limitationsfortheuseofepinephrineincludeavari-abledoseresponseandtheabilityofsomeinflammatorytissueneovascularity torespondsimilarly toneoplasticneovascular-ity. Hepatic and splenic arterial injection results in spasm ofthese vessels but little constriction of normal hepatic, gastric,duodenal,andpancreaticsmallvessels.Also,normalmesentericvesselsdonotrespondappreciablytoepinephrine.

F

1  Imaging Contrast Agents and Pharmacoradiology 17

pancreatic tissue, but became a moot issue when a syntheticproductwasdeveloped.InjectableglucagonisproducedbytheexpressionofrecombinantDNA.Thechemicalstructureofthissyntheticglucagonisidenticaltothatofhumanglucagon.

Glucagonisahormonethathassignificantmetabolicinflu-enceon a number of organs. It binds at specific receptor cellmembranes intargetorgans.Inthe liver, itstimulatesglucoseoutputandhepaticketogenesis.Itlysesadiposetissueandleadstoareductionofcirculatingcholesterolandtriglyceridelevels.Itstimulatesinsulinreleaseandappearstobeinvolvedinliverregeneration,butitsfullroleintheliverisnotclear.Glucagonincreases blood flow to the kidneys. Specific effects are alsopresentintheadrenalglandsandheart.Itismetabolizedintheliverandkidney.Glucagonisdegradedbygastricsecretionsandthereforeisineffectivewhengivenorally.

In smooth muscle, glucagon is a relatively potent spasmo-lyticagent,anditisthisspasmolyticactionthataccountsforitsuseinradiology.Pharmacologicdosesareused.SmoothmuscleindifferentGItractsegmentsvariesinsensitivitytoglucagon(Table1-1).Forexample,0.1mgIVissufficienttoinducegas-troduodenalhypotoniainmostadults,butsuchasmalldoseisinadequateforcolonichypotonia,forwhichadoseseveraltimesgreaterisneeded.

Intravascular glucagon is also a vasodilator. It improvesportal vein visualization during mesenteric arteriography,although it has been replaced by other vasodilators becauseof its propensity to induce nausea and vomiting at the dosesrequiredforvasodilation.

Gastrointestinal  Tract.  Acute esophageal obstruction causedbyfoodimpactionisoftenrelatedtoanunderlyingstrictureorspasm.Spasmolyticdrugshavebeenrecommendedifspasmissuspected,butinamulticenter,double-blindstudyofglucagonand diazepam, no significant difference in disimpaction wasevidentbetweenspasmolyticagentsandplacebo.47Effervescentagentshavealsobeenusedtotreatesophagealfoodimpaction,withvaryingsuccess.

ThemainadvantageofglucagonoveranticholinergicagentsininducingupperGItracthypotonicityisitslackofsideeffects.IntheUnitedStates,glucagonisgenerallyusedtoinducehypo-tonia;insomecountries,theanticholinergicagentscopolaminebutylbromide (Buscopan) isusedmoreoften.One reason forusing an anticholinergic agent is the higher cost of glucagon,although the price ratio of glucagon to Buscopan fluctuatesconsiderablyworldwide.

Glucagon decreases intragastric and intraduodenal meanpressures.Somestudiessuggestthatbariummucosalcoatingin

Dopamine, a potent renal artery dilator, is similar to ace-tylcholine; it has also been studied mostly in renal vessels.Preliminary work has suggested that dopamine decreasescontrast-inducednephrotoxicity,butfurtherstudiesfoundthatit may have a deleterious effect.43

Prostaglandinshavevariablevasculareffects,dependingontheir chemical composition and specific use. Prostaglandintherapy is used in neonates with cyanotic congenital heartdisease.ProstaglandinE1(PGE1)hasasimilareffectonsplanch-nic vasculature as acetylcholine and tolazoline, resulting inincreased portal vein blood flow. Thus, superior mesentericartery injection of PGE1 during CT hepatic arteriographyresults in increased conspicuity between hepatocellular carci-nomanodulesandsurroundingparenchyma.44Also,theuseofPGE1 during CT hepatic arteriography helps reduces thenumberofpseudolesionsaroundthegallbladderbed.45Prelimi-nary work has suggested that it may reduce contrast-inducednephropathy,butthisisassociatedwithanumberofsideeffects,and its role isnotclear.PGF2αdilatesnormalcolonicvesselsbutvasoconstrictsinflammatoryandneoplasticcolonvessels.

Persistent,usuallyasymptomatic,gastricdistention,detectedon radiographs, is a complication of prostaglandin therapy.Distentionusually resolvesaftercessationof therapy.Superfi-cially, this condition resembles pyloric stenosis, althoughimagingshowsgastricmucosalthickeninganddistalantralandpyloricelongation,butnomuscularwallthickening.

Papaverineisavasodilatoroflargeandsmallvessels.Similarto other vasodilators, it improves portal vein visualizationduring mesenteric angiography. It is not degraded in a singlepassthroughtheliver,andrepeatedinjectionscanleadtosys-temichypotension.

Secretin increasespancreaticbloodflow.Attimes,selectivevenous sampling after intra-arterial secretin injection aids indetecting gastrinomas. IV secretin, however, does not appearreliableenoughfordetectingchronicpancreatitisorapancre-aticadenocarcinoma.

GASTROINTESTINAL AGENTS THAT PRODUCE HYPOTONIA

Bowelhypotoniaishelpfulinanumberofsettings.Forexample,aspasmolyticagentdilatesasegmentofspasticcolonthatoth-erwisemightmimicabenignstrictureormalignancy.Similarly,polypsanddiverticulainthesmallbowelcanbedetectedmorereadilyifthebowelisdilatedandatonic.Spasmolyticpharma-cologicagentscanbedividedintohormonalagents(e.g.,glu-cagon)andanticholinergicagents.Agentsevaluatedforuseinthe GI tract include morphine, propantheline bromide (Pro-Banthine),atropine,andrelatedcompounds.Withsomeagents,afterinitialenthusiasm,recognitionoftoxicityandundesirablesideeffectsledtotheirabandonment.

GlucagonHuman glucagon is a single-chain polypeptide containing 29amino acid residues, with a molecular weight of 3483. It isgeneratedbyαcellsintheisletsofLangerhans.Insomespecies,glucagonisalsoproducedinthestomach;whetheranygastricglucagonisproducedinhumansiscontroversial.Theglucagonaminoacidsequenceinanimalsrangesfromonesimilartothatinhumans tocompletelydifferentsequences. Identicalaminoacidsequencesarefoundinhumans,pigs,andcattle;thiswasonce significant, when glucagon was obtained from animal

From Miller RE, Chernish SM, Brunelle RL, et al: Double-blind radiographic study of dose response to intravenous glucagon for hypotonic duodenography. Radiology 127:55–59, 1978.

GLUCAGON DOSE (mg)

Location and Response 0.25 0.5 1 2

Stomach 4.9 8.7 10.1 15.1Duodenal bulb 7.5 10.1 12.5 16.7Duodenum 7.8 10.1 12.5 16.1Proximal small bowel 8.3 9.4 13.7 19.7Distal small bowel 8.6 9.4 14.0 19.7

TABLE 1-1

Spasmolytic Effect of Intravenous Glucagon: Average Duration of Atonicity (min)

18 SECTION I General Radiologic Principles

that glucagon might have a role in ileocolic intussusceptionreduction. A number of reports described intussusceptionreductionaftertheadministrationofglucagon,butsuchempiricusedoesnot implythatanyeventualreductioncanbeattrib-utedtoglucagon;evenasecondorthirdattemptatreductionimprovestheoverallsuccessrate.Controlledstudieshavefoundsimilar success rates for intussusception reduction with andwithoutglucagon.

Current controversy centers not on whether glucagon isuseful in intussusception reduction, but whether the contrastagentshouldbebarium,asalinesolution,orair,andwhetherfluoroscopyorsonographyisthepreferredimagingmodality.

Preferred Imaging Modality with GlucagonIn Computed Tomography. Inspiteofglucagon’sdiverseeffectsintheliver,itdoesnotappeartoinfluencehepaticCTenhance-ment. With older scanners, glucagon and somatostatin wereusedtodecreasemotionartifacts.Thereislittleneedtoinducebowel hypotonia with multidetector scanners, but the abilityto maintain bowel distention with a hypotonic agent aidsgastric and bowel wall evaluation. Although some studiessuggest that glucagon prior to CT colonography does notimprove colonic distention,49 others find an antispasmodicagent useful in maintaining hypotonia during air insufflationand scanning. Spasm developing during the study can bedecreased by the judicious use of glucagon.

In Ultrasonography. Occasionally, bowel atonia is helpfulin abdominal ultrasonography. An acoustic window to thebiliarytractcansometimesbeobtainedbyfillingthestomachwith fluid and inducing hypotonia of the surrounding GIstructures.

In Magnetic Resonance Imaging. Currently, glucagon isused infrequently inMRI, although itmayhavea role iforalMRI contrasts agents and bowel distention are used. Whenevaluatingmuralandserosaldisease,IVglucagonallowsbettervisualization of normal bowel loops and bowel wall thicken-ing.50 Glucagon also helps eliminate ghost images of positivecontrast-opacified bowel.

Contraindications and Side EffectsAmythpersiststhatglucagonshouldnotbegiventodiabeticpatients. It should be pointed out that glucagon is used totreathypoglycemic reactions indiabetics.On theotherhand,in the setting of hyperglycemia and ketoacidosis, temporaryadditional glucose level elevation induced by glucagon is oflimited importance. The diabetic patient can safely receiveglucagon whenever clinically indicated prior to an imagingstudy.

Side effects of glucagon are less than those with atropineorpropantheline. Inone study, the sideeffectswithglucagonwere similar to those seen with a placebo. The prevalenceof nausea and vomiting after injection of glucagon isdose dependent.51 When given IV, glucagon slowly decreasesthese reactions.

Becausecommerciallyavailableglucagonhasanaminoacidresidue similar to human glucagon, allergic reactions shouldnot occur, although rare ones have been reported. In thesepatients,anallergytothepreservativeused,ratherthantoglu-cagon,shouldbeconsidered.Glucagonisanaturallyoccurringpolypeptideand,inpureform,shouldnotresultinhypersen-sitivity reactions. Previously, commercial glucagon containedbovine or porcine insulins, protoinsulins, other nonglucagon

thestomachandduodenumisimprovedmorewithanticholin-ergicagentsthanwithglucacon,therationalebeingthatanti-cholinergicsalsodecreasegastricsecretions,whereasglucagonhasnosucheffect. Inactualpractice,Buscopanandglucagonproduceessentiallyequaldistentionandbariumcoatingofthestomach and duodenum. A more basic question is whetherinducedgastricandduodenalhypotoniaimprovestheabilitytodetectlesions.Thediagnosticqualitywithandwithoutgluca-gondoesnotappeartodiffersignificantly.Radiologists intheUnitedStatestendnotuseahypotonicagentduringupperGIstudies.

Inenteroclysis,bariumisinstilleduntilalesionorobstruc-tion is reached, or the terminal ileum is filled. Glucagon ishelpfulifitisdeemeddesirabletoslowdownbariumprogres-sion,suchaswhenasuspiciousregionisidentified.Ingeneral,0.25mg IV is sufficient to induce hypotonia, permitting aleisurelystudyoftheregioninquestion.

Duringabariumenema,glucagonrelaxestheileocecalvalveand allows easier barium reflux into the distal small bowel.Thus,ifretrogradeileographyisbeingperformedforsuspecteddistalilealdisease,itappearsreasonabletoadministerglucagon.Anticholinergicshavelittleeffectontheileocecalvalve.

Inselectpatients,aperoralpneumocolonexaminationallowsdouble-contrast study of the terminal ileum and right colon.Barium is introduced via a conventional oral examination orenteroclysisapproach,andairisaddedthroughanenematip.Becauseglucagonrelaxestheileocecalvalve,itmayincreasethesuccessrateofthisexamination.

Colonhypotoniaisachievedbyinjecting2mgofglucagonintramuscularly.Hypotoniabeginswithinseveralminutesandlasts about 15 minutes. Hypotonia can also be achieved with0.25to0.5mgofglucagonIV,althoughinsomepatientsupto1.0mg may be necessary; the onset of hypotonia with an IVinjection is almost immediate and lasts approximately 10minutes.Ingeneral,thesmallerIVdoseisusedbecauseofcostconsiderations. In infants and children, an IV dose of 0.8 to1.25µg/kghasbeenrecommended.48

Few studies have evaluated whether the use of glucagonresults inamoreaccuratediagnosis,andwhetherglucagon isinjected during a barium enema varies considerably amongradiologists. It ismorecommonlyused inhospitalized,older,and ill patients. In some practices, it is injected routinely fordouble-contrastbariumenemasbutindividualizedwithsingle-contraststudies.Inanoutpatientsetting,manyradiologistsuseglucagonwhenapatienthaspainfulspasmorspasminterferingwith the examination, is unable to retain the enema, or hassuspectedcolitisordiverticulitis.Glucagondecreasestheextentand severity of colonic spasm during a barium enema andmakespatientsmorecomfortable.

Occasionally,colonicspasmpersistsdespiteadministrationofglucagon.Ithasbeenmyempiricobservationthatpatientswith long-standing diabetes have more glucagon-resistantcolonicspasmthannondiabeticpatients,butthereasonforthisdecreased response indiabetics isnotknown.Manydiabeticsalreadyhavehighbloodglucagon levels,althoughthese levelsare in the physiologic rather than pharmacologic range. Thepresenceofautonomicneuropathyinsomediabeticsmaybeafactor.Attimes,refillingthecolonseveralminuteslaterresultsinamarkeddecreaseofspasm.

Reduction  of  Intussusception.  Because of its spasmolyticeffectandtendencytorelaxtheileocecalvalve,itwasthought

F

1  Imaging Contrast Agents and Pharmacoradiology 19

Oral hyoscyamine sulfate is a potential hypotonic agent,havingactionsandcontraindicationssimilartothoseofatro-pineandotheranticholinergicagents.Itappearstoprovidenobenefit when used as pain premedication during a bariumenema, although it aids in achieving distention during CTcolonography.

Complications. Inpatientspredisposedtoglaucoma,increasedintraocularpressureinducedbyanticholinergicdrugsmaypre-cipitateanacuteattack.Althoughmostpatientswithahistoryofglaucomahavechronicglaucoma,apatientmayhaveacuteangle-closureglaucomaandnotbeawareofit.Acuteglaucomashouldbesuspectedifeyepainorlossofvisiondevelopsafteradministrationof ananticholinergicagent.Buscopanuse canresultinblurredvision.

The effect on the autonomic nervous system can leadto urinary retention. This complication is exacerbated inpatientswithprostatichypertrophyorotherpredispositiontourineretention.Allergicreactionstoanticholinergicagentsareuncommon.

GASTROINTESTINAL AGENTS THAT INCREASE BOWEL MOTILITY

Insomepatients,therateofgastricemptyingisincreasedifthebariumvolumeusedisincreased.Acoldsuspensionisnotonlybettertoleratedbutalsoleadstofastergastricemptying.Fastersmallboweltransitcanbeachievedbyaddingahyperosmolarproducttoabariumsuspension;asmallamountofdiatrizoatemeglumine (Gastrografin) can be added to an oral bariumsuspension.

High-osmolality sorbitol added to oral CT contrast willacceleratebowelopacification.Somemanufacturersaddsorbi-toltotheirbariumsulfateproducts.

MetoclopramideMetoclopramide is an antiemetic agent and is also useful intreatingdiabeticgastroparesis.ItsprimaryeffectsintheGItractare an increase in gastric peristalsis, pyloric relaxation, andincreaseinsmallbowelperistalsis.Ithasnomajoreffectonthecolon. Metoclopramide appears to decrease gastric secretionsbuthaslittleeffectonmucosalbariumcoating.Atypicaldoseis10to20mgparenterallyororally.Itisarelativelysafedrug,althoughextrapyramidalsideeffects,suchasacutedystoniaandtardivedyskinesia,developoccasionally.

Oral metoclopramide reduces small bowel transit time. Itcan be given shortly before a small bowel study or up to 90minutesbeforetheprocedure.AdministeredorallybeforeaCTscan, metoclopramide improves opacification of the ileum,right colon, and transverse colon but not the more proximalbowel. Longitudinal contractions and foreshortening of ilealloops tend to elevate the ileum out of the pelvis. Metoclo-pramidealsoappearstoimprovevisualizationofthepancreasin abdominal sonography, with its primary benefit beingdecreasedgastricandduodenalgasartifacts.

DomperidoneDomperidone,apotentdopamineantagonist,increasesgastricemptying and accelerates small bowel transit. Although itdecreasessmallboweltransittime,itseffectonthesmallbowelappearstobelessthanthatwithmetoclopramide.54Ithasbeenusedforthetreatmentofdiabeticgastroparesis.Domperidone

proteincontaminants,andpreservatives,andanyofthesemaybe associated with a hypersensitivity reaction.A rash, perior-bital edema, erythema multiforme, respiratory distress, andhypotension have been reported. Currently used geneticallyengineeredglucagonshouldbeassociatedwithfewanaphylacticreactions.

Contraindicationstoglucagonincludepriorsensitivity,sus-pected pheochromocytoma, or insulinoma. Glucagon canrelease catecholamines from a pheochromocytoma and resultin the sudden onset of life-threatening hypertension. Suchhypertensioncanbecounteredwiththeα-adrenergicblockingagent phentolamine mesylate (Regitine). In adults, a dose of5mgIVappearsuseful,althoughconsiderablevariabilityexistsintreatmentrequirements.Glucagoncanalsostimulateinsulinrelease froman insulinoma, resulting in severehypoglycemia;thisconditionistreatedwithglucose.

Anticholinergic AgentsAnticholinergic agents as a group are effective in tissues withreceptors supplied by cholinergic postganglionic autonomicnerves. They block the effect of acetylcholine liberated fromnerveendings.TheyreduceGItractmotility,decreasetonicityin theurinary tract, andmayalsohaveahypotonic effectonthebileducts.Theseagentsalsodecreasesalivaryandbronchialsecretions,dilatepupils,andincreaseheartrate,withthedura-tion of action and specific effect on various target organsdependentonthespecificcompoundanddose.Theiractionontonicityandmotilityissimilartoglucagon,but,unlikegluca-gon,theyalsoreducesecretions.Currentevidencesuggeststhatthelattereffectisinsignificantinimagingstudies.

SomeoftheseagentshavebeenusedinpepticulcerdiseasetherapyinconjunctionwithantacidsandH2receptorantago-nists, although these applications have been inconclusive andcontroversial.Theyalsoplayaroleinthetreatmentofirritablebowelsyndromeandhavebeenusedassupplementaltherapyin treating biliary and ureteral colic to relax smooth musclespasm;resultsherealsoareinconsistent.

Useful Agents. Themostwidelyknownanticholinergicagentisatropinesulfate.It isavailable intabletsandasaparenteralinjectable liquid. Radiologists in North America may stillremembertheanticholinergicagentpropanthelinebromide,invogue in the 1960s and early 1970s as a GI hypotonic agent.Currently,atropineandpropanthelinehavebeensupplantedbyotheragents.

In many countries, the short-acting anticholinergic agentscopolaminebutylbromideisused,butitisnotavailableintheUnited States. It is administered IV; the usual dose is 20mgbeforeanupperGIexamination. Itshypotoniceffect lasts for15 to 20 minutes. It does not induce gastroesophageal reflux,nordoesithaveanysignificanteffectonthevisualizationofahiatalhernia.ItmaybeusefulwithCTandMRIinevaluatingsuspected gastric cancers.52 One study found that it improvescolonic distention during CT colonography (compared withcontrols),andtheauthorsrecommendeditsuse.53

Pirenzepine, an antimuscarinic drug, shows promise as ahypotonicagentwithouttheadverseeffectsofscopolamine.

Although other anticholinergic agents are available, theirsideeffectsandlongerdurationofactionlimittheirapplicationinradiology.Forexample,scopolaminehydrobromideisavail-ablebut isnotused inradiologicexaminationsbecauseof itsadversesideeffects.

20 SECTION I General Radiologic Principles

andinducesgallbladdercontraction,duodenalhypoperistalsis,andhyperperistalsisofthejejunum,ileum,andcolon.Itreversesbowelaperistalsisinducedbydrugsactingonentericneuralorsmoothmuscle.

In the early 1980s, ceruletide appeared to be a promisingagent to increase small bowel peristalsis and thus shortenthe duration of a small bowel examination, but radiologistslost interest in thisagent;givenIV,ceruletide inducesnausea,vomiting, and abdominal cramps. For accelerating smallbowel transit, a usual dose of 0.25 to 0.3 µg/kg is admin-istered. Whether the shorter small bowel transit time leadsto a better small bowel study is debatable; pronounced con-tractions tend to obscure anatomic detail, particularly in thedistal ileum.

Becauseceruletide inducesgastrichypotonia, itshouldnotbeadministeredpriortosignificantamountsofbariumreach-ing the jejunum. Such gastric stasis can be overcome by alsoadministeringmetoclopramide.

DRUGS AFFECTING THE BILIARY TRACT AND PANCREAS

Bileflowintotheduodenumisregulatedbybileproductionintheliverandgallbladdertonicity.Agentsaffectingonlythelatterareconsideredinthissection.

Inhibition of gallbladder contractions can be achieved byglucagon,atropine,andothercholinergicdrugs,somatostatin,somecalciumchannelantagonists,andseveralotherlessstudieddrugs, keeping in mind that contractions are also inhibitedin obesity, diabetes mellitus, celiac disease, and autonomicneuropathy.

Gallbladder contraction is stimulated by cholecystokinin,ceruletide,motilin,prostigmine,erythromycin,andpossiblybycisaprideandcholestyramine.

Glucagon relaxes the gallbladder. Glucagon also decreasespapillaofVatermeanpressures.Currently,littleclinicalusehasbeenmadeofthesefindings.Glucagonisoflimiteduseinper-cutaneous transhepatic cholangiography or various biliarydrainage procedures. In an occasional patient with persistentdistal commonbileductnarrowing, in the sphincterofOddiregion,glucagonaidsindifferentiatingatumor,impactedstone,andspasm.Inmostpatients,however,judicioususeoffluoros-copyissufficient.

Glucagon improvesbileductvisualizationduringMRCP.59Because incomplete duct visualization may lead to a repeatstudy or even an invasive procedure such as ERCP, glucagonis routinely used at some centers. Although earlier studiessuggested that glucagon improves the quality of operativecholangiography, a double-blind prospective study found noimprovement.60 Use of fentanyl during surgery is associatedwithsphincterofOddispasmand,inthesepatients,glucagonmay have a role.

Hypotonic agents are commonly used during ERCP toinduceduodenalhypotonia,inhibitcontractionsofthesphinc-ter of Oddi, and aid ampullary cannulation. In the UnitedStates,glucagonisusedalmostexclusivelyforthispurpose;inother countries, an anticholinergic drug such as scopolamineisusedmoreoften.

Neostigmine, together with morphine, has been proposedas a provocative test in hepatobiliary scintigraphy for evalua-tion of sphincter of Oddi dyskinesia in postcholecystectomypatients.61

mayincreaseserumlevelsofprolactininpatientswithapitu-itaryprolactin-releasingtumor.Ithasalsobeenassociatedwithsuddencardiacdeath.55

CisaprideCisaprideisaprokineticsubstancethatinducesantralcontrac-tility, enhances gastric emptying, and promotes small bowelperistalsis.Italsoenhancesloweresophagealsphinctertoneandisarelativelypotentesophagealmotorstimulator. Ithasbeenproposedforthetreatmentofdiabeticpatientswithgastropa-resis and as an antigastroesophageal reflux agent. It has hadlimitedapplicationinradiology.ItsusehasbeendiscontinuedintheUnitedStatesbecauseofassociatedcardiacarrhythmiasanddeaths.56

NeostigmineNeostigmine methyl sulfate is a cholinesterase inhibitor thatpromotesgastricandsmallbowelperistalsisandleadstofastergastricemptyingandshortersmallboweltransittime.Itpro-motesperistalsiswhenactivityisdepressedbycholinergicstim-ulation. It is useful in treating colonic pseudo-obstruction(Ogilvie’ssyndrome),57butiscontraindicatedwithmechanicalbowelobstructionandinsomesettingsofadynamicileus.58Ithasledtocolonperforation.

ErythromycinErythromycin,primarilyanantibiotic,improvesgastricmotil-ityandpromotesgastricemptying.Itisusedasanaidinpost-operative gastroparesis and for the treatment of diabeticgastroparesis,buthashadlittleapplicationinradiology.

MIXED ACTION AGENTS

MorphineSomeradiologistsmayundoubtedlyrememberusingmorphinesulfateforhypotonicduodenography,aprocedurerelegatedtohistory.Currently,morphinehasaroleinnuclearmedicineandapossibleroleinMRCP(seelater).

CholecystokininCholecystokinin,apeptidehormone,hasmyriad functions; itinducesgallbladdercontractionandincreasesbowelperistalsis,resultinginfastersmallboweltransit.Italsoregulatespancre-atic enzyme secretion, inhibits gastric acid secretion, affectssatiety signaling, and acts as a neurotransmitter. It stimulatesaldosteronesecretionfromhumanadrenocorticalcells.Fromaradiologic viewpoint, cholecystokinin induces simultaneouscontractionofthesphincterofOddiandgallbladder;thus,mostof the bile from the gallbladder refluxes into the intrahepaticbileductsandreentersthegallbladderafterhormoneinfusionstops.

Secretionofcholecystokininisimpairedinceliacdiseaseandbulimianervosa.Untreatedceliacshavelowpostprandialcho-lecystokinin levels. It is overexpressed in certain neuroendo-crinetumorsandinmedullarythyroidcarcinomas.

Generally, only the COOH-terminal octapeptide of chole-cystokininisused.Thisfragmentismorepotentthantheentiremolecule.

CeruletideCeruletideisasyntheticcompoundsimilartocholecystokinininitspharmacologiceffects—namely,itdelaysgastricemptying

F

1  Imaging Contrast Agents and Pharmacoradiology 21

Ceruletide has an effect on the gallbladder similar to thatof cholecystokinin or a fatty meal.64 In patients with recur-rent symptoms after cholecystectomy, an ultrasonographi-cally detected increase in extrahepatic bile duct dilationfollowing ceruletide injection suggests sphincter of Oddidysfunction.

The cholecystokinin-secretin pancreatic exocrine functiontestisusedtodetectpancreaticexocrineinsufficiencybutdoesnot differentiate between chronic pancreatitis and pancreaticcarcinoma.MRIcanalsoevaluatepancreaticexocrinefunctionbymeasuringduodenalfillingafterstimulationwithsecretin.65Secretin improves pancreatic duct visualization and aids indetecting abnormalities during MR pancreatography.66 Pan-creasdivisumandotherabnormalitiesaremorereadilydetectedaftersecretin,potentiallyobviatingtheneedforERCP.MRpan-creatographyisbestperformedwithin5minutesaftersecretininjection.67Secretin-augmentedMRpancreatographyandMRIperfusionareuseful fordetectinggraftdysfunctionafterpan-creatictransplantation.68

Analysis of duodenal aspirations of pancreatic juice aftercholecystokinin-octapeptidestimulationcandetectpancreaticinsufficiencyinpatientswithchronicpancreatitis.However,theanalogoussecretintestismorecommonlyused.

MorphinemayhavearoleinMRcholangiography;IVmor-phine constricts the sphincter of Oddi, thus distending thebiliaryandpancreaticducts.62Itmayalsoimproveductvisual-ization inprimary sclerosingcholangitis.An intervalof10 to20minutesbetweenmorphine injectionand imagingappearsuseful.

Cholecystokinin’seffectonthegallbladderhasbeenusedtoincreaseradiographiccontrastduringoralcholecystography;italso aids in evaluating gallbladder function. The gallbladderejection fraction is typically determined with hepatobiliaryscintigraphybut lessoftenwithultrasonography;MRcholan-giography uses an infusion of cholecystokinin as an alternatemethod.

A provocative test using a cholecystokinin derivative (Sin-calide) to identify acalculous cholecystitis patients likely tobenefitfromcholecystectomyhasnotachievedgeneralclinicalacceptance.CholecystokininrelaxesthesphincterofOddiandappearstoassistinthepassageofbileductstones.Ontheotherhand,acholecystokininreceptorantagonistprovidespainreliefin patients with biliary colic.63 Interestingly, cholecystokinindoes not induce its usual sphincter of Oddi inhibitory effectafteracholecystectomy.Cholecystokininisalsousefulasadiag-nostictestofpancreaticfunction.

REFERENCES

1. Krause W, Schneider PW: Chemistry of x-raycontrast agents. In Majoral J-P, editor: Topicsin Current Chemistry, vol 222, Heidelberg,Germany,2002,Springer-Verlag.

2. American College of Radiology: Manual onContrastMedia,ed4,Reston,VA,1998,Ameri-canCollegeofRadiology.

3. European Society of Urogenital Radiology:Guidelines on contrast media, version 4.0,2004. http://www.esur.org.

4. Schabelman E,Witting M: The relationship ofradiocontrast, iodine, and seafood allergies: Amedical myth exposed. J Emerg Med 39:701–707,2010.

5. LawrenceV,MatthaiW,HartmaierS:Compara-tivesafetyofhigh-osmolalityandlow-osmolalityradiographiccontrastagents.Reportofamul-tidisciplinaryworkinggroup. InvestRadiol27:2–28,1992.

6. Park SW, Bae IY, Eun HW, etal: Small-bowelangioedemaduringscreeningcomputedtomographyduetointravascularcontrastmate-rial. J Comput Assist Tomogr 35:549–552,2011.

7. Mikkonen R, Aronen HJ, Kivisaari L, etal:Plasma levels of prekallikrein, alpha-2-macro-globulin and C1-esterase inhibitor in patientswithurticarialreactiontocontrastmedia.ActaRadiol38:466–473,1997.

8. LasserEC,BerryCC,TalnerLB,etal:Pretreat-mentwithcorticosteroidstoalleviatereactionstointravenouscontrastmaterial.NEnglJMed317:845–849,1987.

9. Pagani JJ, Hayman LA, Bigelow RH, etal:Diazepam prophylaxis of contrast media–inducedseizuresduringcomputedtomographyofpatientswithbrainmetastases.AJR140:67–72,1983.

10. Gleedon TG, Bulugahapitiya S: Contrast-induced nephropathy [review]. AJR 183:1673–1689,2004.

11. GoergenSK,RumboldG,ComptonG,HarrisC:Systematic review of current guidelines, and

their evidence base, on risk of lactic acidosisafter administration of contrast medium forpatients receiving metformin. Radiology 254:261–269,2010.

12. MertenGJ,BurgessWP,GrayLV,etal:Preven-tion of contrast-induced nephropathy withsodiumbicarbonate:Arandomizedtrial.JAMA291:2328–2334,2004.

13. Jang JS: Sodium bicarbonate therapy for theprevention of contrast-induced acute kidneyfailure.CircJ76:2255–2265,2012.

14. KramerBK,KamerlM,SchwedaF,SchreiberM:A primer in radiocontrast-induced nephropa-thy. Nephrol Dial Transplant 14:2830–2834,1999.

15. MuellerC:Preventionofcontrastnephropathyincriticallyillpatientsusingacetylcysteineandtheophylline. Internat J Artificial Organs 27:1066–1069,2004.

16. Kelly AM, Dwamena B, Cronin P, etal:Effectiveness of drugs for preventing contrast-induced nephropathy. Ann Intern Med 148:284–294,2008.

17. Weisbord SD, Palevsky PM: Strategies for theprevention of contrast-induced acute kidneyinjury. Curr Opin Nephrol Hypertens 19:539–549,2010.

18. MarenziG,MaranaI,LauriG,etal:Thepreven-tionofradiocontrastagent–inducednephropa-thy by hemofiltration. N Engl J Med 349:1333–1340,2003.

19. KurokohchiK,MasakiT,MiyauchiY,etal:Effi-cacyofcombinationtherapiesofpercutaneousor laparoscopic ethanol-lipiodol injection andradiofrequency ablation. Int J Oncol 25:1737–1743,2004.

20. SkucasJ:Bariumsulfate:Clinicalapplication.InSkucasJ,editor:RadiographicContrastAgents,ed2,Rockville,MD,1989,Aspen,pp14–17.

21. BueckerA,WeinBB,Neuerburg JM,GuentherRW: Esophageal perforation: Comparison ofuseofaqueousandbarium-containingcontrastmedia.Radiology202:683–686,1997.

22. NodaY,OgawaY,NishiokaA,etal:Newbariumpastemixtureforhelical(slip-ring)CTevalua-tionoftheesophagus.JComputAssistTomogr20:773–776,1996.

23. Quagliano PV, Austin RF, Jr: Oral contrastagents for CT: A taste test survey. J ComputAssistTomogr21:720–722,1997.

24. GrandDJ,BelandMD,MachanJT,Mayo-SmithWW:DetectionofCrohn’sdisease:Comparisonof CT and MR enterography without anti-peristaltic agents performed on the same day.EurJRadiol81:1735,1741,2012.

25. OgawaY,NodaY,NishiokaA,etal:Newbariumpastemixtureforhelical(slip-ring)CTevalua-tion of rectal carcinoma. J Comput AssistTomogr21:398–401,1997.

26. SahaniDV,JhaveriKS,D’SouzaRV,etal:Evalu-ationofsimethicone-coatedcelluloseasanega-tiveoralcontrastagentforabdominalCT.AcadRadiol10:491–496,2003.

27. Skucas J:Anaphylactoid reactions with gastro-intestinal contrast media. AJR 168:962–964,1997.

28. BlakeboroughA, Sheridan MB, ChapmanAH:Complicationsofbariumenemaexaminations:AsurveyofUKConsultantRadiologists1992to1994.ClinRadiol52:142–148,1997.

29. ZalevAH:Venousbariumembolization,arare,potentiallyfatalcomplicationofbariumenema:2casereports.CanAssocRadiolJ48:323–326,1997.

30. MurphyKD,PosterRB,MarxWH,etal:Uppergastrointestinal examination complicated byvenousintravasationandportalveinthrombo-sis.AJR169:501–503,1997.

31. ParazellaMA:Currentstatusofgadoliniumtox-icityinpatientswithkidneydisease.ClinJAmSocNephrol4:461–469,2009.

32. ManfrediR,MarescaG,BaronRL,etal:DelayedMR imaging of hepatocellular carcinoma en-hanced by gadobenate dimeglumine (Gd-BOPTA). J Magn Reson Imaging 9:704–710,1999.

22 SECTION I General Radiologic Principles

33. JoarderR,deJodeM,LambGA,GedroycWM:The value of MnDPDP enhancement duringMR guided laser interstitial thermoablation ofliver tumors. JMagnResonImaging13:37–41,2001.

34. Harisinghani MG, Dixon WT, Saksena MA,etal:MRlymphangiography:Imagingstrategiesto optimize the imaging of lymph nodes withferumoxtran-10. Radiographics 24:867–878,2004.

35. WenX,JacksonEF,PriceRE,etal:Synthesisandcharacterizationofpoly(L-glutamicacid)gado-liniumchelate:AnewbiodegradableMRIcon-trast agent. Bioconjugate Chem 15:1408–1415,2004.

36. Lin CY, Yadav NN, Ratnakar J, etal: In vivoimaging of paraCEST agents using frequencylabeled exchange transfer MRI. Magn ResonMed71:286–293,2014.

37. Shi H, Liu C, Ding HY, Li CW: Magnesiumsulfateasanoralcontrastmediuminmagneticresonanceimagingofthesmallintestine.EurJRadiol81:e370–e375,2012.

38. Luboldt W, Bauerfeind P, Wildermuth S, etal:Colonicmasses:DetectionwithMRcolonogra-phy.Radiology216:383–388,2000.

39. Pilkington SA, Nugent KP, Brenner J, etal:Barium proctography vs magnetic resonanceproctographyforpelvicfloordisorders.Colorec-talDis14:1224–1230,2012.

40. TaniokaH,ArakiT,SasakiY,etal:Famotidinefor gastric radiography. Radiat Med 11:12–16,1993.

41. HuwerH,WinningJ,StraubU,etal:Clinicallydiagnosed nonocclusive mesenteric ischemiaafter cardiopulmonary bypass: Retrospectivestudy.Vascular12:114–120,2004.

42. Stoeckelhuber BM, Suttmann I, StoeckelhuberM, Kueffer G: Comparison of the vasodilatingeffectofnitroglycerin,verapamil,andtolazolinein hand angiography. J Vasc Intervent Radiol14:749–754,2003.

43. Chamsuddin AA, Kowalik KJ, Bjarnason H,etal:Usingadopaminetype1Areceptoragonistin high-risk patients to ameliorate contrast-associatednephropathy.AJR179:591–596,2002.

44. YamagamiT,NakamuraT, IidaS, etal:Effectsof prostaglandin E(1) injection through thesuperiormesentericarteryonthehemodynam-

ics of hepatocellular carcinoma. AJR 178:349–352,2002.

45. Yamagami T, Nakamura T, Sato O, etal:Valueof intra-arterial prostaglandin E(1) injectionduringCThepaticarteriography.AJR177:115–119,2001.

46. Miller RE, Chernish SM, Brunelle RL, etal:Double-blind radiographic study of doseresponsetointravenousglucagonforhypotonicduodenography.Radiology127:55–59,1978.

47. TibblingL,BjorkhoelA,JanssonE,StenkvistM:Effect of spasmolytic drugs on esophagealforeignbodies.Dysphagia10:126–127,1995.

48. Ratcliffe JF:Glucagon inbariumexaminationsin infants and children: Special reference todosage.BrJRadiol53:860–862,1980.

49. Morrin MM, Farrell RJ, Keogan MT, etal: CTcolonography:Colonicdistentionimprovedbydualpositioningbutnotintravenousglucagon.EurRadiol12:525–530,2002.

50. LowRN,FrancisIR:MRimagingofthegastro-intestinaltractwithIVgadoliniumanddilutedbarium oral contrast media compared withunenhanced MR imaging and CT. AJR 169:1051–1059,1997.

51. Chernish SM, Maglinte DDT: Glucagon:Commonuntowardreactions—reviewandrec-ommendations.Radiology177:145–146,1990.

52. SohnKM,LeeJM,LeeSY,etal:ComparingMRimagingandCTinthestagingofgastriccarci-noma.AJR174:1551–1557,2000.

53. TaylorSA,HalliganS,GohV,etal:Optimizingcolonic distention for multi-detector row CTcolonography:Effectofhyoscinebutylbromideand rectal balloon catheter. Radiology 229:99–108,2003.

54. Morewood DJW,Whitehouse GH:A compari-son of three methods for performing bariumfollow-throughstudiesofthesmallintestine.BrJRadiol59:971–973,1986.

55. Michaud V, Turgeon J: Domperidone andsuddencardiacdeath:Howmuchlongershouldwe wait? J Cardiovasc Pharmacol 61:215–217,2013.

56. QuigleyEM:Cisapride:Whatcanwelearnfromthe rise and fall of a prokinetic? J Dig Dis 12:147–156,2011.

57. Eisner JL, Smith JM, Ensor CR: Intravenousneostigmine for postoperative acute colonic

pseudo-obstruction. Ann Pharmacother 46:430–435,2012.

58. StJohnPH,RadcliffeAG:Contraindicationforthe use of neostigmine in colonic pseudo-obstruction [letter]. Brit J Surg 84:1481–1482,1997.

59. Dalal PU, Howlett DC, Sallomi DF, etal:Does intravenous glucagon improve commonbile duct visualisation during magnetic reso-nancecholangiopancreatography?Resultsin42patients.EurJRadiol49:258–261,2004.

60. CoferJB,BarnettRM,MajorGR,etal:Effectofintravenousglucagononintraoperativecholan-giography.SouthMedJ81:455–456,1988.

61. Madacsy L,Velosy B, Lonovics J, etal: Evalua-tionofresultsoftheprostigmine-morphinetestwith quantitative hepatobiliary scintigraphy:AnewmethodforthediagnosisofsphincterofOddi dyskinesia. Eur J Nucl Med 22:227–232,1995.

62. SilvaAC,FrieseJL,HaraAK,LiuPT:MRchol-angiopancreatography:Improvedductaldisten-tionwithintravenousmorphineadministration.Radiographics24:677–687,2004.

63. Malesci A, Pezzilli R, D’Amato M, Rovati L:CCK-1 receptor blockade for treatment ofbiliarycolic:Apilot study.AlimentPharmacolTher18:333–337,2003.

64. MuracaM,CianciV,VileiMT,etal:Ultrasonicevaluation of gallbladder emptying with ceru-letide.ItalJGastroenterol28:38–39,1996.

65. HaverhagenJT,MullerD,BattmannA,etal:MRhydrometry to assess exocrine function of thepancreas:Initialresultsofnoninvasivequantifi-cationofsecretion.Radiology218:61–67,2001.

66. Monill J, Pernas J, Clavero J, etal: Pancreaticduct after pancreatoduodenectomy: Morpho-logic and functional evaluation with secretin-stimulated MR pancreatography. AJR 183:1267–1274,2004.

67. Fukukura Y, Fujiyoshi F, Sasaki M, Nakajo M:Pancreatic duct: Morphologic evaluation withMR cholangiopancreatography after secretinstimulation.Radiology222:674–680,2002.

68. Heverhagen JT, Wagner HJ, Ebel H, etal:Pancreatic transplants: Noninvasive evaluationwith secretin-augmented MR pancreatographyandMRperfusionmeasurements—preliminaryresults.Radiology233:273–280,2004.

F

23

2 Barium Studies: Single and Double ContrastMARC S. LEVINE | DAVID J. OTT | IGOR LAUFER

CHAPTER OUTLINE

Single-Contrast StudiesDiagnostic PrinciplesEquipmentBarium SuspensionsQuality ControlsEsophagographyUpper Gastrointestinal SeriesSmall BowelBarium Enema

Double-Contrast StudiesPerformanceInterpretationArtifacts

In the past, there was considerable controversy about therelative virtues of single-contrast and double-contrast tech-niques.4,5Currently,however,mostauthorsbelievethatdouble-contrasttechniquesprovidesuperiormucosaldetailandallowearlier detection of subtle lesions than single-contrast tech-niques.As a result, it is generally recommended that double-contrast studies be performed on patients who are youngenoughandhealthyenoughtoundergothistypeofexamina-tion. In contrast, single-contrast barium studies are mostappropriate inolderordebilitatedpatientswhoareunabletocooperateforadouble-contrastexamination.6,7

This chapter discusses the principles for performing andinterpretingsingle-anddouble-contrastbariumstudies.8TheseprinciplesareillustratedwithexamplesdrawnfromthroughouttheGItract.

Single-Contrast StudiesDIAGNOSTIC PRINCIPLES

Dependingontheorganexamined,single-contrasttechniquesmay include observation of function (e.g., pharyngeal andesophagealmotility),compressionimaging,full-columndisten-tion, mucosal relief views, and limited air-contrast images.9-11Theuseofcompressionduringfluoroscopyisacriticalcompo-nent of the single-contrast examination. Small lesions (e.g.,smallulcers,polypoidneoplasms)areoftenvisibleonlywhenthebariumpoolisadequatelythinnedordisplacedbymanualcompression.Thebariumsuspensionmustalsobeadequatelydilutediflesions(especiallysmalllesions)aretobedetectedinthethinned-outbariumpool.

Full-column distention of the lumen is ideal for showingstrictures,largeneoplasms,andlesionsprojectingtangentially,suchasulcersordiverticula.Full-columnviewsofbarium-filledstructuresinvariousprojectionsenabledepictionoflargeulcersand tumors. However, small lesions may be visible on full-column views only when viewed in profile. In such cases,mucosal techniques are required to supplement the barium-filledviews.

EQUIPMENT

Fluoroscopic equipment has evolved dramatically with thetransitionfromanaloguecassette-basedradiographytodigitalimagingandviewingonpicturearchivingandcommunicationssystem (PACS) workstations.12-15 Regardless of the imagingtechnology, single-contrast examinations can be performedwithconventionalorremotecontrolfluoroscopicunits.12Theabilitytoobtainoptimalcompressionviewsisamajorprereq-uisite of any well-designed fluoroscope. Compression can beperformedmanuallyorwithavarietyofhand-helddeviceson

Sincethe1980s,advancesincross-sectionalimagingandendos-copyhaveledtoagradualbutsteadydeclineinthenumberofbarium studies performed in the United States.1,2 Althoughbarium studies no longer reign supreme in the diagnosis ofgastrointestinal(GI)disease,single-anddouble-contrastexam-inationscontinuetohavearoleinmodernradiologypractice.In general terms, barium studies can demonstrate GI abnor-malitiesinthreeways:

1. Mucosalreliefviewsofthecollapsedorpartiallycollapsedlumen obtained with a small volume of barium. Theseviewsenablevisualizationofthefoldsinvariousportionsof the GI tract (Fig. 2-1A). Because the folds contain asubmucosal core, these views are particularly useful forshowingabnormalitiesinvolvingthesubmucosa,suchasesophagealvarices.

2. Single-contrastviewsofthefilledlumenobtainedwithalarge volume of low-density barium (Fig. 2-1B). Theseviewsenablevisualizationofcontourabnormalities,stric-tures,andlargepolypoiddefects.

3. Double-contrastviewsobtainedafterthemucosalsurfacehasbeencoatedwithathinlayerofhigh-densitybariumandthe lumenhasbeendistendedwithgas(Fig.2-1C).Theseviewsenablevisualizationofsubtlemucosallesions,suchastheearlychangesofinflammatoryboweldiseaseandearlyneoplasticlesions.

Although these three types of views are incorporated tovaryingdegreesinbothsingle-anddouble-contrastexamina-tions, single-contrast studies rely more heavily on diagnosticfluoroscopy, mucosal relief, and barium filling,3 whereasdouble-contraststudiesemphasizetheinterpretationofdouble-contrast images supplemented by barium filling and mucosalrelief.

24 SECTION I General Radiologic Principles

which compression views, mucosal relief views, and limiteddouble-contrast views are required. A standard peroral smallbowel follow-through study can be performed with the samebarium suspension used for the upper GI series. A single-contrastenteroclysisstudyrequiresa15%to20%w/vbariumsuspension, although somewhat denser solutions have beenrecommended.18,19Finally,a15%to20%w/vbariumsuspen-sion is used for a single-contrast barium enema, because thisis the optimal suspension for obtaining compression viewsof the colon.

QUALITY CONTROLS

Qualitycontrol for single-contrastexaminations requiresbal-ancingthebariumdensity,kilovoltage,andwidthofthebariumcolumntoachieveadequate translucencyof thebarium-filledbowel. This permits radiographic penetration of the bariumsuspensiontovisualizelesionsthatmightotherwisebeobscuredbybariuminthelumen.6,10,11Thevisibilityofskeletalshadowsthroughthebariumcolumnindicatesthatsmallfillingdefectsaremorelikelytobeseen,whichisparticularlyimportantforthedetectionofcolonicpolyps.

Anotherqualitycontrolconsideration,especiallyduringthebariumenemaandsmallbowelfollow-through,istheabilitytoseethroughoverlappingloopsofbowel.6,20Onbariumenemas,a tortuous sigmoid colon may have overlapping loops, evenwith appropriate compression. Similarly, on small bowel

aconventionalunit;alternatively, theplasticconeonthespotimage device of the fluoroscope can be used for applyingcompression.

On remote control equipment, compression is easily per-formed with a vertical movable device incorporated into themachine, facilitated by the ability to angle the x-ray tube.Remotecontrolunitsoftencontainacompressiondevicethatallowsgradedcompressionandtubeangulationwithcompres-sion. Nevertheless, many radiologists prefer standard fluoro-scopic units rather than remote control units because it isgenerallyeasiertomove,turn,andcompressthepatientwhenthefluoroscopistperformingtheprocedureisatthetableside.

BARIUM SUSPENSIONS

Numerous barium products are available commercially; anumber are formulated for specific purposes, whereas otherscanbeusedforavarietyofexaminations.16,17Bariumsuspen-sionsforsingle-contraststudiesshouldbeofmoderatedensity(50%-100% w/v) when not diluted. The optimal bariumsuspension for a particular study depends on the structurebeingexaminedandthetypeofexaminationbeingperformed.For example, an esophagogram requires a moderately densebarium suspension that provides full-column and mucosalrelief imaging; a high-density barium suspension or pastemay also be needed for optimal mucosal coating. A similarbarium suspension can be used for an upper GI series, in

Figure 2-1  Three types of views for visualizing the gastrointestinal tract, as illustrated in the esophagus. A. Mucosal relief view. With the esophagus collapsed and coated, the normal longitudinal folds are seen. B. Single-contrast view. With the patient continuously drinking barium in the prone position, the barium-filled esophagus is demonstrated.  C. Double-contrast view. With the patient in the upright position, the smooth, featureless surface of the esophagus is seen. 

A B C

F

2  Barium Studies: Single and Double Contrast 25

Full-column images of the esophagus are usually obtainedwith the patient on the fluoroscopic table in the prone, rightanterior oblique position; a bolster may be used to increaseintra-abdominal pressure. Esophageal peristalsis is inhibitedby rapid swallowing of barium, allowing the esophagus todistend fully. Multiple images of the esophagus should beobtained at all levels; these images may be full-length viewsof the esophagus or coned-down views at different levels,depending on the imaging options of the fluoroscopicequipment.

Maximaldistentionoftheesophagogastricregionisrequiredfor optimal detection of hiatal hernias and lower esophagealrings(Fig.2-3).22,24Rapid ingestionof thebariumsuspensionfollowedbydeepinspiration(oraValsalvamaneuver)promotesdistention of the gastroesophageal junction. Careful fluoro-scopic observation is required to visualize lesions only seenwhenthisregionisoptimallydistended.

Full-column views are complemented by mucosal reliefviewsofthecollapsedesophagus,withcoatingofthelongitu-dinalfoldsbythebariumsuspension.23Ahigh-densitybariumsuspension (e.g., that used for a double-contrast upper GIseries)isidealforthispurpose.Thepatienttakesoneorseveralswallows of high-density barium to coat the esophageal foldswithbarium.Thesemucosalreliefviewsmayrevealthickened,irregularfolds,smallesophagealneoplasms,andrefluxorinfec-tiousesophagitis(Fig.2-4).Nevertheless,double-contrastviewsarebetterforshowingtheplaquesofCandidaesophagitis,thesmallulcersofherpesesophagitis,andthegiantulcersofcyto-megalovirus(CMV)orhumanimmunodeficiencyvirus(HIV)esophagitis.25

Single-contrastmucosalreliefviewsarealsobestfordetect-ing esophageal varices.23,26 The patient takes several swallowsof the barium suspension, which coats the lower esophagus,and is then asked not to swallow to inhibit peristalsis. Inter-mittent fluoroscopic observation is performed for severalminutestovisualizethevaricesoptimallyastheybecomemoredistended.

Fluoroscopic observation is an integral part of the radio-graphicevaluationoftheesophagusandisusuallyadequatetoassess esophageal function.9,23 Motion recording methodsgreatly aid in evaluating oropharyngeal swallowing disordersbecauseoftherapideventsthatoccurwithdeglutition,andmayalso be used to assess esophageal motility. Motion recordingsshould be obtained while the patient takes multiple discreteswallows of barium, because rapid swallowing causes reflexinhibitionofesophagealperistalsis.Withtheuseofsingleswal-lows, the single-contrast esophagogram is an excellent tech-niqueforevaluatingesophagealmotility.9

UPPER GASTROINTESTINAL SERIES

Thesingle-contrastupperGIseries isacomplexexaminationrequiring fluoroscopic observation, abdominal compression,andtheuseofmultipletechniquesforexaminingtheesopha-gus,stomach,andduodenum.11,27Thestudycanbeperformedquicklyandistoleratedevenbypatientswhoareimmobileorunabletocooperatefully.

The examination starts with the table upright. After thepatientingestsseveralswallowsofbarium,thestomachiscom-pressedwithacompressionpaddleortheconeonthefluoro-scopetodemonstratetherugalfolds,assessthepliabilityofthegastric wall, and detect focal areas of rigidity secondary to

follow-throughs, overlapping small bowel loops in the pelvismay compromise the fluoroscopist’s ability to detect abnor-malities in thedistal ileum.Insuchcases,pelvic loopscanbebettervisualizedbyplacingthepatientinapronepositionwhileheorsheliesonabolsterorinflatedballoontolifttheseloopssuperiorlyfromthepelvis.21

ESOPHAGOGRAPHY

Routinesingle-contrastesophagography includesfluoroscopicobservation of the esophagus supplemented by motionrecordings, full-column views, and mucosal relief views.22,23Full-column (barium-filled views) and mucosal relief viewsconstitutethesingle-contrastphaseoftheexamination.Motionrecordingscanbeusedtodocumentpharyngealfunctionandesophagealmotilityusinganalogordigitalrecordingsorrapidsequence solid-state recordings built into modern digitalfluoroscopes.12-15

Depending on the imaging options of the fluoroscopicequipment, full-column technique is performed by obtainingpartial or full-length views of the esophagus distended withbarium.Theseviewsallowdetectionofesophagealcarcinoma(Fig.2-2)andotherabnormalitiesatthegastroesophagealjunc-tionsuchashiatalhernias,pepticstrictures,andloweresopha-gealrings.22Loweresophagealringsarebestvisualizedonproneviewsofthebarium-filledloweresophagus,sometimessupple-mented with a solid bolus such as a marshmallow or bariumtablet.22-24

Figure 2-2  Annular carcinoma of the midesophagus. The lesion is well shown on the full-column portion of the barium esophagogram. 

26 SECTION I General Radiologic Principles

SMALL BOWEL

Thesmallbowelcanbeexaminedbysingle-contrasttechniquewithasmallbowelfollow-through,enteroclysis,orretrogradestudy via an ostomy or reflux from the colon.18,19,21A peroralsmall bowel follow-through may be performed after a single-contrastupperGIseriesorasaseparatestudy.Alargevolume(≥500mL)ofbariumisrecommendedtopromotegastricemp-tying, accelerate small bowel transit, and optimally distendsmall bowel loops. Fluoroscopic imaging and compression ofallsmallbowelloopsisacriticalcomponentoftheexaminationbecausefocal lesionsareeasilyobscuredbyoverlapping loopsofsmallbowelunlesscompressionisappliedtoseparatetheseloopsandensureanadequateexamination.

Asmallbowelfollow-throughtypicallyrequiresaminimumof 500 mL of orally ingested barium, which can be the sameproductusedforasingle-contrastupperGIseries.Proneover-headorlow-magnificationdigitalimagesofthesmallintestineare taken at timed intervals (e.g., every 30 minutes) untilbarium fills the right side of the colon. Depending on thebariumproductused, transit timethroughthesmallbowel istypically 60 to 90 minutes. Fluoroscopic spot imaging withmanualcompressionshouldbeperformedatregularintervalsduringtheexamination.Compressionofall loopsandappro-priateimagingoftheentiresmallbowelisrequiredtooptimizedetection of abnormalities (Fig. 2-8). When the barium sus-pensionhas reached the colon, compressionviewsof the ter-minalileumareobtained(typicallywiththepatientinasupineorleftposteriorobliqueposition)foroptimalvisualizationofthis region.

tumor or scarring. If the stomach empties, the barium-filledduodenal bulb can also be examined by compression in theuprightposition.

The table is then lowered to the horizontal position, andmucosalreliefviewsofthestomachareobtained,firstwiththepatient in a supine position and then after the patient turnsinto a prone position. These views supplement the uprightcompressionviewsandcansometimesshowsmallpolyps,ero-sions, or ulcers that are later obscured in the barium-filledstomach(Fig.2-5).4,28,29

With thepatient in theprone, rightanteriorobliqueposi-tion,theesophagusisthenexaminedusingthesametechniquesdescribed previously (see earlier, “Esophagography”). Thegastricantrumandduodenalbulbusuallyfillwithbariuminthisposition, sopronecompressionviewsof theantrumandduodenalbulbshouldbeobtainedusinganinflatableballoonplacedbeneaththepatienttothinoutthebariumpool,enablingdetectionofanteriorwalllesionsintheantrumandduodenalbulb(Fig.2-6).27

With thepatient in the supine, leftposteriorobliqueposi-tion, air in the stomach rises into the gastric antrum andduodenalbulb,solimiteddouble-contrastimagesoftheseareascan be obtained (Fig. 2-7). Compression can be used to dis-place the barium suspension, separate antroduodenal struc-tures,andimprovedistentionwithair.Thefluoroscopicportionof the examination is completed at this time. Some radiolo-gists may then choose to obtain a standard set of overheadradiographs of the stomach and duodenum with the patientin prone, supine, right anterior oblique, and right lateralpositions.

Figure 2-3  Hiatal hernia and lower esophageal ring seen only on prone single-contrast esophagogram. A. Upright double-contrast view of  the esophagus shows no abnormalities. B. Prone full-column view of the esophagogastric region in the same patient shows a hiatal hernia (hh) and a widely patent lower esophageal mucosal ring (arrows). 

A B

hh

F

2  Barium Studies: Single and Double Contrast 27

Figure 2-4  Reflux esophagitis on mucosal relief view of the esophagus. Mucosal relief view from single-contrast esophagogram shows crenulated, irregular folds in the distal esophagus, suggesting esophagitis. Reflux esophagitis was confirmed at endoscopy. 

Figure 2-5  Recumbent compression views for detecting lesions in the stomach. A. Compression view of the gastric antrum shows antral erosions as multiple tiny nodules containing punctate collections of barium. B. In another patient, compression view of the antrum shows several small polyps (P) as filling defects in the barium pool. The findings in A and B were not well shown on double-contrast views but were confirmed at endoscopy. 

A B

PP

Clear delineation of pelvic loops of small bowel, whichoften overlap, is not always possible, but several maneuversmay be performed to improve visualization of these loops.21First,thepatientshouldbetoldnottovoidduringtheexami-nationbecauseafullbladderelevatespelvicsmallbowelloops,enabling them to be separated with manual compression. A

similareffectmaybeachievedbyinstillingairintotherectum.ThepatientcanalsobeplacedinaproneTrendelenburgposi-tionwithaninflatedballoonbeneaththelowerabdomenandpelvistodisplacepelvicsmallbowelloopssuperiorlyforbettervisualization of these loops. If a remote control unit is avail-able,thetubecanbeangulatedtofurtherseparatetheseloops(Fig. 2-9).

EnteroclysisAn intubation small bowel study (enteroclysis) may be per-formed via a tube placed in the duodenum or jejunum.19,21,30Several enteroclysis catheters are available commercially. Thepatient can be intubated via the mouth or the nose, witheach approach having advantages and drawbacks. Whensingle-contrast enteroclysis is performed, jejunal intubationis preferred to prevent duodenogastric reflux and emesis ofbarium.

Approximately800mLofa15%to20%w/vbariumsuspen-sionisplacedinanenemabag,whichishungonanadjustableverticalstandorIVpole.Awater-solublecontrastagentmaybeaddedtostimulateintestinalperistalsisandshortenthelengthoftheexamination.16,17,21Thebariumsuspensionisallowedtoflowthroughthetubebygravity,andtherateofflowisregu-latedbyadjustingtheheightof theenemabag. If thebariumsuspensionflowstooslowly,adequatedistentionisnotachieved.Conversely, rapid flow rates may cause reflex paralysis of thesmallintestine,withslowtransitandexcessiveduodenogastricreflux.Initially,theenemabagisplacedabout2feetabovethetable;thebagcanberaisedorloweredduringtheexaminationto adjust the flow rate for optimal distention of small bowelloops(Fig.2-10).

The examination is performed under fluoroscopic guid-ance with the patient in a supine position. Careful compres-sion spot images of all loops of small intestine are obtainedunder fluoroscopic guidance as bowel segments become fullydistended to depict even subtle abnormalities (Fig. 2-11).When the entire small intestine has been opacified, overheadradiographs or low-magnification digital images of the smallbowel are obtained; the patient can then be placed in aprone position to aid in separating small bowel loops in thepelvis.

28 SECTION I General Radiologic Principles

and for evaluation of inflammatory bowel disease,6,20,32 thesingle-contrast barium enema can be performed quickly andisusuallyabetterchoiceforpatientswhoareimmobile,older,or incontinent.6,7

Preparationofthelargebowelisthemostimportantprereq-uisite for an accurate single- or double-contrast bariumenema.4,27Thediagnosisofneoplasms,includingsmallpolyps,

BARIUM ENEMA

Fluoroscopic observation, careful imaging with graded com-pression, and knowledge of appropriate technical factorsenables detection of a variety of lesions in the barium-filled colon on single-contrast examinations.20,31 Althoughsingle-contrastbariumenemasare less sensitive thandouble-contrast examinations for detecting small polypoid lesions

Figure 2-6  Prone compression view of the duodenal bulb showing an anterior wall ulcer. A. Prone view of the duodenal bulb (a balloon paddle compression device was used) shows an anterior wall ulcer (U) with surrounding edema.  B. Supine oblique air-contrast view of the duodenal bulb in  the same patient shows a ring shadow (R) because of barium coating the rim of the unfilled anterior wall ulcer crater. 

A B

U

R

Figure 2-7  Posterior wall duodenal ulcer on air-contrast view of the bulb. Supine oblique air-contrast view of the duodenal bulb with compression shows a small posterior wall ulcer (arrow), emphasizing the importance of obtaining limited double-contrast views as part of  a thorough single-contrast upper gastrointestinal examination. 

Figure 2-8  Compression spot image of the small bowel showing a Meckel’s diverticulum. A peroral small bowel follow-through study was performed in a patient with gastrointestinal bleeding. Compression spot image of the right lower quadrant shows a Meckel’s diverticulum (M) as the cause of the patient’s bleeding.  The diverticulum was removed at surgery. 

M

F

2  Barium Studies: Single and Double Contrast 29

iseasierandmorereliableinawell-cleansedcolon.Conversely,thepresenceof stool invariably limits thedetectionofpolypsandisthemostcommoncauseoferrorswheninterpretingtheimages.4,31

Avarietyofcolon-cleansingprotocolscanbeusedtoobtaina thoroughly clean colon in the vast majority of patients.31,33One recommended bowel preparation regimen includes thefollowing:

1. A24-hourclearliquiddiet2. One glass of water hourly the day before the

examination3. Asalinecathartic suchasmagnesiumcitrateat4:00PM

thedaybeforetheexamination4. 60mLofaflavoredcastoroilorother irritantcathartic

at8:00PMthedaybeforetheexamination5. An optional 1500-mL tap water cleansing enema the

morningofthebariumenemaexamination,althoughtheneedforawaterenemaiscontroversial34

If a tap water enema is administered, the patient needs towait at least 30 minutes before the single-contrast bariumenema is performed to avoid excess fluid in the colon, whichmight further dilute the barium suspension and degrade thequalityofthestudy.31,33,35

A thorough examination protocol must be followed toensure that an adequate single-contrast barium enema isobtained.6,20,31 All portions of the colon must be adequatelyvisualizedandimagedwithoutoverlappingsegmentstoincreasethe fluoroscopist’s confidence that suspected lesions are real

Figure 2-9  Peroral small bowel follow-through study in a posthysterectomy patient with pelvic small bowel loops lying deep within the pelvis. A prone image of the pelvis with a bolster placed beneath the patient and x-ray tube angulation clearly shows the cecum (C) and ileocecal junction. Note how pelvic loops of ileum are well separated and visualized. 

C

Figure 2-10  Single-contrast enteroclysis. Normal small bowel loops are well distended, with the folds in a parallel arrangement. The use of a dilute barium suspension permits a see-through effect for visualization of overlapping loops of small bowel. 

Figure 2-11  Nonobstructing adhesions on single-contrast enteroclysis with compression. Compression view of the mid small bowel shows focal nonobstructing adhesions (arrows) with angulation of the affected bowel and an inability to separate adjacent loops. 

30 SECTION I General Radiologic Principles

Apostevacuationradiographisgenerallyobtainedattheendofthestudytodocumentcolonicemptyingandruleoutgrosscolonic dysmotility. The postevacuation radiograph may alsoshow that a filling defect seen on earlier barium-filled viewspersistsordisappears, thereby indicatingwhether thisfindingwascausedbyatruepolyporresidualstool(Fig.2-14).

Double-Contrast StudiesPERFORMANCE

The yield of diagnostic information from double-contraststudies can be maximized only with meticulous attention tothetechnicalaspectsoftheexamination.Themajorprinciplesof performance include mucosal coating, distention, andprojection.

Mucosal CoatingThediagnosticqualityofdouble-contrast studiesdependsonthequalityofmucosalcoating.Intheabsenceofgoodcoating,lesions can be missed or patchy coating can be mistaken fora lesion. Good mucosal coating requires optimal interactionbetween the barium suspension and mucosal surface. Anappropriate barium suspension must be chosen; it must beprepared properly,36 and the mucosal surface must be cleanenough to enable adequate coating. Even when the mucosalcoating is only slightly impaired, major abnormalities can bemissed (Fig. 2-15).

(Fig.2-12).Asinthesmallbowelfollow-through,manualcom-pressionofthecolonisacriticalcomponentoftheexaminationbecausepolypsandpolypoidcancersprotrudingintothelumenmayonlybevisualizedwithadequate thinningof thebariumcolumn(Fig.2-13).20,31Overheadradiographsmaybeobtainedat the end of the examination, followed by a postevacuationradiograph.

The following technique can be used for performing athorough single-contrast barium enema.31 After insertion oftherectaltip,thepatientisplacedintheleftposteriorobliqueposition, the flow of the barium suspension is started slowly,and a spot image of the rectosigmoid region is obtainedwhile distention is minimal. Because the rectum cannot becompressed, this early image allows smaller lesions to bedetectedmoreeasily.Therectosigmoidregionisagainimagedwhen fully distended. An appropriate number of views areobtained to demonstrate the sigmoid colon without overlap-ping loops. The entire colon is then opacified to the cecum,avoiding ileal reflux, if possible. Compression spot imagesof the remaining segments of colon are then obtained. Afterthe fluoroscopic examination has been completed, overheadradiographs may be obtained. When using a remote controlunit, the overheads may be taken during the fluoroscopicportion of the examination. A reasonable sequence of over-heads includes a left lateral view of the rectum, prone andsupine views of the colon, supine left and right anterioroblique views of the colon, and a prone angled view of therectosigmoid.

Figure 2-12  Colonic carcinoma on single-contrast barium enema. This oblique compression spot image of the splenic flexure shows a polypoid, ulcerated carcinoma (arrows) of the distal transverse colon. Careful patient positioning and the use of compression are critical components of this examination. 

Figure 2-13  Small colonic polyp on compression spot image from a single-contrast barium enema. Oblique compression spot image of the splenic flexure shows an 8-mm colonic polyp (arrow). This small polyp was not seen on other images of the same area when compression was not applied. 

F

2  Barium Studies: Single and Double Contrast 31

Figure 2-14  Value of postevacuation radiographs for detecting colonic polyps. A. Compression spot image shows a small (<1 cm) filling defect (arrow) in the descending colon on a single-contrast barium enema. Diverticula are also present in this region. The filling defect was not clearly present on other views. B. Postevacuation compression spot image of the descending colon (note the location of the previously seen diverticula) again shows this small filling defect (arrow), indicating that it is a true polyp. A small adenoma was removed at colonoscopy. 

A B

Figure 2-15  Risk of suboptimal coating. A. On the initial radiograph, an ulcer crater is barely recognizable along the lesser curvature of the stomach. B. With additional rotation and improved coating, the large ulcer crater (arrow) is clearly seen. 

A B

Distention

Normalfoldsaresoftandpliableandarethereforeeffacedwithmoderatedistention.Theoptimaldegreeofdistention is thatwhichjusteffacesthenormalfolds.Inadequatedistentionmayconceallesions,butoverdistentioncanalsoobscurelesionssuchasshallowulcers.Varyingdegreesofdistentionmaythereforebe required for optimal visualization of complex or subtlelesions.Overdistentionmayaccentuateareasofrigidity,whereaspartialcollapsemayaccentuateabnormalitiesofthefolds.Thefinal diagnosis represents a synthesis of the informationobtainedwiththesevariousviews(Fig.2-16).

ProjectionAnadequatenumberofviewsshouldbeobtained,soeachloopof bowel is projected free of overlapping loops. Ideally, eachsegment of bowel should also be demonstrated in profile. Inpractice,however, thesegoals cannotalwaysbeachieved. It is

therefore important to assessoverlapping loopsofbowel andbe able to recognize abnormalities that are viewed en face aswellasinprofile.Thisisparticularlyimportantforrecognitionofshort,annularlesionsinthecolonbecauseitmaybedifficulttodemonstrateeverycolonicbendinprofile(Fig.2-17).

INTERPRETATION

Aftereveryeffortismadetoobtainexcellentimages,itisimpor-tant to extract all of the diagnostic information available ontheseimages.Theinterpretationofdouble-contraststudiesalsodiffers substantially from the interpretationof single-contraststudies.

Dependent and Nondependent SurfacesThedistinctionbetweenthedependentandnondependentsur-facesmustbeunderstood.Thenondependentsurfacehasathincoating of barium because all the free barium falls onto the